1. Kane's Concurrent-efforts View

Kane distinguishes among ‘three freedoms’.Footnote ⁶ He asserts that ‘Free acts may be’:

1. (1) acts done voluntarily, on purpose and for reasons that are not coerced, compelled or otherwise constrained or subject to control by other agents.

2. (2) acts [free in sense 1 that are also] done ‘of our own free will’ in the sense of a will that we are ultimately responsible (UR) for forming.

3. (3) ‘self-forming’ acts (SFAs) or ‘will-setting’ acts by which we form the will from which we act in sense 2.Footnote ⁷

Kane observes that free actions of type 1, as he conceives of them, are compatible with determinism and that free actions of types 2 and 3 are not.Footnote ⁸ All free actions of type 3, as Kane conceives of them, are indeterministically caused by their proximal causes, and only agents who perform free actions of type 3 can perform free actions of type 2.

Here, as in Free Will and Luck, I call any free A-ings that occur at times at which the past (up to those times) and the laws of nature are consistent with the agent's not A-ing then basically free actions.Footnote ⁹ My focus in this article is on basically free actions that are indeterministically caused by their proximal causes. If there are such actions, it is possible that all actual actions of this kind are self-forming actions. But whether an action is self-forming or not depends on its effects on the agent's ‘will’; and if basically free actions are possible, we can imagine basically free actions that immediately precede the agent's death and therefore have no effect on the agent's will.

An agent performs a basically free action A at a time t only if there is another possible world with the same past up to t and the same laws of nature in which he does not do A at t.Footnote ¹⁰ In some cases, the A-ing is an action of deciding (or choosing) to do something or other – deciding to cheat right then, for example. And in many such cases, there is another possible world with the same past up to t and the same laws of nature in which, at t, the agent decides to do something else instead. Reflection on such pairs of worlds has led some philosophers to worry that what the agent decides to do is too much a matter of luck for the agent to be morally responsible for the decision and to have made it freely. In my own formulation of the worry, the cross-world difference in decisions at t is just a matter of luck, and, for example, the agent's deciding at t to cheat is partly a matter of luck.Footnote ¹¹

I have never claimed that the luck here is incompatible with basically free action or with the agent's being morally responsible for the action. Instead, I have offered a solution to the worry – a solution that acknowledges the presence of luck at the time of action.Footnote ¹² Kane offers another solution.Footnote ¹³ His proposed solution also acknowledges the presence of luck: ‘The core meaning of “He got lucky”, which is implied by indeterminism, I suggest, is that “He succeeded despite the probability or chance of failure”; and this core meaning does not imply lack of responsibility, if he succeeds’.Footnote ¹⁴

Elsewhere, I have explained how Kane's libertarian view evolved over the years in response to some worries about luck.Footnote ¹⁵ Here I cut to the chase more expeditiously. Kane's proposed solution to one such worry, in cases in which the actions at issue are decisions, features the idea that the agent simultaneously tries to make each of two or more competing choices or decisions.Footnote ¹⁶ In this article, to keep things relatively simple, I restrict attention to cases in which only two competing choices are in the running. Regarding such cases, Kane claims that because the agent is trying to make each choice, she is morally responsible for whichever of the two choices she makes and makes it freely, provided that ‘she endorse[s] the outcome as something she was trying and wanting to do all along’.Footnote ¹⁷ If Kane is right, he has provided a successful answer to a certain challenge about luck) – at least in scenarios of a certain kind.Footnote ¹⁸

Part of the inspiration for Kane's position is the observation that ‘indeterminism [sometimes] functions as an obstacle to success without precluding responsibility’ and free action.Footnote ¹⁹ In one of his illustrations, ‘an assassin who is trying to kill the prime minister … might miss because’ his indeterministic motor control system leaves open the possibility that he will fire a wild shot. Suppose the assassin succeeds. Then, Kane says, he ‘was responsible’ for the killing ‘because he intentionally and voluntarily succeeded in doing what he was trying to do – kill the prime minister’.Footnote ²⁰ It may be claimed, similarly, that the indeterminism in the scenario does not preclude the killing's being a free action. If these claims are true, they are true even if the difference between the actual world at a time during the firing and any wild-shot world that does not diverge from the actual world before that time is just a matter of luck.

