A Pragmatic Approach to Libertarian Free Will 9781138498037, 9781351017275

Citation preview

A Pragmatic Approach to Libertarian Free Will

A Pragmatic Approach to Libertarian Free Will argues that the kind of free will required for moral responsibility and just desert is libertarian free will. It is a source of great controversy whether such a libertarian view is coherent and whether we should believe that we have such free will. This book explains and defends Robert Kane’s conception of libertarian free will while departing from it in certain key respects. It is argued that a suitably modified Kanean model of free will can be shown to be conceptually coherent. In addition, it is argued that while we lack sufficient epistemic grounds supporting belief in the existence of libertarian free will, we may still be justified in believing in it for moral reasons. As such, the book engages critically with the works of a growing number of philosophers who argue that we should jettison belief in the existence of desert-grounding free will and the practices of praise and blame and reward and punishment which it supports. John Lemos is the Joseph McCabe Professor of Philosophy at Coe College in Cedar Rapids, Iowa. He is the author of Commonsense Darwinism: Evolution, Morality, and the Human Condition (2008) and Freedom, Responsibility, and Determinism: A Philosophical Dialogue (2013). He has also published over 30 articles in various philosophical journals, such as The American Catholic Philosophical Quarterly, Dialectica, Law and Philosophy, Metaphilosophy, Philosophia, and The Southern Journal of Philosophy.

Routledge Studies in Contemporary Philosophy

Nature and Normativity Biology, Teleology, and Meaning Mark Okrent Formal Epistemology and Cartesian Skepticism In Defense of Belief in the Natural World Tomoji Shogenji Epistemic Rationality and Epistemic Normativity Patrick Bondy From Rules to Meanings New Essays on Inferentialism Edited by Ondřej Beran, Vojtěch Kolman, and Ladislav Koreň Toleration and Freedom from Harm Liberalism Reconceived Andrew Jason Cohen Voicing Dissent The Ethics and Epistemology of Making Disagreement Public Edited by Casey Rebecca Johnson New Directions in the Philosophy of Memory Edited by Kourken Michaelian, Dorothea Debus, and Denis Perrin A Pragmatic Approach to Libertarian Free Will John Lemos For more information about this series, please visit: www.routledge.com

A Pragmatic Approach to Libertarian Free Will

John Lemos

First published 2018 by Routledge 711 Third Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2018 Taylor & Francis The right of John Lemos to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-1-138-49803-7 (hbk) ISBN: 978-1-351-01727-5 (ebk) Typeset in Sabon by Apex CoVantage, LLC

For those who raised me, my father, Ramon, and mother, Mamie Lou, as well as my three older brothers—Noah, Bill, and Chris

Contents

Acknowledgments

viii

1

Introduction

1

2

Kane’s Libertarian Theory

8

3

Libertarianism, Compatibilism, and Manipulation Arguments

29

4

Kanean Libertarianism Examined

57

5

A Consideration of Alternative Event-causal Libertarian Models of Basic Free Actions

83

Moral Responsibility Denial and the Problem of Punishment

113

7

Libertarian Free Will and Moral Obligation

133

8

Hardheartedness and Libertarianism

151

9

Conclusion

170

Bibliography Index

173 181

6

Acknowledgments

This book could not have been written without the help and support of various institutions and persons. Were it not for the dedication of various teachers at various institutions, I could not have written this book. So thanks are due to the Miami-Dade County Public Schools and the many fine teachers and coaches I had in that school system. Thanks are also due to the University of the South where I received my B.A. and Duke University where I received my Ph.D. I am especially grateful to the wise teaching and counsel of my undergraduate advisor, Professor Bill Garland, and my doctoral dissertation advisor, Professor Michael Ferejohn, both of whom did much to inspire my love for philosophy while patiently showing me how to do it well. I have also benefitted greatly by holding a full-time teaching position at Coe College for many years, where I work with two other philosophers, Jeff Hoover and Peter McCormick, who have offered me various kinds of support and encouragement over the years. At Coe, I’ve been supported by its generous sabbatical program as well as an endowed chair, the Joseph McCabe Chair in Philosophy. Besides these institutions and persons, I’ve also benefitted greatly from advice and encouragement from many other philosophers. There is no doubt that I owe my greatest debt to Robert Kane. I believe I began email correspondences with him in March 2007. I had always been attracted to libertarian views of free will and had even defended the libertarian view in one of the chapters of my first book, Commonsense Darwinism: Evolution, Morality, and the Human Condition (2008). After completing that book, my attention turned to full-time consideration of the problem of free will and I am so fortunate to have found a wise and gracious kindred spirit in Bob Kane. While we don’t agree on all matters surrounding the free will issue, we do agree on much and I am forever in his debt for the patient and thoughtful advice he has given me over the years. In addition to my correspondences and conversations with Bob Kane, I have also benefitted from conversations and correspondences with many other philosophers; among these are Joseph Campbell, Gregg Caruso, Randolph Clarke, Michael Corrado, Richard Double, Laura Ekstrom, Ryan Lake, Neil Levy, Michael McKenna, Alfred Mele, Timothy O’Connor, Derk Pereboom, Saul Smilansky, and Bruce Waller. I

Acknowledgments

ix

have probably left out some philosophers who deserve my thanks here, but I know that I have spoken with or corresponded with each of these I list here and have benefitted from them in various ways, whether through their advice or encouragement or both. Perhaps my greatest debts are owed to family. For instance, there is the love and support I was offered by my parents and brothers growing up. It is to them that I dedicate this book. However, I owe as much gratitude to my wife, Laura, and my three children—Rosa, Billy, and Michael—who so patiently put up with my work schedule and my philosophical pursuits. Oddly enough, they do so with good cheer! Finally, it should be noted as well that permission has been granted for the use of all or parts of the following previously published works of mine: “Kanian Freedom and the Problem of Luck,” The Southern Journal of Philosophy 45 (2007): 515–532; “Kane’s Libertarian Theory and Luck: A Reply to Griffith,” Philosophia 39 (2011): 357–367; “Wanting, Willing, Trying and Kane’s Theory of Free Will,” Dialectica 65 (2011): 31–48; “Hardheartedness and Libertarianism,” Philo 16 (2013): 180–195; “Self-forming Acts and the Grounds of Responsibility,” Philosophia 43 (2015): 135–146; “A Kantian Defense of Libertarian Blame,” Journal of Cognition and Neuroethics 3 (2015): 251–263; “Moral Concerns About Responsibility Denial and the Quarantine of Violent Criminals,” Law and Philosophy 35 (2016): 461–483; and “Hardheartedness and Libertarianism Again: A Rejoinder to Double,” Journal of Philosophical Research 42 (2017). doi:10.5840/ jpr2017511104.

1

Introduction

Jack is twenty years old and he lives at home with his parents. He has friends that are going on vacation far away and they have invited him to go with them, if he can pay his own way. But Jack doesn’t have the money for the trip, and so he decides to break into a neighbor’s house and steal property worth several thousand dollars to pay for a vacation with his friends. It is my view that Jack is morally responsible for stealing this property if and only if (1) he knew or should have known that stealing is wrong and (2) he stole the property of his own free will. Let us call the first condition of his moral responsibility the knowledge condition and the second condition the free will condition. What I am saying then is that meeting the knowledge condition and the free will condition is both necessary and sufficient for being morally responsible for the commission of an act. Furthermore, when I say that a person, P, is morally responsible for an action, X, what I mean is that if X is a good/right act then P is a deserving subject of our praise for doing X and if X is a bad/wrong act then P is a deserving subject of our blame for doing X. Thus, on my view, assuming that Jack’s stealing of the property was wrong and assuming that he was morally responsible for stealing the property, then Jack is deserving of blame for stealing the property. Additionally, if he is deserving of blame for stealing the property, he may also, though not necessarily, be deserving of punishment for stealing the property. It is my view that normal adult human beings often engage in actions for which they are morally responsible. Thus, it is often the case that when normal adult human beings act rightly they deserve praise for their behavior and when they act wrongly they deserve blame for their behavior. When they act wrongly but they fail to meet either the knowledge condition of moral responsibility or the free will condition, then they are not morally responsible for their action and do not deserve blame for their action. On my view, acting with free will is a necessary condition for moral responsibility and, thus, deserved praise or blame. It is a matter of great and long-standing controversy what free will consists in and whether or not human beings possess the kind of free will that can make us morally responsible for our actions. Most human beings and

2

Introduction

most philosophers for that matter believe that we do have the kind of free will which can make us morally responsible for our actions. But some people deny that we do have this kind of free will. Philosophers who deny that we have the kind of free will which makes us morally responsible give different kinds of reasons for saying this. In what follows I will speak about some of the different reasons that philosophers have given for the view that we lack the kind of free will which makes us morally responsible. However, before doing so I want to make a terminological point. In this book, I am interested in questions about the nature and existence of the kind of free will that makes us morally responsible. There may be some other kinds of free will, such as the freedom to do what one wants. But some deeper kind of freedom is required if one is to be morally responsible for one’s action. For instance, suppose Jack is a kleptomaniac and when he steals the property from his neighbor he knows he is acting wrongly but he is moved to steal by an overwhelming, compulsive desire to steal which he would rather not act on. If his free will is simply the freedom to do what he wants, then, since he does desire to steal the property, we’d have to say that Jack is morally responsible for stealing the property. However, his desire to steal is an alien desire and he’d rather not be moved to act on it. So, it is wrongheaded to say he’s morally responsible in this case. Or suppose Jack has never in his life wanted to steal anything. But last night an evil and savvy neuroscientist secretly manipulated his brain functioning so that today he wants to steal his neighbor’s property, and so he does. Again, if free will is simply the freedom to do what one wants, then in this case we’d have to say Jack steals of his own free will. But, again, this is wrongheaded, as he is subject to the manipulations of the evil neuroscientist. Thus, the kind of free will required for moral responsibility must be more complex than merely having the freedom to do what one wants. Again, to be clear let me just say that in this book when I speak of free will I mean the kind of free will that is required for moral responsibility. Thus, by using the term “free will” I am referencing something more complex than the mere freedom to do what one wants. Rather, I am referencing that kind of freedom which when exhibited by an agent who knows that his act is wrong or should know it is wrong, he is then morally responsible for his action and a deserving subject of blame for his action. Now, as noted some philosophers say that we lack the kind of free will that I am talking about and they give different reasons for saying this. Some of them are hard determinists. Hard determinists say that determinism is true and because of this we lack the kind of free will that can make us morally responsible. Determinism is the view that all events, including human actions and decisions, are the necessary consequences of prior states and events and the laws of nature. As such, determinists are committed to the view that at any time the universe has exactly one physically possible future. If determinism is true, then just as striking fragile glass with a solid, fast-moving object necessitates its breaking, so too all human actions and

Introduction

3

decisions are necessitated by the mental states of the agents in the moments leading up to their actions and decisions. Furthermore, on the determinist view, the mental states which causally determine an agent to act at time t were themselves causally determined by some earlier states and events at t-1 which were themselves causally determined by earlier states and events at an earlier time t-2 and so on, going back in time to events which occurred prior to the birth of the agent. Regarding our case of Jack who stole from his neighbors, the determinist might say that given the laws of nature and given the state of Jack’s mind in the moments leading up to his decision to steal he was necessitated to steal the property from his neighbor’s home. This by itself need not be taken to imply that Jack lacked free will in acting or that he is not responsible for stealing the property. But the hard determinist will be quick to note that if all events are determined in this sort of way and if every action that Jack has ever committed has been determined in this way then it is hard to see how he could ever act freely or ever be morally responsible, since everything he will have ever done will have been a necessary consequence of prior states and events and the laws of nature.1 Some other philosophers who deny that we have free will are hard incompatibilists. Unlike hard determinists, the hard incompatibilist’s denial of free will and moral responsibility is not grounded in the belief that determinism is true. Rather, the hard incompatibilist says that free will is incompatible with determinism and there is no free will. That is, according to the hard incompatibilists, if determinism is true then there is no free will and if determinism is not true there is still no free will. The hard incompatibilist thinks no one has free will whether determinism is true or not. The reader might initially find hard incompatibilism to be an odd view. The reader might think if determinism is not true and if some of my decisions/actions are causally undetermined, then not all of my actions are necessitated by prior states and events and so there may be a number of occasions in which I act where I could have done otherwise and so I can act with free will. But this is to move too quickly and the hard incompatibilist would be quick to note this, pointing out that just because determinism is false it would not necessarily mean that there are any undetermined human actions and even if there were it would not necessarily mean that undetermined human actions are the kind of actions that could rightly be described as free willed actions—undetermined human actions might just be random happenings that are not under our control in such a way as to be free willed.2 In response to those who say that we lack the kind of free will which makes us morally responsible, philosophers tend to take one of two different possible paths, defending either compatibilism or libertarianism. Compatibilism is the view that free will and moral responsibility are compatible with determinism. Thus, according to compatibilism, even if all events, including human actions and decisions, are causally necessitated by the laws of nature and prior states and events, human beings may still possess the

4

Introduction

kind of free will which makes them morally responsible for their actions and decisions. Furthermore, to be morally responsible for an action is to be a deserving subject of blame if the action is bad and a deserving subject of praise if the action is good. Thus, if compatibilism is true, then praising and blaming agents for what they do may be appropriate even if determinism is true. On this view, even if every action that Jack ever performed was causally determined by the joint effects of the laws of nature and prior states and events, he could still rightly be said to have freely stolen the property from his neighbors; and assuming he does so freely and meets the knowledge condition he would then be morally responsible for the theft and deserving of our blame.3 In contrast, libertarians believe that free will is incompatible with determinism and that we have the kind of free will which makes us morally responsible. As such, the libertarian is committed to the view that some of our free willed acts are causally undetermined acts. On this view, not all of our free willed acts must be causally undetermined; rather, only some of them must be undetermined. For instance, suppose Jack’s stealing of his neighbor’s property was causally determined by his character or motivational states in the moments leading up to his decision to steal. Even so, the libertarian will say that this could nonetheless be a free willed act if the act is a product of a character or motivational states which he has formed through prior causally undetermined free willed acts.4 In this book, I will argue that the libertarian view of free will is the view of free will that we should accept. This view is often criticized on two main grounds. First, it is often said that it is an incoherent view—that is, there is no way to make sense of how causally undetermined actions can be free willed acts. Second, it is said to be inconsistent with our empirically supported modern scientific world view. In response, I will argue that the libertarian view is indeed a coherent view and that it can be made to fit with our scientific understanding of the world. Besides the previous two challenges that my defense of libertarianism must face, there is a third: the existence problem, the problem of knowing whether in fact anyone has libertarian free will. One of the great benefits of adopting the compatibilist theory of free will is the fact that it does not face the same obstacles regarding the existence question. Both compatibilists and libertarians believe that many of our decisions and actions are causally determined. Given the kinds of conditions which compatibilists believe are sufficient for free willed action, it is indisputable that we sometimes meet such conditions. As such, if they are correct in their account of the conditions of free will, then it will follow that we have free will. But libertarians believe that free will requires that some of our free actions be causally undetermined. Thus, since it’s not clear that any of our actions are causally undetermined, libertarians face a difficulty establishing that we have free will even if they can give a coherent account of the nature of free will. I will argue that the libertarian view is coherent and that the occurrence of causally

Introduction

5

undetermined free acts is consistent with a scientific world view, but this does not amount to evidence that we actually have such free will. It is my view that we simply don’t have the empirical, scientific evidence for the existence of such free will yet nor do we have any logical or metaphysical evidence of its existence. Nonetheless, I will argue that (1) compatibilism is false and (2) given the moral importance of the belief in free will and moral responsibility, we should accept a libertarian view. In this book, I will proceed as follows: in Chapter Two I will provide a preliminary sketch of the kind of libertarian view I seek to defend. I favor the kind of libertarian view which has been developed and defended by Robert Kane. His theory has been much discussed in the literature and subjected to many published criticisms. I believe that many of these criticisms are answerable while others do raise some serious problems which suggest that Kane’s view needs modification. In Chapter Two, I merely try to give a clear and accurate general account of the nature of Kanean libertarianism and I contrast it with alternative kinds of libertarian views in hopes of making more clear some of the distinctive features of the view. In later chapters, I consider various criticisms of the view and I suggest ways in which the view should be modified to avoid the more significant problems which the view faces. Chapter Three discusses manipulation arguments. In his writings, Kane motivates his consideration of the libertarian view by arguing that compatibilist views of free will fall prey to manipulation arguments. Some critics of Kane’s view have argued that his own view falls prey to manipulation arguments. In Chapter Three, I explain how Kane uses manipulation arguments to criticize compatibilist views and I also consider various manipulation arguments which have been made against Kanean libertarianism. I go on to argue that his view can be defended against such manipulation arguments and that compatibilism does, indeed, continue to be plagued with problems of manipulation. My defense of Kane’s view in this chapter helps us better see why it is that compatibilism is plagued with manipulation problems while libertarian views are not. Chapter Three is intended to help motivate the more extensive discussion of libertarianism in later chapters of the book. In Chapters Four and Five I look at various criticisms of Kanean libertarianism and I respond to them. In Chapter Four I begin by considering various criticisms which can be answered without modifying the Kanean view. In the later parts of this chapter, I consider two criticisms which present more serious worries for the Kanean view and I show why it is that, despite his attempts to answer them, such criticisms continue to raise serious doubts about the coherence of his view. In Chapter Five, I examine some other libertarian views which are most similar to Kane’s and which avoid the problems Kane’s view faces. I argue that each of these views suffers problems of their own which are instructive in helping us understand how Kane’s view would need to be modified to make it more defensible. In the later parts of Chapter Five, I explain two different new ways in which Kane’s theory could

6

Introduction

be modified to avoid the problems originally considered in the latter half of Chapter Four. I provide some support for each of the new models I propose, but I remain uncertain as to which of them is more promising. The last three chapters of the book develop what I call a “pragmatic justification” of libertarianism. It is my view that at this point in time we simply don’t have sufficient experimental/empirical evidence nor sufficient metaphysical nor logical nor intuitive evidence to establish that libertarian free will exists. At the same time, for reasons considered in Chapter Three of this book, we have reason to think that compatibilism cannot give an adequate conception of free will. Thus, as I see it, we are faced with a choice between living and acting as if we have the kind of free will which supports moral responsibility and the practices of praise and blame or living and acting as if we lack such free will. This is a kind of existential question, and in response to it I argue that there are strong pragmatic/moral reasons for us to live and act as though we have libertarian free will. In the last three chapters of the book, I argue that, even though we lack the empirical/metaphysical/intuitive evidence to prove we have it, to deny that we have the kind of free will which makes us morally responsible in the just desert sense threatens human dignity and weakens our concept of moral obligation. In Chapter Six, I consider the ways in which denying just deserts responsibility threatens human dignity in the realm of criminal justice. In Chapter Seven, I consider why the belief in free will is important to defending a strong sense of moral obligation which can help motivate moral behavior. And in the last chapter of the book I consider a significant challenge to the kind of pragmatic libertarianism I endorse—the problem of hardheartedness—and I respond to it.

Notes 1. Hard determinism is rarely defended these days. However, the view has been defended in the seventeenth century by Baruch Spinoza ([1677] 1992), in the eighteenth century by Baron Paul d’Holbach (1770) and in the twentieth century by Paul Edwards (1958). Most contemporary philosophers who deny that we lack free will reject hard determinism because they believe it is no longer clear that determinism is true. The findings of quantum physics raise serious doubts about the truth of determinism. 2. This view has, perhaps, most famously been defended by Derk Pereboom (2001, 2007a, b, 2014), as well as Pereboom and Caruso (forthcoming). In Pereboom (2001) when he describes his view as “hard incompatibilist,” he seems to want to separate his view from other figures that I would include under this heading. I wish to include other figures in the hard incompatibilist camp, such as Gregg Caruso (2016), Saul Smilansky (2000), Galen Strawson (1986), and Bruce Waller (1990, 2011, 2015). I include these others, because like Pereboom, they all believe (1) free will is incompatible with determinism and (2) we lack free will. Pereboom wants to use “hard incompatibilism” as a distinct term for a view like his which (a) allows that there is a coherent concept of free will which could make sense of moral responsibility but also maintains (b) given the scientific evidence we have reason to believe such free will does not exist. In contrast, someone like Strawson finds the very concept of free will to be incoherent. Also, I would note that Bruce

Introduction

7

Waller, actually thinks, we have a kind of freedom of will, but it is not a kind of free will that makes us morally responsible for our actions. Thus, given the way I am defining free will in this book, as the kind which makes us morally responsible, I include him among these other figures whom I list as hard incompatibilists. 3. There are very many defenders of the compatibilist view on the nature of free will and moral responsibility. I cannot possibly list them all here. Some historically famous defenses of the view are Thomas Hobbes ([1651] 1958), David Hume ([1748] 1974), and John Stuart Mill (1874). Some notable recent and contemporary defenders of the view are A. J. Ayer (1954), Daniel Dennett (1984a, b, 2003), Owen Flanagan (2002, Ch.4), Harry Frankfurt (1969, 1971), G.E. Moore (1912), Kai Nielsen (1971), J.J.C. Smart (1963), Gary Watson (1975), and Susan Wolf (1987, 1990). I would also note that a related but slightly different view is that of John Martin Fischer and Mark Ravizza. They defend a view they call “semicompatibilism.” This is the view that free will is not compatible with determinism, but moral responsibility is compatible with determinism. See Fischer (1994, 1999, 2006) and Fischer and Ravizza (1998). 4. Libertarian views on free will have been defended by a number of philosophers. In the eighteenth century, the view was defended by George Berkeley ([1710] 1998) and Thomas Reid ([1788] 1969). The view comes in different varieties. The agent-causal libertarian view was defended by both Berkeley and Reid, and in more recent times the view has been defended by Roderick Chisholm (1964, 1976), Timothy O’Connor (2000), William Rowe (1991, 2000, 2006), Richard Taylor (1966, 1992). I would note as well that while he does not endorse the agent-causal view, Randolph Clarke (2003) does, nonetheless, provide an extensive defense of the view. The noncausal (or simple-indeterminist) libertarian view has been defended by Carl Ginet (1990, 1997), Stewart Goetz (1988, 1997), and Hugh McCann (1998). And the event-causal variety has been defended by Mark Balaguer (2010), Laura Ekstrom (2000, 2003), David Hodgson (2012), and Robert Kane (1985, 1996, 2002a, 2007a, b, 2011, 2014a, b, 2016a, b). Also, Daniel Dennett (1978) and Alfred Mele (1995, 2006) present what they regard as plausible models of event-causal libertarianism without endorsing them.

2

Kane’s Libertarian Theory

My views on the nature of free will and moral responsibility are grounded in an understanding of Robert Kane’s libertarian theory of free will. He has explained and defended his theory in a number of published works. He has given two full-length defenses of his view in Free Will and Values (1985) and his classic The Significance of Free Will (1996). Over the years, he has further developed and defended his view in numerous articles and book chapters.1 In later chapters, I will explore the depths of his view further by considering and responding to various criticisms of it. Given the path that my discussion will lead in later chapters, it is fundamental that we start with a clear and accurate account of the fundamental elements of Kane’s view.

Kane’s Theory Contrasted  With Agent-causal and Noncausal Libertarianism Libertarian theories of freedom and responsibility are committed to the idea that normal human beings sometimes act freely and that, typically, such free acts are acts for which they are responsible. Libertarians also believe such freedom and responsibility are incompatible with causal determinism, the belief that all events, including human actions, are causally necessitated by earlier events. Consequently, as a libertarian, Robert Kane is committed to the view that sometimes human beings act freely and responsibly and that such freedom and responsibility are incompatible with determinism. As a libertarian, Kane believes that some of the free acts we perform and for which we are responsible must be causally undetermined. Kane’s theory is often referred to as an “event-causal” libertarian theory. His theory gets this name because it differs with traditional agent-causal libertarian theories and noncausal libertarian theories. Traditional agentcausal libertarian theories understand our causally undetermined free acts as actions caused by the person, the agent, who performs the act; and such theories understand agent-causation of an act as inconsistent with eventcausation of the act. So, for instance, on this view if I agent cause my act X, then I cause X and no event internal or external to me causes it. Thus, on traditional agent-causal views neither my beliefs nor my desires nor my

Kane’s Libertarian Theory

9

efforts of will nor any combination of these causes my action when I commit a causally undetermined free act—rather, I alone cause it and I do so as an uncaused cause of my act.2 In contrast, noncausal libertarian theories view causally undetermined free acts as entirely uncaused. On this view, when a person commits a causally undetermined free act he acts for certain reasons but neither he nor his reasons nor anything else causes the action. According to such theories, uncaused free acts are still performed for certain reasons and because of this they are not random occurrences; rather, uncaused free acts are rationally explicable and we are responsible for them.3 The difference between Kane’s theory and the noncausal theories is straightforward; unlike the noncausal theories, Kane’s theory views our causally undetermined free acts as caused acts. The difference between Kane’s theory and traditional agent-causal theories is more complex. This is where the title used to describe Kane’s view—“event-causal libertarianism”—can be misleading. For Kane does not believe it is correct to say that our causally undetermined free acts are merely caused by events going on internal to the agent. Rather, he believes our causally undetermined free acts are caused by the agents who perform them, but they are not caused by them in such a way that excludes event-causation altogether. That is, unlike traditional agent-causal theories, Kane does not believe that in the causally undetermined free performance of an act, X, the agent’s causing of X is inconsistent with the fact that his doing of X is at the same time caused by events internal to him, such as his exerting of the effort to X. To better understand this last point, let’s consider one of Kane’s favorite examples of a basic, underivatively free, causally undetermined act—the kind of act he calls a “self-forming act (SFA).” Imagine a businesswoman is on her way to an important business meeting. While on her way, she sees an assault taking place in an alley. Upon seeing this, she feels torn between the desire to go on to her meeting and the desire to stop and prevent the assault. Since she cannot do both she must deliberate and choose between them. It is Kane’s view that in situations like this in which we are strongly motivated to do each of the options we are considering, we exert effort or exercise willpower to do each of the things we are considering doing. So, as the businesswoman deliberates between her two options, she exerts effort or willpower to go on to her meeting and exerts effort or willpower to prevent the assault. Her ultimate choice between these two options will be the resolution of a conflict in her will and whichever way she decides her choice will be caused by one of those efforts. If she decides to prevent the assault, that choice will be caused by her effort to do so. If she decides to go on to her meeting, that will be caused by her other competing effort. On Kane’s view, causally undetermined, basic free acts (SFAs) are caused by mental events (efforts of will) occurring in the agents in the moments leading up to choice. Thus, in a sense it is correct to say that the businesswoman’s causally undetermined free choice will be caused by an event going on internal to her. But to say this can be misleading, because on Kane’s view it is also true

10

Kane’s Libertarian Theory

that the businesswoman’s choice will be caused by her, since her choice will be a product of her effort either way. There is agential involvement in the decision here reflected in the fact that she exerts the effort that leads to her choice; the choice is not something that merely happens to her, she brings it about through her effort. Thus, Kane embraces a kind of agent-causation, albeit one that is nontraditional since it allows that such agent-causation is consistent with event-causation. Kane is explicit about this agent-causal aspect of his view in some of his writings. For instance, consider the following passages: Agents on my view cause or bring about their self-forming choices or SFAs and they do so voluntarily by making efforts of will. These efforts of will are in turn causally influenced by the agent’s motives, reasons, and other states of mind, but their outcomes are not determined by these motives and reasons. The efforts may fail. But if the efforts succeed, as I argued, the agents will have voluntarily brought about the choices as a result of their efforts and hence can be held responsible for doing so. (Kane 2007b, 173) And, [I]n making efforts of will to choose in terms of their reasons and motives, agents do play a causal role in bringing about their choices over and above the causal role played by their reasons, motives, intentions, and other mental states alone. In other words, it’s not as if the agents sit back and watch while the reasons or motives cause the choice. It is rather that, by making efforts, the agents actively bring about the choices for the reasons and motives. Efforts are different from desires and other motives in this respect . . . because efforts are actions of the agents and not merely states. (Kane 2007b, 174) These ideas are developed even further in Kane (2011, esp. 395–397). As I’ve just shown, Kane’s view can be understood as a kind of nontraditional agent-causal view, but I will follow custom and refer to his view as an “event-causal” libertarian view. It is best to do so, as he clearly thinks our basic free willed actions are caused by events and traditional agent-causal libertarian views deny this insisting that our basic free willed actions are caused solely by the persons, the agents, who perform them. It is nonetheless important to understand that when Kane says our basic free willed acts are caused by events he does not intend to suggest that they are not agent-caused. On his view, such acts are agent-caused; they are just not agent-caused in a manner inconsistent with their also being event-caused. Kane rejects traditional agent-causal and noncausal libertarian theories, because he wants to show that libertarian free will can be made intelligible and

Kane’s Libertarian Theory

11

consistent with what is known in modern science about human nature and the cosmos without having to appeal to mysterious forms of agency or causation as past libertarian theories have done. A modern scientific world view understands the world as operating in accordance with cause and effect. For any event that occurs, it is assumed there is a cause for that event and that the cause is another event. Both traditional agent-causal libertarianism and noncausal libertarianism are inconsistent with such a scientific perspective. The former conflicts with the scientific assumption that all events are caused by other events and the latter conflicts with the scientific assumption that all events are caused. In contrast, Kane believes that in order for us to possess free will and to be ultimately responsible for our actions some of our free actions must be the product of indeterministic causal processes occurring in the brains of the agents who perform them. He believes that we are responsible for such free actions and for many other actions that are causally determined by our motivational states and/or character. On his view, we are responsible for the latter determined actions when they stem from a character that has been formed by earlier indeterministically caused free acts. Such earlier actions he calls “self-forming actions” (SFAs) or “selfforming willings” (SFWs).