Kane contends that ‘libertarian views in general must try to show that whatever chance may be involved in undetermined choices need not undermine free agency and responsibility’.Footnote ²¹ He also contends that to show this one must go beyond my daring libertarian view (to be described in sections 3 and 4) and defend the claim that ‘the agent makes one set of reasons win out over the other at the moment of choice, so that the agent can be fully responsible for causing it to be the case that one choice rather than the other is made, despite the indeterminism’.Footnote ²² Here he appeals to his concurrent efforts idea: ‘the agent makes one set of reasons prevail over the other by making an effort to do so against the competing effort to make a contrary choice’.Footnote ²³ The assassin succeeds at something he is trying to do, despite the luck involved. Similarly, Kane says, the person who makes dual efforts to choose may succeed – whichever of the two competing choices he makes – at doing something he is trying to do, despite the luck involved. And his succeeding includes his making one set of reasons prevail over the other.

2. Probing Kane's View

As I pointed out elsewhere, although Kane is ordinarily pretty sensitive to the phenomenology of agency, trying to choose to A while also trying to choose to do something else instead seems remote from ordinary experience.Footnote ²⁴ We may occasionally have an experience of trying to bring it about that we choose to A. (For example, someone who knows that it would be best to quit smoking but who has not yet chosen to quit may try to vividly represent to himself the most important reasons for quitting, including the dangers of not quitting, with a view to bringing it about that he chooses to quit.) But how many of us have experienced simultaneously trying to bring it about that we choose a particular course of action and trying to bring it about that we choose a competing course of action instead? If such dual efforts never occur, they never underwrite free choices.

Obviously, someone who grants that we never experience dual efforts to choose may wish to posit such efforts anyway in an attempt to solve a theoretical problem.Footnote ²⁵ But alternative routes to a solution should also be explored. If concurrent efforts of the kind at issue never happen, is an event-causal libertarian unable to explain why some choices that are indeterministically caused by their proximal causes are free? I take up this question in sections 3 and 4.

In response to the phenomenological worry, Kane asserts that ‘introspective evidence cannot give us the whole story about free will’.Footnote ²⁶ But, of course, no one who expresses the worry claims that such evidence can give us the whole story. Instead, people like me have their doubts about whether normal human beings ever simultaneously try to make each of two (or more) competing choices or decisions, and phenomenological considerations are among the considerations we cite in support of the doubts.

I turn from this empirical issue to a theoretical question. Even if people sometimes do make dual efforts of the kind at issue, would that give Kane the result he wants? Consideration of some cases featuring dual efforts of some other kinds will prove useful.

Here is a warm-up case. Ann knows that either of a pair of targets will suddenly disintegrate before a bullet fired at it hits it and that it is undetermined which target will do this. She is promised a prize of ten dollars for hitting target 1 and twenty dollars for hitting target 2. Ann is ambidextrous and an expert with firearms. She fires simultaneously at each target, shooting at target 1 with the pistol in her right hand and at target 2 with the pistol in her left hand. As luck would have it, target 2 disintegrates and Ann hits target 1.

This story differs from stories about Kane-style dual efforts to choose in three potentially noteworthy ways. It is not a story about alternative choices, the agent's efforts end before success is achieved (Ann's efforts end when she pulls the triggers), and the efforts do not hinder one another.

Here is a story in which the second difference is eliminated. Beth is promised a prize of ten dollars for fully depressing the v key on a computer keyboard with her left index finger and twenty dollars for fully depressing the m key on the same keyboard with her right index finger. She knows that one key or the other will stick (and so will not fully depress) and that it is undetermined which. She also knows that if she opts to press both keys she must press them simultaneously in order not to be disqualified. Her plan is to press each key simultaneously. She tries to press the v key all the way down with her left index finger while also trying to press the m key all the way down with her right index finger. As luck would have it, the v key sticks and Beth fully depresses the m key.