Alternative Possibilities, Self-forming Acts, and Efforts of Will Unlike libertarians, compatibilists believe that moral responsibility and causal determinism are compatible. That is, they think that even if all events, including all of our decisions and actions, are necessitated by earlier events we can still be morally responsible for our actions. In recent decades, much emphasis has been placed on supporting this view by showing that alternative possibilities, the capacity to do otherwise, are not necessary for moral responsibility.4 Kane acknowledges that in a certain sense alternative possibilities, the capacity to do otherwise, are not necessary for moral responsibility. On his view, a person can be morally responsible for a certain act even if his current character causally determines him to form certain motives and even if these motives in turn causally determine him to act as he does. Borrowing an example used by Daniel Dennett (1984b), Kane notes that perhaps Luther spoke truthfully when, breaking from the Roman Catholic Church, he said “Here I stand, I can do no other.” In other words, it may well have been the case that given Luther’s character and the motives shaped by this character he could not do anything other than break from the Church of Rome. But, as Dennett and Kane note, this would not mean that Luther lacked moral responsibility for his action. However, unlike Dennett, Kane believes that there is a sense in which Luther must have been able to do otherwise if he is morally responsible for this break with the Roman Church. In particular, he believes that to be morally responsible in breaking off from Rome Luther must have been able to have formed a different character from the one he has and which leads

12

Kane’s Libertarian Theory

him to break off from Rome. If genetic and environmental factors wholly dictate the shaping of Luther’s character, then, according to Kane, it makes no sense to say that he is responsible for his later actions which are dictated by that character. His point is that, even if Luther is presently determined to act a certain way now due to the kind of character he has, he can still be responsible for this action as long as he is ultimately responsible for the kind of character he has. To be responsible for the character he currently has, it must have been possible for Luther to have done otherwise at various points in his past when engaging in various actions that led him to have the character he has now. The kind of actions in which we must be able to do otherwise and which shape our character and which are necessary for us to have ultimate responsibility for actions later caused by our character are what Kane calls “self-forming actions (SFAs)” or “will-setting actions.” SFAs “occur at those difficult times of life when we are torn between competing conceptions of what we should be or become” (Kane 2002a, 417). They involve making a difficult choice between competing motives or wants that cannot both be satisfied. Kane believes that choices made in these kinds of situations are indeterministically caused by events in our brains but that we are nonetheless responsible for whichever choice we make. Because they are indeterministically caused, in committing SFAs we could do otherwise. The example of the businesswoman which I discussed earlier is one of Kane’s favorite examples of the kinds of decisions he calls “SFAs.” In situations, like that of Kane’s businesswoman, the agent is torn between the desire to do two different actions and she can only do one of them. Kane says that in such situations when the agent is deliberating about what to do, she actually exerts efforts to do each of the actions she is contemplating doing and one of these efforts will cause her particular choice in an undetermined way. Thus, when she chooses, whichever choice she makes, it will be causally undetermined but it will also be a consequence of her prior effort. Thus, whichever choice the agent makes she will succeed at something she was trying to do and, thus, be responsible for her decision. In saying that her choice is indeterministically caused Kane is suggesting that it is caused by indeterministic causal processes occurring in her brain. He goes on to suggest a way in which such causally undetermined free decisions could possibly fit within a modern scientific world view, saying that when faced with an SFA: there is tension and uncertainty in our minds about what to do . . . that is reflected in appropriate regions of our brains by movement away from thermodynamic equilibrium—in short, a kind of stirring up of chaos in the brain that makes it sensitive to micro-indeterminacies at the neuronal level. The uncertainty and inner tension we feel at such

Kane’s Libertarian Theory

13

soul-searching moments of self-formation would thereby be reflected in the indeterminacy of our neural processes themselves. What is experienced personally as uncertainty corresponds physically to the opening of a window of opportunity that temporarily screens off complete determination by influences of the past. (Kane 2002a, 417) According to Kane, the capacity to do otherwise which is characteristic of SFAs might be a result of micro-indeterminacies at the neuronal level. He goes on to suggest that what stirs up the “chaos” in the brain, making it sensitive to micro-indeterminacy at the neuronal level, is the occurrence of competing neural networks. He says that in cases like that of the businesswoman it may be that: two competing neural networks are involved. (These are complex networks of interconnected neurons in the brain circulating impulses in feedback loops of a kind generally involved in high-level cognitive processing.) The input of one of these networks is coming from the woman’s desires and motives for stopping to help the victim. If the network reaches a certain activation threshold (the simultaneous firing of a complex set of “output” neurons), that would represent her choice to help. For the competing network, the inputs are her ambitious motives for going on to her meeting, and its reaching an activation threshold represents the choice to go on. Now imagine further that these two competing networks are connected so that the indeterministic noise that is an obstacle to her making one of the choices is coming from her desire to make the other. Thus, as suggested for SFAs generally, the indeterminism arises from a tensioncreating conflict in the will. (Kane 2002a, 419) According to Kane, what makes a person become sensitive to microlevel indeterminacy in the brain, leading to the indeterministic causation of action in an SFA, is the presence of competing efforts of will that cannot both be satisfied. It is important to understand that on Kane’s view when we engage in an SFA we don’t merely want to do one act, A, and another act, B, and then choose one or the other. Rather, on his view in deliberating about whether to A or to B, we actually are trying/exerting effort to A and simultaneously trying/exerting effort to B. The physical correlates of these competing efforts are competing neural networks in the brain. In these situations the choice or action is caused by processes in the brain, but the choice is not deterministically caused. So, in the case of the businesswoman, whichever course of action she takes, her action will be caused but undetermined by one of her efforts of will—her effort to prevent the assault or her effort to go on to her meeting.

14

Kane’s Libertarian Theory

In more recent writings, Kane has invoked the concept of “volitional streams” in talking about the neural foundations of these efforts of will. He writes: in such cases of self-formation, where we are faced with competing motivations, whichever choice is made will require a striving or effort of will to overcome the temptation to make the other choice. I thus postulate, in such cases, that multiple goal-directed cognitive processes (volitional streams as I call them) would be involved in the brain, corresponding to the different choices that might be made—in short, a form of parallel processing in the free decision-making brain. One of these neural processes or volitional streams would have as its goal the making of one of the competing choices (say, a moral choice), realized by reaching a certain activation threshold, while the other has as its goal the making of the other choice (e.g. a self-interested choice). The competing processes or volitional streams would have different inputs, e.g. moral motives (beliefs, desires, etc.), on the one hand, selfinterested motives, on the other; and each of them would be the realizer of the agent’s efforts or striving to bring about that particular choice (e.g. the moral choice) for those motives (e.g. moral motives). In such circumstances, if either process succeeds in reaching its goal (the particular choice aimed at) despite the indeterminacy involved, the resulting choice would be brought about by the agent’s effort or striving to bring about that choice for those motives. This would be so because the process itself (the volitional stream) would be the neural realizer of this effort and it succeeded in reaching its goal, despite the indeterminism. (Kane 2016a) Here Kane is suggesting that the neural correlates of our efforts of will would be these volitional streams, which are goal directed cognitive processes aimed at achieving certain ends given certain motivational inputs. In our SFAs, it will be causally undetermined which choice we make because it will be causally undetermined which volitional stream governs our choice. But whichever choice is made it will be a reflection of our will and we will be responsible for our choice, because whichever choice we make it will be the result of one of these volitional streams (the neural realizers of our efforts) having “succeeded at reaching its goal.” Thus, if our businesswoman makes the undetermined choice to stop and prevent the assault, then she will be responsible for this choice because it will have been a result of her effort to bring about that result; and this effort is realized in the cognitive processes of her brain in the neural processing which constitutes a volitional stream aimed at making that choice. It may be wondered how a person can be responsible for an action that is indeterministically caused. Due to the indeterminacy of the cause of such action, they would seem to be random, mere chance events, suggesting that they are not really under the control of the agent nor a product of her will.

Kane’s Libertarian Theory

15

Kane responds to this point by noting that the indeterminacy of the choice and action in an SFA stems from the fact that the agent engages in competing efforts of will in the moments leading up to the choice. It is, then, causally undetermined which effort will cause the choice. So, for instance, in the case of the businesswoman in deliberating she exerts effort to go on to her business meeting and she exerts effort to stop and prevent the assault, and it is causally undetermined what she will end up doing. But, according to Kane, whichever choice she makes her action will be a product of an effort that she has made and because of this she will control what happens and, thus, be responsible for her choice either way. Some people feel that if an agent’s choice is causally undetermined then the agent cannot be responsible for it. Thus, if the businesswoman’s choice is causally undetermined, then she cannot be responsible for her choice. But Kane presents various arguments from analogy which suggest that such thinking is misguided. In one of these arguments, he has us imagine that an assassin tries to shoot the prime minister of his country, and he has us suppose that, as the assassin pulls the trigger, it is causally undetermined as to whether he will have a nervous twitch leading him to miss his target. Kane has us suppose that the bullet does reach its target, killing the prime minister. He then notes that even though the killing was causally undetermined, the assassin is still fully responsible for the killing. He goes on to maintain that if the assassin is fully responsible for killing the prime minister despite the causal indeterminacy involved in the act, then so too our businesswoman can be fully responsible for her decision to stop and prevent the assault despite the causal indeterminacy involved in her decision. In supporting this analogy, he notes that, while the killing of the prime minister was causally undetermined, the assassin exerted effort to kill the prime minister and he succeeded at what he was trying to do and this suffices for his responsibility in killing him. Analogously, while the businesswoman’s choice is causally undetermined, she exerted effort to prevent the assault and in choosing to do so, she succeeds at what she was trying to do. Thus, she is responsible for this decision (Kane 2002a, 418, 2007a, 27).

On the Arbitrariness Involved in SFAs It is controversial whether Kane’s theory of the dual efforts involved in SFAs actually establishes how agents can be in control of these causally undetermined choices and responsible for them. I will return to this issue later. Now, I want to take up the issue of the arbitrariness of these choices. Since the choices made in SFAs are causally undetermined, this means that prior to the choice there are not sufficient reasons for one choice being made as opposed to another. Thus, it may seem that whatever an agent does in such situations is arbitrary and this may be taken to be a problem with Kane’s view. Do we really want to say that the decisions we make which establish our ultimate responsibility are arbitrary decisions?

16

Kane’s Libertarian Theory

In response to this issue, Kane writes: Even if one granted that persons, such as the businesswoman, could make genuine self-forming choices that were undetermined, is there not something to the charge that such choices would be arbitrary? A residual arbitrariness seems to remain in all self-forming choices, since the agents cannot in principle have sufficient or overriding prior reasons for making one option and one set of reasons prevail over the other. There is also some truth to this charge, but again I think it is a truth that reveals something important about free will. I have argued elsewhere that such arbitrariness relative to prior reasons tells us that at every undetermined self-forming choice is the initiation of what might be called a “value experiment” whose justification lies in the future and is not fully explained by the past. In making such a choice we say, in effect, “Let’s try this. It is not required by my past, but is consistent with my past and is one branching pathway my life can now meaningfully take. Whether it is the right choice, only time will tell. Meanwhile, I am willing to take responsibility for it one way or the other.” (Kane 2002a, 425) Here Kane acknowledges that there is arbitrariness in SFAs, but he states that this is the way things must be. Things must be this way because SFAs involve making decisions in situations where we could have done otherwise. If we make decisions in these circumstances, then our decisions cannot have “sufficient or overriding prior reasons” which make one option prevail over the other(s). Were there such sufficient or overriding prior reasons, then our choice would be deterministically caused and we would not be involved in an SFA. Notice how Kane also says that in an SFA the choice made is “not required by [the agent’s] past, but is consistent with [her] past.” In terms of the businesswoman, prior to being faced with her dilemma she already valued helping others and she also valued succeeding in business. These two values come to her from her past, but in the circumstances they now lead into conflict. But, whichever choice she makes—(a) to prevent the assault or (b) to go on to her meeting—the choice will be consistent with her past since either course of action will be motivated by values that have shaped her past self. Following up on this theme, Kane states: [A]gents who exercise free will are both authors of and characters in their own stories at once. By virtue, of “self-forming” judgments of the will (arbitria voluntatis ), they are “arbiters” of their own lives, “making themselves” out of the past that, if they are truly free, does not limit their future pathways to one. If we should charge them with lacking a sufficient or conclusive prior reason for choosing as they did, any one of them might reply: “True enough. But I did have good reasons for choosing as I did, which I’m willing to endorse and take responsibility

Kane’s Libertarian Theory

17

for. If they were not sufficient or conclusive reasons, that’s because, like the heroine of [a] novel, I was not a fully formed person before I chose (and am still not, for that matter). Like the author of [a] novel, I am in the process of writing an unfinished story and forming an unfinished character who, in my case, is myself.” (Kane 2002a, 425) Here Kane is suggesting that human freedom and moral responsibility require a certain amount of arbitrariness in our decision-making. The alternative to this would involve having all of our choices causally determined by overriding prior reasons. But then there could be no SFAs and without these there could be no genuine freedom or moral responsibility. Kane also wants to say that, despite the arbitrariness, it is not as though the actions we commit in SFAs are totally random and/or beyond the control of the agent, since what the agent does is ultimately a product of one of her competing efforts of will. The agent has certain reasons or motives which lead to these efforts of will. Since the motives or reasons she ends up acting on are partially constitutive of whom she is as a person and since it is she who engages in these efforts of will, it is she who controls what she does in making her causally undetermined choice.

Plural Voluntary Control and Its Significance A key element of Kane’s theory which is fundamental in his attempt to account for agent control over causally undetermined free decisions is his theory of the dual efforts involved in them. However, according to Kane, a person’s responsibility for an SFA is not established merely by the fact that one engages in dual efforts of will and makes a causally undetermined decision in which one of these efforts succeeds. Kane says that since the businesswoman’s choice will be caused but not determined by one of her efforts, this establishes that what she does will not be “inadvertent,” “accidental,” “capricious,” or “merely random.” Furthermore, each of her dual efforts are motivated by separate sets of reasons which she endorses, and when she decides she does so without coercion or control by other agents. All these factors contribute to her being responsible for whichever choice she makes. Indeed, they do so because, according to Kane, when these conditions are met in the performance of a causally undetermined free choice, then one acts with “plural voluntary control.” Agents have plural voluntary control over a set of options (such as the woman’s choosing to help the victim or to go on to her meeting), when they are able to bring about whichever of the options they will, when they will to do so, for the reasons they will to do so, on purpose, rather than accidentally or by mistake, without being coerced or compelled in doing so or willing to do so, or otherwise controlled in doing or willing to do so by any other agents or mechanisms. Each of these conditions

18

Kane’s Libertarian Theory can be satisfied by SFAs, like the businesswoman’s, as I have described them. The conditions can be summed up by saying that the agents can choose either way at will. (Kane 2007a, 30)

According to Kane, since we can meet all of these conditions in the performance of causally undetermined acts involving dual efforts of will, we can be in control of our choices in such situations and morally responsible for them. Some critics of Kane’s view (Pereboom 2001, 2007a, b, and Clarke 2003) note that these kinds of conditions, such as acting as one wills, for reasons, on purpose as opposed to accidentally or by mistake, without compulsion or coercion, can be met by agents acting in a deterministic world. Such critics then argue that it is unclear how or why Kane’s view establishes that we have more control over our behavior than we would have on a compatibilist view. This is supposed to be a problem, since libertarian views are intended to establish that we have additional kinds of control over our behavior which compatibilist theories cannot account for. This is often referred to as the “no-further-power” objection. Kane responds: I submit that, if agents can exercise plural voluntary power or control in the above sense over more than one undetermined alternative, as I have argued they can in SFAs, then the agents do have more power than compatibilists can give us in a determined world. For the most compatibilists can say of agents in a determined world who choose voluntarily, intentionally, and rationally is that the agents may have chosen otherwise (or in some other way) voluntarily, intentionally, and rationally, only if the past or the laws of nature had been different in some way (if the agents had different thoughts, desires, beliefs, etc.). Compatibilists cannot say that the agents may have chosen otherwise voluntarily, intentionally, and rationally, given the actual laws of nature and the past as it actually was at the moment of choice. (Kane 2007b, 177) Furthermore, not only is such a plural voluntary power a further power that compatibilists cannot give us in a determined world, but I submit that it is just the kind of power that libertarians have always demanded for free will and moral responsibility—that is, a power that can be voluntarily (non-coercively), intentionally (purposefully), and rationally (for reasons) exercised in more than one way here and now, given the actual laws of nature and the actual past at the moment of action. (Kane 2007b, 177) Let’s go back to our example of the businesswoman. Suppose she makes the choice to stop and prevent the assault instead of going on to her business

Kane’s Libertarian Theory

19

meeting and suppose this is a causally undetermined choice. Suppose as well that in making the choice she engages in dual efforts of will in the moments leading up to her choice, trying to stop and prevent the assault and simultaneously trying to go on to her business meeting. Suppose as well that she has reasons for each of her efforts, she wills these different actions intentionally and not by accident, and she is not coerced or compelled in exerting these efforts. This means she has voluntary control over each of these efforts. Thus, she has plural voluntary control, and when she makes her undetermined choice she will have control over what she does either way. Now, the compatibilist says that even if her choice to prevent the assault is determined, her choice to do so would still be made intentionally, for reasons, and not coerced or compelled. This is what motivates the “no-further-power” objection. But, as Kane notes, in SFAs there is a further power involved. Because when she has plural voluntary control, she has the power to make either choice given the way the world is at the moment of her choice. Compatibilists think that we can exhibit free will in a deterministic world. But in a deterministic world the businesswoman would not have as much control over her choice, because at the moment of her choice the joint effects of the past and laws of nature would necessitate the choice she makes. She could only choose otherwise in the sense that had she had different beliefs or desires, then she would choose differently. But, Kane’s view allows her the ability to choose otherwise given both (a) the way the world is and (b) the past and the laws of nature are what they are. As he says, “Plural voluntary power, given the actual past and laws of nature, is more power than compatibilists can give us in a determined world” (Kane 2007b, 177).

More on Luck and Control Despite Kane’s arguments from analogy and his theories of plural voluntary control, it still might be thought that it is just a matter of luck which decision gets made in SFAs—the basic, causally undetermined free acts that are the source of our ultimate responsibility. It might be thought that since these actions are causally undetermined right up to the moment of choice, then there is nothing about the agent which explains why she makes one choice, A, instead of some other choice, B, which she was also considering as an option. Without a contrastive explanation for why the agent chooses A instead of B, it may be thought that what the agent ends up choosing in an SFA is just a matter of luck and the agent has no control over what is chosen. Kane’s response to this is that the lack of such a contrastive explanation for why in an SFA the agent chooses one thing instead of another only entails that the agent lacks “antecedent determining control” over the choice he makes. In our SFAs, we lack, “the power to guarantee or determine which of a set of outcomes is going to occur before it occurs” (Kane 1996, 2016a).

20

Kane’s Libertarian Theory

Because the businesswoman’s choice is causally undetermined, there is nothing about her or her situation which determines or guarantees which choice she will make. However, it does not follow from this that in our SFAs we lack control over our decisions at the moment we make those decisions. [I]t does not follow that, because you lack control over which of a set of outcomes occurs before it occurs, you lack control over which of them occurs, when it occurs. When conditions of plural voluntary control are satisfied, agents exercise control over their future lives then and there in a manner that is not antecedently determined by their pasts. (Kane 2016a) According to Kane, in SFAs we lack antecedent determining control over the choices we make, but we still have plural voluntary control over such choices when we make them. For when we make such causally undetermined choices, they are still caused in an indeterministic way by one of our efforts. As such when we commit SFAs, we act intentionally, voluntarily, on purpose, for certain reasons, and not accidentally or by mistake. Thus, regardless of which choice gets made we will be responsible for the choice either way. Had the businesswoman made the selfish choice to go on to her business meeting, think how lame it would be for her to say, “Don’t blame me, it was a matter of luck that my effort to go to my meeting won out.” This is lame, because in choosing to go to her meeting she has, despite the indeterminacy, succeeded in doing something she was intentionally and voluntarily trying to do. As such, she has a sufficient level of control over her choice to be held responsible for it. Furthermore, since she was simultaneously trying to prevent the assault in the moments prior to choosing, she would have been responsible for that choice if she had made it. While the fact that there can be no contrastive explanation for why agents make the choices they make in SFAs may not show that such choices are matters of luck and beyond the control of agents, it still might be thought that the view suffers from a significant problem of control-robbing luck. It might be felt that since the particular choice made in an SFA will be the consequence of how certain quantum level indeterminacies impact neural functioning during deliberation and since we don’t have control over how these processes play out, then we lack the requisite kind of control over our SFAs to be responsible for them. Kane responds that this kind of objection is based on confusion. He says we don’t have to be in control of all the particular neurophysiological processes involved in an action to be responsible for it. We do not have to micro-manage our individual neurons one by one to perform purposive actions and we do not have such micro-control even when we perform ordinary actions, such as swinging an arm down on a table. (Kane 2016a)

Kane’s Libertarian Theory

21

When we engage in purposive, intentional behaviors for which we are clearly responsible, we don’t directly control each of the particular neurophysiological processes involved in them nor can we do this. Thus, we need not control the way in which the quantum indeterminacies affect our decision to be in control of and responsible for the causally undetermined choices we make in SFAs. Kane says: What we need when we perform purposive activities, mental and physical, is macro-control of processes involving many neurons—processes that may succeed in achieving their goals despite the hindering effects of some recalcitrant neurons. (Kane 2016a) Borrowing ideas from the neuroscientist Marius Usher (2006), Kane argues that the decisions we make in SFAs are the result of our exhibiting macro-level control over our efforts or volitional streams in the moment leading up to our decisions. He calls this kind of control “teleological guidance control (TGC).” TGC is exhibited in complex dynamical systems, as are found in the human brain. The behavior of a complex dynamical system exhibits TGC when it tends through feedback loops and error correction mechanisms to converge on a goal (called an attractor) in the face of perturbations. The goal or attractor in the case of these efforts or volitional streams is a particular choice (e.g. to cheat for gain or not to cheat) and their input consists of the motives for making that particular choice (e.g. self-interested motives or moral motives, as the case may be). (Kane 2016a) The idea here is that while we don’t control all of the particular neurophysiological processes involved in making the dual efforts involved in SFAs, we nonetheless control the direction these efforts—the volitional streams—take. Analogously, I control my reaching for my tea cup without micro-managing all of the neurophysiological processes which enable me to do this. Kane and Usher argue that TGC is compatible with indeterminism. Complex dynamical systems can tend through feedback loops and error correction mechanisms to converge on a goal in the face of perturbations, even when it is undetermined whether the goal will be achieved. Moreover, in such cases, if the goal is achieved, it will have been brought about by the teleological guided activity of the agent. (Kane 2016a) While the choice made in an SFA is causally undetermined and while the agent lacks micro-level control over the way in which the quantum indeterminacy impacts neural functioning, the agent involved in an SFA does nonetheless have

22

Kane’s Libertarian Theory

control over the larger neurological systems at work in exerting each of his dual efforts of will. Since the choice made in an SFA will be a consequence of one of these efforts, which is neurologically realized as a volitional stream, and since the agent has TGC over these efforts, then whichever choice the agent makes in an SFA it will be a product of his control and he will be responsible for it. Furthermore, Kane says this kind of control over our SFAs is exactly the kind of control over choices and actions which libertarians have always wanted—“the power plus opportunity at a time of an agent to cause or bring about any one of more than one alternative possible choices at the time, and to do so voluntarily, intentionally and rationally either way” (Kane 2016a).

Kane’s View Contrasted With Some Alternative Event-causal Libertarian Views Earlier I contrasted Kane’s event-causal libertarian theory with agent-causal and noncausal libertarian theories. I would also note that there are rival event-causal theories of free will which differ from Kane’s in significant respects. For instance, while not endorsing these theories, Daniel Dennett (1978) and Alfred Mele (1995) have developed event-causal libertarian theories which place the causal indeterminacy in libertarian free choices farther back in the deliberative process leading up to choice. This version of libertarianism is what Mele calls, “modest libertarianism,” and it is distinct from another kind of libertarianism he has described called “daring libertarianism.” (See Mele 2006, 2017 for his discussions of daring libertarianism.) Kane believes that in causally undetermined SFAs our choices remain causally undetermined right up to the moment of choice. On the alternative models of Dennett and Mele, causally undetermined choices would not be undetermined right up to the moment of choice. Rather, on their models the indeterminism would occur earlier during deliberation. They suggest it may be causally undetermined what ideas occur to us as we think about what it would be best to do. For instance, as one begins to think about whether to go to the movies tonight or to the baseball game it may be causally undetermined which ideas or reasons come to mind as one deliberates. But, depending on which ideas or reasons come to mind this can then determine which choice one makes. This theory allows for there to be causal indeterminacy in which choice gets made by allowing for indeterminacy early in the deliberative process leading up to choice. At the same time, such choices would still be determined in the sense that after ideas or reasons come to mind indeterministically the reasoning from those considerations to a decision would be deterministic. Mele and Dennett find this view more attractive than Kane’s view because they think that allowing indeterminism right up to the moment of choice as Kane does makes the choice a matter of control-robbing luck. In contrast, on their model once all ideas and reasons have come to mind—a process which, again, may be undetermined—the actual choice would still be determined by how the agent weighs those ideas and reasons in the moments leading up

Kane’s Libertarian Theory

23

to choice. In this way, we can see how differences in the agent’s reasoning in later stages of deliberation leads to the differences in their choices, putting them in control of their choices and making them responsible for them. As noted, Kane acknowledges that there will be some arbitrariness in our causally undetermined SFAs, but he doesn’t think there will be any controlrobbing luck. On his view, the agent engages in dual efforts of will in the moment leading up to choice such that regardless of which choice the agent makes he will be responsible for it either way. For either way, the agent’s choice will be a consequence of her effort, making her responsible for it. Again, it is controversial whether this approach adequately resolves worries about luck, control, and responsibility. I will take this issue up later in Chapter Four. For the reasons just given Kane doesn’t think the event-causal libertarian theories offered up by Dennett and Mele solve any problems which his own view cannot solve. Furthermore, Kane would argue that their models suffer from a control-robbing luck problem of their own. The problem their view faces concerns the manner in which ideas or reasons indeterministically come to mind in deliberation. They don’t arise as a result of our efforts of will. Consequently, it may be wondered how we are in control of which thoughts come to mind in the earlier stages of deliberation. If we don’t control this and if, as Mele’s and Dennett’s models suggest, in the later stages of deliberation our choice is a determined result of what considerations come to mind and how we weigh them, then it is puzzling how such models account for agent control over and responsibility for such choices.5 Another event-causal libertarian view which is distinct from Kane’s view is that of Laura Ekstrom (2000, 2003). Unlike Mele and Dennett, her view does not place the indeterminacy in free decisions merely in the coming to mind of certain ideas or reasons in the beginning of the deliberative process leading up to choice. Rather, she views free willed acts as caused by our preferences or intentions, and she thinks such acts are free when the decision process leading to the formation of the relevant preference or intention is uncoerced and causally undetermined. The idea here is that we sometimes deliberate about what to do (what intention to form) and sometimes we deliberate about what we most want (what preference to have). She thinks our actions are often the result of forming such intentions or preferences which then cause us to overtly act in certain ways. When these overt actions are the consequence of intentions or preferences which were formed by causally undetermined deliberative processes which are also uncoerced, then they are free actions. Her view is clearly different from that of Mele and Dennett, because she does not think the causal indeterminacy is moved so far back in the deliberative process. Rather, she thinks that the indeterminacy occurs in the weighing of the competing reasons for thinking that one course of action is preferable to another and this indeterminacy persists until one makes a decision. She regards a decision as the forming of either an intention or a preference (Ekstrom 2000, 106–116). It is unclear just how different her view is from that of Kane.6 But one clear difference lies in her rejection of Kane’s theory of dual efforts of will. She

24

Kane’s Libertarian Theory

finds the dual efforts theory to be problematic and she doesn’t think we need to appeal to it to make sense of agent control over causally undetermined free willed action. She argues that the dual efforts theory conflicts with the phenomenology of decision-making. We don’t experience ourselves as engaging in dual efforts of will when we deliberate; rather, we experience ourselves as wanting to do two or more different things, and seeing that we can only do one of them, we deliberate about which of these things we most want to do or which we think is all things considered best to do. She also argues that the dual efforts view renders our underivatively free acts, our SFAs, irrational. It is irrational to exert efforts to do A and to do B when one knows that one can only do one of them. Recall that on Kane’s view as she deliberates the businesswoman is to be understood as trying to go on to her business meeting and trying to prevent the assault, but she also knows that she can only do one of them. Ekstrom finds such dual efforts irrational. Since the dual efforts view disagrees with the phenomenology of decision-making and since it renders underivatively free actions irrational, she thinks we have good reason to reject such a view in favor of an event-causal view that lacks these problems.7 Such critiques of Kane’s view are worthy of serious consideration. I will take these issues up later in Chapter Four. Here I will just note that Kane would likely have doubts about whether Ekstrom’s view can adequately account for agent control of causally undetermined decisions. She wants to account for agent control over free action in the following way: Consider again the problem of control. In response, although the notion of control does require elucidation in some causal terms, it does not require deterministic explanation. In the case of deliberation preceding a free act, which decision-settling event it will be that occurs is not determined until it occurs. The formation of preference x or preference y or preference z may happen at t-1. At t, suppose that preference x is formed. Its formation has a causal (and rational) explanation in terms of the agent’s considerations. Furthermore, the act resulting from the preference, in the case of free action, is controlled by the agent in being causally determined by the agent, since the agent on my view is identified with a set of preferences and acceptances, together with the power to fashion and refashion the character. (Ekstrom 2000, 115) In the causally indeterministic process of deliberation moving towards a decision, our preferences are weighed and considered in hopes of establishing what one most wants to do or what one thinks it is best to do. Since the agent just is “a set of preferences and acceptances together with the power to fashion and refashion the character,” and since such deliberation is the exertion of this power through the ordering of our preferences, it follows that whatever one ends up deciding it will be under the control of the agent.

Kane’s Libertarian Theory

25

The process of deliberation just is the set of preferences with the power to fashion and refashion itself—the agent—exerting that power upon itself. So, the deliberation and the decision are under the control of the agent, even if the decision is causally undetermined. One must remember that Ekstrom rejects agent-causation. So, when she says the agent controls such undetermined decisions, this must be reducible to events occurring internal to the agent. Presumably, it will be the agent’s own weighing of her competing preferences which will dictate which decision gets made, but this will itself be a causally indeterministic process. We might wonder what establishes control over these weighings. Sure, they may go on internal to the agent and then yield a decision, but the mere fact that they go on internal to the agent does not establish agent control over them. Lots of things go on internal to us over which we have no control. So, we might be left here wondering how or why the agent has control over how the preferences get weighed, and since how they are weighed will dictate which decision gets made, we may be left wondering how it is that agent control over such undetermined choices is established. How her preferences get weighed and consequently what the agent decides may in the end just be something that happens to the agent as opposed to something she does. Kane will think his own view is preferable here, because of its inclusion of the dual efforts. We engage in efforts; they don’t just happen to us. On Kane’s view, in deliberating the businesswoman tries to prevent the assault and tries to go on to her meeting and it is undetermined which effort wins out, dictating what she ultimately decides and does. Thus, whichever act she does it will be a result of her effort—a result of something she does—and not merely something which happens to her. In this way, he will regard his view as offering a more satisfactory account of agent control over causally undetermined free choices. A different kind of event-causal libertarian theory has been developed more recently by Mark Balaguer (2010). His theory is more similar to Kane’s in certain key respects. Most notably, like Kane, he thinks we can be morally responsible for causally undetermined choices even if they are undetermined right at the moment of choice. However, unlike Kane, he doesn’t think we need to appeal to a theory of dual efforts of will leading up to causally undetermined choice. Also, unlike Kane, Balaguer does not believe that our responsibility for actions causally determined by our character or motivational states hinges upon our making earlier undetermined free choices which shape our character or motivational states. He is not as pessimistic as Kane about the capacity of compatibilist theories of free will to make sense of free and responsible action, so he doesn’t feel the need to appeal to something like Kanean SFAs to account for our responsibility in the performance of causally determined acts. Balaguer’s theory is a hybrid theory which allows that there may well be libertarian free actions, but he doesn’t see them as necessary for human free will and moral responsibility because he thinks compatibilist free will and responsibility make sense.

26

Kane’s Libertarian Theory

Kane would reject the hybrid nature of Balaguer’s view. Kane doesn’t think we can make adequate sense of our responsibility for determined choices unless they issue from a character or motivational states which are formed by prior free undetermined SFAs. As we will see in Chapter Three, he thinks compatibilist views fall prey to manipulation arguments which exhibit precisely why there must be some causal indeterminacy at play in the history of any actions for which we are morally responsible. This, of course, does not mean that all of the acts for which we are morally responsible must be undetermined acts; rather, it’s just that determined acts for which we are morally responsible must stem from a character or motivational states formed by prior causally undetermined free acts (SFAs). Kane would also find fault with Balaguer’s assertion that we can get by without dual efforts of will. On Kane’s view, event-causal libertarian theories need to appeal to the dual efforts idea to make sense of agent control over and responsibility for undetermined choices. In Balaguer (2010), he explains that if we exhibit any libertarian free will this would occur in what he calls “torn decisions.” A torn decision is a decision in which the person in question (a) has reasons for two or more options and feels torn as to which set of reasons is stronger, that is, has no conscious belief as to which option is best, given her reasons; and (b) decides without resolving this conflict—that is, the person has the experience of “just choosing.” (Balaguer 2010, 71) Balaguer contends that when we make torn decisions which are undetermined in the appropriate way they are libertarian free decisions; and, again, he says these torn decisions need not involve efforts of will (Balaguer 2010, 74). He also says that in our undetermined free decisions our choices are “settled by undetermined brain events that are parts of the neural/mental events that are our torn decisions” (Balaguer 2010, 108). In a published response to Balaguer’s view, Kane refers to these undetermined neural events which Balaguer claims settle our choices as “neural coin tosses.” And he states: But then even if these neural coin tosses are internal to the torn decisions, the agent does not have control over how the neural coin tosses come out, or they would not be undetermined events; and if how these coin tosses come out “settles” which option is chosen, the agent would not have control over that either. (Kane 2014a, 55) What troubles Kane is that if it is just undetermined neural events occurring in the brains of agents which settle the choices made by agents, then they don’t have control over their choices; rather, these undetermined neural

Kane’s Libertarian Theory

27

processes over which they have no control settle which choices get made (Kane 2014a, 55). Kane thinks that even if the undetermined neural events which settle which choice is made are internal to the process of deciding there is still a problem of control-robbing luck here, since “these undetermined parts of the torn decisions (the neural coin tosses), which are not themselves agential doings, are what ‘settle’ which tied-for-best option is chosen, and hence which way the decision goes” (Kane 2014a, 55). In contrast, on the Kanean model what settles the choices in causally undetermined SFAs are “agential doings”; they are the efforts of will of the agent. Either way what the agent chooses will be a direct result of an effort she has made, establishing agent control of and responsibility for the choice. I will reconsider some of these alternative event-causal libertarian views, as well as other such views, in Chapter Five. I only discuss them here to let us better understand some of the distinctive features of Kane’s particular version of event-causal libertarianism, as well as the reasoning behind his view.

On the Existence of Libertarian Free Will My exposition here of Kane’s view has focused on his conception of the nature of free will and the differences between his view and other libertarian views of free will. Thus, I have focused on his account of what conditions must be met if we are to have the kind of freedom required to make us morally responsible for our actions. In doing so, I was led to explain the nature of his event-causal libertarian theory and explain some of the key points he has made in defense of the coherence of this view. But now it might be wondered what Kane has to say about the existence of free will. Regarding this, Kane states: I reject the view (amply refuted by John Stuart Mill and others) that we can know we have free will by introspection or by a priori reasoning alone. There are empirical aspects of the free will issue that mere philosophical speculation cannot co-opt. If free will of a nondeterminist kind should exist in nature, then the atoms must somewhere “swerve” to make room for it, and they must swerve in places where it matters— in the brain, for example. But neither can philosophical speculation be indifferent to the empirical questions. (Kane 1996, 17) And, I think libertarians must accept the empirical challenge of determinism (that it might turn out to be true), if libertarians are going to be serious about finding a place for free will in the natural order where we exist and exercise our freedom. This is the “Existence Question” for free will, and . . . it cannot be finally settled by armchair speculation, but only by future empirical inquiry. What philosophers can do with

28

Kane’s Libertarian Theory regard to the Existence Question is to suggest models about how free will might work in the brain so that future researchers have some idea of what to look for when they search for it. (Kane 1996, 184)

Thus, he believes that introspection and a priori argument cannot tell us whether we have the kind of free will which he thinks is necessary for moral responsibility. Rather, he thinks this is an empirical question. To resolve that issue, we will need more empirical knowledge of how the human mind and the human brain work. Having said this, he also believes that there is not at this point sufficient empirical evidence to rule out the possibility that we possess libertarian free will. It is an open empirical question. As noted in the second of the above quoted passages, he does think philosophers can provide accounts of how libertarian free will could possibly fit with a scientific understanding of the mind/brain, and I have touched on some of his thoughts on this in my exposition. In the end, it is his view that at this point we do not know whether in fact free will exists.

Notes 1. Kane’s publications on the subject are too numerous to list here. I will just note that excellent summary accounts and defenses of his view can be found in Kane (2002a, 2005, 2007a, b, 2011, 2016a). 2. The agent-causal libertarian view has been defended by various figures, see Roderick Chisholm (1964, 1976); Timothy O’Connor (1996, 2000, 2009, 2011); Thomas Reid ([1788] 1969); and Richard Taylor (1966, 1992), among others. Also, see Randolph Clarke (2003), who explains and defends the view at length without endorsing it. 3. This noncausal, or simple-indeterminist, view of libertarian free will has been defended by Carl Ginet (1990, 1997), Stewart Goetz (1988, 1997), and Hugh McCann (1998), among others. 4. The seminal paper on this issue is Harry Frankfurt (1969). For criticisms of Frankfurt’s position see Carl Ginet (1996), Robert Kane (1996), Timothy O’Connor (2000), Peter Van Inwagen (1983). For defenses of Frankfurt’s position see Daniel Dennett (1984b), John Martin Fischer (1982, 1994), Ishtiyaque Haji (1998), Alfred Mele (1995), Alfred Mele and David Robb (1998). 5. While she does not endorse Kane’s notion of efforts of will, Laura Ekstrom (2000, 136–137) expresses similar concerns about the event-causal theories of Mele and Dennett. She also is concerned that there is a kind of control-robbing luck involved in their account. 6. For instance, consider the issue of “Valerianism.” Non-Valerian views of libertarian free will allow for our free willed decisions to be causally undetermined right up to the moment of choice, as in Kane’s view. Valerian libertarian views only allow for indeterminism in earlier stages of deliberation and deny that free willed choices are undetermined right up to the moment of choices, as in Mele’s modest libertarian view (1995) and Dennett (1978). Randolph Clarke (2002) interprets Esktrom as holding a Valerian view, but Balaguer (2010) does not. Balaguer notes that in private correspondence Ekstrom self-identifies as a non-Valerian (Balaguer 2010, 175n5). 7. These criticisms of Kane’s view can be found in Ekstrom (2003, 163–164). Randolph Clarke (2003, 88–89) raises similar concerns about the rationality of free willed choices on the Kanean model.

3

Libertarianism, Compatibilism, and Manipulation Arguments

In this book, I defend an event-causal libertarian theory of free will which is very similar to that of Robert Kane, who is one of the best-known advocates of libertarianism.1 In earlier chapters, I noted that one persistent problem libertarian views face is the existence question. How can we know whether we have libertarian free will? Both Kane and I believe that at this point in time we don’t have sufficient scientific, metaphysical, or logical evidence to establish that we have such free will. In later stages of this book, I will argue that despite this fact there are nonetheless good moral or pragmatic reasons to believe we have such free will or at least to justify living as if we have it. Those who believe in free will are either libertarians or compatibilists. The latter believe that even if all human decisions and actions were causally determined, then human beings could still possess free will. One of the chief benefits of the compatibilist understanding of free will is that if this view can be shown to be coherent—that it makes adequate sense of how free will can exist in a deterministic universe—then the existence question will automatically be solved. For there is little doubt that many of the causally determined actions which the more plausible compatibilist theories describe as free willed do exist. Thus, if these theories are coherent, then we have every reason to believe we possess free will. If I am going to defend a libertarian theory which makes it so hard to know whether we have free will, then dialectically it makes sense to begin by explaining why we should reject compatibilism. For the reasons given, a coherent compatibilism has no problem dealing with the existence question, whereas a coherent libertarian view does have this problem. As such, anyone interested in defending the existence of free will has good reason to have an initial preference for compatibilism. Thus, I want to start out by making a case against compatibilism. It is my view that other philosophers have already given strong reasons to think compatibilism is a deeply problematic approach to understanding the nature of free will and moral responsibility. Peter Van Inwagen’s consequence argument (1983, 2002), Derk Pereboom’s four-case manipulation argument (2001, 2014), and Bruce Waller’s argument from fairness (2011) all make strong cases against compatibilism. However, while I’ve always

30

Compatibilism and Manipulation Arguments

found these arguments compelling, the debate on the merits of these arguments is ongoing with compatibilist rivals continuing to challenge the merits of these arguments. What I propose to do here is to revisit Kane’s manipulation argument and develop it further in new ways in hopes of providing a different way of seeing why compatibilism is so problematic. Those more informed about the contemporary literature on free will may think it problematic for a libertarian to endorse manipulation arguments in attacking compatibilism. After all, in the recent literature libertarian views have themselves been subjected to manipulation arguments. This fact informs and shapes my approach in the present chapter. In what follows, I begin by explaining Kane’s manipulation argument against compatibilism. In doing so, I also appeal to key ideas from Michael McKenna (2004, 2012) which help clarify how Kane’s points can be generalized to create a general argument against any compatibilist theories of free will and moral responsibility. Next, I address a variety of recent attempts to show that Kane’s libertarian view does itself fall prey to manipulation problems. I will show that these arguments fail. Some of these failures will then offer some insights into why manipulation arguments against compatibilism succeed. In the latter stages of the chapter I return to the manipulation argument against compatibilism, arguing that neither “hard-line” nor “soft-line” responses to it suffice.

Compatibilism and the Problem of CNC Manipulation Compatibilists give different accounts of the necessary and sufficient conditions of free and responsible action. In a classic essay by the compatibilist Walter Stace ([1952] 1993), he states that a person acts freely if and only if his actions are caused by psychological states internal to the agent. Such a view is problematic because persons can be moved to act by internal psychological states and still not act freely. For instance, consider an agent who acts under the grip of overwhelming desires which feel foreign to himself, as when one has an addiction or an obsessive-compulsive disorder. Later thinkers have come up with more sophisticated compatibilist accounts which avoid this problem in Stace’s account. For instance, Harry Frankfurt (1971) maintains that we have both firstorder and second-order desires. First-order desires are desires to do certain actions, such as the desire to eat lunch or to clean the kitchen. Second-order desires are desires to be moved to act by certain first-order desires. According to Frankfurt, a person acts with free will when she is moved to act by a firstorder desire which agrees with her second-order desires. So, for instance, if I have a first-order desire to steal from my neighbor and this desire moves me to steal from my neighbor and I want to be moved to act by this desire, then I freely steal from my neighbor and I am morally responsible for doing so. In contrast, if, as is often the case with kleptomaniacs, I am moved to steal from

Compatibilism and Manipulation Arguments

31

my neighbor by a desire that I would rather not act on, then I don’t freely steal from my neighbor and I am not morally responsible for doing so. An alternative compatibilist view comes to us from the semi-compatibilists, John Martin Fischer and Mark Ravizza (1998). Semi-compatibilists deny that we have free will but claim that we are often morally responsible for our actions and that moral responsibility is compatible with determinism. According to Fischer and Ravizza, a person is morally responsible for his action when he exhibits guidance control over his action. One has guidance control over one’s action when one’s behavior is controlled by one’s intentions and when one is reasons-responsive, such that if one had better reasons to do something else then one would have acted on those better reasons. Typically, when one steals one exhibits guidance control, because one does what one deliberately intends to do and one would act differently if one felt that one had better reasons to do something else. But the kleptomaniac typically does not act with guidance control when he steals. For while he intentionally steals, he is not reasons-responsive. Even if the kleptomaniac felt he had better reasons not to steal, his conduct would not be guided by those reasons, as the overwhelming urge to steal would take charge of him despite his belief that it would be more reasonable to act differently. There are yet other compatibilist accounts of free will and moral responsibility that could be presented here, such as Daniel Dennett’s self-control conditions (1984a, 2003), but I’m just trying to give the reader some general understanding of the kinds of criteria compatibilists have given. My primary concern here is to explicate Kane’s manipulation argument against such compatibilist views, but to understand his critique we must first have some sense of the kinds of criteria for free will and moral responsibility which compatibilists have offered. In developing his critique of compatibilism, Kane talks about covert nonconstraining control (CNC). CNC can be contrasted with constraining control (CC), which occurs when an agent is knowingly forced to do something against his or her will, such as when someone is threatened at gunpoint to do something or forcibly locked in a room against his or her will. In contrast to CC, CNC involves secretly “manipulating the will of others so that the others (willingly) do what the controllers desire” (Kane 1996, 64). When subject to CNC, The controlled agents consequently do not feel frustrated or thwarted. They act in accordance with their own wants, desires, or intentions. Yet, they are controlled nevertheless by others who have manipulated their circumstances so that they want, desire, or intend only what the controllers have planned. In the most interesting cases, such control is “covert” nonconstraining control—or CNC control, as I have previously put it— in which the controlled agents are unaware of being manipulated or perhaps even unaware of the existence of their controllers. (Kane 1996, 64–65)

32

Compatibilism and Manipulation Arguments

Kane notes that fictional tales, such as Aldous Huxley’s Brave New World or B.F. Skinner’s Walden Two, give excellent descriptions of how CNC might look when carried out on an entire society. Frazier, the fictional founder of Walden Two, gives a clear description of CNC control when he says that in his community persons can do whatever they want or choose, but they have been conditioned since childhood to want and choose only what they can have or do. (Kane 1996, 65) In The Significance of Free Will (1996), Kane goes on to argue that the possibility of CNC presents serious problems for a view like Frankfurt’s. The problem is that people can be manipulated by external controllers not only into wanting to do certain acts but into wanting to be moved by such firstorder desires. It is conceivable that a society could be constructed in which everyone’s desires and subsequent actions are manipulated in such a way that whenever they act they act on first-order desires which agree with their second-order desires. The problem is that according to Frankfurt’s theory this would mean that everyone in such a society would act with the kind of free will that would make them morally responsible; and this is wrongheaded, since a society in which persons’ desires are manipulated in this way would not be a society in which persons exhibit free will and are morally responsible for their actions. While Kane clearly uses the possibility of global manipulation to argue against Frankfurtian compatibilism, he also makes comments in The Significance of Free Will (SFW) which suggest that he thinks the argument can be extended to any compatibilist theories.2 In more recent work, Michael McKenna clarifies how manipulation arguments can be generalized to all forms of compatibilism. In McKenna (2012), he talks about “Compatibilist-friendly Agential Structures (CAS).” CAS are any compatibilist accounts of the necessary and sufficient conditions of free or morally responsible action, whether Stace’s or Frankfurt’s or Fischer’s. Additionally, we may use X as a placeholder for any manner of global manipulation. Then we can build a generalized version of the manipulation argument which would target any kind of compatibilism, whether Frankfurt’s or any other. 1. If an agent, S, is manipulated in manner X to perform act A, then S does not A freely and is therefore not morally responsible for A-ing. 2. Any agent manipulated in manner X to A is no different in any relevant respect from any normally functioning agent determined to do A from CAS. 3. Therefore, any normally functioning agent determined to do A from CAS does not A freely and therefore is not morally responsible for A-ing. (McKenna 2012, 147)

Compatibilism and Manipulation Arguments

33

The idea here is that we can imagine agents being globally manipulated not only in ways that show Harry Frankfurt’s criteria fail but also in ways that show that other types of CAS, such as Dennett’s or Fischer’s and Ravizza’s, fail. For any CAS proposed, it can be shown that globally manipulated agents can meet those conditions of free or responsible action and yet not be free or responsible. Further, in line with the argument outlined above, it can be argued that normal, nonmanipulated agents acting in a deterministic universe do not act freely or possess moral responsibility for the same reasons as the manipulated agents—namely they don’t have sufficient control over the formation of their characters to make them free or responsible for what they do. In a deterministic universe, the role of the global manipulator is played by the joint effects of the laws of nature and the prior states and events which begin shaping our character and conduct prior to our birth. McKenna goes on to note that compatibilists have adopted two different general types of strategies in responding to manipulation arguments. Some provide a “soft-line reply” in which they try to argue that the proposed example of global manipulation doesn’t really allow the manipulated agent to meet their preferred compatibilist criteria (CAS) of free or responsible action. This amounts to challenging the second premise of any such general argument. In contrast, a “hard-line reply” rejects the first premise, arguing that while agents may be globally manipulated into doing A, they still act freely as long as they meet the relevant CAS. This hard-line reply is taken by Harry Frankfurt, who rejects the first premise on the grounds that as long as his own conditions of CAS are met it does not matter that the agent has been manipulated. So, for instance, while I may be manipulated into A-ing from a first-order desire to A and even if I have been manipulated into desiring to be moved by this first-order desire, I am still responsible for A-ing and I do A freely. McKenna calls such a response “hard-line,” because it comes at the price of ignoring a widely shared and strong intuition that such manipulated agents would not act freely and not be morally responsible for their action.3 In the second half of this chapter, I will develop responses to these hardline and soft-line compatibilist replies to the manipulation argument. Before doing so, however, I want to consider the charge that libertarian views, like Kane’s, are just as vulnerable to the problem of manipulation. In the recent literature, various philosophers have made such arguments. I will look at several of these arguments and show why they fail. In doing so, my objectives are two-fold. First, since I am ultimately out to defend something like a Kanean libertarian view in this book and since I am using a manipulation argument to critique rival compatibilist theories, I must have an adequate response to the charge that the Kanean view is itself falsified by manipulation arguments. Second, in examining why Kane’s view is immune to these arguments, we will gain some insights which will help us in seeing why the hard-line and soft-line defenses of compatibilism are inadequate.

34

Compatibilism and Manipulation Arguments

Manipulation Arguments and Libertarianism: A Defense of Kane’s View (i) Haji’s and Cuypers’s Argument In Haji and Cuypers (2001), it is argued that Kane’s view has just as much of a problem with CNC manipulation as compatibilist views do. They present a manipulation case which is supposed to prove this. They have us imagine that Max is a neurologist who can and does secretly manipulate the thinking and actions of others. They write: Imagine that Max turns Jenny into a psychological twin of Val who is struggling with his smoking. Post-surgery Jenny finds herself in a conflict situation analogous, down to minute details, to a conflict situation in which Val found himself: she is tempted to smoke and she is also tempted to refrain. Suppose that she reasons just like Val, exerts an indeterminate effort of will type-identical to that exerted by Val, decides to smoke, and smokes. Why should the mere fact that her decision to smoke is undetermined, given that it was caused by an indeterminate effort of will, render her responsible for her choice to smoke when she is simply one more victim of Max? (Haji and Cuypers 2001, 232) The idea behind this example is that Jenny has secretly had her psychological states—her beliefs, desires, motivational states, etc.—significantly altered by Max. Presumably, prior to this manipulation she was not a smoker and had no smoking problem; but now she finds herself with a strong desire to smoke and a strong desire not to smoke and she must make a choice as to whether she will or won’t. What they suggest is that on Kane’s view if she now makes the causally undetermined choice to smoke and if she does so while possessing plural voluntary control, then she is morally responsible for that choice. However, Haji and Cuypers believe this is fundamentally wrongheaded, for she would not be making such a choice nor choose to smoke were it not for the manipulations of Max. Thus, Kane’s view leads to the wrong conclusion in this manipulation case. For this reason, they believe Kane’s view is no more immune to manipulation arguments than are compatibilist views. This critique of Kane’s view is problematic. There may be good reason for Kane to say Jenny, our manipulated agent who makes the undetermined choice to smoke, has some very minimal responsibility for her choice in this situation. To see this, it helps to know that on Kane’s view there are two ways in which we may be responsible for our actions: (a) from our performance of SFAs and (b) by acting from a character or motivational states that are the product of prior SFAs. Because of this Kane believes that young

Compatibilism and Manipulation Arguments

35

children bear some, but very little, responsibility for their earliest SFAs. In these earliest SFAs the conflicting efforts of will involved in them stem not from a character that they have built themselves; rather, in early childhood our efforts are too much a direct result of genetic and environmental factors beyond our control. In the very first SFA of a child, the child does for the first time make some contribution of her own to the formation of her character by making an undetermined choice between two efforts that were not of her own making. Since those efforts are themselves wholly the product of factors beyond her control, she has only very minimal responsibility for this first SFA.4 The implications of this are that by analogy in Jenny’s case it would make sense to say that she bears a very minimal responsibility for the undetermined choice she makes between the values and desires which have covertly been placed into her by Max. Max’s covert implanting of these values plays the same role as genetic and environmental factors do in the child making her first SFA. In making her choice to smoke, she acts not from a motivational state that was shaped by her prior SFAs but from a motivational state which has been implanted in her by the manipulator, Max. Thus, the only responsibility she has for her choice to smoke stems from the fact that the choice is an SFA. Again, Jenny’s responsibility will however be minimal here because she is not responsible for the motivational states which give rise to this choice. This is not an unreasonable conclusion to draw. Furthermore, suppose that in one day Max had covertly manipulated all manner of values in Jenny which she didn’t have before and suppose these values sometimes conflict. If so, then, as I’ve just argued, the first undetermined decisions she makes when encountering competing efforts of will are decisions in which she exhibits very little freedom and for which she bears very little responsibility. But years later, as she engages in more and more SFAs, she will become more and more responsible for them even though she will have gotten her initial set of values through Max’s manipulation. This is not unreasonable. By contrast, on compatibilist views, like Frankfurt’s, we’d have to say that right after Max’s manipulation Jenny is fully responsible for her decisions as long as they meet his compatibilist criteria of free and responsible action and even if her decisions are necessitated by the values foisted upon her by Max. This really is problematic.5 (ii) Pereboom’s Argument Derk Pereboom also believes that Kane’s theory falls prey to manipulation arguments. In building his argument he refers to Kane’s businesswoman who must choose between stopping to prevent an assault or going on to her business meeting. Recall that Kane believes that when she commits an SFA she has reasons to do both acts and exerts efforts to do both acts as she

36

Compatibilism and Manipulation Arguments

deliberates; then, whichever choice she makes it is a result of her effort and done with reasons. Pereboom writes: [I]magine businesswoman**, whose capacities, character, motives, and efforts are qualitatively identical to those of businesswoman. Businesswoman** also makes the choice to help the victim. But intimately involved in this choice is not the usual determining manipulator who spins a dial that will land on one of two positions. Imagine that the spinning of the dial is a metaphysically indeterministic event, and that it is not merely indeterministic relative to the spinner’s knowledge and power. In addition, suppose that the dial’s landing on a position just is the crucial indeterministic component of the neurophysiological realization of the choice to perform the action, and that it is not simply an indeterministic event that deterministically causes one choice rather than the other. Also, suppose the dial’s landing on a position does not supplement the agent’s normal neurophysiological processes, but rather replaces the crucial indeterministic component in these processes. Businesswoman**’s effort, from the psychological point of view, is exactly qualitatively identical to businesswoman’s. The position the dial lands on therefore makes the salient difference as to which decision the agent makes. If the dial lands on one position, her efforts of will result in one choice, and if it lands on the other, they then produce the alternative choice. Imagine that her efforts of will are indeterminate in the sense that their causal potential does not become determinate until the choice occurs. Before the choice occurs, there are various ways in which this causal potential can be resolved. (If one is tempted to reply that the two agents will not be neurophysiologically the same, and thus will not feature the same efforts, assume instead that the various indeterministic components of the microphysical realization of businesswoman**’s effort are suitably supplied by the randomizing manipulator.) Suppose, in addition, that businesswoman**, just like businesswoman, has reasons for making the choice, she chooses for these reasons, she wants to choose for these reasons more than any others, and she is not being coerced or compelled in choosing. It is yet implausible that indeterminacy of the sort exhibited by businesswoman**’s decisions provides for moral responsibility, and this is intuitively because she lacks the control over her decision that moral responsibility demands. More fundamentally, she is not the source of her decision in the sense that moral responsibility requires. It is incumbent on a defender of Kane’s view to provide a principled difference between businesswoman** and businesswoman that would explain why an agent could be responsible for the decision that fits Kane’s theory but not for the one produced by the manipulator. This I believe cannot be done. Pereboom notes that Peter Van Inwagen (1983, 132–134) and Alfred Mele (1999b, 277) make similar sorts of randomizing manipulation arguments

Compatibilism and Manipulation Arguments

37

against event-causal libertarian views, like Kane’s. I would note as well that a similar sort of argument is made by Daniel Dennett in Freedom Evolves (2003, 132–133). The kinds of arguments made by Pereboom and these others are different from that of Haji and Cuypers. The latter have a manipulator putting beliefs and values into the mind of the agent and allow that the agent indeterministically resolves conflicts in these values herself. But the manipulation argument of Pereboom and these other philosophers allow the agent to have her own beliefs and values while manipulating the actual decision process with a randomizing decision generator. The strategy endorsed by Pereboom and these others works no better than that of Haji and Cuypers, because there is a relevant difference between having one’s decision generated from a randomizing conflict resolution device and what happens in a normal SFA. It is significant that in normal SFAs the agent continues to exert effort to do A and effort to do B while meeting compatibilist requirements in making those efforts. That is, in making each of these competing efforts, the agent is neither forced nor coerced into doing so; she wants to be making these efforts and she desires to be moved by her desires to do so (Frankfurtian compatibilist conditions) and she is reasonsresponsive in exerting these efforts, such that if she felt she had better reasons to not be making these efforts she wouldn’t be making them (Fischer’s and Ravizza’s compatibilist conditions), and so forth. Meeting such compatibilist requirements in the efforts establishes that it is she who will be causally responsible for the choice either way and not some external factor beyond her control, such as the manipulator’s random decision output generator. Sure, in normal SFAs the decision that is made will be undetermined as in Pereboom’s manipulation case; but in one—normal SFAs—the decision is made by the agent (not in some agent-causal nonreductive sense, but in some sense established by the agent involving events that are her efforts of will) and in the other the decision is made by the manipulator’s random decision output device. This is a very important difference! For in normal SFAs the agent is both the source of the indeterminacy and her choice. This is because in normal SFAs the indeterminism arises through the efforts the agent makes and the agent’s choice either way is a product of one of her efforts. Whereas in Pereboom’s manipulation case the indeterminism and the choice are both a product of something external to the agent—the randomizing manipulation device. It might be objected here that according to Pereboom the randomized, dial’s landing on a position does not supplement the agent’s normal neurological processes but rather replaces the crucial indeterministic component in these processes. Businesswoman**’s effort from the psychological point of view is exactly qualitatively identical to businesswoman. The position the dial lands on therefore makes the salient difference as to which decision the agent makes. (Pereboom 2001, 53)

38

Compatibilism and Manipulation Arguments

It might be thought that if there is no difference in the psychological states of businesswoman and businesswoman** and the randomizing device is simply providing the indeterminacy that would otherwise be provided by quantum phenomena in the neurons of her brain, then Kane would have to rationally concede to the problematic conclusion that there’s no significant difference between businesswoman and businesswoman**. Such a response will not suffice however. For if in the case of businesswoman** it is the position the dial lands on which makes “the salient difference as to which decision the agent makes,” then from the psychological point of view businesswoman and businesswoman** are not qualitatively identical. For in businesswoman** it is the position the dial lands on which indeterministically causes the choice, but in businesswoman it is one of her efforts which indeterministically causes the choice. This is a crucial difference, for in the latter case it is her own effort which indeterministically causes the choice and this is what gives her agency over what she does and what makes her responsible for what she does. When something else besides her efforts causes the decision, she loses agency in the making of the decision. A fan of Pereboom’s argument may here want to draw attention to his parenthetical remarks in the passage I cited above, in which he states, “(If one is tempted to reply that the two agents will not be neurophysiologically the same, and thus will not feature the same efforts, assume instead that the various indeterministic components of the microphysical realization of businesswoman**’s effort are suitably supplied by the randomizing manipulator.)” It might be argued that Pereboom’s point is that we can imagine businesswoman and businesswoman** as the same neurophysiologically and psychologically as they make their decisions, while in businesswoman the indeterminism is supplied by quantum indeterminacy in her neurons and in businesswoman** the indeterminism is supplied by the “randomizing manipulator.” In this way, it might be thought that Pereboom can build his argument while preserving the crucial qualitative identity of businesswoman’s and businesswoman**’s psychological states. However, to achieve qualitative identity of the psychological states of businesswoman and businesswoman** it must in both cases be the effort of the agent which indeterministically causes the choice to help the victim. If the randomizer dictates the choice as Pereboom asserts—“The position the dial lands on therefore makes the salient difference as to which decision the agent makes (as quoted above)”—then it is not the agent’s effort that causes the choice to help the victim. Now, if the “randomizing manipulator” is just providing some indeterminacy to the decision process that is functionally equivalent in every way to the Kanean view of the role of quantum indeterminacy in the neurons of the brain, then businesswoman and businesswoman** could be both neurophysiologically and psychologically identical as they make their decisions. But, were this the case then when businesswoman** makes her decision it would not be where the dial on the manipulator lands which dictates the