The next story eliminates two of the three differences. Cathy's situation is like Beth's except that her index fingers are linked together by a fancy collection of plastic strings and pulleys. Moving either finger downward makes it harder to move the other finger downward. Her fingers are an inch above the keys, and her plan is to try to press each key at the same time. As luck would have it, the m key sticks and she fully depresses the v key.

Here is an obvious point about these cases. It is not up to the agent which target she hits or which key she fully depresses. Is it up to agents what they choose in Kane-style cases of dual efforts to choose? Or is what happens pretty well understood on the model of my third case? In that case, we have two simultaneous attempts, each of which hinders the other, and one of them happens to succeed while the other one happens to fail.

Recall Kane's claim that ‘the agent makes one set of reasons prevail over the other by making an effort to do so against the competing effort to make a contrary choice’.Footnote ²⁷ One might try to do justice to this claim by representing the agent as trying to make reasons R1 prevail over reasons R2 while also trying to make reasons R2 prevail over reasons R1. But this does not capture the idea that the agent is making the former effort ‘against’ the latter effort, if what is being claimed is that the agent represents the former effort as being undertaken against the latter.Footnote ²⁸ If that is what is being claimed, we can say the following: the agent is trying to make reasons R1 prevail over reasons R2 and to prevent his contrary effort from making R2 prevail over R1 while also trying to make reasons R2 prevail over reasons R1 and to prevent his contrary effort from making R1 prevail over R2.

My third case can be modified accordingly. This time, Cathy performs her task while in a brain scanner. She is told that one thing she must do to win either prize is to represent what she is trying to do in a certain way and that the scanner will reveal how she represents her attempts. The relevant part of her instructions read as follows:

After thinking about this for a while and then trying to represent her options to herself in terms of making one reason prevail over another, Cathy reports that she is almost ready. After a bit more thinking – specifically about the idea of her trying to prevent an effort of hers from being successful – Cathy reports that she is ready. According to the scanner, she represents what she is up to in the specified way and, as luck would have it, the m key sticks and Cathy fully depresses the v key.

In this case, as in the earlier story about her, it is not up to Cathy which key she fully depresses. Which key she fully depresses depends on which key gets stuck, and she has no say at all about which key gets stuck.

Here, Kane may say, we have hit on an important difference between Cathy's story and a case of Kane-style dual efforts to choose. In Cathy's story, the outcome hinges on an external event over which she has no control; but in a Kane-style story, all the work is done by the agent's own activity – his dual efforts. How much mileage can one get out of this difference?

Considering the following story from my Free Will and Luck will help answer this question.Footnote ²⁹ Bob lives in a town in which people make many strange bets, including bets on whether the opening coin toss for football games will occur on time. After Bob agreed to toss a coin at noon to start a high school football game, Carl, a notorious gambler, offered him fifty dollars to wait until 12:02 to toss it. Bob was uncertain about what to do, and he was still struggling with his dilemma as noon approached. Although he was tempted by the fifty dollars, he also had moral qualms about helping Carl cheat people out of their money. He judged it best on the whole to do what he agreed to do. Even so, at noon, he chose to toss the coin at 12:02 and to pretend to be searching for it in his pockets in the meantime.

Imagine, if you can, that Bob is trying to choose to toss the coin at noon, as promised, while also trying to choose to cheat and that these efforts are or include efforts to make pertinent reasons prevail over other pertinent reasons and are made ‘against’ each other. In possible world W1 Bob's attempt to choose to cheat succeeds at t. But there is another possible world, W2, with the same past up to t and the same laws of nature in which, at t, Bob's attempt to choose to toss the coin on time succeeds. We get two very different outcomes with no antecedent difference at all.