Compatibilism and Manipulation Arguments

39

decision, it would be her indeterminate effort. In essence, if, as Kane’s view demands, the manipulation device was responsive to the path her indeterministic efforts take and not vice versa and her efforts ultimately dictate the choice she makes, then and only then would the psychological states of businesswoman** be identical to the psychological states of businesswoman.6 However, and crucially, if this were how the randomization device was working in the case of businesswoman**, so that it was responsive to the undetermined mental exertions of the agent, then we would have no reason to think that when the device is operative and she makes a decision, then she lacks freedom and responsibility. (iii) Clarke’s Argument A third kind of manipulation argument against libertarian views like Kane’s has been made recently by Roger Clarke (2012). In his manipulation case, he has us suppose it is wrong to buy eggs produced by caged hens and that it is morally acceptable to buy eggs from free-range hens. He also has us suppose Brown wants Smith to make a causally undetermined choice to do the morally wrong act of buying eggs produced by caged hens. Brown sends Smith to the grocery store to buy eggs and right in front of Smith are eggs produced by caged hens and more expensive eggs produced by free-range hens. Brown has a remote controller and he is watching Smith. If Smith makes the causally undetermined free choice to buy eggs from free-range hens and begins to do so (say, reaches for them), then Brown can press a button erasing Smith’s memory of the choice and make her relive the experience of seeing the two kinds of eggs and make her choose which to buy all over again. Brown will make Smith redo the choice until she makes the causally undetermined choice to buy the eggs from caged hens. Once the choice is made to buy the eggs from caged hens, then Brown allows Smith to go ahead and act on the choice and purchase them. According to Clarke, each choice is a free causally undetermined choice and, thus, it seems the ultimate purchase of the eggs from caged hens should be viewed by libertarians as something for which Smith is morally responsible, since it is the direct result of her free causally undetermined choice. But, says Clarke, this is a problem since Smith is manipulated in such a way as to rob her of moral responsibility for buying the eggs from caged hens. According to Clarke, even though it is true that Smith could have done otherwise when she chose to buy them, ultimately, her purchasing of these kinds of eggs was unavoidable, as Brown would keep repeating the scenario until Smith made such a choice (Clarke 2012, 140–141). Clarke goes on to talk about how a libertarian might best avoid the problematic conclusion that Smith is responsible for her purchase of the caged hen eggs. In considering this, he insists that, “The metaphysics of Smith’s choice at each iteration of the simulation is exactly as it would be without Brown’s influence” (Clarke 2012, 144). Additionally, he contends that

40

Compatibilism and Manipulation Arguments

the best strategy for the libertarian to take in response to this case would involve conceding that while Smith’s decision to purchase the caged hen eggs is a free choice, she is still not free in her purchase of them since it was inevitable that she would end up purchasing them. However, he argues that such a libertarian move is inconsistent with the best libertarian response to Frankfurt cases, which are intended to show, contra libertarians, that alternative possibilities are not necessary for free and responsible action. Clarke presents the following Frankfurt-style case derived from Frankfurt (1969): Black is a Republican mad scientist, and there is an election in progress. Black believes he owes his party more than just one vote, and so yesterday he implanted Jones with a microchip to ensure that Jones also votes Republican today. But though Black is mad, he is also careful, and so prefers not to interfere with Jones’s mind more than necessary, for fear of detection. Thus, the implanted chip is a mere fail-safe device: if Jones would have voted Republican without Black’s interference, the chip will do nothing; but if Jones would otherwise vote Democrat, the chip is activated and compels Jones to vote Republican instead. As it happens, Jones decides on his own to vote Republican, and the chip is never activated. It seems, then, that Jones should be judged responsible for his vote, since he acted as he did without Black’s influence. (We may assume that the process of implanting Jones with the chip did not itself affect Jones’s decision.) But Jones could not have done otherwise. Since it was impossible for him not to vote Republican, it seems the libertarian must not consider him responsible for his vote. (Clarke 2012, 144) The case is intended to show that, contra libertarianism, having the ability to do otherwise is not necessary for free and responsible action, since it seems intuitively obvious that Jones would be morally responsible for voting Republican in this case even though he could not have done otherwise. Clarke believes the best libertarian response to the Frankfurt case is that of David Widerker (1995), who uses a strategy akin to that of Kane (1985). Clarke argues that this best libertarian response to the Frankfurt case— Widerker’s response—is inconsistent with the best libertarian response to his Smith/Brown case. Thus, Clarke thinks libertarians are caught in a bind where they can only give a good answer to one of these criticisms of their view. Ultimately, I will show that Clarke’s argument here is based on a confusion about the nature of the best libertarian reply to the Frankfurt case as well as confusion about how libertarians should respond to Clarke’s own Smith/Brown case. To see this, we first need to get clear about how Clarke understands the best libertarian reply to the Frankfurt case. Clarke says that according to Widerker in assessing the responsibility of Jones in the Frankfurt case we should be concerned with whether he could

Compatibilism and Manipulation Arguments

41

have chosen not to vote Republican—was it in his power to choose to do otherwise than vote Republican? In explaining the response, he writes: It is difficult to see how Black could devise a microchip to act in the desired way with regard to such a metaphysically singular action, as opposed to a complex one such as voting. The chip must be designed either to be activated before Jones’s decision, or to be activated after the decision. If it is designed to act before Jones’s decision, there must be some cue that will tell the chip whether to activate or not. That is, there must be some sign S such that if S occurs at t1 then at some later t2 Jones will decide not to vote Republican. But if there is such a sign, then Jones’s decision at t2 is causally determined by a prior state of the world. To insist that this decision is free is simply to beg the question against the libertarian. On the other hand, if the chip is to be activated after the decision, then it does not genuinely remove the possibility of deciding not to vote Republican. If Jones decides not to vote Republican, he will be overridden by the chip before he can act on that decision—but he has made the decision. The libertarian can certainly account for Jones’s freedom in this case. (Clarke 2012, 145–146) In explaining the best libertarian reply, Clarke notes that our libertarian free choices should be causally undetermined right up to the moment of choice. Thus, according to libertarians, if Jones makes a free choice his decision must be causally undetermined right up to the moment of choice. If so, then there can be no prior sign telling Black how Jones intends to vote. Thus, if Black is to control Jones when he makes a causally undetermined free choice, then the chip in Jones’s brain must only allow Black to intervene after Jones has already made a decision one way or the other. As quoted above, this leads Clarke to say that in such a situation, “The libertarian can certainly account for Jones’s freedom in this case.” The problem here, however, is that this assertion is vague. In this case, can the libertarian account for Jones’s freedom in choosing or freedom in voting? That is, if Black with his controller is poised to make Jones vote Republican and if he allows Jones to make an undetermined free choice to vote Republican or Democrat, only intervening if need be after the choice, then does Clarke think the point is that Jones is free only in choosing or that he is also free in voting Republican if that is his choice? Ultimately, where Clarke stands on this is clarified when he describes the contradiction between what he regards as the best libertarian answer to his Smith/Brown case and what he regards as the best libertarian solution to the Frankfurt case. Referring to the Frankfurt-style case and the best libertarian response to it, he writes: But this response runs directly contrary to what was above called the most promising libertarian account of the Smith/Brown case. To account for

42

Compatibilism and Manipulation Arguments Smith’s lack of responsibility, it was recommended that the libertarian focus on the fact that she has no alternative but to act as she did, despite the fact that she could have chosen otherwise. Widerker’s response to Frankfurt cases says that Jones is responsible because he has the possibility of deciding to vote Democrat—if only for a short time, before Black overrides that decision—despite the fact that Black removes his ability to act otherwise. If Widerker-style libertarians are to hold that Jones is responsible for his vote, then, it seems they must also hold that Smith is responsible for her eventual buying of eggs in Brown’s simulation. The libertarian must bite the bullet and accept the counterintuitive upshot of one of the two types of cases: either Smith is responsible or Jones is not. (Clarke 2012, 146)

Clarke clearly thinks the best libertarian response involves the belief that when Jones makes a causally undetermined free choice to vote Republican and then Black allows him to go on and vote Republican, then Jones is responsible for voting Republican. The suggestion here is that since the undetermined choice to vote Republican was freely made, then Jones also freely votes Republican and is, thus, responsible for doing so. This, of course, is the source of what he regards as the problem for libertarians. Clarke thinks his own manipulation case involving Smith and Brown works with the Frankfurt-style case to force libertarians into a dilemma where either option illustrates the inadequacy of their position. He thinks the best libertarian solution to his Smith/Brown case conflicts with the best libertarian solution to the Frankfurt-style case. The point is that to answer the Smith/ Brown case the libertarian will need to say that Smith is not responsible for her purchase of the eggs since she couldn’t have done otherwise. To answer Frankfurt’s Jones/Black case the libertarian will need to say that Jones can be responsible for voting Republican even though he could not have done otherwise. Clarke thinks a good answer to one of the cases will exclude a good answer to the other. Clarke’s argument is based on significant confusion. First, a libertarian should readily admit that in Frankfurt’s Jones/Black case Jones can be responsible for voting Republican even though he could not have done otherwise. That’s because free will and responsibility only require that some of our actions, the SFAs, allow for alternative possibilities. Many of the free acts for which we are responsible are causally determined by our character and we cannot do otherwise. We are nonetheless free and responsible if such acts are determined by motivational states issuing from a character we have formed by prior undetermined SFAs for which we were responsible. Thus, if Jones’s choice to vote Republican is causally determined by a character he formed through prior SFAs for which he was responsible, then when he chooses to vote Republican now and Black lets him do so he is responsible

Compatibilism and Manipulation Arguments

43

for voting Republican even though he could not do otherwise. This is consistent with the most reasonable libertarian position. Additionally, Clarke’s understanding of the Frankfurt scenario differs in significant respects from the scenario to which libertarians like Kane and Widerker originally responded. The original Frankfurt case involves a controller poised to block a choice to vote for a Democrat, not the act of voting for a Democrat subsequent to choice. Kane’s and Widerker’s point was that free will requires that we engage in some free causally undetermined choices. Such free undetermined choices require plural voluntary control and alternative possibilities right up to the moment of choice. Thus, if Jones is choosing to vote for a Republican or a Democrat and if this is to be a free choice, then it must be undetermined right up to the moment of choice. If it is undetermined up to the moment of choice, then the intervener will have no prior sign telling him which choice Jones will make. So, if he wants to ensure a Republican choice, then he must intervene forcing Jones to choose a Republican. In this case, Jones’s choice would not be a free choice. On the other hand, if the intervener allows the agent to make his own undetermined choice, then as long as he has plural voluntary control in making the choice it will be a free choice. In this way, Kane and Widerker argued that the Frankfurt case poses no threat to a libertarian conception of free choice. In contrast, Clarke understands the Frankfurt case as a situation in which the intervener allows his subject to make the choice he wants to make, but then he will intervene if he begins to act on a choice he disagrees with. He envisions Jones having the freedom to choose as he wants, but if he begins to vote for a Democrat then Black will make him vote for a Republican. A libertarian should readily admit that the choice made in this scenario can be a free act for which the agent is responsible; indeed, the agent’s choice may be an SFA in Kane’s sense if it is causally undetermined and the agent exerts plural voluntary control over it. Furthermore, in this scenario if the agent freely chooses to vote for a Republican and then does so, then the libertarian should say he is responsible for voting for the Republican. What he should not say, however, is that his voting for the Republican is an SFA. For this is not an act over which he exerts plural voluntary control; the intervener will not allow him to do otherwise than vote for a Republican. Having sorted through these matters, let’s turn to Clarke’s Smith/Brown case. This is a new kind of case, and we should consider what a libertarian should say about this and whether the best libertarian response conflicts with anything I’ve said above. Recall that in the Smith/Brown case, Brown is the intervener who wants Smith to make a causally undetermined choice of caged hen eggs and then purchase them. If Smith chooses free-range hen eggs, then Brown will hit a button forcing a replay, making Smith redo her causally undetermined choice. Brown will keep forcing replays until Smith makes the choice he wants and then he will let her go on and purchase the caged hen eggs.

44

Compatibilism and Manipulation Arguments

The libertarian should allow that every causally undetermined choice Smith makes in this scenario may be a free choice for which she is responsible, if she has plural voluntary control (PVC) over these choices. Indeed, every choice she makes could count as an SFA if she has PVC over them and they shape her character. But her purchase of the eggs should never be counted as an SFA, because it is just the determined result of her prior choice of which eggs to buy. Additionally, the libertarian should say that in the scenario in which Smith’s first choice is a free choice of caged hen eggs her subsequent purchase of those eggs is a free act for which she is responsible. For in this case it is the determined consequence of her undetermined free choice to purchase them. But suppose her first choice is for free-range hen eggs, and, thus, Brown forces a replay. At this point, despite the freedom of any subsequent choice she makes and no matter how many replays occur, Smith’s purchase of the caged hen eggs should then be seen as an unfree act for which she is not responsible. The latter claim may seem odd. The reader may respond, If it takes a number of replays, then why should that matter to the agent’s responsibility? If we allow that she purchases them freely when on the first try she freely chooses caged hen eggs and purchases them, then we should count her later replayed choices as free and her later purchase of the caged hen eggs as free. In answer to this, I would note that there is a relevant difference here. When the first undetermined free choice is for caged hen eggs and she is allowed to go on and purchase them, Brown never intervenes. Thus, as her purchase is just the unhindered consequence of a free choice she made, there is no actual manipulation involved and we should see such a purchase as a free act for which she is responsible. In contrast, if the first choice goes the other way and Brown forces a replay onto Smith, then while any of the subsequent choices will be free the later purchase of caged hen eggs will not be free. This is because the later purchase based on a replay choice will be the result of Brown’s intervention. Had he not forced the replays on Smith, then Smith would not be purchasing the caged hen eggs. Clarke maintains that the best libertarian response to the Smith/Brown case is to say that any of the choices would be free but none of the purchases would be free since the intervener makes it such that caged hen eggs must be purchased. But I’m suggesting the best libertarian response should be more nuanced, allowing that when the first choice is for caged hen eggs the purchase of them will be a free act and any purchase based on a replay choice will be unfree because the manipulator would then have actually intervened to guarantee the purchase of caged hen eggs. Clarke also maintains that the best libertarian response to the Smith/ Brown case is inconsistent with the best libertarian response to the Frankfurt case. Is this true? Well, I’ve just described the best libertarian response

Compatibilism and Manipulation Arguments

45

to the Smith/Brown case. Is it inconsistent with the best libertarian reply to the Frankfurt case? The best libertarian reply to the Frankfurt case is the Kane/Widerker reply. But, recall that they understand the Frankfurt case as concerned with freedom of choice, not the freedom of action after choice. Their point is that libertarian free choices must be causally undetermined up to the moment of choice. Thus, if a controller wants to ensure the choice of a Republican, he will have to force that choice upon his subject, not letting him freely choose. Thus, the Frankfurt case carries no weight for those committed to a libertarian conception of free choice. Given this correct understanding of the Kane/Widerker (the best) libertarian reply to the Frankfurt case, it is utterly unclear how this conflicts with what I’ve described as the best libertarian reply to Clarke’s Smith/Brown case. Despite these points, it might be argued that it’s still legitimate for Clarke to run the Frankfurt scenario in a different way, making it more relevantly similar to his Smith/Brown case. In effect, this is what he does in his essay. That is, he has us suppose that the controller, Black, lets Jones make a choice for himself regarding whom to vote for. If Jones chooses a Democrat, Black forces him to switch his vote; if the choice is for a Republican, Black lets Jones vote for the Republican. This is different from the original Frankfurt scenario to which Kane and Widerker responded, but let’s grant it. Clarke says that when the agent makes an undetermined libertarian free choice of the Republican and is then allowed to go on and vote for the Republican, the best libertarian response should be that the agent acts freely in voting Republican. Why? Because voting Republican is a determined consequence of his libertarian free choice. Thus, even though he couldn’t do otherwise than vote Republican, the libertarian should in this case see it as a free act for which he is responsible. What should we make of this? First, let me say that Clarke is right about what the proper libertarian response to his modified Frankfurt case should be. However, he is wrong to think this conflicts with the best libertarian response to his Smith/Brown case. Clarke’s modified Frankfurt case involves no replay. If in the modified Frankfurt case, Jones makes a libertarian free choice to vote for a Republican and Black then lets him vote for a Republican, then the libertarian should say Jones acts freely in voting for the Republican. This is akin to the situation in the Smith/Brown case where Smith makes the choice Brown wants on the first go around. Here when Brown lets Smith go on and purchase the caged hen eggs, then, as I argued, the best libertarian response is to say the purchase of the eggs is a free act. Why? Again, in this situation Smith makes the purchase as a result of a free choice without being subjected to Brown’s manipulation. Now, indeed, the libertarian should see any purchases based on replays as unfree, because those choices will be due to Brown’s manipulation and Brown will be responsible for the purchase, not Smith, since Brown will have intervened to ensure that it would occur. But this is entirely consistent with the libertarian’s best response to Clarke’s modified Frankfurt case. For

46

Compatibilism and Manipulation Arguments

as I note above that case is only akin to the situation where Smith freely chooses as Brown wants on his first choice; and in this situation the best libertarian response is in line with the best libertarian response to Clarke’s modified Frankfurt case. For all of the above stated reasons, then, we should see that Clarke is mistaken when he says he has caught the libertarian on the horns of a dilemma. Contra Clarke, the best libertarian response to his Smith/Brown case is consistent with the best libertarian response to the original Frankfurt case as well as Clarke’s modified Frankfurt case. (iv) Reflections In this section, I’ve tried to show that despite recent attempts to prove otherwise, Kane’s libertarian view seems to be immune to manipulation arguments. Perhaps some better manipulation argument could be made against Kane’s view, but absent such an argument we have reason to think his view is immune to this problem. Thus, if it can be shown that compatibilism really does fall prey to manipulation problems, then we have good reason to give further consideration to libertarian views.

Manipulation Cases: An Intractable Problem for Compatibilism As I noted earlier in this chapter, compatibilists have given both hard-line and soft-line responses to the manipulation arguments against their views. In what follows, I will try to show that either approach is problematic in significant respects. (i) Problems With the Hard-line Response The hard-line response involves biting the bullet of intuition and maintaining that, contra the manipulation argument, it just doesn’t matter how one came to possess the beliefs, values, motivational states, and character one has. Suppose one acquires these through manipulation by others, as in Kane’s examples of the CNC manipulation we find in Huxley’s Brave New World or Skinner’s Walden Two. Even so, according to the hard-liners, as long as the agents acting from these beliefs, values, and so forth do so while meeting the most plausible Compatibilist-friendly Agential Structures (CAS), we have good reason to believe they act freely and are morally responsible for what they do. So, for instance, as McKenna (2012, 152–153) notes, Frankfurt has taken this approach. On Frankfurt’s view, even if you or I have been manipulated into having the beliefs, values, and characters we have through the regimented social programming of others and we now do morally bad deeds as a necessary consequence of this programming, we are still rightly regarded as acting freely and morally responsible for this wrongdoing if we

Compatibilism and Manipulation Arguments

47

meet his compatibilist criteria of free and responsible action, his conception of the CAS. This kind of hard-line response is wrongheaded. Manipulation by others can be designed so as to lead us to act in ways that satisfy whatever CAS a compatibilist favors, whether Frankfurt’s or Fischer’s and Ravizza’s or some other’s. If one then acts with necessity from a character one has had no hand in making and even if one meets plausible CAS standards, it still doesn’t make sense to view one’s actions as free willed or as actions for which one is morally responsible. Suppose the manipulated character one has leads one to act badly while satisfying a plausible conception of CAS. To now regard this person as having acted with free will and as morally responsible is to regard this person as a fitting object of our blame. But this hardly seems fair when she has had no hand in shaping the character which now determines her conduct. Perhaps, my point can be better understood if we consider an interesting example from McKenna (2004)—the case of Suzie Instant. Consider . . . the case of Suzie, who is created by a god at an instant and who is placed in a determined world. Suzie is created to be a psychologically healthy woman indistinguishable from any other normally functioning thirty-year-old person whom any of us might encounter. To get this result, she is given a huge set of beliefs according to which she has lived a normal human life for thirty years. For instance, she believes (falsely) that she had a twelfth birthday and that her daddy bought her a pony. Furthermore, Suzie has some range of values and principles that are unsheddable. In Suzie’s case, she has a set of false beliefs about how she came to acquire her now unsheddable values. She thinks that she acquired them through a process of sustained effort over the years leading up to what she thinks is her thirtieth. She takes pride in this fact and believes that she is responsible for this process and that she engaged in it freely. (On this point, she is clearly mistaken.) Also, assume that she is a richly self-controlled person who is able to resist the inclination to act with weakness of will (a nonhistorical requirement Mele specifies for CAS), that she satisfies something like Frankfurt’s hierarchical account of freely willed conduct. Furthermore, she is reasons-responsive in the manner that Fischer and Ravizza and also Haji require (sans any historical component). In short, Suzie satisfies the juiciest nonhistorical demands of the Compatibilist-friendly Agential Structure. Add to this the fact that she has a robust and relatively consistent range of beliefs about her history (which for her are false), like those that any psychologically healthy person would have. (180) In the case of Suzie Instant, Suzie has been popped into existence by a god and she has the psychological properties of the typical adult human being

48

Compatibilism and Manipulation Arguments

and she meets the kinds of conditions for freedom and responsibility that are expressed in a variety of the contemporary Compatibilist-friendly Agential Structures (CAS). McKenna adds on to the case, having us imagine that soon after popping into existence, she does a bad thing. She carelessly and selfishly knocks down one of your small children in her rush to get to a taxi. McKenna has the reader imagine that she sees Suzie do this and that the reader knows (a) Suzie has just been popped into existence and (b) Suzie has all of the mental abilities of the typical adult human being, which in part means that she has all kinds of beliefs and desires and values and that she has a sense of right and wrong and she knows what it means to be blamed and held responsible for one’s actions. Then, he asks whether it would be wrong to hold Suzie responsible for her action in this case and blame her for what she has done. As my preceding comments would suggest, my view is that if one knows that she was just moments ago popped into existence with beliefs, values, and a character that are not of her own making and she now is causally determined to do what she does due to this character, then she should not be blamed. If she doesn’t have control over the construction of that character and it now determines her to act as she does and you know this about her, then it is unfair to blame her for what she does. Hard-liners, like Frankfurt, will disagree with my stance on this but, again, that just seems wrongheaded. I would stress here as well that the reason why Suzie lacks responsibility for her bad behavior is not simply due to the fact that she’s had no control over the formation of her character, it is also due very much to the fact that her unchosen character now determines her to act as she does. Recall that Haji and Cuypers give a similar but significantly different example in which beliefs and values have been implanted into an agent by a manipulator and these values conflict, leading the agent to make a causally undetermined choice resolving this conflict. They argue that one would not be responsible for such a choice since it issues from a character not formed by the agent but from a character formed by the manipulator. In response, I noted that in this scenario if the undetermined choice met the specifications of a Kanean SFA, resulting from dual efforts of will and exhibiting plural voluntary control, then the agent would have some very minimal responsibility for the choice. So too, if Suzie Instant were to engage in a causally undetermined choice constituting an SFA soon after coming into being, then she would bear a very minimal responsibility for her choice. But, if Suzie is necessitated to act from a character she has not formed for herself, as in McKenna’s case, then she is not responsible at all for her action and deserves no level of blame for what she does. Interestingly, McKenna offers up a way of defending the hard-line response by providing a theory about the nature of moral responsibility and the nature of blame. He talks about the nature of blame and the conditions that must be met for justified blame. He says he understands moral responsibility in terms of the suitability of praise and blame. On his view, persons who act

Compatibilism and Manipulation Arguments

49

wrongly and are morally responsible for their action are suitable objects of blame. And, borrowing from P.F. Strawson (1962), he says, To blame a person is to respond to her or to judge it appropriate to respond to her, in light of a perceived moral wrong reflected in the quality of her will, with a morally reactive attitude of resentment or indignation. For a person to be blameworthy is for such a response to her to be appropriate (other things being equal). (2004, 187) McKenna also understands the act of blaming as a kind of engagement in a conversation. To blame someone for wrongdoing “is a manner of responding to what she had done as on analogy with a conversation in which the blameworthy person’s conduct has a significance” (2004, 187). He considers the placing of blame as inviting a response from the person blamed, where the person blamed may respond with “an apology, a defiant unwillingness to acknowledge the wrong, a justification, a plea for forgiveness, and so on” (2004, 188). His conversational theory of blame is derived from the work of Gary Watson (1987). McKenna goes on to note that on this understanding of blame there is no essential connection between justified blame and justified retribution and punishment. Just because it is fitting to blame persons and express that blame does not mean that one is justified in subjecting them to retribution or punishment. What is essential to blame on the model I endorse is that to blame a person is to morally communicate with her by way of one’s attitudes so that the person is morally engaged, appreciates the demand that she not behave in that manner, and acknowledges that she will not be regarded as one who is (otherwise) in good standing within the moral community. (2004, 188) He also says that for a person to be a fitting object of blame depends on whether she can “engage in appropriate moral ‘dialogue’ with us by way of our moral responsibility practices” (2004, 188). The idea here is that the person blamed must be able to understand the significance of being blamed and understand that she is then called upon to apologize for her action or justify her action or ask for forgiveness, and so forth. With these points in mind, McKenna returns to the case of Suzie Instant, writing: Imagine that you were to have a chance encounter with Suzie Instant, well aware that, despite her strange history, as she is in her current state, she is a completely competent moral “interlocutor” in the fashion relevant to a conversational model of moral responsibility. Suppose she

50

Compatibilism and Manipulation Arguments pushed your child down to get to a taxi in a hurry. Knowing that Suzie Instant is fully articulate in this way, what would speak against judging her to be a morally responsible agent? Being able to think of her in this way, what would speak against “conversing” with her under the warranted belief that she satisfies the proper status as a being who can be so engaged? In keeping with Strawson’s Quality of Will Thesis and Watson’s development of it by way of a conversational model, I confess, I cannot see what good reasons there would be for not thinking of engaging with Suzie Instant in relevant ways. Furthermore, there are transparently good reasons for doing so—she knocked over a child to get to a taxi, fully capable of appreciating the moral significance of what she did, and fully capable of engaging in a “moral transaction” with respect to it. (2004, 188)

Here McKenna makes the case that blame would be appropriate even though you know that Suzie Instant was popped into existence by a god who has given her all of her beliefs, values, and psychological characteristics which have now led her to knock down a child in her haste to catch a taxi. That is, you know that she has not chosen to have these beliefs, values, psychological traits, and so forth. As I’ve argued this fact gives one grounds to believe that blaming her for her actions is misplaced, as it is unfair to her. But McKenna responds to this suggestion as follows: One might think it is unfair. Poor Suzie Instant was fashioned all in an instant without her knowledge. If she now does morally wrong, just after being so fashioned, it seems unfair to blame her. But I suspect the worry about unfairness is largely based upon the thought that blame involves harm brought on by deserving a retributive response. What is unfair, it might be thought, is that Suzie Instant is now to be harmed for acting from actional resources that she had no part in making. I agree that any time it is judged that a person should be harmed, moral considerations regarding fairness and desert must take center stage. But if blaming is as I have characterized it, then matters of fairness are misplaced. It is, instead, a matter of fittingness. If to blame a person is to communicate with her in a certain fashion in response to a morally salient will and conduct, then either the will and conduct makes the reply (the blame) fitting or it does not. It is a different matter whether, if a person is worthy of blame and is indeed blamed in some manner, she should also be harmed or otherwise punished in any way. (2004, 189) I find McKenna’s argument problematic. He doesn’t sufficiently attend to the nature of blame and, thus, doesn’t see that a kind of harm to the person blamed is the typical consequence of the act of blame. Recall that, as McKenna

Compatibilism and Manipulation Arguments

51

does himself define it, to blame a person is to respond to her or to judge it appropriate it to respond to her, in light of a perceived moral wrong reflected in the quality of her will, with a morally reactive attitude of resentment or indignation. Resentment and indignation involve expressing displeasure or anger towards someone for what they have done. Typically, we feel bad when others are displeased or angry with us. Feeling bad in reaction to being blamed amounts to a kind of psychic harm which should not be taken lightly or overlooked. To blame persons for wrong actions is not simply to point out to them that they have acted wrongly. Rather, the act of blame carries with it certain attitudes and to express one’s blame, to communicate it to another, requires conveying not just the sense that this other person has acted wrongly but also that her wrong action makes you displeased or angry with her. McKenna wants to draw a line between the notions of retribution and punishment on the one hand and blame on the other that is not so easily drawn, as there is harm—psychic harm—that is the typical accompaniment with being the target of blame. As quoted above, he states, “I agree that any time it is judged that a person should be harmed, moral considerations regarding fairness and desert must take center stage. But if blame is as I have characterized it, then matters of fairness are misplaced.” The problem is that he has rightly characterized blame as responding with resentment or indignation, however, those subjected to this are typically hurt by this. So, moral considerations of fairness and desert should take center stage in judging whether blaming is appropriate. At this point, it might be wondered what the appropriate response to Suzie Instant’s wrongdoing should be. To this I would suggest that she should not be blamed, she should not be the object of our resentment or indignation, rather she should simply be told not to act in this way and it should be explained to her why she should not act in this manner. Perhaps, in the future if she does not conform to advice and recommendations on how to act appropriately, then at that point blame may be justified. But, even then one should proceed with caution because there may be fixed elements of her character which she had no choice in making that would make her an undeserving object of our blame. Ultimately, the case of Suzie Instant gives us strong reasons to believe that the hard-line response to CNC manipulation arguments is problematic. Even though Suzie Instant may meet the most plausible nonhistorical standards of compatibilist free will or moral responsibility, we should not regard her as morally responsible. For to view her as morally responsible implies that she would be a fitting target of our blame if, soon after popping into existence, she is determined by her character to commit bad actions. This is plainly unfair given that she is determined to act badly from a character which she had no hand in making. Perhaps, some better defense of the hard-line approach could be given than the one McKenna offers, but absent such an argument we should conclude that the hard-line response to CNC manipulation arguments is inadequate.

52

Compatibilism and Manipulation Arguments

(ii) Problems With the Soft-line Response Compatibilists who endorse a soft-line response to manipulation arguments believe it matters how agents came to possess the character and motivational states which determine their behavior. In his writings, McKenna (2012) distinguishes between positive historical compatibilism and negative historical compatibilism, each of which provide soft-line responses to the manipulation argument. Historical compatibilists, in clear opposition to Frankfurt’s contention, have argued that among the conditions necessary for free and morally responsible action is some kind of history requirement. Either the agent must have come to be in the state she is in just prior to her free action through some kind of freedom-and-responsibility-enabling historical process, or she must not have come to be in the pertinent state from a history that included any freedom-and-responsibility-undermining process. Call the former sort positive historical compatibilists and the latter sort negative historical compatibilists. For instance, Fischer and Ravizza advance a positive historical thesis. They require, roughly, that when an agent acts freely, she acts from agential springs that are hers—that is, that she owns—by virtue of the fact at an earlier time she took responsibility for those springs of action. She did this by going through a process whereby she came to see herself as an agent in the world, one who is a fair target of blame and praise, and she came to so see herself as such on the basis of appropriate evidence. Agents who have not come to acquire their springs of action at a time through such a history do not act freely while, other things being equal, agents who have such a history do act freely. Mele instead recommends a negative historical thesis. Unlike the positive historical compatibilists, on the view he advises compatibilists to adopt, there is no particular kind of history an agent must possess relative to her actions for them to be free. An agent acts freely and is morally responsible for what she does only if she lacks a history of a certain sort relative to the causes of her action. Roughly, if her act is generated in a certain way by pertinent attitudes, such as unsheddable values, and if she came to acquire those attitudes through coercive means, means that bypassed her ability to critically assess, endorse, and sustain them, she does not act freely. (McKenna 2012, 153–154) Historical compatibilists can respond to CNC manipulation cases in one of two ways. First, if they feel the manipulation case does not allow the relevant agent to meet their preferred conception of the relevant historical compatibilist conditions of responsibility, then they will conclude that the case does not lead them into a mistaken conclusion about the responsibility of the relevant agent. So, for instance, consider Suzie Instant. As I’ve

Compatibilism and Manipulation Arguments

53

argued, her case can be used to argue that a nonhistorical compatibilist view, like Frankfurt’s, leads to the wrong conclusion that she is responsible for selfishly knocking down a child. However, an historical compatibilist, like Mele, will see this case as no threat to his view. Mele will say the values Suzie Instant possesses which now causally determine her to act badly were not acquired in such a way as to make her responsible for what she does. Rather, her values were foisted upon her and she begins acting in accord with them immediately without her having the opportunity to critically assess them, endorse them, and sustain them. For similar but different reasons, Fischer’s and Ravizza’s historical compatibilist view would lead to a similar conclusion about the case of Suzie Instant. On the other hand, if the manipulation does allow the relevant agent to meet their preferred conception of the relevant historical compatibilist conditions of responsibility, then once again they will say they are not led to the wrong conclusion. For instance, suppose that in Kane’s Walden Two case the adults living there have been manipulated to have the values they have but at the same time over the years and as they’ve matured they’ve been able to critically assess those values, endorse them, sustain them. In essence, suppose that, despite the manipulation, the kinds of historical compatibilist conditions endorsed by, say, Mele and Fischer and Ravizza are met. In this case, the historical compatibilist can simply say, “Yes, there’s manipulation involved here, but it is not of the kind to threaten responsibility.” Here they will admit that the manipulated agents are responsible but do so with no embarrassment, since in meeting their historical conditions there’s nothing unreasonable or counter-intuitive about coming to this conclusion. While historical compatibilists have good reason to reject manipulation arguments grounded on cases like that of Suzie Instant, they will have a hard time providing a good response to more sophisticated cases in which their various historical compatibilist conditions are met. For instance, suppose there was an all-knowing, all-powerful being who for some reason liked the way all events transpired on our planet Earth for the past 100 years. Suppose he decided to create a replica of our planet and its inhabitants as it was 100 years ago, and then using his knowledge and power he made it so that on that planet, call it “Twin Earth,” all our lives were played out again so that our doppelgangers there thought, reasoned, behaved in exactly the same ways the residents of our planet did for the past 100 years. Now, presumably in the past 100 years many of us humans have engaged in many actions which have met the various historical compatibilist requirements for free and responsible action. It’s not as though Mele and Fischer and Ravizza think their conditions of responsible action are never or seldom met, quite the contrary. But this being the case, they will have to say that our doppelgangers on Twin Earth, who have been designed and determined to reason, think, decide, and act in exactly the same ways we have reasoned, thought, decided, and acted here, are morally responsible for the same acts

54

Compatibilism and Manipulation Arguments

there that we are responsible for here. So, for instance, suppose that Jane on our Earth contemplates doing either good act A or bad act B, and she chooses and does bad act B, and this choice meets the kind of historical compatibilist conditions for free action articulated by Mele or Fischer and Ravizza. If so, then they will have to say she is responsible for choosing and doing bad act B. But now her doppelganger, Jane*, on Twin Earth with an entirely similar life history reasons in exactly the same way and chooses B as well, and all of this, of course, is a result of the all-powerful, all-knowing being’s designed plan for Twin Earth. The historical compatibilists will have to say that Jane* is responsible for her choice as well, even though it is the necessary consequence of the planning and manipulation of Twin Earth’s designer. In McKenna (2004) he points out that Fischer and Ravizza respond in this way to a similar case which has been proposed by Carl Ginet. He also notes that Mele has responded in similar fashion to a related case of his own construction.7 While my own intuition is that our doppelgangers on Twin Earth would not possess free will and would not be morally responsible for their actions, I do not think that the historical compatibilist has any reason to think this case alone gives them good reason to doubt the truth of their view. They may simply feel that my take on the Twin Earth case just reflects a prior commitment to incompatibilism, but that the case itself gives the compatibilist no reason to doubt his view. Indeed, this is exactly how Fischer (2000, 415) has responded to Ginet’s own analogous case. At this point, however, I want to revisit the case of Suzie Instant. Recall that she is popped into existence with a full set of beliefs, memories, values, desires, and some bad character traits. Also, soon after being popped into existence she does some bad things, like selfishly knocking down a child while trying to get to a taxi cab. As I’ve said, the historical compatibilists will say she is not responsible for these bad things she does right after coming into existence, because she has not formed the character which determines these bad actions in the right sort of way for her to be responsible for what she does. Given that right after popping into existence Suzie Instant’s bad actions are determined by her character and she had no control over the formation of her character, the historical compatibilist is absolutely correct in saying that she is not responsible for what she does. But this conclusion about the case of Suzie Instant can be seen to conflict with the historical compatibilists’ take on a case like my own Twin Earth case. As noted, in responding to my Twin Earth case and to be consistent with their principles, historical compatibilists must say the manipulated agents on Twin Earth, such as Jane*, are responsible as long as they meet their historical compatibilist conditions for morally responsible action. However, the designer of Twin Earth has set things up so that everyone on Twin Earth shall reason, think, believe, desire, and act as they do. Just like many people on Earth, many of these doppelgangers on Twin Earth will critically reflect on their values, embrace their values, take responsibility for themselves. But

Compatibilism and Manipulation Arguments

55

which doppelgangers do this, how they do this, when they do this will have all been antecedently established by the designer of Twin Earth. What I want to suggest here is that since it is correct to say Suzie Instant is not responsible for what she does after popping into existence and since she is not responsible because she lacks control over what she does, then so too our doppelgangers on Twin Earth are not responsible. For they too lack control over what they do. Sure, many of them will often meet historical compatibilist criteria of free and responsible action, but they still won’t have the kind of control over what they do to be genuinely responsible for what they do. For everything they do will just be done as required by the planning of the designer. My point is that when we reflect on why Suzie Instant is not responsible for what she does right after popping into existence, we see that it is because she’s not in control over what she does. She’s not in control because she’s determined to act as she does from a character she had no hand in making. But, once we see this, we should see as well that our doppelgangers on Twin Earth have no control over their actions either, even though they often meet historical compatibilist conditions for free and responsible actions. Historical compatibilism will be lead to the wrong conclusion about the freedom and responsibility of those on Twin Earth, and seeing why they are right about the Suzie Instant case helps to see why they are wrong about the Twin Earth case. In reflecting further on these matters, I would note as well that it is determinism that really presents the block to the agent control and, thus, the free will and the moral responsibility of Suzie Instant and our doppelgangers on Twin Earth. As I’ve said earlier, if right after popping into existence Suzie Instant’s bad acts were causally undetermined acts which also met Kane’s criteria for being self-forming acts (SFAs), then she would act with free will and possess a very minimal kind of responsibility for her action. As it is, however, in the example she is determined to act as she does from a character she had no control in forming. Thus, she does not act freely and is not responsible. Additionally, as we turn to the Twin Earth case, we should see that were the designer to merely give our doppelgangers an initial set of values, beliefs, characters and let them live out their lives operating with these and making some causally undetermined SFAs, then they could still possess free will and be morally responsible. But, as it is, on Twin Earth the designer has set things up so that everything our doppelgangers do must be in accordance with his plan to have them act and decide as we have done here on Earth for the past 100 years. Further, while the Twin Earthers may critically reflect on their values, endorse their values, take responsibility for them, this will not suffice for giving them free will or for making them morally responsible. For the path their critical reflections follow and how and whether they endorse their values or take responsibility for them will have been preordained by the designer. So, too, if we on planet Earth are subject to determinism in all aspects of our lives whether by a divine plan or by the blind operations of natural scientific laws operating in conjunction with prior states and events, then we never act with free will or moral responsibility either.

56

Compatibilism and Manipulation Arguments

Notes 1. Kane has articulated and defended his libertarian theory in many published writings. For a sampling, see Kane (1985, 1996, 2002a, 2007a, b, 2011, 2016a, b). 2. For instance, see SFW, Ch.5, Sect 7. 3. See McKenna (2012, 151–152). See Frankfurt (2002, 27–28) for his expression of a hard-line response. McKenna notes that Kane (1996, 67–68) acknowledges a similar distinction between these compatibilist strategies. Kane refers to them as “hard compatibilist” strategies and “soft compatibilist” strategies; and he identifies Thomas Hobbes and B.F. Skinner as advocates of the hard compatibilist/ hard-line response. 4. See Kane (2007b, 174–175 and 2011, 398–399). 5. One way compatibilists have tried to avoid the problem mentioned here is to include some kinds of historicist constraints on what counts as a free action. So, for instance, they hold that it matters how one came to possess the motivational states one has which lead one to act now, such that if one has been manipulated to have the motivational states one has and did not forge for oneself the character from which these motivational states issue, then one is not responsible for one’s action. See McKenna (2012) for a nice discussion of the issues. He notes that Fischer and Ravizza (1998) and Alfred Mele (1995, 2006) endorse such historicist principles. 6. In Dennett’s discussion of his own manipulation case against Kane’s view, he notes that in correspondence Kane has said that, “The indeterminacy-producing mechanism must be responsive to the dynamics within the agent’s own will and not override them or it would be making the decisions and not the agent” (Dennett 2003, 133). For a defense of Kane’s view against Dennett’s related kind of manipulation argument, see Lemos (2017b). 7. See McKenna (2004, 177–179). See also John Martin Fischer (2000, 408–417, 414– 415) and Alfred Mele (2006, 188–193).

4

Kanean Libertarianism Examined

In the previous chapter, I showed that while Kane’s libertarian theory of free will and responsibility can be adequately defended against manipulation arguments, the plausibility of compatibilist theories of free will and responsibility are seriously challenged by such arguments. In this chapter, I want to give further consideration to Kane’s theory by considering a variety of different kinds of objections which have been made against the view. I will start with a brief review of the central tenets of Kane’s libertarian view. I will then consider three different lines of criticism against which Kane’s theory can be adequately defended. Then I will go on to consider two other lines of criticism, which present more serious challenges to Kane’s position raising concerns about the dual efforts he posits in his conception of SFAs. In Chapter Five, I will go on to critically examine different event-causal theories which are not grounded on a theory of dual efforts and I will suggest some new ways of thinking about SFAs which might solve problems I find in Kane’s view.

A Brief Review of Kane’s View (i) The Basics Robert Kane believes that many of the free acts for which we are responsible are causally determined. Using the example of Martin Luther breaking his association with the Roman Catholic Church, Kane notes how Martin Luther said, “Here I stand, I can do no other.” Kane says that even if it were literally true at the time of his break from Rome that Luther could do no other, Luther could still rightly be said to act freely and responsibly in breaking off his association with the Roman Church. That is, even if his decision to break from Rome is causally determined by his character or his motivating states, such as his beliefs desires, preferences, or intentions, Luther could still be regarded as acting freely in doing so and as being responsible for this decision. However, as a libertarian Kane believes that any such determined free actions have only a derivative freedom. That is, such causally determined free acts get their freedom from the fact that they proceed from

58

Kanean Libertarianism Examined

a character or a motivational state that was shaped by earlier causally undetermined free acts. Kane calls these causally undetermined free acts which shape our character or motivational states “self-forming acts” (SFAs). Kane says SFAs are causally undetermined acts which we perform at those difficult times in life when we are strongly motivated to perform two or more different actions and we can only perform one of them. These situations often occur when we are torn between the desire to do what is morally right and the desire to serve our own self-interest or when we are torn between the desire to satisfy some immediate desire and the desire to sacrifice now for some end to be achieved later. Kane often illustrates the nature of his view on SFAs with the example of a businesswoman who encounters an assault taking place while on her way to an important business meeting. Upon seeing the assault occur she immediately acquires the desire to stop and prevent the assault, but she also desires to go on to her important meeting. She realizes she cannot do both and she feels torn as to what she should do. According to Kane, in this kind of situation the agent actually exerts efforts to choose to do each act she is contemplating. That is, she tries to make the choice to stop and prevent the assault and she tries to make the choice to go on to her business meeting. Kane has us suppose that, whichever choice she ends up making, her ultimate decision is causally undetermined, and he says that such causally undetermined acts are SFAs. He also notes that since in SFAs we try to make each choice and since we are responsible for what we try to do, it follows that in SFAs whichever choice we make we will be responsible for that choice. Again, Kane’s view is that many of the acts we perform and for which we are responsible are causally determined. He also thinks that when we are responsible for such determined actions, it is only because we have shaped the character or motivational state from which they proceed by earlier SFAs, which are causally undetermined actions that involve making dual efforts to do two different actions at the same time. Kane believes these SFAs are the underivatively free actions that are the basis of the freedom and responsibility we have in later actions we perform which are causally determined.1 (ii) The Problem of Luck, Dual Efforts, and a Key Argument From Analogy Since he believes the underivatively free actions must be causally undetermined, Kane’s view is often criticized on the ground that it suffers from a problem of luck. The idea here is that if it is causally undetermined what the agent does in acting freely then what the agent does would seem to be a matter of luck, a random happening uncontrolled by the mental states of the agent. The luck objection is often phrased in terms of a counterpart agent in a nearby possible world. To develop this point, imagine that our businesswoman decides to do the moral thing and stop and prevent the assault. Moreover, suppose that her decision is not causally determined. If

Kanean Libertarianism Examined

59

the businesswoman’s decision to do the moral thing is not deterministically caused, then there will be a nearby possible world with the same laws of nature and with the same past up to the moment of decision in which the businesswoman (or her counterpart) decides differently—she decides to do the self-interested thing and carry on to her meeting. But since the laws of nature and the past up to the time of the decision are the same in both worlds, then—as Alfred Mele (1998) puts it—“there is nothing about the agent’s powers, capacities, states of mind, moral character, and the like that explains this difference in outcome” (583). There is nothing to explain why she decided one way in one world but decided the other way in the other world. And if there is nothing that explains this difference, then it seems that the businesswoman’s decision to do the moral thing (in the actual world) is just a matter of (good) luck, and so it cannot be free or responsible action.2 Kane believes the dual efforts/willings involved in SFAs are crucial to solving the problem of luck. Recall that, according to Kane, in our underivatively free actions, we will/exert effort to do two or more competing acts and ultimately end up doing only one of them. Kane maintains that we are in control of and responsible for what we will to do. Thus, while what we do in SFAs is causally undetermined, what we ultimately end up doing is something that we willed to do and thus in deciding and acting we are responsible for our decision. Think of our businesswoman. She wills to both go on to her business meeting and to stop and prevent the assault. Suppose she ultimately stops and prevents the assault. On Kane’s view, the decision to prevent the assault is causally undetermined, but the businesswoman is responsible for it as it is a consequence of her willing it. Further, had she gone on to her business meeting she would have been responsible for that choice for similar reasons.3 Critics have complained that despite the dual willings/efforts involved in the Kanean picture of SFAs there is still a problem with the view. Sticking with the example of our businesswoman, critics note that since it is causally undetermined whether she stops to prevent the assault or go on to her meeting, she does not then exert control over which of these acts is performed. Because of this it is believed she cannot be responsible for doing the one thing as opposed to the other. To answer this sort of worry Kane has given various arguments from analogy intended to show that causal indeterminacy in the performance of an action does not entail a lack of responsibility for the act. One of them proceeds as follows: A husband, while arguing with his wife, in a fit of rage swings his arm down on her favorite glass-top table top intending to break it. Again, we suppose that some indeterminism in his outgoing neural pathways makes the momentum of his arm indeterminate, so that it is undetermined whether the table will break right up to the moment when it is

60

Kanean Libertarianism Examined struck. Whether the husband breaks the table is undetermined and yet he is clearly responsible, if he does break it. (It would be a poor excuse to offer his wife, if he claimed: “Chance did it, not me.” Though indeterminism was involved, chance didn’t do it, he did.) (Kane 2007a, 27)

Here the husband is clearly trying to break the table by striking it, but it is causally indeterminate as to whether the table will break. If it does break, the husband is clearly responsible for its breaking as he will have succeeded at what he was trying to do. Kane is arguing that if the husband can be responsible for this causally indeterminate breaking of the table then so can persons involved in causally indeterminate decisions made in the context of SFAs. Thus, even if it was causally undetermined whether the businesswoman would stop and prevent the assault, since she willed doing so and this is ultimately what she did, she is responsible for doing so. Furthermore, had she gone on to her business meeting, she would have been responsible for that choice as well, since she simultaneously also willed to do that in the moments leading up to choice.4

Neil Levy and the Problem of Dual Efforts Canceling Out Responsibility Neil Levy (2005) points out that Kane’s theory is typically criticized because it cannot explain why an agent does one act A as opposed to another act B when involved in a dilemmatic choice that is supposed to involve free will. Absent such a contrastive explanation the choice made seems fundamentally arbitrary or random so that what the agent does cannot properly be understood as an exertion of free will. Levy contends that to avoid this objection Kane states that in the moments leading up to an SFA the agent both tries to choose A and tries to choose B. Since we are responsible for what we try to do and since whatever the agent ends up doing—A or B—he will be doing what he tried to do, he will be in control over his action and thereby act freely and with responsibility. In response to Kane, Levy maintains that it will ultimately be a matter of luck whether the agent As or Bs. Thus, all that the agent is responsible for here is whatever he tries to do. Levy then has us suppose that A is a moral act and B is an immoral act. He notes that if the agent tries to do both, then he is equally responsible for both efforts. But, if the agent is equally responsible for trying to do both the moral and the immoral acts, then whatever he does in this situation his responsibility cancels out—he is deserving of neither praise nor blame for what he does. This is problematic because acting in these situations calls for SFAs which are to be the paradigm cases of free and responsible action, and action in these situations provides the foundation for our responsibility in our deterministically caused actions at later stages of life. Thus, Levy views Kane’s theory as fundamentally flawed.

Kanean Libertarianism Examined

61

Kane has never published a reply to this criticism. In defending Kane, I think we need to notice that there is a presumption in Levy’s thinking which Kane would reject. Why does Levy think Kane places responsibility for SFAs solely on our responsibility for the efforts leading up to choice? I suspect he thinks this because he thinks that if there is not a contrastive explanation for why the agent chooses A instead of B, then the agent does not have control over the actual choice made but only over the efforts which occur antecedent to choice. Thus, he thinks that on Kane’s view our responsibility for the choice is completely derivative upon the control and responsibility we have over the efforts. However, Kane would reject this picture of things. We need to look at the fact that in our SFAs our dual efforts are extended over time throughout the deliberation process and on up through choice. When the agent makes an effort to do moral act, A, and a simultaneous effort to do immoral act, B, she continues to exert those efforts on up through the act of choosing. Thus, in making the choice of A, the agent does through her effort make that choice. And, while the choice is causally undetermined and, thus, she could have made the other choice, she nonetheless is responsible not just for the effort to A but for choosing A as well. It is not as though the agent makes an effort to A and an effort to B and then just sits back and waits to see which effort will win out in making a choice, as though the choice is not within her control. On this picture of things it would make sense to say her responsibility does cancel out and she is not responsible for her choice. But, according to Kane, the reality is that the efforts are sustained by the agent up to and through the choice and the choice is a direct consequence of the efforts made. Thus, if the agent chooses A instead of B, she has through her effort made the reasons for A win out over the reasons for B and in doing so she endorses those reasons, and she is responsible for choosing A and not just for each of her efforts. As Kane states, when an agent commits an SFA, she endorses one set of competing reasons over the other as the one she will act on. But willing what you do in this way, and doing it for reasons that you endorse, are conditions usually required to say something is done “on purpose,” rather than accidentally, capriciously, or merely by chance. Moreover, these conditions taken together (that the choices were willed either way, were done for reasons and the agents endorsed them) rule out each of the reasons we have for saying that agents act, but do not have control over their actions. . . . Of course, for undetermined SFAs, agents do not control or determine which choice outcome will occur before it occurs. But it does not follow, because one does not control or determine which of a set of outcomes is going to occur before it occurs, that one does not control or determine which of them occurs, when it occurs. When the above conditions for SFAs are satisfied, agents exercise control over their future lives then and there by deciding. (Kane 2007a, 29–30)

62

Kanean Libertarianism Examined

According to Kane, the mere fact that we cannot provide a contrastive explanation for why in an SFA the agent chooses A instead of B does not mean that the agent has no control over choosing A instead of B. It only means the agent has no antecedent control over this. But this still allows the agent to have control over this at the moment of choice. Kane also argues that because of the dual efforts of the agent, the agent actually has a kind of control over his choice, plural voluntary control, which is richer than what any compatibilist theory can account for. At the moment of choice, the agent has the power of choosing A instead of B or B instead of A, and this is not a power which compatibilists can account for. Furthermore, whichever choice is made the agent will have control over it because whichever choice is made it will be a result of his effort. Thus, in SFAs we are not solely in control of and solely responsible for each of our efforts leading up to choice, we are also in control of and responsible for the particular choice we make, for instance the choice of A instead of B.

Meghan Griffith’s Reply to Kane: Is the Analogy Apt? Meghan Griffith argues that event-causal libertarian theories do indeed suffer from a problem of luck. In making her case for this she focuses on the views of Kane in particular, and she attacks the arguments from analogy he uses to support his view. Regarding SFAs, she writes: While the agent can be said to cause the “action”, she seems not to have control over the crucial element for which she is responsible; that she has decided to A rather than B. This can be compared to the husband. Although the husband causes the table’s breaking, he does not completely control whether the table breaks. In this sense, its breaking happens to him. He is held responsible for this nonetheless, but that is because he intentionally performed an action (bringing his arm down upon it) that he knew could lead to its breaking. The decision case is different. Although free will theorists often claim that responsibility for a current decision can be traced back to prior decisions, there must be decisions that stop an infinite regress. In fact, it is the regress stopping decisions that must not be determined, according to the event-causal libertarian. (Griffith 2010, 50) Kane has argued that causally undetermined decisions made in the context of SFAs are decisions for which we are responsible because they are sufficiently akin to the situation of the husband who is responsible for his causally indeterminate breaking of the table. But Griffith is arguing the analogy here is not strong enough to support the conclusion. She thinks there are significant differences between the cases. One unfortunate aspect of Griffith’s critique of the argument is that it is not entirely clear what she thinks the significant differences between the

Kanean Libertarianism Examined

63

cases are. In what follows, I will suggest a couple of interpretations of her point and argue that either way her argument is problematic. Maybe Griffith’s point is that in the table-breaking case the man can be responsible for his causally indeterminate breaking of the table because he was in clear control of his striking of the table and he clearly knew (or should have known) that this might break the table. But in the context of a Kanean event-causal free decision between A or B and where the agent chooses A it’s not clear that the agent is in control of doing anything in the moments leading up to the decision which makes her responsible for it. Thus, Kane’s argument from analogy does not show how agent control and/ or responsibility are present in Kanean event-causal free decisions. In defense of Kane, it can be said that in the moments prior to the choice of A the agent wills the doing of A and she wills the doing of B. We are in control of what we will to do. Our willings don’t just happen to us; we will things intentionally and with a purpose. Thus, in the moments leading up to the choice of A the agent was in control of willing A. Her simultaneous willing of B makes her ultimate choice of A causally indeterminate, but when she chooses A she is still responsible for the choice because she was in control of willing A and had she not willed A she would not have chosen A. In her defense, Griffith might contend that I have not really captured the point of her criticism. She might say that in Kanean free choice the agent decides to A rather than to B. She might say there’s nothing about the agent which explains why she decides to A rather than B. On the Kanean view the agent wills the doing of each, causal indeterminacy ensues, and a decision occurs; but there’s neither an explanation for nor a reason why A is chosen over B. In this sense, the choice of A over B just happens to the agent and is not controlled by her. Griffith might also add that if the agent does not control the decision to A instead of B then the agent cannot be responsible for this decision. In contrast, in the case of the husband’s breaking of the table there’s no indeterminacy in the choice to strike the table. Rather, in the latter case he simply formulates the intention to break the table and strikes the table with his arm. The indeterminacy lies not in the making of a choice of doing one thing instead of another, but in the attempt to break the table.  This is why in the husband’s case we can plainly see that he is responsible for breaking the table. Since there is no indeterminacy in his choice to strike the table, he has clearly chosen to do so and, thus, he is in turn responsible for its breaking. Perhaps, we are now closer to the point that Griffith is trying to make, but I believe that even on this interpretation of her argument Kane can still defend the view that his event-causal theory of free choice can make sense of agent responsibility. In fact, I think that Kane would concede that on his view agent control is in some sense diminished in the free choices made in the contexts of SFAs, but that, despite this, agents involved in these choices are still fully responsible for the choices they make. Before considering how

64

Kanean Libertarianism Examined

responsibility is preserved and accounted for on the Kanean view, consider the following passages taken from one of Kane’s writings on the subject: But can’t we say that it is a “matter of chance” whether one of these efforts leading to SFAs succeeds or not? For isn’t it true that whether or not an effort succeeds in producing a choice depends on whether certain undetermined neurons involved in the agent’s cognitive processing fire or do not fire (perhaps within a given time frame)? And whether these neurons fire or not is by hypothesis undetermined, is it not, and therefore not under the control of the agent? Well, yes, we can say all of these things: whether an effort succeeds does depend upon whether certain undetermined neurons fire or not; and whether these neurons fire is not under the control of the agent; and we can consequently say it is a matter of chance whether the efforts leading to SFAs succeed or not. But the really astonishing thing is that, while all these things can be truly said, it does not follow that the agent is not responsible for the choice, if the effort succeeds. (Kane 2007a, 37) Here Kane begins to acknowledge that in Kanean event-causal undetermined free choices there is a kind of diminished control over the choice but no loss of responsibility. The point is developed further in the next passage. But does not the presence of indeterminism or chance at least diminish the control persons have over their choices or actions? . . . The answer is yes, again. But the further surprising point worth noting—a point that I think is so often missed—is that diminished control in such circumstances does not entail diminished responsibility when the agents succeed in doing what they are trying to do. (Kane 2007a, 38) Kane would say that when the businesswoman from our earlier example is torn between the desire to (A) stop and prevent the assault and the desire to (B) go on to her business meeting and when she wills to do both of these things, this makes her choice of doing one or the other causally undetermined. Supposing again that she chooses to do A, Kane’s point is that the indeterminacy involved in this choice does, indeed, reduce her control over the choice but it does not reduce her responsibility for stopping to prevent the assault. As I’ve noted already, Kane believes that in SFAs the causally indeterminate nature of the choices we make entails that we lack antecedent determining control over the decisions we make in these situations. Suppose our businesswoman decides to stop and prevent the assault instead of going on to her meeting. Well, because her decision is causally undetermined, she does not make this happen until the moment of choice. At the time of her choice she has the power to choose either option and she exercises this power at the time of choice; in doing so, she exerts plural voluntary control over it.

Kanean Libertarianism Examined

65

But prior to that she did not exert control over the outcome of her deliberation; that could not be done due to the causally undetermined nature of her choice. While there is some reduction of agent control over our SFAs (a loss of antecedent determining control), there is still sufficient agential control over them (plural voluntary control) to make us fully responsible for such choices. After all, as I have also noted, such plural voluntary control is a rich kind of control that cannot be captured on a compatibilist view of free choice. For on the compatibilist view of things, one could choose otherwise only if the past or the laws of nature had been different. But on Kane’s view in SFAs the agent is understood as having the power to choose otherwise given that her character and motivational states and the past and the laws of nature are what they are at the moment of her choice. Furthermore, we are responsible for such causally undetermined choices, because regardless of which choice we make it will not be a random happening, rather it will be an undetermined causal consequence of one of the agent’s efforts of will. As such, regardless of which choice is made, the agent will act with reasons and on purpose without being forced or coerced and not by accident or mistake. Despite the Kanean points noted above, Griffith would likely argue that the Kanean position I am defending is still problematic. In talking about decisions made in the context of SFAs, she says: While the agent can be said to cause the “action”, she seems not to have control over the crucial element for which she is responsible: that she has decided to A rather than to B. (Griffith 2010, 50, my italics) She thinks that in the context of an SFA if the agent does A instead of B, then it is “crucial” that the agent have control over the contrastive fact that A is chosen instead of or rather than B. Later she states: It is already granted that the event-caused libertarian decision is caused by the agent. The problem is that the agent is not able to control which decision is made. The extent to which the agent is causally involved leaves this open. So the agent in such a picture does disappear to an important extent. And this problem is not constituted by the unavailability of an explanation. It is constituted by a lack of control. It just happens to the agent that A occurs instead of B, since the agent has no involvement beyond those states or events that leave it causally open. So in the end, the event-causal libertarian cannot give the agent the control she needs. (Griffith 2010, 51) What concerns Griffith here is that, as she sees it, on Kane’s view in an SFA decision between A and B the agent exerts control over each of her efforts, but, then, since the choice she makes is undetermined, the agent lacks control of the particular choice she makes, say the choice of A instead of B.

66

Kanean Libertarianism Examined

Notice how she says, “It just happens to the agent that A occurs instead of B, since the agent has no involvement beyond those states or events that leave it causally open.” However, this view of things is confused. We have to remember that on Kane’s view it is not as though in the moments leading up to choice the agent exerts an effort to A and an effort to B and then just sits back and lets chance take over. Rather, in SFAs those efforts are exerted up to and through the act of choice and whichever choice is made it is a result of one of these efforts. Additionally, it is the agent who makes the choice as a consequence of one of her efforts and, again, in doing so she acts with reasons, on purpose, intentionally, and not by accident or mistake. As such, on Kane’s view it is not true that the agent has no involvement beyond those states and events which leave her decision causally open. The agent is on the scene and involved throughout the deliberation process and up through the moment of choice. Further, notice that the choice of A instead of B could not happen without the agent’s continued effort to choose A all the way up to the moment of choice. For if, at any moment, she stopped trying to choose A, then her strong desire for B would have led her to make that choice instead. For all of the above stated reasons, Griffith’s critique of Kane’s eventcausal libertarian theory is problematic. She does not exhibit any weakness in his argument from analogy to the causally indeterminate breaking of a table top and she overlooks the ways in which Kane’s view can account for a sufficient level of agential involvement and agent control over the choices they make in SFAs.