In W1, what is the status at t of Bob's attempt to choose to toss the coin on time? One possibility is that it has not yet stopped. It is still underway at t, but it is not successful. Another possibility is that this attempt stopped just then. We can say, if we like, that Bob's attempt to choose to cheat rendered this competing, persisting attempt ineffective in the former scenario without stopping it and that it stopped the competing attempt at t in the latter scenario. (Recall Kane's claim that ‘the agent makes one set of reasons prevail over the other by making an effort to do so against the competing effort to make a contrary choice’.)Footnote ³⁰ But we should not lose sight of the point that what is going on is such that, in another possible world with the same laws of nature and the same past right up to t, exactly the opposite happens. There is no difference in the efforts – or anything else, for that matter – before t; and, even so, in W2 Bob's attempt to choose to toss the coin on time succeeds (and, if you like, stops the competing attempt at t or renders it ineffective at t without stopping it then). The difference at t between W1 and W2 seems to be just a matter of luck. And readers would understandably have doubts about claims that the following things were up to Bob: which of his two efforts to choose succeeded; which reasons prevailed; what he wound up choosing when his dual efforts to choose ran their course.

As I see it, to assert that the difference at issue is just a matter of luck is not to assert that Bob's decision is not a basically free action or not something for which he is morally responsible. In fact, one plank in my response to the problem about luck at issue is the following thesis: (LD) Even if the difference between what an agent decides at t in one possible world and what he decides at t in another possible world with the same past up to t and the same laws of nature is just a matter of luck, the agent may make a basically free decision at t in both worlds.Footnote ³¹ In Free Will and Luck, I present what I called the problem of ‘present luck’ in the same spirit that someone who hopes for an adequate explanation of why a perfect God would allow all the pain and suffering that exists may vividly present the problem of evil, and I do it without formulating any argument for a conclusion that is incompatible with LD.Footnote ³² Part of what I asked for, in effect, was a plausible explanation of the truth of LD. Kane has offered the answer that I have been discussing. I will get to the alternative answer that I offered pretty soon.

LD resembles the following thesis about actions in general: (LG) Even if the difference between what an agent does at t in one possible world and what he does at t in another possible world with the same past up to t and the same laws of nature is just a matter of luck, the agent may perform a basically free action at t in both worlds. Recall my stories about Ann, Beth, and Cathy. It is not up to Ann whether she hits target 1 or target 2, and it is not up to Beth and Cathy whether they fully depress the v key or the m key. Even so, readers who believe that basically free actions are possible may well regard the actions at issue – Ann's hitting target 1, Beth's fully depressing the m key, and Cathy's fully depressing the v key – as basically free. And, of course, if these actions are basically free, the same goes for Ann's hitting target 2, Beth's fully depressing the v key, and Cathy's fully depressing the m key in worlds with the same laws and past in which that is what happens.

There are versions of my stories about Ann, Beth, and Cathy in which their actions have moral significance. For example, we can imagine a version of the keyboard stories in which key presses are means of administering painful shocks to kittens. Fully depressing the m key administers a shock to an adorable gray kitten, and fully depressing the v key does the same to an equally adorable white kitten. Beth and Cathy press the m key in an attempt to shock the gray kitten, and they press the v key in an attempt to shock the white kitten. If the kitten-shocking actions are performed freely, might the agents be morally responsible for them? In Kane's view, as I mentioned, the assassin with an indeterministic motor control system is morally responsible for killing the prime minister, something he was voluntarily trying to do. Beth and Cathy also voluntarily try to do what they succeed in doing (but while also trying to shock the other kitten). So when Beth and Cathy shock the gray (or white) kitten, Kane may be happy to say that they are morally responsible for doing that. But what he should not – and presumably would not – say is, for example, that Cathy is morally responsible for the fact that she shocked the gray kitten rather than the white kitten. She lacks moral responsibility for this contrastive fact.Footnote ³³ Her being morally responsible for that fact would require that she is morally responsible both for the fact that she shocked the gray kitten and for the fact that she did not shock the white kitten. And although Cathy's attempt to shock the white kitten failed, she is not morally responsible for its failing nor for her not shocking this kitten. I return to this matter in section 4.

3. Comparing the Competing Views

Libertarianism has a negative and a positive side. The negative side is incompatibilism, and the positive side is the claim that at least some human beings sometimes exercise free will. The positive claim is not merely that free will is possible; it is that free will is actual and actually exercised by real people. Now, any typical libertarian view makes it a necessary condition for a directly free action that there is no time at which it is determined (in the ‘deterministic causation’ sense of determined) that the action will occur.Footnote ³⁴ And any typical event-causal libertarian view holds that although directly free actions are caused, they are not deterministically caused by their proximal causes. Given the positive side of libertarianism, we have here a commitment about actual human beings: namely, that at least some of us sometimes perform actions that are not deterministically caused by their proximal causes. Because, in any case of action, at least some of the proximal causes of actions are internal to agents, the commitment here for typical event-causal libertarians is to there being indeterminism in some action-producing causal streams right up to the time of action (including choice or decision).

Kane certainly seems to hold that dual (or multiple) efforts of the kind he posits are required for directly free choices. This adds a second commitment about actual human beings – at least, those who make directly free choices. This commitment is not merely to what is possible. It is a commitment about what some actual human beings actually do. Kane mentions evidence about dual processing in perception; but, to the best of my knowledge, there is no direct evidence of Kane-style dual attempts to choose in normal human beings.Footnote ³⁵

Two things have kept me from being a libertarian. First, I have not been persuaded by any argument for incompatibilism. Second, for reasons I have set out in Free Will and Luck, I take (a naturalistic) event-causal libertarianism to be the most promising brand of libertarianism, and I do not know of strong evidence that human brains work as they would need to work if a theoretically attractive event-causal libertarian view is true.Footnote ³⁶ Both of these things stand in the way of my endorsing the daring libertarian view that I floated. But, as I observed, all I want to argue here is that event-causal libertarians should prefer that view (DLV) to Kane's view.

My DLV (in Free Will and Luck) is similar to Kane's view. The main difference is that where Kane postulates concurrent competing indeterministic efforts to choose, I postulate an indeterministic effort to decide (or choose) what do. That effort can result in different decisions, holding the past and the laws of nature fixed. For example, in Bob's story, as I tell it, there are no concurrent competing efforts to choose. Instead, there is a possible world in which Bob's effort to decide what to do about the coin toss issues at t in a decision to cheat, and in another world with the same past up to t and the same laws of nature, that effort issues at t in a decision to toss the coin right then. Bob has competing reasons at the time, and the decision he makes – whether it is to cheat or to do the right thing – is made for the reasons that favor it. The cross-world difference at t in what Bob decides seems to be a matter of luck. But it does not seem to be any more a matter of luck than a cross-world difference that I identified in a version of Bob's story in which he is trying to choose to cheat while also trying to choose to do the right thing: namely, the difference between the former effort succeeding and the latter effort succeeding.

Kane contends that choices of the sort at issue – choices that issue from one member of a pair (or group) of competing efforts to choose – ‘are “up to the agent” in the strong sense that the agents have plural voluntary control over whether or not they are made’.Footnote ³⁷ He comments on the nature of plural voluntary control earlier in the same article:

As Kane grants, when an agent satisfies the conditions set out in DLV, ‘we can … say that the choice that results is made by the agent; and we can even say it is voluntary (since uncoerced) and intentional (since knowingly and purposefully made)’.Footnote ³⁹ Moreover, we can correctly say these things both about the choice the agent makes at t in the actual world and about a competing choice he makes at t in another possible world with the same laws of nature and the same past up to t.Footnote ⁴⁰ And from this we can infer that DLV accommodates ‘plural voluntary control’ over choice making. If, as Kane says, plural voluntary control in this connection is sufficient for the choice the agent makes to be up to the agent, then DLV also accommodates it sometimes being up to agents what they choose in scenarios of the sort at issue. The point is that, given Kane's own account of plural voluntary control and his own claim about what is sufficient for an agent's choice to be up to the agent, an agent whose making of a particular choice fits my DLV makes a choice that it was up to him to make.