The Clarke/Mele Objection: How Are We Responsible for the Dual Efforts? Randolph Clarke (2002) and Alfred Mele (2006) raise a different kind of objection to Kane’s arguments from analogy. Their objection views the arguments from analogy, such as the one discussed in the previous section, as crucial to Kane’s defense of his position, and it seeks to show that such arguments from analogy are flawed. In particular, Clarke and Mele contend that we are only willing to accept that the husband is responsible for breaking the table because we presume that he freely willed to break it. Thus, if the analogy is to work, we must also presume that the businesswoman freely willed to prevent the assault. The problem here, as Mele notes, is that Kane does not claim that in cases of dual efforts to choose, the choices made are products of freely made efforts. Nor has he put himself in a position to claim this, for he has not offered an account of what it is for an effort to choose to A to be freely made. (Mele 2006, 52)

Kanean Libertarianism Examined

67

What Mele is saying is that Kane has told us the businesswoman is responsible for stopping the assault because (a) she freely chose to do this as opposed to going on to her business meeting and (b) despite the causal indeterminacy involved in her decision, she is responsible because she willed the stopping of the assault. But, for the table-breaking analogy to support these claims about the businesswoman’s responsibility, she must have freely willed to prevent the assault prior to choosing to do it. Kane does not offer up any account of how the dual willings involved in SFAs can be understood as free. Thus, we have no reason to think his argument from analogy works. Mele says Kane has suggested that in the context of SFAs as long as the agent’s willing/effort to choose A satisfies a good compatibilist set of conditions for free action then it should be regarded as free (Mele 2006, 53). Thus, as long as her willings of (A) preventing the assault and (B) going on to her meeting satisfy good compatiblist criteria of free action, then each of those willings/efforts should be regarded as free. For instance, we might say that, as long as her willings of (A) preventing the assault and (B) going to her meeting are in accord with her wishes and not the product of external force or coercion nor internal compulsion, then they are freely willed.5 However, Mele sees this as presenting a new problem for Kane’s view. He reminds us that Kane’s view is libertarian. That is, Kane thinks human freedom and responsibility require that we engage in some causally undetermined free actions. The SFAs are to be the causally undetermined free acts that establish the freedom of later causally determined actions. But, as Mele notes, Kane appeals to compatibilist criteria of freedom in accounting for how we can be free in and responsible for our SFAs. On the Kanean view, any human being’s first free actions should be SFAs, as they provide the basis for all other human freedom and responsibility. However, we are now told that to be responsible for SFAs agents must freely engage in the dual willings leading up to choice and that meeting some good compatibilist criteria of freedom suffice for making these dual willings free. The suggestion that meeting such compatibilist criteria suffices for the freedom of the dual willings appears to be inconsistent with the spirit of Kane’s overall theory. Furthermore, Mele states, Now, if a compatibilist account of the freedom of these actions is true, the actions may be free even if they are deterministically caused. And, if we believe that that is true of these actions, why should we not believe that it is true of all free actions that do not depend for their freedom on earlier indeterministically caused free actions? (Mele 2006, 53) This latter point suggests that if meeting compatibilist criteria suffices for the freedom of the dual willings/efforts involved in SFAs, then it becomes

68

Kanean Libertarianism Examined

unclear why causal indeterminism is required for human freedom and responsibility. Given the way Kane has expressed his theory in his own writings, his view would seem to be that the first free acts of any agent should be causally undetermined SFAs. But the criticisms of Clarke and Mele have pushed him to say that prior to the performance of these SFAs the dual willings/efforts of the agent must be freely made, and Kane has suggested that such willings may be free merely by satisfying good compatibilist criteria. As Mele notes, the problem here is that if the dual efforts/willings in SFAs are free by meeting good compatibilist criteria, then it is no longer clear why we need to meet any libertarian criteria at all to be free. The Clarke/Mele objection seems unanswerable, but it isn’t. In what follows, I will argue that Kane’s appeal to compatibilist criteria in supporting his contention that we may be fully responsible for causally undetermined acts of choice is neither inconsistent with his overall libertarian perspective nor does it make it unclear why meeting any libertarian criteria is needed to be free. Let’s go back to the example of the businesswoman. The Clarke/Mele objection is that if we are to regard her as responsible for her causally undetermined decision to prevent the assault, then we must presume that she freely willed/tried to prevent the assault. Recall this kind of presumption is what makes the assassin analogy work. As Mele notes, Kane has argued that for her to be responsible she must have freely willed the prevention of the assault prior to the undetermined decision to prevent it, and he seems to think that meeting a good set of compatibilist criteria suffices for the freedom of this willing. As noted in the preceding section, Mele then pounces on this Kanean appeal to compatibilist criteria, arguing that it is inconsistent with Kane’s libertarianism and that it creates the worry that we no longer have any reason to believe libertarian criteria are even needed to make sense of human freedom and responsibility. However, I think it is important to consider what the appeal to compatibilist criteria does for Kane’s view depending on whether we are talking about the SFAs of an adult, like our businesswoman, or the first SFAs of a young child. In the case of the businesswoman, she has a developed character which presumably has been shaped by many SFAs performed in the past. On Kane’s view these past SFAs are causally undetermined decisions that are the basis for her having responsibility for the character she has as an adult. Consequently, when she wills/tries to prevent the assault and this willing/ effort issues forth from her character in a way that is fully consistent with good compatibilist criteria, she does so freely. But notice she does so freely not simply because compatibilist criteria are met, but also because the character she has, which causes the willing/effort, is a product of past free, undetermined SFAs. The dual willings/efforts involved in adult SFAs may be regarded as free by meeting good compatibilist criteria but they also have a

Kanean Libertarianism Examined

69

derivative libertarian freedom because they issue forth from characters that have been shaped by prior undetermined free SFAs. In the case of the first SFAs of childhood, things are quite different. As Kane has said in various writings, children have very little responsibility for their earliest SFAs.6 This is because in our earliest SFAs only the compatibilist criteria of freedom are met in the formation of the dual willings/efforts involved in them. In these early SFAs there is no past history of SFAs allowing for the agent herself to contribute to the construction of her character. Instead, in the earliest SFAs, the desires or reasons which motivate the dual willings/efforts involved are only products of genetic and environmental factors. Given this fact about the source of the dual willings/efforts in the earliest SFAs, the agent’s contribution to what is done is so minimal that there is very little responsibility on the part of the child for what is done.7 Mele argues that if the freedom of the dual efforts involved in SFAs and our responsibility for these dual efforts is explained using compatibilist criteria, then we have reason to wonder why compatibilist criteria should not suffice for explaining the nature of freedom and responsibility in general. Here I draw attention to the difference between our responsibility for childhood and adult SFAs to show that on the Kanean view meeting good compatibilist criteria in the formation of dual willings/efforts does not suffice for making the agent fully responsible for them. What makes the adult fully responsible for her dual willings/efforts is the fact that they meet compatibilist criteria and they issue from a character formed by prior causally undetermined SFAs. Thus, while meeting compatibilist criteria of freedom is part of the story about why adults are fully responsible for their SFAs, it is not the whole story. The causal indeterminacy of past SFAs in shaping character plays a key element as well. Thus, and again contra the Clarke/Mele objection, even though Kane sees an important role for compatibilist criteria to play in making sense of responsibility, the appeal to libertarian free choice is still essential to any adequate account. In response to the preceding considerations, critics may grant that there is very little responsibility for the earliest SFAs and they may grant that full responsibility for the dual willings/efforts in adult SFAs is not established merely by meeting a good set of compatibilist criteria. However, they may note that at some point in the life of agents they transform from possessing very little responsibility for SFAs in early childhood to having full responsibility for SFAs in adulthood. The charge could be made here that Kane has not done enough to explain how this happens. Without an adequate account of this, it might be argued that we have no reason to take the Kanean view seriously. In what follows, I will address this concern in basically Kanean terms. Let’s consider an example. Imagine Rosa is quite young—three or four years old. She has been told not to eat from the cookie jar unless she has her parents’ permission. She has already had her daily allotment of cookies, but now she wants more. She sees that she might be able to sneak a cookie from

70

Kanean Libertarianism Examined

the jar while no one is looking, but she fears she may be caught and reprimanded. She’s torn, and she wonders what she should do. Suppose this is the occasion of her first SFA. If so, then she tries/wills to (A) sneak a cookie and to (B) obey her parents. Let’s suppose she ultimately caves to temptation and sneaks a cookie from the jar. If so, and assuming this is an SFA, then the choice to do so is causally undetermined. On Kane’s view, even if each of the dual willings/efforts meet compatibilist criteria of freedom, she has very little responsibility for her choice, because she has not formed the character from which her dual efforts arise; rather, the desires or reasons which motivate these efforts are at this point in her life too much a product of genetic and environmental factors. She has not yet performed enough of the free undetermined SFAs to give her much responsibility for her choice. Here it may be wondered how Rosa can have any responsibility for her choice. After all, if the decision is undetermined right up to the moment of choice and if she’s not responsible for the character from which her dual efforts arise, then what is the source of her limited responsibility? In the preceding paragraph it is said that she has some minimal responsibility for her choice to sneak a cookie, but what is the source of this responsibility? In response Kane will say that while her decision to sneak a cookie is undetermined and while she’s not responsible for the character from which her dual efforts arise, Rosa herself—the agent—still plays a causal role in determining whether the decision to sneak a cookie occurs. It is fundamentally important to note that if she did not continue to exert the effort to sneak a cookie while simultaneously trying to obey her parents then the sneaking of the cookie would never occur. Rather, without that effort she makes to sneak the cookie, she would not do so and would, instead, opt to obey her parents. In this way, Rosa, through her own effort, plays a significant causal role in making the choice to sneak a cookie. Indeed, the choice is undetermined but she nonetheless plays a significant causal role in the choice to sneak a cookie, for had she not exerted the effort to sneak a cookie then the choice to do so would never be made.8 Here it might be wondered how Rosa’s effort to sneak a cookie can play any role in making her responsible for the choice to sneak one. After all, she is not responsible for the character from which this effort arises—her character at this early stage of life is too much a product of genetic and environmental factors beyond her control and she does not have a history of SFAs upon which to ground responsibility for her character. In response to this concern, the Kanean should note that this is where meeting the compatibilist criteria of freedom is important. Given that Rosa exerts effort in accord with compatibilist criteria of freedom, there is a sense in which the efforts made in her SFA are hers as opposed to alien external forces, such as would be the case were she acting due to physical force, or threat, or as a consequence of external mind control. As Kane has said in his own writings, even in a deterministic world there would be a kind of freedom and responsibility corresponding to the kinds of conditions outlined

Kanean Libertarianism Examined

71

by compatibilists. But, he also believes such compatibilist freedom is inadequate to ground ultimate responsibility.9 Thus, in exerting the effort to sneak a cookie and then making the causally undetermined decision to do so, Rosa bears some level of responsibility. Since she meets compatibilist criteria of freedom in exerting the effort, it is established that she and not external factors controls the effort. At the same time, given that at this early stage of life she is all too much a product of genetics and environment as opposed to her own causally undetermined SFAs, the responsibility she has for her effort and choice is minimal. So, here in our first SFAs we make undetermined choices between competing desires that arise in us not from a character shaped through our own choices but from a character shaped through genetics and environment. However, we are nonetheless responsible for these choices because it is through our own sustained efforts that these choices get made. As noted above, had Rosa not sustained the effort to sneak a cookie she would not have then made the choice to do so. Thus, while her choice is undetermined she has some control over its occurrence and she is, therefore, to some extent responsible for the choice. It should also be added that the notion of endorsement can be invoked here to clarify another sense in which the child is responsible, albeit minimally so, for this first SFA. It should be noted that by making the decision to sneak a cookie and acting on this decision, Rosa then and there endorses her desire to sneak the cookies over her competing desire to obey her parents; and she is not determined to do so. She did not determine her prior character in this case, but she did determine which aspect of her pre-formed character would win out and govern her subsequent behavior. By making an undetermined decision and acting on it, the child endorses a certain desire— a certain aspect of her character, if you will—and takes on responsibility for the action which emanates from it.10 The choices made in SFAs contribute to shaping one’s character over time. Thus, as she makes more of these undetermined choices over time she will contribute more and more to the shaping of her own character. This will increase her level of responsibility for later SFAs as those later SFAs will become more and more a reflection of a character that she has shaped herself through prior undetermined decisions which she was at least partly responsible for making through her own efforts. While children have some responsibility for their earliest SFAs, they have very little responsibility due to the fact that the dilemmatic choices young children face are thrust upon them by genetic and environmental factors which are beyond their control. As a child grows older, she becomes more responsible for her character because she has through her previous free choices done more to shape the character which gives rise to the dilemmas she faces in her SFAs. There is no doubt that our character plays a significant role in shaping what are dilemmatic situations for us. Going back to our example of the businesswoman, imagine that she was a person of very high

72

Kanean Libertarianism Examined

moral character with very little ambition for her career. Were this her character, the dilemmatic choice to (A) prevent the assault or (B) go on to her meeting would not arise. She wouldn’t feel torn between the two; instead, she’d stop and prevent the assault without thinking about it. As it is, she does feel torn between the two options, and this is a consequence of her character. In addition, my comments above suggest the responsibility she bears for her decision in this scenario stems not merely from the causally undetermined free choice she makes in this situation but also from the fact that she has through many of her past free decisions come to have the character she has now which makes the situation a dilemma for her. In contrast, little Rosa chooses freely but any responsibility she has for her choice to sneak a cookie is grounded solely in the effort she makes in the face of the temptation to do otherwise—an effort which is rightly said to be hers as she meets compatibilist criteria of free action in exerting that effort. Before concluding let me note that on the view I am defending here incompatibilist, libertarian freedom and control presupposes compatibilist notions of freedom and control, since compatibilist “absence of constraint” conditions (no coercion, compulsion, incapacity, etc.) must also be satisfied for incompatibilist freedom and control. What gives Rosa minimal responsibility for her first SFAs is the fact that she exerts the efforts in these SFAs; and what establishes the fact that she exerts the efforts is the fact that in doing so she meets compatibilist “absence of constraint” conditions. Meeting such compatibilist conditions of responsibility is fundamental to becoming a libertarian free agent bearing ultimate responsibility for actions done in adulthood. Kane himself has acknowledged this point in his own writings, stating, “incompatibilist freedom and control presupposes compatibilist freedom and control. We cannot get to incompatibilist freedom and control in one fell swoop in the real world. That is one leap too far” (Kane 2011, n.14, 404). Let’s take stock of where we are and how we got here. Kane has used arguments from analogy to show how one can be responsible for an SFA even though it is causally undetermined. Mele and Clarke respond that for this argument to work we must assume that dual efforts in SFAs are freely made and they contend that Kane needs an account of how they can be freely made. In reply, Kane has suggested they are free as long as they satisfy some good compatibilist criteria of freedom. In turn, Mele has responded that this suggests our first free acts are not undetermined SFAs, but the freely formed dual efforts involved in our first SFAs. According to Mele, this is problematic as it seems inconsistent with the overall libertarian nature of Kane’s view and it makes it unclear why causally undetermined SFAs are needed for responsibility and freedom. What I have tried to show here is that on the Kanean view there is indeed a sense in which our first free acts are not SFAs. Prior to performing her first SFA, Rosa must in some sense freely will to sneak a cookie and freely will to obey her parents. According to Kane, she may do this freely in some

Kanean Libertarianism Examined

73

compatibilist way by trying, for instance, to do them as a consequence of her own desires and not as a consequence of external force, threat, or coercion. But to acknowledge this is not inconsistent with the overall libertarian nature of Kane’s view. For to get from the minimal compatibilist responsibility Rosa has for these dual efforts/willings to the full libertarian responsibility she has for her adult SFAs she will have to engage in some causally undetermined decision-making for which she is responsible. As I have argued here she does this in her first SFA. Since it is she that exerts and sustains the effort to sneak a cookie and since this plays a significant role in her undetermined choice to sneak a cookie, this undetermined choice is something for which she is at least minimally responsible. Further, were it not for these undetermined choices we make at earlier stages of our lives we could not then have ultimate responsibility for our character; that is, without these undetermined choices our characters would then be all too much a product of genetic and environmental factors and not authored by ourselves. Further, since ultimate responsibility for our own characters is vital to our having full responsibility for later actions and decisions made in adult life, it follows that our performance of causally undetermined SFAs remains vital to an adequate account of fully responsible adult agency.

Ekstrom, the Phenomenological Problem, and Unconscious Efforts Some critics (Ekstrom 2003, 163–164) contend that Kane’s account of the nature of SFAs just doesn’t fit with the phenomenology of dilemmatic decisions. Critics note that when faced with a difficult choice between doing two things that we want to do, we don’t experience ourselves as trying or willing to do both. Rather, we experience ourselves as simply wanting both and then we decide which want we are going to satisfy. In Kane (2007a) he responds to this kind of objection. He admits that in SFAs we are not introspectively aware of the dual willing of two incompatible acts. He argues, however, that our lack of introspective awareness of such gives us no reason to think that we aren’t in fact involved in such dual willings. He states: I am not claiming that agents are conscious of making dual efforts. What they are introspectively conscious of is that they are trying to decide about which of two options to choose and that either choice is a difficult one because there are resistant motives pulling them in different directions that will have to be overcome, whichever choice is made. In such introspective conditions, I am theorizing that what is actually going on underneath is a kind of parallel processing in the brain that involves separate efforts or endeavorings to resolve competing cognitive tasks. The point is that introspective evidence does not gives us the whole story . . .

74

Kanean Libertarianism Examined It is now widely believed, for example, that parallel processing takes place in the brain in such cognitive phenomena as visual perception. The theory is that the brain separately processes different features of the visual scene, such as object and background, through distinguishable and parallel, though interacting, neural pathways or streams. Suppose someone objected that we are not introspectively aware of such distributed processing in ordinary cases of perception. That would hardly be a decisive objection to this new theory of vision. For the claim is that this is what we are doing in visual perception, not necessarily that we are introspectively aware of doing it. And I am making a similar claim about free will. (Kane 2007a, 34)

In a more recent publication, Kane acknowledges that some might find this argument from analogy to visual perception to be problematic. He writes: It might be argued that this analogy fails because the distributive processes involved in perception are not efforts, whereas, on the view proposed, the distributive processes involved in deliberation would be efforts; and we are normally consciously aware of making efforts when we make them. But such an argument would undermine the analogy only if it were assumed to entail that all of our effort-making must be conscious, that effort-making can never occur unconsciously. And that this is the case cannot be settled by introspective evidence in which we are consciously aware of making efforts. In complex cognitive processes such as theoretical and practical reasoning or deliberation, much of the processing that is going on, it is now widely held, occurs unconsciously and some of what occurs unconsciously may involve effort-making, e.g. efforts to access memories, associations or considerations that may have a bearing on a decision, efforts to overcome temptations to suppress other information we may not want to think about, efforts to resist strong inclinations, or to avoid rationalizations or self-deception (any of which may fail), and so on. (Kane 2016a) Here Kane acknowledges that we are typically aware of making efforts when we make them. But he goes on to note that it does not follow from this that we never unconsciously make efforts. He goes on to talk about and cite recent empirical psychological research in which it is demonstrated that much of the mental processing going on in theoretical and practical reasoning does occur unconsciously. In a prior publication (2011b), I myself have defended the phenomenological objection against Kane’s view. In doing so, I argued that the problem the phenomenological objection raises concerns the fact that it drives Kane to the admission that the dual efforts involved in our SFAs are unconscious.

Kanean Libertarianism Examined

75

My own view is that Kane is right when he says it is entirely possible that we engage in unconscious dual efforts of will. As he notes, there is a lot of cognitive processing which we engage in that goes on unconsciously. Further, I agree with him that just because we do not experience ourselves as engaged in dual efforts when making the kinds of difficult decisions we make in SFAs does not mean that we are not engaging in them. However, as I see it the real problem raised by the phenomenological objection is that it gets us to see that if we engage in these dual efforts in SFAs it must be that we do so unconsciously, and this I think is a problem—a problem which Kane has yet to address in his writings. The problem here is that to be responsible for unconscious actions these unconscious actions must issue from a character or motivational states which we have formed through prior conscious free choices we’ve made and for which we were responsible. For instance, consider someone, call him “Jim the Lech.” Now suppose Jim is an older businessman who frequently irritates and makes uncomfortable the women he works with in his office. What he does is he unconsciously ends up staring at their breasts and behinds, and this makes his female coworkers uncomfortable. If Jim is doing this unconsciously, he can nonetheless be responsible for his conduct if he has made conscious free choices in the past to do certain things, like watching pornography or ignoring the polite requests of others to not stare, which he should have known or did know might contribute to his being the lecherous man he is today. Additionally, if he has been told by his coworkers or his boss that he needs to be mindful of his stares, then he also may take on responsibility for his conduct. However, if he has made no conscious free decisions in the past which have led him to be the lecherous man he is now and if no one has told him to stop with his uncomfortable staring, then he cannot rightly be held responsible for these unconscious stares. The problem these thoughts present is that SFAs are to be the free acts we perform which are to make us responsible for other actions we perform which are determined by our character or motivational states. Furthermore, as Kane himself acknowledges, to be responsible for our SFAs we must be responsible in at least some compatibilist sense for each of the dual efforts involved in our SFAs. Well, if I am right that responsibility for unconscious actions requires that those actions proceed from a character formed by prior conscious free decisions for which we were responsible, then we have a problem here. Think about a child’s first SFA. How is she to be responsible for the efforts involved in them if they are unconscious efforts? To be responsible for an unconscious effort it must issue from a character formed through prior conscious free decisions for which the agent was responsible. But in the first SFA this is not possible. So, it is hard to see how we will ever build up responsibility for our character through commission of SFAs if the efforts involved in them are unconscious efforts. Before moving on, it is only fair to consider some possible Kanean replies to the concerns I raise here. First, Kane might argue that as long as the

76

Kanean Libertarianism Examined

efforts in our first SFAs (or any other SFAs, for that matter) meet compatibilist requirements of free action then we are at least minimally responsible for those efforts. So, for instance, if in a first SFA where an agent tries to decide between doing A or B and in trying to A and trying to B she is unforced, uncoerced, unmanipulated, then she acts freely in making these efforts and is at least minimally responsible for them in some compatibilist sense whether she is conscious of them or not. Such a response would be problematic, however. It’s too hard to see how the agent in her first SFA could even be minimally responsible for such efforts if they are unconscious. In the first SFA, there is no backlog of SFAs present to make one ultimately responsible (UR) for each of one’s efforts. Rather, in such a situation each of one’s efforts is a determined consequence of genetic and environmental factors. Furthermore, if they are determined in this way and they occur unconsciously, then it’s unclear that the agent has the kind of control over them that even a compatibilist would require for free and responsible action. On this picture, in the first SFA each of the efforts is a determined consequence of genetic and environmental factors; and unconscious to the agent these efforts occur and one of them does in some undetermined way dictate her choice. Here the agent will no doubt be consciously aware of making a choice, but the choice is so much a product of (a) factors beyond her control and (b) efforts made unconsciously that it’s hard to see how the agent could even meet plausible minimal compatibilist standards of responsibility for her choice. Perhaps a better Kanean reply might instead take note of the fact that in our first SFA and subsequent SFAs (1) we are consciously aware of making these choices and (2) we are (or at least should be) aware that the choices we make in these situations lead to the formation of our character. It might be thought that as long as conditions (1) and (2) are met, we can become responsible for the unconscious efforts made in our SFAs over time. Perhaps, our businesswoman doesn’t know that when deliberating she is unconsciously trying to prevent the assault and unconsciously trying to go on to her meeting. But still these efforts issue from a character formed by prior undetermined choices which she consciously made and she knew (or should have known) these choices were shaping her character in such a way as to affect her later choices. In this way, the businesswoman could be thought to be UR for the choice she makes even though the efforts she makes in the moments leading up to her choice occur at an unconscious level. However, we might wonder whether having a backlog of undetermined choices which we’ve consciously made that involved unconscious dual efforts is enough to establish our UR for the decisions we make in later life. Indeed, in my youth I might be told that these conscious decisions I’m making are shaping my character and it may be true that they are. But we have to remember that each SFA is caused in an undetermined way by these efforts and we are to be responsible for our SFAs by being responsible for the efforts we make in the moments leading up to our SFAs. If these efforts

Kanean Libertarianism Examined

77

occur unconsciously, this still seems to be a problem. For merely being consciously aware of making these choices and knowing that they are shaping my character does not suffice for establishing my UR for them. Rather, what is also needed is conscious control over the things that make us responsible for these choices and that is the efforts themselves! If they occur unconsciously, then it’s hard to see how an agent can have the conscious control over them that is necessary to be responsible for her SFAs. For all the above stated reasons, I continue to think Kane’s assertion that the dual efforts of will involved in SFAs are unconscious remains a problem for his theory. The phenomenological objection drives him to assert that the dual efforts are unconscious. But if they are unconscious, then that makes it difficult to see how our SFAs can be the source of our ultimate responsibility.

Ekstrom and Clarke and the Problem of Irrationality Another criticism of Kane’s view notes that to exert effort to do two different actions which cannot both be performed is irrational. Thus, it is said that Kane’s theory describes the paradigm cases of free and responsible action as a kind of irrational action. Since this is not any kind of free will worth wanting, critics urge the rejection of Kane’s theory (Ekstrom 2003, 163–164; Clarke 2003, 88–89). In Kane (2007a) he also responds to this charge, arguing that in the contexts of SFAs it can be rational to exert effort to do two incompatible acts that cannot both be performed. He states: [T]here are special circumstances in which it is not irrational to make competing efforts: These include circumstances in which: (i) we are deliberating between competing options; (ii) we intend to choose one or the other, but cannot choose both; (iii) we have powerful motives for wanting to choose each of the options for different and incommensurable reasons; (iv) there is a consequent resistance in our will to either choice, so that (v) if either choice is to have a chance of being made, effort will have to be made to overcome the temptation to make the other choice; and, most importantly, (vi) we want to give each choice a fighting chance of being made because the motives for each choice are important to us. The motives for each choice define in part what sort of person we are; and we would be taking them lightly if we did not make an effort in their behalf. These conditions are, of course, the conditions of SFAs. (Kane 2007a, 34–35) In Kane (2016a), he makes the same points, but he goes on to talk about how our wills are not settled when we are deliberating in the context of making an SFA. He says that when we are torn inside about what to do, say the moral thing or the self-interested thing, and we care about both and our will is not settled as to what we shall do, it can make sense to exert effort

78

Kanean Libertarianism Examined

to do each of the things we are considering doing so as to give each course of action a “fighting chance” of being realized. In contrast when we are not torn and we know darn well what we want to do, then engaging in dual efforts of will is irrational. In [will-settled situations] it would be irrational to make incompatible efforts because our wills are already set on doing what we are trying to do. There are, in other words, “rationality constraints” on making efforts in will-settled situations because it is irrational to attempt to do contrary things when one’s will is already set on doing one of them. But will-settling situations of the kind that occur in SFAs represent another alternative in which one’s will is not yet set on doing either of the things one is trying to do, but where one has strong reasons for doing each (e.g., deciding to A or deciding to B) and neither set of reasons is as yet decisive. Because most efforts in everyday life are made in will-settled situations in which our wills are already set on doing what we are trying to do, we tend to assimilate all effort-making to such situations. But we thereby fail to consider the uniqueness of will-setting which is of a piece, in my view, with the uniqueness of free will. (Kane 2016a) I find this reply to be problematic in several respects. First, when we are torn inside about what to do and our will is not settled, we have conflicting motives and we are not sure which is more important to us. If we are faced with conflicting motives and we don’t know which we care about more, then the rational thing to do is to try to sort out which motive is more important to us and then act on that one. This is rational on at least two fronts. First, we cannot do both acts so it is futile to exert efforts to do each of the acts. Second, in determining which motive is more important to us before choosing and acting, we can then choose and act in a way that is more consistent with the hierarchy of values that is partly constitutive of our self. It might be that Kane would agree that when you don’t know which motive is more important to you, then it is rational to try to figure out which is stronger before acting. But Kane might say that in SFAs the problem is that the competing motives are either equally strong or we don’t really have the time or the ability to determine which is more important to us. He might contend that these are the circumstances in which SFAs occur, and when these are the circumstances, willing both of two incompatible acts is rational since you care deeply about both and you want to give both courses of action a fighting chance of being realized. Even here the argument is unconvincing. Look at it this way: Either you know the relative strengths of your competing motives or you don’t. If you do know and you value one more than the others, then it is rational to act on that motive; and it is not rational to will both acts. Further, if you know the strengths of your competing motives and you know you value them

Kanean Libertarianism Examined

79

equally, it is still not rational to exert effort to do each of the acts because you cannot do each of them, you can only do one of them. In this rare kind of case, the rational thing to do is to arbitrarily choose which motive to act upon and then will to do that act. If you don’t know the relative strengths of your competing motives, then as I argued above you should try to sort through them and determine which motive is more important to you. Now, as noted, Kane might say that in SFAs you don’t have the time or the ability to do this. But while this may often be the case, it still doesn’t make trying to do each of them rational. I say this because (1) in SFAs you should want to choose the act that is most consistent with your hierarchy of values and (2) in SFAs you also generally don’t know whether you have the time or ability to figure out which motive is stronger. So, when you don’t know which motive is stronger or whether they are equal, the rational thing to do is to engage in a deliberative process in which you try to discern which motive is more important. This is distinct from exerting effort to do each act and it is more rational in its acknowledgment that both acts cannot be done and in its recognition of the importance of willing in a manner that reflects the hierarchy of one’s values. I would also note that when ignorant of the relative importance of one’s motives and doubtful of one’s time or ability to figure out which is more important, the rational thing to do is still to sort through the relative value of one’s motives and then make a decision, leading to the willing of one act or the other. Again, this is more rational than trying (exerting effort) to do each of the acts, because it acknowledges that both acts cannot be done and it shows respect for the importance of acting in accord with the hierarchy of one’s values. In the end, one may have to decide and act before one has sorted it all through and, consequently, one will have to decide from a point of relative ignorance; but one will have still done all that one could to have rendered his act consistent with his values and this is more rational than trying to do each of the acts. In response to my arguments, Kane might say that his view is in some ways consistent with my view. He might say that in deliberating we do and should try to figure out which option we value more highly and then act accordingly. However, he might also note that on his view in doing this we are at an unconscious level exerting efforts to make each of the choices we are considering making. Perhaps his point is that in the deliberative process we are consciously aware of trying to discern which option we value more highly, but this is just how we consciously experience ourselves as acting in these situations. Whereas beneath the surface of conscious awareness there are these dual efforts being made to make each choice and our consciously reflecting on the merits of each option is just how we manifest the dual efforts in our conscious deliberation. Kane might say the dual efforts occurring at the unconscious level are rational insofar as we want to do both options very much and if we do not, say, try to A while simultaneously trying to B, then the desire to do one of them, say, B will just overpower us,

80

Kanean Libertarianism Examined

not giving us the chance to choose A. At the same time, if we don’t exert effort to choose B then the desire to do A will overpower us robbing us of the chance to choose B. In deliberating we experience ourselves not as trying to A and trying to B but as trying to figure out which we value more highly, but, again, this is just our conscious awareness of what is going on in deliberation; there may nonetheless be dual efforts involved at the unconscious level and these may themselves be rational. Bearing these points in mind, we still might wonder whether such unconscious dual efforts would be rational. What is it that makes them seem rational? It is the presumption that the agent desires A and desires B and that without making an effort to do each of them then the desire for one of them will overpower him not giving him the chance to satisfy his other desire. In this way, the dual efforts of will are seen as rational. For by making these dual efforts we keep alive the possibility of making either choice. Whereas if one fails to make an effort for one of these options, the agent runs the risk of having his choice dictated by an overpowering desire for the other option. Notice that here the goal of the efforts is to keep one’s options alive, to give them “a fighting chance”, as Kane says, at being realized in the agent’s choice. But, does this make it rational to try to choose A and to try to choose B in the moments leading up to choice? No, it doesn’t; rather, what it would be rational to do is to merely exert effort to keep alive the possibility of making either choice and this does not require trying to choose A and trying to choose B. To keep both options alive, making it possible to make either choice, all that is necessary is exerting an effort to keep option A alive and exerting another effort to keep option B alive. Making these dual efforts would be rational as (1) it is not inconsistent to keep both of two valued options alive and (2) in exerting them the agent would not be trying to do two different things—to choose A and to choose B—when he knows only one of them can be done. In contrast, engaging in dual efforts to choose A and to choose B, as Kane postulates, is irrational because (a) they involve efforts to do two different things when one knows only one of them can be done and (b) they are misdirected in the sense that the goal is merely to keep each option alive so that each has a chance of being chosen. It might be wondered why Kane insists that the dual efforts be directed at choosing A and choosing B instead of being directed at keeping option A alive and keeping option B alive. Ultimately, he conceives the efforts as directed at choosing A and choosing B, because he thinks we are responsible for undetermined choices when we succeed at what we were antecedently trying to do. Recall that while it was undetermined whether the assassin would kill the prime minister when he tried to do so, when he does kill him he is responsible for doing so because he succeeded at what he was trying to do. In this way, even though the killing is causally undetermined, the assassin is still responsible for it. Now suppose in an SFA I choose A instead of B.

Kanean Libertarianism Examined

81

Well, if I was not trying to choose A instead of B prior to making this choice but only trying to keep option A alive and trying to keep option B alive, then it becomes hard to see how my choice of A is something for which I am responsible. For on this model I am not antecedently trying to choose A, I’m only trying to keep the option/the possibility of choosing A alive. It seems that in an SFA I must be trying to choose A before choosing it to be responsible for the choice of A. This is why Kane envisions the dual efforts involved in SFAs as being directed at the choice of A and the choice B and not merely at keeping option A alive and option B alive. At the same time, for the reasons I’ve given, the dual efforts as he envisions them are irrational. In Chapter Five, I will argue that we can make sense of our responsibility for our causally undetermined SFAs using a dual efforts model wherein those efforts are conceived as directed merely at keeping the choice options alive and not as efforts to choose A and to choose B, as Kane envisions them. One virtue of such an approach is that it allows us to view our SFAs as rational while also making sense of our responsibility for them.

Conclusion As this and the previous chapter shows, a variety of objections have been made against Kane’s theory. It is my view that many of these objections can be answered and I have tried to explain in both this chapter and the previous chapter how they can be answered. However, as the last two sections of this chapter show, I do think some of these objections have force against Kane’s view. Thus, despite his efforts at rebutting the last two kinds of objection, we have good reason to give consideration to a modified kind of eventcausal libertarian theory. In the next chapter, I will develop and defend such a modified view.

Notes 1. Kane has articulated this view of the nature of freedom and responsibility in various publications. See Kane (1996, 1999a, b, 2002a, 2007a, b, 2011, 2016a). 2. For different expressions of the luck objection, see Allen (2005), Clarke (2002), Haji (1999, 2005), Strawson (2000), and Waller (1988). 3. In dealing with the problem of luck, Kane appeals to the theory of dual willings/ efforts in various writings. See, for instance, Kane (1999a, 1999b, 2002a, 2007a, 2011, 2016a). 4. Kane appeals to such arguments from analogy in various writings. See, for instance, Kane (2002a, 418 and 2007a, 27). 5. In Mele (2006), he says it has only been in conversation that Kane has appealed to compatibilist criteria as the ground of the freedom of dual willings/efforts. In Kane (2011), he puts this view in writings. See especially note 14, 403–404. 6. See Kane (2007b, 174–175 and 2011, 399). 7. Kane has made similar remarks on why children bear little responsibility for their first SFAs. Again, see Kane (2007b and 2011). 8. See Kane (2007b, 173–174).

82

Kanean Libertarianism Examined

9. See, for instance, Kane (2011, 382–383). See also Kane (2011, n.14, 404), where he again acknowledges the value of compatibilist freedom and says it plays a significant role in making sense of incompatibilist freedom. He states: incompatibilist freedom and control presuppose compatibilist freedom and control. We cannot get to incompatibilist freedom in one fell swoop in the real world. That is one leap too far. We must get there step wise, by exercising compatibilist guidance control over cognitive processes aimed at making choices. (2011, n.14, 404) 10. I would also note that reflections on the practices of wise parents can be brought in here to provide additional support for this point. The wise parent realizes that it is a mistake to never hold a child responsible for these earliest choices. Offering very mild punishments for these wrong choices offers incentive for the child to endorse the proper motive in future situations. As Aristotle was well aware, responsibility for character is a function not only of the efforts of the individual but also of the way in which she is raised.

5

A Consideration of Alternative Event-causal Libertarian Models of Basic Free Actions

In the previous chapter it was argued that, despite Kane’s efforts to rebut them, some of the criticisms of his views raised by Ekstrom (2003) and Clarke (2003) do present serious challenges to his view. Recall how Ekstrom (2003) argues that insofar as we don’t experience ourselves as engaging in dual efforts in the moments leading up to choice, Kane’s suggestion that dual efforts of will must be involved in our self-forming acts (SFAs) conflicts with the phenomenology of dilemmatic choices. Also, both Ekstrom (2003) and Clarke (2003) argue that Kane’s appeal to dual efforts of will would have us understand SFAs as irrational actions. The idea here is that it is irrational to try to do two different actions, A and B, when you know that you can only do one of them, yet Kane thinks that is exactly what we do in the moments leading up to an SFA. This is problematic as SFAs are to be the underivatively free acts that serve as the foundation for all our other freedom and responsibility and it would be odd if all our freedom and responsibility was based on irrational actions. These criticisms give us some motivation for considering whether a more plausible event-causal libertarian view could be constructed which makes no appeal to dual efforts of will. However, I would also note that even if I am wrong in thinking Kane’s replies to Ekstrom and Clarke are problematic, there is still another good reason to search for an alternative libertarian view that makes no appeal to dual efforts of will. This other reason concerns the fact that we have no evidence that we actually ever engage in dual efforts of will in the moments leading up to our causally undetermined SFAs. It is plain that when we are torn between doing two different actions and we cannot do both, we often want to do both and we often have reasons to do both; but it’s not at all clear that in deciding between two options we ever engage in dual efforts of will in the moments leading up to choice. Because of this, Kane’s appeal to the theory of dual efforts is highly speculative. Thus, even if his view can be reasonably defended against all of the criticisms of it considered in the prior chapter, we should still consider whether event-causal libertarianism can be adequately defended without making an appeal to dual efforts of will.

84 Alternative Models of Basic Free Actions This chapter has two parts. In the first part, I consider some other eventcausal libertarian theories which make no appeal to dual efforts of will in their understanding of basic, underivatively free undetermined actions. Each of the views considered in the first part has problems with them. However, by studying them we learn what will be needed in order for an eventcausal libertarian view to succeed if it makes no appeal to the kinds of dual efforts of will which Kane posits. In the second half of the chapter, I will go on to present two new models for thinking about SFAs which reject Kanestyle dual efforts. One of them is more Kanean than the other, insofar as it embraces dual efforts while understanding them in a significantly different way than Kane does. The other view rejects dual efforts altogether. At present I am unsure as to which of these models is rationally preferable, but I offer both of them as plausible event-causal libertarian models which overcome weaknesses found in other such theories.

Part One A Simple Dual Wantings Model Considered Reflecting on the criticisms raised by Ekstrom and Clarke, one might be led to think that Kane’s appeal to the dual efforts of will is simply an unnecessary extravagance which creates problems for his view. It might be thought that we could get by just talking about the fact that in SFAs the agent wants to do A and wants to do something else B and cannot do both and then makes a causally undetermined choice to do one or the other. It might be thought that in making such a choice as long as (1) the agent is not coerced, manipulated, or somehow forced in making her choice and (2) her choice is a reflection of what she wants to do, then even if her choice is causally undetermined she will be responsible for her choice either way. After all, either way she will be doing what she wants to do and acting free from external manipulation, coercion, or force.1 Such an approach has an initial plausibility and it avoids the problems raised by Ekstrom and Clarke. Consider Ekstrom’s point about the phenomenology of choice. She thinks Kane’s view is problematic because we don’t experience ourselves as engaging in dual efforts in the moments leading up to choice. Well, on the model I am proposing here, our SFAs would not involve dual efforts of will; rather, they would only involve dual wantings. So, for instance, our businesswoman would not try to go on to her meeting and simultaneously try to stop and prevent the assault. She would merely want to do the one thing and simultaneously want to do the other and then make a causally undetermined choice one way or the other. Ekstrom is right that in situations of dilemmatic choice we don’t experience ourselves as trying to do both, but we do experience ourselves as wanting to do both. As such the model I am proposing here fits better with the phenomenology of dilemmatic choice. Additionally, the model considered here avoids the problem of the irrationality of dual efforts. The model under consideration does not require that dual efforts be involved in SFAs. It only requires that the agent want to do

Alternative Models of Basic Free Actions

85

one thing A and want to do another thing B and then make an uncoerced, unforced, manipulation-free causally undetermined choice from among these conflicting wants. There is nothing irrational about wanting to do one thing A and another thing B when you can only do one of them. It is simply a fact of life that we have various wants, goals, and values and sometimes in life these come into conflict with one another such that we cannot always satisfy all our desires or meet all of our goals. Despite the initial attractiveness of such a solution to the problems raised by Ekstrom and Clarke, this alternative model is problematic. It can plausibly be argued that the view under consideration falls right back into the problem of luck. The problem of luck notes that according to event-causal libertarianism two agents with exactly the same pasts and exactly the same character and motivational states facing the same choice in two identical but distinct possible worlds could choose differently. So, for instance, suppose in the actual world our businesswoman (BW) chooses to stop the assault. Well, in another possible world just like this world, businesswoman* (BW*), who is just like BW in all respects, chooses to go on to her business meeting. The point here is that if two people who are alike in all respects right up to the moment of choice can choose differently, then the difference in choice is not a consequence of any difference in their characters, beliefs, motivations, and so forth. Consequently, what the agent does in the context of an SFA would seem to be a matter of luck—it would seem to be random—and not something over which the agent exerts the kind of control sufficient for freedom and responsibility. Kane’s solution to this problem begins by noting that we are responsible for what we try or will to do. Thus, if in SFAs the agent tries or wills to do both acts—(a) tries to help prevent the assault and (b) tries to go on to her business meeting—then whichever act the agent does she will be responsible for what she does. In this way, the notion of dual willing/trying gives a way to answer the problem of luck. But, as we’ve seen, the notion of dual efforts of will has its own problems. The dual wantings model under consideration here suggests dropping the notion of dual willings/efforts and replacing it with the notion of dual wanting, but now it can be argued that this model still faces the luck problem. After all, we are not typically responsible for our wants. Our wants arise within us without our willing them and they often arise contrary to our conscious will, interfering with what we are consciously trying to do. It might be argued that if we are not responsible for our wants, then saying that the businesswoman does what she wants either way will not establish her responsibility for what she does. Thus, it may be argued that the dual wantings view leads right back into the problem of luck.

A Modest Libertarian Model Considered Another way of defending event-causal libertarianism without dual efforts has been developed by both Daniel Dennett (1978, 294–299) and Alfred

86 Alternative Models of Basic Free Actions Mele (1995, Ch.12). Mele calls the approach “modest libertarianism.” Such a modest libertarian view could be used to develop an alternative way of conceiving of the basic, underivatively free actions which Kane refers to as SFAs. Indeed, it could give us a way of conceiving of these actions as undetermined free acts for which we are responsible without invoking the concept of dual efforts of will. While neither Dennett nor Mele actually endorses modest libertarianism, they both describe a view which they think can make coherent sense of our responsibility for and the freedom of actions that are the product of indeterministic deliberative processes. However, on this view the indeterminism in the deliberative process must have a specific character if the agent’s choice is to be a free choice for which he is responsible. In particular, the indeterminism must occur in the earlier stages of the deliberation in the coming to mind of certain ideas or reasons for or against the options considered in the deliberation. Then once these reasons come to mind it requires causal determinism in the later stages of deliberation as the deliberation moves closer to the moment of choice. They believe such a view could make sense of how our choices could result from causally indeterministic processes while we still have enough control over our choices so that they are not random happenings but products of our free will. The idea here is that while the agent could have chosen otherwise because there is indeterminacy in the coming to mind of certain ideas or reasons, what is ultimately chosen is not a random happening since the choice will be a determined consequence of the way in which the agent deliberates in the later stages of deliberation. Thus, on this view, while the choice is undetermined and the agent genuinely could have done otherwise, the character of the agent and how she deliberates still dictates what is chosen once all reasons she considers have been accounted for. Here some difference in the agent in the moments prior to choice explains her choice, such that what is chosen is not just a matter of luck. To better understand this view, let’s consider an example. Suppose Billy is trying to decide which school to attend for college. Suppose he has narrowed his choice down to (A) a certain Christian school in Malibu, California, and (B) a private liberal arts college in Wisconsin. Suppose that in his thinking he has noted that the schools are equally strong academically and one of them, A, has a location he prefers, while the other, B, has a stronger music program. With these considerations in mind he still feels undecided as to what to do. But, then, suppose that it comes to his mind in an undetermined way that at the Malibu school, option A, he would have to attend required chapel meetings and he would not have to do so at the Wisconsin school. Further, as it gets closer to his time of decision, Billy’s distaste for chapel combined with his understanding that option A requires chapel as well as the other considered reasons determine his choice of B, the Wisconsin school. Here Billy could have done otherwise. Had the understanding that the Malibu school requires chapel not occurred to him, he might have made

Alternative Models of Basic Free Actions

87

a different choice. At the same time, his ultimate choice is not a random occurrence, since his choice is explicable in terms of the beliefs and desires he had at the time of the choice. Using this modest libertarian conception of free choice, Billy’s choice in this case could be understood as a basically free choice for which he is responsible. However, a Kanean would not be willing to accept that such actions could suffice for the basic, underivatively, free acts he calls SFAs. For on Kane’s view free will and ultimate responsibility (UR) require more than the modest libertarian view can give us. If this is how we understand basic free choices and if such actions were to be the ground for our free will and UR, the view would be inadequate. The problem is that such choices would not be a reflection of the agent’s free will. Notice that in Billy’s case, he does not will the coming to mind of the fact that the Malibu school has required chapel; rather, he just happens to remember this. Additionally, he just happens to dislike going to chapel. These factors then lead him to choose in favor of the Wisconsin school. Since it was causally undetermined whether he’d remember that chapel is required at school A, he could have genuinely done otherwise. Despite this, his ultimate choice is something that happens to him and not something he does of his own free will. For these kinds of reasons, the modest libertarian view cannot give an account of the agent control over undetermined choices that is necessary to make sense of them as free willed choices for which we are ultimately responsible.

Mark Balaguer’s Model Considered In recent work Mark Balaguer (2004, 2010), has proposed a different kind of event-causal libertarian view which makes no appeal to dual efforts of will, and he argues that it avoids the luck problem. In doing so he discusses what he calls “torn decisions.” A torn decision is a decision in which the person in question (a) has reasons for two or more options and feels torn as to which set of reasons is strongest, i.e., has no conscious belief as to which option is best, given her reasons; and (b) decides without resolving this conflict—i.e. the person has the experience of “just choosing”. (Balaguer 2004, 382) The businesswoman in our example above must make a torn decision, as she is torn between the desire to go on to her business meeting and the desire to stop the assault. Kane defines the decisions we make in the contexts of SFAs as causally undetermined. What Balaguer calls “torn decisions” are clearly similar to the sorts of decisions that Kane describes as SFAs, but Balaguer does not believe that all torn decisions are causally undetermined so he does not define them as undetermined decisions.2 Nonetheless, he does maintain that there may

88 Alternative Models of Basic Free Actions well be causally undetermined torn decisions, noting that this is an open empirical question. He, also, argues that if and when such decisions are undetermined in the right sort of way they are not the result of mere luck, rather they are rational decisions authored and controlled by the agent. In supporting his position, Balaguer considers the example of a woman at a restaurant who wants a dessert and has no more reason to order one or the other of the two options—tiramisu or a fruit plate—presented to her. It is time for her to order and she wants a dessert and these are her options, so she just chooses one of them while feeling no compelling reason to choose one instead of the other. He has us imagine that this torn decision is causally undetermined at the moment of her choice. He notes that even if this decision were undetermined at the moment of choice the agent would still be consciously, intentionally, and purposefully choosing from among her options. Additionally, if the choice is undetermined, then nothing external to her causes her to choose as she does. From here Balaguer goes on to conclude that were such a decision undetermined and consciously, intentionally, and purposefully performed then we have every reason to regard it as a decision that is authored and controlled by the agent, and, as such, it is not a decision which results from mere luck. Critics may balk at such a view, arguing that if the choice is undetermined then women in different possible worlds who are exactly alike right up to the moment of choice may choose differently. Since there is nothing about the nature of these agents in different possible worlds which explains their different choices, it follows that their choices would just be a matter of luck. In reply to this criticism, Balaguer notes that this reasoning is confused. In the context of causally undetermined torn decisions, we should expect that agents with exactly the same pasts and character will make different decisions. If similar agents faced with the same torn decisions in different possible worlds were led to make the same decisions in all of these worlds, then we would have reason to think that either they were not actually facing torn decisions or that something besides the agents was causing them to choose as they do. For these reasons, Balaguer says that the mere fact that similar agents across different possible worlds would choose differently does not imply the presence of control-robbing luck.3 Here it might be objected that, while all of this is fine and good, there is still a question about why we should think that in undetermined decisions it is the agent that decides or chooses, as opposed to a decision or choice just occurring as a consequence of indeterministic neural processes. Balaguer’s answer to this is that in the context of undetermined decisions nothing external to the agent causes the choice. If the woman ordering dessert is not caused to choose by external factors and the choice is undetermined and she opts for the tiramisu, then the only event relevant to her so choosing is her decision. But this decision is a mental event and it’s her mental event. In a scenario like this we procure as much authorship and control as we can, given that the agent is making a torn decision.4

Alternative Models of Basic Free Actions

89

According to Balaguer, the presence of dual efforts of will in our torn decisions is inessential for agent control over them; rather, undetermined torn decisions are simply the result of undetermined neural events occurring in the brains of the agents who perform them. In contrast, Kane argues that unless torn decisions are seen as the result of antecedent efforts of will made by the agent then we cannot make proper sense of agent control over the decision made. In his defense, Balaguer argues that the undetermined neural processes which result in torn decisions are parts of or internal to the agent’s decision. He says that when this happens and there are no factors external to the agent shaping the choice, then the agent’s control over his choice is enhanced. For in such situations, the agent makes his choice for certain reasons without any other causal factors interfering and he could have done otherwise. Consider our businesswoman, who makes the torn decision to stop and prevent the assault instead of going on to her business meeting. Balaguer believes that even if her decision is the result of causally undetermined neural processes occurring in her brain, she will still be in control of and responsible for her choice if those undetermined neural events are part of her decision process and she makes the decision while uncoerced nor manipulated by covert neural controllers. He believes that in such cases agent control is enhanced, for not only is the decision made by the agent for certain reasons, but since the decision is causally undetermined she could have done otherwise. Robert Kane finds this problematic. He argues that if a torn decision is not the result of an agent’s effort of will but is merely the result of undetermined neural events which constitute the decision process, then such decisions are the result of mere “neural coin tosses.” But then even if these neural coin tosses are internal to the torn decisions, the agent does not have control over how the neural coin tosses come out, or they would not be undetermined events; and if how these coin tosses come out “settles” which option is chosen, the agent would not have control over that either. (Kane 2014a, 55) If it is just undetermined neural processes occurring during the deliberation of the agent which dictates which choice is made, then the agent won’t have control over the choice. For such undetermined neural events which settle which option is chosen are not themselves agential doings. Thus, Kane believes there is nothing in Balaguer’s view of causally undetermined torn decisions which establishes agent control over them (Kane 2014a, 55). In a published response to Kane, Balaguer (2014) reiterates that if an agent makes an undetermined decision that is uncoerced nor the result of manipulation by external neural controllers and if that decision is intentional and reasons-based and the indeterminacy is constituted by neural indeterminacy in her decision process, then she controls the decision made. Again, even if

90 Alternative Models of Basic Free Actions it is causally undetermined, the decision is a result of her own uncoerced, unmanipulated decision process. Thus, she will have control over it and be responsible for it (2014, 91). In addition, he adds: Let me say one more thing about Kane’s version of the luck objection. I don’t understand why he thinks this objection applies to my view but not his. Suppose a torn decision satisfies the Kane-conditions. Isn’t the outcome of the decision still going to be settled by quantum events? And if it is, then why doesn’t the objection that Kane raises against my view apply equally to his own view? (2014, 91) Here Balaguer notes that Kane should not press this criticism against his view, because Kane’s own view suffers from a similar problem. As such, if this is the only problem with Balaguer’s view, it gives us no reason to prefer Kane’s view over Balaguer’s. In my “Kane, Balaguer, Libertarianism, and Luck (forthcoming),” I argue that Balaguer’s response here is problematic. In what follows, I will summarize some points I make in that essay and extend upon them in new ways. I would first note that, in contrast to Balaguer, Kane postulates dual efforts of will as essential elements in causally undetermined free choice. Because of this his view has an advantage which Balaguer’s lacks. If an agent is making an uncoerced and unmanipulated undetermined choice between two options, A and B, and the agent makes the effort to choose A and the effort to choose B in the moments leading up to the choice and the choice is the undetermined result of one of these efforts, then it is much easier to see how the agent has the kind of control over his choice which would make him responsible for it. In contrast, Balaguer views dual efforts of will as inessential to free undetermined choice. But if dual efforts are not involved, it is too hard to see what it is the agent does that gives her control over the particular choice she makes. Balaguer says she makes the decision and so she controls which particular choice is made, but the problem here is that the agent is on the “sidelines” on this view rather than actively involved in the decision process. On Balaguer’s view the ultimate choice—say, the choice of A—is not made because the agent tried to choose A but as a result of a “neural coin toss.” On this view, it is enough that the agent engages in an uncoerced decision process and then, whichever reasons happen to win out due to neural processes over which the agent exerts no control, the agent will nonetheless be said to act freely and be responsible for her choice. The problem with Balaguer’s view can, perhaps, be seen more clearly if we think about the conditions which initiate a decision process. Oftentimes, agents don’t freely decide to make decisions; that is, they don’t freely choose to engage in a decision process. Rather, external circumstances beyond the control of the agent along with the beliefs and desires of the agent causally

Alternative Models of Basic Free Actions

91

necessitate the initiation of the decision process. Consider, again, our businesswoman. She doesn’t choose to deliberate and decide whether to prevent the assault or go on to her business meeting. Rather, the confluence of seeing the assault occur and her desire to stop it and her simultaneous desire to go on to her meeting precipitate her engaging in deliberation. We rarely choose to engage in deliberation, rather situations calling for deliberation are often foisted upon us. But when they are foisted upon us, we may still make free undetermined choices. Now suppose that through no choice of her own our businesswoman is confronted with a situation calling for choice and she doesn’t engage in dual efforts of will but just satisfies the Balaguerian conditions for free undetermined choice. Could this make her choice a free willed act for which she is responsible? She doesn’t initiate the decision process; rather, the confluence of events and her desires bring it about. But she does decide and this choice is the result of undetermined processes internal to her decision. The problem here is that she doesn’t control those processes; they simply happen and then a decision occurs. Now, yes, she decides here, but the process is not initiated of her free will nor is the outcome of the process a product of her will. So, how are we to make sense of her choice as a free willed choice? In contrast, on the Kanean model, even though the businesswoman may not have controlled the initiation of the deliberation process, we can still see how she is actively involved in the outcome of her decision process and how her decision is a reflection of her will. For on Kane’s view the agent exerts dual efforts of will in the moments leading up to choice and so either way the choice will be a reflection and consequence of her will. As such Kane’s view makes it clear how our causally undetermined, torn decisions can be products of our will and, indeed, be free willed acts for which we are responsible, whereas Balaguer’s view cannot make this clear.

Mele’s Daring Libertarian Model The weakness in Balaguer’s view is instructive. Notice that the problem with his view is that there is not enough agential involvement in the moments leading up to an undetermined choice in order for us to make sense of the agent control that is required for free action and moral responsibility. Whereas on the Kanean view, the agent is trying to A and trying to B in the moments leading up to the choice. If the agent makes the causally undetermined choice to A, then he will be responsible as it was what he was trying to do. Furthermore, had he chosen B, he’d have been responsible for that choice as well, since he was simultaneously trying to do B. Since we are responsible for what we try to do, it follows that in SFAs we will be responsible for what we do either way. In contrast, on Balaguer’s model the agent need not be doing anything in the moments leading up to the choice of A or the choice of B. Rather, on his model the agent might merely be conflicted about what to do—for instance, the agent might merely be consciously

92 Alternative Models of Basic Free Actions aware of the desire for tiramisu and the desire for the fruit plate—and then a “neural coin toss” dictates which choice is made. A similar problem occurs if we try to understand basic (or underivatively) free actions, SFAs, on the model of modest libertarianism. For on this model as well there is not enough agential involvement in the deliberation leading up to choice. Recall the case of Billy. In some causally undetermined way during deliberation, he just happens to remember that school A requires attendance at chapel and it also so happens that he dislikes going to chapel. These factors then tip the balance of his reasoning in favor of choosing school B and he does so. Here too the choice just happens to him without sufficient agential involvement in the process. Thus, on this model, like Balaguer’s, there doesn’t seem to be enough agential involvement in the choice to make sense of it as a free willed choice for which the agent is responsible. What this means is that if we are to salvage event-causal libertarianism while avoiding the pitfalls of the theory of dual efforts, we will need a theory which includes a sufficient level of agent involvement in the moments leading up to choice so as to establish responsibility for the causally undetermined choices which constitute our SFAs. In recent writings, Alfred Mele (2006, 2017) has developed another kind of event-causal libertarian view, “daring libertarianism,” which tries to make sense of basic (or underivatively) undetermined free actions without an appeal to dual efforts of will. This view differs from the modest libertarian view he developed in earlier writings. On the modest libertarian view, it was said that in making undetermined free choices the indeterminacy had to occur only in the earlier stages of deliberation in the coming to mind of certain ideas or reasons which would then be considered in the moments leading up to choice. The idea was that in the later stages of deliberation there would be causal determinism in moving from the considered reasons to a decision. However, on the daring libertarian (DL) view, it is said that in undetermined free choice there may be indeterminacy in the deliberation right up to the moment of choice. In allowing for indeterminism right up to the moment of choice in basically free actions, Mele’s view is similar to Kane’s view, as Kane also allows for this in his conception of SFAs. However, as noted, Mele does not believe dual efforts of will are necessary to make sense of the agent control of and responsibility for basically free undetermined actions. In distinguishing his view from Kane’s, Mele writes: The main difference is that where Kane postulates concurrent competing indeterministic efforts to choose, I postulate an indeterministic effort to decide (or choose) what to do. That effort can result in different decisions, holding the past and the laws of nature fixed. (Mele 2017, 60) What Mele suggests is that in basic undetermined free choices in the moments leading up to choice we engage in a causally indeterminate deliberative process

Alternative Models of Basic Free Actions

93

which constitutes an effort to decide what to do. As such, we will be responsible for whichever choice we make, since whichever choice we make it will be a product of our single effort to choose. An attractive feature of this view is that it provides an event-causal view without dual efforts of will, while also doing something to bring more agential involvement into the causal sequence leading up to choice. As I note above, the weaknesses of the modest libertarian view and Balaguer’s view are that in their conceptions of basic, underivatively free undetermined choices there is not enough agential involvement in the moments leading up to choice to make sense of the agent’s control of and responsibility for such choices. But on Mele’s DL view, the agent’s deliberative process leading up to choice is conceptualized as an effort to decide. In deliberating, the agent is trying to decide between two or more alternative courses of action. By conceptualizing our basic, underivative free choices in this way, the DL view includes an element of agential involvement in the moments leading up to choice which might lead us to think the agent’s choice is more a reflection of what he does as opposed to something which happens to him. This helps us see the agent as having control over his choice. Mele thinks DL gives a coherent conception of basic, underivatively free undetermined choices. However, as was the case with his modest libertarian view, it is not a view which he endorses. Mele does not endorse DL, because (1) he does not believe there are sufficiently strong reasons to think incompatibilism is true and (2) he does not believe there is sufficiently strong evidence that human brains work in the kinds of ways necessary for libertarianism to be true (Mele 2017, 60). At the same time, he thinks that if one is inclined to embrace an event-causal libertarian view then one should prefer DL over Kane’s view. He says this because he thinks the DL view is every bit as coherent as Kane’s view, but it makes no appeal to dual efforts of will. Like me, Mele finds the appeal to dual efforts of will to be problematic. In particular, Mele thinks we don’t experience ourselves as engaging in such dual efforts of will; also, there is no evidence that we do engage in them. Thus, he thinks that if the DL view provides an account of basic, underivatively free action which is just as coherent as Kane’s dual efforts model, then we should favor the DL view (Mele 2017, 53–54, 65). In Kane (2014b), he criticizes Mele’s DL view, noting that on this view there is still not enough agential involvement in the moments leading up to choice to make sense of the agent’s control over the particular choice he makes. He states: [O]ne needs to say more than that certain reasons (for doing A) win out in the process of deliberation at the moment of choice, or certain other reasons (for doing B) win out, and it is undetermined which ones win out. One must also be able to say 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

94 Alternative Models of Basic Free Actions choice rather than the other is made, despite the indeterminism. And this requires more resources. To do it, I argue, the activity of the agent prior to choice must not merely be imagined to be a single process (deliberation) with multiple possible outcomes (deciding to A or deciding to B), one of which occurs by chance. One must also consider that there is an internal complexity to the preceding process of deliberating that gives the agent more of an active role in settling which reasons win out over the others. This can be done if, instead of merely imagining a single process (deliberation) with multiple possible goals, one also imagines that within that larger process, there are multiple processes, each with a specific goal the agent is trying to realize. These processes, of course, are my efforts of will or volitional streams . . . In such a scenario, it would not merely be the case that one set of reasons “wins out” over the other, but rather 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. (Kane 2014b, 208) Kane feels that without dual efforts involved in the deliberation process leading up to undetermined choice, then there won’t be enough agential involvement in what happens to make sense of agent control over and responsibility for the choice. Without the dual efforts, which set of reasons wins out in the deliberative process just happens to the agent and the agent doesn’t make a set of reasons win out over other reasons through his effort. In this way, Kane thinks his view gives a better account of agent control over and responsibility for undetermined choices. According to Kane, in the moments leading up to the choice the agent tries to choose A and tries to choose B. Thus, whichever choice he makes it will be something he tried to do—it will be a result of his own effort of will to do that very thing. Thus, either way the particular choice he makes will be what he willed to do and he will be responsible for his chosen action either way. In contrast, on Mele’s DL view the agent doesn’t try to choose A and try to choose B, he merely tries to decide which of them to do and then a choice of one or the other gets made. Suppose the agent chooses A. On the DL view, the agent merely tries to decide but never tries to choose A prior to choosing A. Thus, when he chooses A it is not because he willed it. What he tried to do was to decide. For all we know, the choice of A is merely a consequence of the fact that one set of reasons just happens to win out in the causally undetermined process of deliberation leading up to choice. On Kane’s view, in an SFA when the agent makes his undetermined choice either way (A or B) he will end up doing something he was trying to do all along because he was exerting an effort to choose A and an effort to choose B throughout the deliberation process leading up to the choice. Thus, on his view the choice either way will be a reflection of his will, making him clearly responsible for what he chooses.