Even so, Kane finds fault with DLV. He writes: ‘The agent will indeed make one choice or the other at t, but which choice the agent makes depends on which reasons “win out”; and this is undetermined. That the agent decides to do A at t in one world and B in another seems therefore to be a matter of luck or chance’.Footnote ⁴¹

A pair of observations are in order. The first is about victorious reasons. Call Bob's reasons for cheating RC and his reasons for tossing the coin at noon RT. In the present context, what it is for RC to win out is for Bob to choose for those reasons; and if that happens, then, of course, Bob chooses to cheat. Now, in order for Bob's choice to cheat to be basically free – according to Kane's view and my DLV – there must be no time at which it is determined that he will choose to cheat and so no time at which it is determined that he will choose for RC.Footnote ⁴² So Kane cannot, given his own view, treat the point that which reasons will win out is undetermined as incompatible with Bob's making a basically free choice to cheat. The upshot, of course, is that he cannot consistently claim that this point falsifies DLV. If the point falsifies DLV, it falsifies Kane's view too.

My second observation is predictable, given some remarks I have already made. When Bob's choice-making occurs in a way that fits my DLV, the cross-world difference in what he chooses is, in Kane's words, ‘a matter of luck or chance’.Footnote ⁴³ But when Bob's choice-making occurs in a way that fits Kane's concurrent-efforts view, the cross-world difference in which of his efforts to choose wins out is no less a matter of luck or chance. Picture Bob's dual efforts to choose as (mutually interfering) processes aimed at targets or goals.Footnote ⁴⁴ The target at which his effort to choose to cheat aims (T1) is his choosing to cheat, and the target at which his effort to choose to do the right thing aims (T2) is his choosing to do the right thing. In W1 the former process wins out at t: T1 is hit then and T2 is not. And in W2, which has the same past all the way up to t and the same laws of nature, the latter process wins out: T2 is hit then and T1 is not. This difference is no less a matter of luck than the featured cross-world difference when Bob's choice-making accords with DLV. More is going on in Kane's vision of things than in mine: He represents the agent as trying to make each of two different competing choices, and I represent him as simply trying to decide what to do. But the more that is going on in Kane's concurrent-efforts picture does not yield less cross-world luck.

I will say more about DLV shortly, and I will compare the main costs of the two views at issue. But I would like to linger for a while over Kane's idea that, in a dual-efforts scenario, the agent makes some reasons prevail over others.

In my story in which Cathy is in a brain scanner, readers may have a hard time imagining her trying to make her ten dollar reason prevail over her twenty dollar reason. Such an attempt would be perverse, and it seems that all she is really up to is simultaneously trying to fully depress the ten dollar key and trying to fully depress the twenty dollar key while also attempting to represent what she is doing in a certain reason-featuring way. So consider the following case. Donna replaces Cathy in the scanner, each full key press is worth fifteen dollars, and the two keys assign money to two different charities – one for stray dogs and the other for stray cats. The representation instructions that Donna receives reflect this fact, of course. Further details: Donna is very fond of cats and dogs and very interested in helping them, and she believes that the two charities are equally proficient at achieving their aims.

Donna attempts to follow her instructions about representations. The strategy that she endeavors to implement includes her vividly imagining the plight of an adorable stray kitten in an attempt to make her reasons for helping stray cats prevail and vividly imaging the plight of an equally adorable stray puppy in an attempt to make her reasons for helping stray dogs prevail. Her former attempt also includes rehearsing reasons for helping stray cats and the latter includes rehearsing reasons for helping stray dogs. Donna simultaneously presses both keys. In the actual world, she fully depresses the cat key at t, and in another possible world with the same laws and the same past up to the sticking point, she fully depresses the dog key at t. We can say, if we like, that in fully depressing the cat key, Donna made her reasons to help cats prevail. And this claim can be counted as true, if we do not read too much into it. But the truth of the claim is utterly compatible with the difference in the two worlds at the time at issue being just a matter of luck. And, in a Kane-style dual-efforts scenario, the same is true of the difference between Bob's making his reasons to cheat prevail in choosing at t to cheat and his making his reasons to do the right thing prevail in choosing at t to toss the coin. Bear in mind that Bob does not make either set of reasons prevail before he makes his choice. Which reasons prevail is up for grabs until he makes his choice, and the prevailing of a collection of reasons is precisely a matter of Bob's choosing for those reasons – that is, his choosing for reasons RC to cheat or his choosing for reasons RT to do the right thing. Again, in one world one set of reasons prevails at t, and in another world a competing set of reasons prevails at t – and there is no cross-world difference in the reasons, Bob's efforts, or anything else before t.Footnote ⁴⁵

As I observed, Kane contends that his view secures the making of choices that ‘are “up to the agent” in the strong sense that the agents have plural voluntary control over whether or not they are made’.Footnote ⁴⁶ And, as I explained, my own DLV fares no less well in this regard: Agents who fit the latter view can satisfy Kane's sufficient conditions for plural voluntary control over what they choose. Kane can reply that exercising plural voluntary control over what one chooses is not, after all, sufficient for acting freely because an additional requirement for choosing freely is that the agent was trying specifically to choose what he chose.Footnote ⁴⁷ But what contribution does a choice's satisfying this requirement make to its being a free choice? Kane might, by way of analogy, point to the contribution that the assassin's trying to kill the prime minister (in an example mentioned earlier) makes to the killing's being a free action. However, the contribution here seemingly consists in the support the occurrence of the trying offers for the claim that the assassin intentionally killed his victim, presumably as a means to an end;Footnote ⁴⁸ and it is commonly recognized that choosing to A is essentially intentional.Footnote ⁴⁹ Agents do not need to try to choose to A nor to try to bring it about that they choose to A in order to choose to A; and the nature of choosing is such that, whenever they choose to A, they intentionally do so: there are no nonintentional choosings to A. If trying to choose to A is supposed to make a contribution to freely choosing to A that goes beyond its contribution to exercising plural voluntary control over what one chooses, Kane has not said what that contribution is.

4. The Two Views: Relative Costs

I have made two main points that bear directly on this article's thesis.

1. 1. On cross-world luck. When an agent's choice-making occurs in a way that fits my DLV, the cross-world difference in what he chooses is, in Kane's words, ‘a matter of luck or chance’.Footnote ⁵⁰ When the agent's choice-making occurs in a way that fits Kane's concurrent-efforts view, the cross-world difference in which of his efforts to choose wins out is no less a matter of luck or chance.

2. 2. On empirical burdens. Kane's concurrent-efforts view requires more for basically free decisions than DLV does. It requires that the agent simultaneously makes competing efforts to choose. Ordinary experience supports the claim that normal human agents sometimes make an effort to decide what to do. The same cannot plausibly be said for the claim that agents sometimes make concurrent efforts to choose of the kind featured in Kane's view. And, to the best of my knowledge, there is no direct evidence of any kind that normal agents ever make Kane-style concurrent efforts to choose.

The conjunction of 1 and 2 is motivation for this article's thesis – that event-causal libertarians should prefer DLV to Kane's concurrent-efforts view. Kane's view has a significantly heavier burden – and therefore carries a significantly higher cost – on the empirical front, and, as far as I can see, it has no advantage over DLV on the issue of cross-world luck at the time of choice or decision.

Regarding point 1, some people may worry that DLV leaves more room for wildly improbable choices than Kane's view does. But the worry is unfounded. Presumably, a typical Kane-style agent's character, values, beliefs, learning history, and the like constrain the options that it is open to him to choose, and there is nothing to prevent these factors from being equally constraining in an agent who fits DLV.

Regarding point 2, I have heard it said that Kane is not seriously claiming that agents make simultaneous competing efforts to choose – that all he really has in mind is that the agent has reasons or motives for two or more competing choices and chooses for reasons. Quotations from Kane's work that I have provided here should make it clear that this interpretation is far off the mark, as should a little reflection on the use to which Kane puts his assassin analogy (discussed earlier).Footnote ⁵¹ Moreover, on DLV, agents have reasons or motives for two or more competing choices and choose for reasons in cases of the kind at issue; and Kane makes it clear that he is not willing to settle for this, as I have observed.

Another difference between DLV and Kane's concurrent-efforts view merits mention. Kane tries to find a solution to a worry about present luck in what is happening at and around the time of a decision. According to DLV, not everything that is needed for a solution to the worry can be found there; we should also look back in time. A detailed discussion of what a daring libertarian hopes to find in agents’ histories is beyond the scope of the present article. My DLV finds in reflection on agents’ pasts a partial basis for an error theory about why some people may view cross-world luck at the time of decision as incompatible with making a basically free decision.Footnote ⁵² Brief attention to this issue is in order.

My error theory is for a limited audience – people who are attracted to libertarianism and reject agent-causal and non-causal libertarianism. When some such people reflect on stories like that of Bob and the coin, they may ignore the sources of the antecedent probabilities of Bob's choosing to cheat and his choosing to do the right thing. If it is imagined that these probabilities come out of the blue, Bob may seem to be adrift in a wave of probabilities that were imposed on him, and, accordingly, he may seem not to have sufficient control over what he chooses to be morally responsible for his choices. But, as I have explained elsewhere, it is a mistake to assume that ‘indeterministic agents’ probabilities of action are externally imposed’ or that such agents ‘are related to their present probabilities of action roughly as dice are related to present probabilities about how they will land if tossed’.Footnote ⁵³ If it is known that Bob's pertinent probabilities shortly before noon are shaped by past intentional, uncompelled behavior of his, one may take a less dim view of Bob's prospects for being morally responsible for the choice he makes and his prospects for making it freely.

This is a long story that carries us all the way back to candidates for young agents’ earliest basically free actions, and I cannot do justice to it here.Footnote ⁵⁴ I will touch upon just one further issue and then wrap things up.

Recall my observation at the end of section 2 that Cathy is not morally responsible for the contrastive fact that she shocked the gray kitten rather than the white kitten because she is not morally responsible for the fact that she did not shock the white kitten. Cathy deserves no moral credit for the failure of her effort to shock the white kitten. What about Bob in a story in which what goes on is captured by my DLV? Might he be morally responsible for deciding to cheat rather than deciding to do the right thing. Well, if, past intentional, uncompelled behavior of his played a significant role in shaping his character and the antecedent probability that he would decide to cheat, and if better behavior was open to him on many relevant occasions in the past, behavior that would have given him a much better chance of deciding to do the right thing on this occasion, then maybe so. But this, as I say, is a long story that I have spun elsewhere.Footnote ⁵⁵

I mentioned that my error theory is for a limited audience. Some people who regard DLV as lacking the resources to provide what is needed for basically free action may require something for free action that no event-causal theory can give them: luck-excluding control over what they choose in a scenario in which it is at no time determined what they will choose.Footnote ⁵⁶ But Kane is not such a person. He means to get by with an event-causal view of action (including choice) production, and he acknowledges the presence of luck in cases of basically free actions.Footnote ⁵⁷ What I have argued here is that anyone with Kane's aspirations – and any event-causal libertarian – should prefer my daring libertarian view to Kane's concurrent-efforts view.Footnote ⁵⁸