Alternative Models of Basic Free Actions

95

In Mele (2017), he addresses this kind of reply, stating: 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. But what contribution does a choice 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 . . . makes to the killings 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; and it is commonly recognized that choosing to A is essentially intentional. Agents do not need to try to choose to A nor to try to bring it about that they choose 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. If trying to choose A is supposed to make a contribution to freely choosing A that goes beyond its contribution to exercising plural voluntary control over what one chooses, Kane has not said what that contribution is. (Mele 2017, 64–65) Here Mele contends that we don’t need dual efforts of will, because all that they provide is clarification that whichever choice is made it is intentionally made. As Mele notes, however, even on his DL view the choice will be intentionally made, because choices are “essentially” intentional. This reply is inadequate. While it may be that all choosings are intentional acts, the recognition of this won’t help in defending Mele’s view. It’s important to note that intentional doings are not necessarily under our control. When I suddenly see an object flying towards my head, I intentionally duck to avoid it; but this is done instinctively. I intentionally duck, but I cannot help but do so and I am not responsible for doing so. On the DL view, I might intentionally choose A at the end of an indeterministic deliberative process, but my intentional choice of A may still be something that merely happens to me as it is caused not by anything I did but by certain reasons winning out in the moments leading up to choice. In contrast, if I engage in an effort to choose A and an effort to choose B in the moments leading up to choice and then I make a causally undetermined choice of A due to my effort to choose A, then my choice of A will be a product of something I purposefully did and it won’t be something which merely happens to me. The point of the dual efforts is not to establish that the agent will make an intentional choice at the end of a causally indeterminate process of deliberation, rather the point is to ensure that there is sufficient agent involvement in the moments leading up to choice so as to ensure that the particular choice made, whether the choice of A or the choice of B, is a product of the agent’s will, his control, as opposed to something which just happens to the agent.

96 Alternative Models of Basic Free Actions Another problem with the DL view is that in undetermined basic free actions the agent is said only to engage in a single effort to choose in the moments leading up to the choice between two options, A or B. This can only make the agent responsible for engaging in the intentional act of choosing. Suppose the agent chooses to A. Engaging in the mere effort to choose between A or B does not establish that such an action counts as an SFA or what Mele calls a “basic free action.” While his choosing is an intentional act and he is responsible for choosing because he tried to choose, it does not follow that what is chosen is under the control of the agent in such a way as to make him ultimately responsible for it. For the particular choice made, in this case the choice of A, is not something he antecedently tried to do, rather it is just the consequence of whichever set of reasons wins out in the undetermined process of deliberation leading up to choice. In reply to this Mele might say but even on the dual efforts view the particular choice made, say the choice of A, will also be something which just happens to the agent, since it will be a consequence of whichever effort of will happens to win out. However, this would be a mistake. To see this let’s go back to Kane’s example of the assassin’s causally undetermined killing of the prime minister. As the assassin takes aim at the prime minister and begins to pull the trigger, it is causally undetermined as to whether he will have a nervous twitch and fail to kill the prime minister. As it is, he doesn’t twitch and he, instead, succeeds at what he was trying to do. Notice here that while the killing was causally undetermined, we don’t say the assassin was robbed of control over killing the prime minister in such a way that he was not responsible. This is because we recognize that he succeeded at what he was trying to do. So, too, even if it is causally undetermined which effort of will wins out in making a choice, the agent will choose of his own free will and be responsible either way he chooses. For regardless of which choice gets made, he will have succeeded at what he was trying to do. In contrast, on the DL view in the moments leading up to the choice of A or B the agent merely exerts effort to decide from among the options and then makes an undetermined choice of one or the other. Here one set of reasons, either the reasons for A or the reasons for B, may dictate in some undetermined way which particular choice is made, but if so, it will not be up to the agent or a reflection of the agent’s will. Thus, on the DL view the particular choice made, the choice of A or the choice of B, may just happen as a result of one set of reasons winning out; if so, the particular choice will not be a product of the agent’s will and he cannot be ultimately responsible for it. Additionally, in response to this discussion of Mele’s DL view, Kane writes: What would it say of a man’s moral character if he were willing to resolve an important conflict in his life (say, choosing to cheat or not to cheat on his spouse) by accepting the result of an indeterministic process, whether internal, or external (like a coin flip), without having

Alternative Models of Basic Free Actions

97

any control over which result might occur? The spouse would certainly not be impressed by his moral character or his commitment to her, if he pleaded that the indeterministic process came out in her favor, and so he did not cheat on her. What would we say of the character of an agent all of whose important torn decisions, moral and otherwise, were resolved in this way? Yet this is what Mele’s DL would seem to allow for libertarian free choices. It seems a very impoverished view of what libertarian free will involves. (Kane in correspondence) Kane’s concerns here are rooted in concerns that I have already raised in my earlier remarks. He thinks Mele’s DL view allows the decisions we make in SFAs to be merely the undetermined consequences of whichever set of reasons win out in the deliberation process leading up to the choice. But in these remarks, he adds something new for our consideration. In particular, he makes an important link between these concerns and our perception of the moral character of agents involved in SFAs. Kane is right that the woman in his example would hardly be impressed if she found out it was a matter of luck which choice her husband made and he was willing to accept either result of his luck ridden deliberations. If this is how his deliberations proceed then the husband’s making of his decision is no better than it would be had he flipped a coin and decided on the basis of his coin toss. In thinking further about this example, it is important to note that a key factor which leads Mele to think event-causal libertarians should prefer his view is that he believes the actual choices made in SFAs on either his DL view or Kane’s view will be matters of luck (Mele 2017, 57–58). Thus, in response to this example, Mele might say, Well, even if the decision involves dual efforts of will, it would still be a matter of luck what happens in the man’s decision to cheat. On Kane’s view, in SFAs the agent is of two minds while deliberating and it is still causally undetermined what path the deliberation goes. Thus, it will be a matter of luck which part of the self—the cheating self or the faithful self—will win out and dictate the particular choice made. Again, Mele thinks we neither perceive ourselves as engaging in dual efforts of will nor do we have any evidence that we engage in them. Thus, he is led to look for an alternative way to make sense of causally undetermined basic (underivatively) free actions. Since he believes his DL view is coherent and it renders SFAs no more lucky than they would be on Kane’s view, he is then led to conclude that event-causal libertarians should prefer his view to Kane’s. However, if we do engage in dual efforts of will, then it is not just a matter of luck which effort of will wins out because, as I’ve argued, the agent makes one set of reasons win out over the other set. Suppose the husband decides

98 Alternative Models of Basic Free Actions to be faithful to his wife. If so, on the dual efforts view his choice will be undetermined but he will have made this choice because of his own effort to be faithful. What leads people to think that the Kanean view makes our SFA decisions lucky is that they tend to think that on his view there are two parts of the self involved—say, the faithful self and the cheating self—and then quantum level indeterminacies in the brain dictate which choice gets made. Were this Kane’s view then, indeed, Kanean SFAs would be subject to luck in the way that Mele suggests. However, this is a mistaken interpretation of the Kanean view. Rather, Kane sees the quantum level indeterminacy occurring in the neural structures of the brain not as dictating which choice gets made but as opening up a window of opportunity for one part of the husband’s self to establish dominance over the other part of himself through its own effort. The choice is not a product of randomness where the agent sits back and sees where the quantum indeterminacy will lead. Rather, the indeterminacy creates the opportunity for one part of our self to establish dominance over another part of our self. Now, Mele might say, “Even here, should the wife feel any better knowing that the husband’s better self just happened to win out? After all, the decision might have gone the other way.” But, here, Kane’s view really does seem advantaged over Mele’s DL view. For Kane can say that, while she might rightfully prefer that her husband not have felt tempted to cheat at all, she will at least know that his faithfulness is a result of strivings within him that reflect who he is and in this case he is someone who has made his better self take precedence over his inferior self. While he could have chosen otherwise, he didn’t because through his own effort he has made the better part of himself win out. This man’s wife can feel some consolation in the fact that her husband did not leave the decision to chance; rather, through the efforts of his better self he made it the case that he did not cheat. In an SFA, on the Kanean model when we choose rightly we are better than we are on Mele’s DL view, because (1) we don’t leave the choice up to chance and (2) our choice reflects the fact that we have made the better part of ourselves take precedence in our lives. Mele’s DL view cannot capture these important moral dimensions of self-formation.

Part Two The discussion in the first half of this chapter suggests that if we are to defend an event-causal libertarian view of causally undetermined basic, underivatively free choice without an appeal to Kane-style dual efforts, then we will need to develop a plausible view in which there is sufficient agential involvement in the moments leading up to choice. Without the right kind of agential involvement in the moments leading up to choice, the agent’s particular choice from among his options, such as the choice of A instead of B, will be something that happens to him as opposed to something he does. As we’ve seen, failure to establish the right kind of agent involvement leading

Alternative Models of Basic Free Actions

99

up to choice has been a major stumbling block for the various alternative event-causal views we’ve considered. In what follows, I will present two new models for understanding the nature of SFAs both of which reject Kane-style dual efforts of will. One of these is a more Kanean view, as it embraces the notion of dual efforts but understands them in a way very different from the way Kane understands them. The second of these views tries to understand SFAs without making any appeal to dual efforts at all, making a more significant break from the Kanean view. I am undecided as to which of these models is preferable, but I present them both here as options for event-causal libertarians, as I think both solve problems inherent to other event-causal libertarian theories.

A Non-Kanean Dual Efforts View As we’ve seen in Part One of this chapter, event-causal libertarian theories which reject Kane-style dual efforts of will have a difficult time making sense of how agents can be in control of and, thus, responsible for the particular choices they make when engaged in SFAs—basic, underivatively free undetermined decisions. For instance, on Mele’s daring libertarian view, which posits a single effort to decide prior to choice, we can make sense of the agent’s control over the making of a choice, but, as I argued, this view does not have the resources to explain agent control over the particular choice she makes—the agent’s control of, say, her choice of A instead of B. Kane makes sense of the agent control over this in terms of his theory of dual efforts of will. On Kane’s view, since the agent was trying to choose A in the moments leading up to choice, then this particular choice is directly a product (a causally undetermined one) of her effort to do so. Had she not made the effort to choose A her strong desire to choose B would have won out and she’d have chosen B. On Kane’s view, while the agent’s choice is undetermined, it is still a product of her will because it is something which she tried to do, and had the choice gone the other way, she would have been responsible for that choice as well because she was simultaneously trying to choose B. In this way, says Kane, agents can have plural voluntary control over their choices in SFAs. Now, as we’ve seen Kane’s theory of dual efforts has certain problems. I’ve argued that the view has three problems in particular; these are: (1) the phenomenological problem, we don’t perceive ourselves as engaging in dual efforts of will in the moments leading up to choice; (2) the irrationality problem, to engage in dual efforts to choose A and to choose B when you know you can only do one of them is irrational; and (3) the problem of being overly speculative, we don’t have sufficient evidence that we engage in Kane-style dual efforts of will. In the previous chapter, I noted how Kane has responded in the literature to the first two of these problems; but I also went on to present problems with Kane’s replies to these issues. Regarding the third problem, Kane would admit that his view is speculative in the

100 Alternative Models of Basic Free Actions manner noted. However, he doesn’t think that at present there are any rival libertarian theories which can make better sense of agent control over the particular choices we make when engaging in SFAs. He remains open to the possibility that some such theory could be constructed, but as of now he doesn’t see how this can be done. In what follows, I will present a theory similar to Kane’s which embraces the idea of dual efforts but which understands them in a manner different from that of Kane. In doing so, I hope to present a view which avoids the three problems noted above, while also providing for enough agential involvement in the moments leading up to choice to make sense of the agent’s control of and responsibility for the particular choice made in an SFA. What I suggest is that we understand agents involved in SFAs not as trying to choose A and trying to choose B in the moments leading up to choice but as trying to keep alive the option of choosing A and trying to keep alive the option of choosing B. In the deliberation process, as the agent visits and revisits and reflects on the reasons for choosing A and the reasons for choosing B as well as the reasons against these choices, she is keeping her actional pathways open; she is sustaining the possibility of choosing either one, and she does this through her own efforts. We might think here of the weighing of the reasons for and against choosing A and the weighing of the reasons for and against choosing B as the agent committing dual efforts to keep these possible courses of action alive. Here the dual efforts are not efforts to choose A and to choose B; rather, they are distinct efforts to keep both actional pathways open. Suppose the agent makes the causally undetermined choice of A. Now, on this view the agent will not antecedently try to choose A. So, her control over this choice and her responsibility for it will not be grounded in such an effort. Rather, on this view the agent’s control of and responsibility for the particular choice she makes, the choice of A, is grounded in her effort to sustain the choice of A as a possible actional pathway. Had she not done this, then she could not have made this choice; rather, had she not exerted the effort to keep that pathway open, then her strong desire to choose B would have led her to choose B instead. Through continuing to contemplate the reasons for choosing A in the deliberative process, the agent knowingly and intentionally sustains the possibility of choosing A. Thus, when she chooses A, she is responsible for that choice. Additionally, since the agent simultaneously exerts effort to keep option B alive and since this is necessary to keep her from being overwhelmed to act on her desire to do A, then should she choose B instead of A she will be responsible for that choice as well. Thus, on this view the agent will be responsible for her causally undetermined choice either way. Notice that if this view is adequate, it will avoid the three problems which I have argued plague Kane’s view. Consider the problem of irrationality. When you very much want to do A and you very much want to do B and you can only do one of them and you are trying to choose which one to do,

Alternative Models of Basic Free Actions

101

is it irrational to try to keep both actional pathways open? Certainly not. Since you very much want to do both acts, it is rational to keep both options open until you can either decide which you prefer or, at least, arbitrarily pick one if you cannot form a rational preference for one or the other. And, what of the phenomenological objection? While we certainly don’t experience ourselves as trying to choose each of the options we consider when deliberating, something which Kane himself concedes, we do clearly experience ourselves as contemplating the reasons for and against each of the options under consideration. What I am suggesting is that in contemplating the reasons for choosing A and against choosing B, the agent is trying to keep the option of choosing A alive. At the same time, in contemplating the reasons for choosing B and against choosing A, the agent is simultaneously trying to keep the option of choosing B alive. Thus, in experiencing ourselves contemplating the reasons for and against each option we are in effect experiencing ourselves engaged in the effort to keep each option open. The agent strongly desires each option and wants to give herself a chance at doing either and so by continuing to contemplate the reasons for and against each option she is trying to keep her options open and she experiences herself as doing such as she deliberates. Regarding the latter point, a critic might find this to be a stretch. The critic might say just as we don’t experience ourselves as trying to choose A and trying to choose B, so too we don’t experience ourselves as trying to keep option A alive and trying to keep option B alive. We just try to decide what to do, and that’s what we experience ourselves as doing. The contemplating of reasons is not seen as an attempt to keep options open; rather, it is seen as a necessary step in making a decision which fits best with our hierarchy of values. In response to this possible objection, I’d note it’s true that we experience ourselves as trying to decide; but, nested within this effort there may yet be other efforts which are essential to the process of deciding. We may not initially see ourselves as engaged in the kinds of dual efforts which I’m saying we commit, but if we attend to the deliberative process we may come to see that indeed we do engage in dual efforts to keep our options open. To see this think of a choice where you strongly desire to A and you strongly desire to B and you can only do one of them. If you don’t continue to weigh the reasons in favor of each, keeping those reasons in focus, you may become overwhelmed with the desire to choose one of the options prematurely and make a choice you will regret. Thus, to avoid hasty action you continue to deliberate, but in doing so you are trying to keep both options alive—trying to keep option A alive and trying to keep option B alive—so as to make the most reasonable choice given your set of values. From this perspective, I think we can see that we do engage in such dual efforts, and, as I’ve argued, doing so is rational insofar as we have strong desires for each of the options and we want to give both a fighting chance to be chosen. Regarding the last of the three objections, it seems to me that we have more evidence that we engage in the kinds of dual efforts I describe than

102 Alternative Models of Basic Free Actions we have for the kinds of dual efforts which Kane describes. As I’ve just argued, if we attend in the right way to the phenomenon of deliberation, then we will come to see ourselves as trying to keep both options alive in the weighing of the reasons for and against each of the options under consideration. While we might not see right away that this is what we are doing, we can nonetheless come to see that this is the case. As such, there is at least some introspective data for the truth of the kind of dual efforts that I am postulating. In contrast, as Kane himself admits, there doesn’t seem to be any introspective evidence for the kind of dual efforts to which his view is committed. Admittedly, introspective data is not the most reliable kind of evidence. To have sufficient empirical evidence for my view, it would help to have further support from empirical disciplines, such as psychology and neuroscience. But, I would add here as well that the little empirical support which Kane provides for his view, such as his references to the fact that our brains engage in parallel processing (as in visual perception) lends as much credence to my view as it does to his. Thus, the view I am describing has as much empirical support as Kane’s, as well as the added benefit of fitting with our introspective data. A fan of Kane’s view might note that even if the non-Kanean dual efforts view I’ve described has the advantages noted above, it still should not be preferred over Kane’s view if it fails to establish agent control over and responsibility for the particular choices we make in our SFAs. A critic might argue that on the non-Kanean dual efforts view the agent still does not have enough control over his choosing A instead of B, because he doesn’t make the reasons in favor of A win out over the reasons in favor of B. On Kane’s view, when the agent tries to choose A in the moments leading up to choice and then he chooses A, it follows that even though the choice is undetermined he makes the reasons for choosing A win out by trying to choose A. This happens only through his effort to choose A. For he was simultaneously trying to choose B and, thus, had he stopped trying to choose A he would not have chosen it but would have chosen B instead. In this way, the Kanean critic may feel that engaging in mere dual efforts to keep options alive in the moments leading up to choice does not do enough to establish agent control over the particular choices made in SFAs. Additionally, it may be wondered what kind of positive argument can be made in support of my view. Kane can support his view with his various arguments from analogy, such as the assassin analogy. Kane thinks we have reason to believe that in an SFA where the agent makes an undetermined choice to A instead of B, he can be responsible for this choice for the same reason that the assassin is responsible for his undetermined killing of the prime minister. The assassin is responsible because he succeeded at what he antecedently tried to do. So, too, on Kane’s view in an SFA when one antecedently tries to choose A one will be responsible for that choice because one will have succeeded at what he was antecedently trying to do. On the non-Kanean dual efforts view that I have described, in an SFA the agent is

Alternative Models of Basic Free Actions

103

no longer viewed as trying to choose A; rather he merely tries to keep this option alive. As such, the argument from analogy here no long has application. If the agent merely tries to keep his options alive and then chooses from among them, then he may have succeeded at keeping those options alive; but, the argument from analogy would not contribute to showing that he was responsible for choosing A, because on the non-Kanean dual efforts view he is not understood as having tried to choose A prior to making the choice. Admittedly, the arguments from analogy which Kane uses to support his view won’t be adequate to support the view I’ve described here. Despite this, we can build comparable analogies in support of the non-Kanean dual efforts model. For instance, suppose I have two dogs. One of them, Spot, is sweet and gentle, and the other, Spike, is mean and dangerous. Suppose, as well, that I keep them on a leash to which they are both connected and they constantly strain at this leash, each trying to run loose in my neighborhood. Also, suppose that I know one of them will eventually break free in the neighborhood and it is causally undetermined which one will do so. Suppose that I keep them both alive by feeding them daily while keeping them on this leash, knowing that eventually one will break free. Finally, imagine that Spike the dangerous one breaks free, causing injury to some children in the neighborhood. Here it is plain that I am responsible for the injuries caused by my dog. For I did knowingly and through my own efforts sustain him in this condition, keeping alive the possibility that this may happen. While it was causally undetermined that he would break free and cause harm, I am nonetheless responsible for the harm caused by knowingly and actively keeping this possibility alive. Additionally, had the kind and gentle dog, Spot, broken free instead and brought joy to the children of the neighborhood, then I would have been responsible for this as well, as I also knowingly and through my own effort sustained this possibility. In our daily lives, most of our conduct is causally determined by the stronger motivations, by the stronger desires, we have at any given moment. I desire to scratch my itchy head, so I do so; I desire to sip my coffee, so I do so; I desire to go back to reading my book, so I do so; and so on. But when we simultaneously possess strong desires to do two or more actions and can only do one of them, we must then deliberate about what to do. Like Kane, I think our undetermined SFAs which serve as the ground of libertarian freedom and ultimate responsibility occur in these situations where we are torn between conflicting desires which cannot both be satisfied. As I see it, when an agent has a desire to do some good act, A, and some bad act, B, and he deliberates, considering and reflecting on the competing reasons for each option, the agent is in effect exerting effort to keep each of his options viable. He exerts effort to sustain the possibility of choosing A and he exerts effort to sustain the possibility of choosing B. In deliberating he doesn’t, as Kane would have it, try to choose A and try to choose B in the moments leading up to choice; rather, through considering the reasons for each option

104 Alternative Models of Basic Free Actions he tries to keep each option alive as an open possibility. Since the agent sustains the possibility of doing B while simultaneously sustaining the possibility of doing A, knowing that one of these is likely to be realized in his act of choosing, he will be responsible for his choice either way. Even though my position can be offered support through the analogy provided above, it’s likely Kane will still be resistant to this modified dual efforts view. As I noted above, he thinks if we are to be ultimately responsible for our SFAs, it is fundamental that we make one set of reasons win out over another set of reasons when making our choice. Recall that this is his complaint with Mele’s daring libertarian view. Kane feels that on the latter view when we make underivatively free undetermined decisions between two options, A and B, one set of reasons just wins out over a competing set of reasons without the agent making this happen through a prior effort of will to, say, choose A instead of B. Kane will likely argue that my proposed non-Kanean dual efforts view suffers from a similar problem, contending that on this view the choice of A instead of B would also just be the consequence of one set of reasons winning out. In response to this issue, I will argue that while Mele’s daring libertarian view does face a significant problem regarding this issue, my non-Kanean dual efforts model avoids this problem because of its similarity to Kane’s view. To see this it will be necessary to reexamine elements of both Kane’s view and Mele’s daring libertarianism. First, while on Mele’s daring libertarianism there is a problem with the fact that certain reasons just win out over competing reasons in undetermined free choice, it needs to be acknowledged that on Kane’s view there remains a sense in which the choice of A instead of B also involves a winning out which just happens to the agent. Remember on Kane’s view it is causally undetermined which effort of will wins out. As such, when the agent chooses A instead of B there remains a sense in which the decision just happens to the agent. It’s not as though the agent has complete control of the fact that he chooses A instead of B. Because the choice is causally undetermined, he certainly lacks antecedent control over this. Nonetheless, Kane’s view is still preferable to Mele’s daring libertarianism, because Kane’s reference to the antecedent effort of the agent to choose A instead of B lets us see the agent’s particular choice of A as a product of his will in a way that Mele’s view cannot capture. For had the agent not made this prior effort to choose A his strong desire to B would have taken hold of him leading him to choose B instead. Thus, on Kane’s view while there remains a sense in which the particular choice of A instead of B just happens, it is still properly seen as a result of the agent’s prior effort in a way that Mele’s daring libertarian view cannot capture. For on the daring libertarian view the prior effort is merely an effort to decide and not dual efforts to choose A and to choose B. As such, Mele’s view only allows us to see the act of choice as a product of the agent’s effort (will) but not the particular choice made, the choice of A instead of B.

Alternative Models of Basic Free Actions

105

Nonetheless, when we turn to my non-Kanean dual efforts model and examine it in these respects, we will see that it enjoys the same advantages Kane’s view has over Mele’s daring libertarianism. Like Kane’s view, the non-Kanean dual efforts view allows us to see the agent’s particular choice of A instead of B as a product of his will. While on this view the agent makes no prior effort to choose A instead of B, he still makes the prior effort to keep this option alive as well as a simultaneous effort to keep option B alive. Additionally, had the agent not made the effort to keep option A alive his strong desire to choose B would have won out leading him to make a different choice. As on Kane’s view, on my non-Kanean model the agent doesn’t have complete control over the particular choice made (again, due to the undetermined nature of the choice there is no antecedent control), but still when the agent chooses A instead of B this is a reflection of the agent’s will in a way that Mele’s daring libertarian view does not capture. For on Mele’s daring libertarian view there is merely a single effort to decide and this only allows for the agent’s will to be reflected in the act of choice but not in the particular choice made. Kane may still object that his view can give us more than my non-Kanean dual efforts view can give. It might be argued that on Kane’s view of SFAs the choice of A instead of B is more fully a reflection of the will of the agent, because on his view the agent actually exerts effort to make this particular choice prior to choosing. It could be argued that merely exerting effort to keep an option alive is weaker in the sense that in making such an effort the agent makes no commitment to choosing or doing A. In contrast, on Kane’s view in SFAs when the agent chooses A he does something which he antecedently willed to do, and neither my non-Kanean view nor Mele’s daring libertarian view can give you this. Here it must be conceded that Kane’s view does in the sense noted above give us more than my non-Kanean dual efforts view can give. There is a sense in which Kane’s view allows for the particular choice made to be more of a reflection of the agent’s antecedent will, since the agent is seen as actually trying to choose A in the moments leading up to choosing A. All that I can say in response to this is that such a view comes at a cost. Recall that, as I’ve argued, Kane’s kind of dual efforts view suffers from the phenomenological problem, the problem of irrationality, and problem of being overly speculative. I’ve argued as well that the non-Kanean dual efforts view is less problematic than Kane’s view when it comes to these matters. It could be said that the non-Kanean dual efforts view is preferable to Kane’s view, because (a) it gives you a sufficient level of agential involvement in the moments leading up to SFAs to make adequate sense of the agent’s responsibility for them, while (b) doing a better job addressing these three issues. Additionally, we might wonder about at least one more thing, namely the counter-intuitive nature of Kane’s conception of SFAs. In the argument of the preceding paragraph, it is noted how Kane’s view can give us more by establishing that in an SFA the agent’s antecedent effort to

106 Alternative Models of Basic Free Actions choose A establishes his antecedent commitment to doing A and as such makes his choice of A more a reflection of his own will. But this is counterintuitive. In the moments leading up to choice, when we are deliberating, we certainly don’t feel committed to choosing either option. Rather, we want to do both and we keep both options alive by continuing to deliberate; but it is not until the moment of choice that we make a commitment to doing either A or B. It might be thought that the mere fact that Kane’s view is counter-intuitive in this way is not a significant problem or if it is it is not really any different from the issues raised by Ekstrom’s phenomenological objection. But I would add something else to this point. Notice that there is a difference in our moral assessment of someone who is merely keeping an option, A, alive and someone who is committed to doing A. Suppose I’m in the deliberations leading up to an SFA and you know that I am considering doing a bad act, A, or a good act, B. It reflects badly on one’s character when one takes measures to keep his option of doing a bad act open to him. So, even if I am merely keeping the possibility of choosing A alive, that speaks badly of my character. But, it is even worse to be trying to choose A, as this expresses a level of commitment to doing A that is not there in the person who is merely keeping the option of choosing A alive. The Kanean view would entail a commitment to a more negative assessment of those who contemplate doing bad acts than I think is warranted.

SFAs Without Dual Efforts: A Modified Nozickian Model As noted, an adequate conception of basic, underivatively free undetermined choices, what Kane calls “SFAs,” must be able to account for agent control over and responsibility for the particular choice made. Kane’s dual efforts view can provide this whereas the alternative event-causal models considered in the first half of this chapter cannot. I’ve just explained a different kind of dual efforts view which overcomes the problems of both Kane’s view as well as the event-causal views considered in this chapter. At this point, I want to consider another new model which rejects the notion of dual efforts altogether. This model is grounded in a concept found in Robert Nozick’s event-causal libertarianism, but it employs the relevant concept in a manner different from the way in which Nozick uses it. The relevant concept is Nozick’s concept of the “assigning of weight to reasons” (Nozick 1981, Ch.4). Nozick holds that our causally undetermined free decisions occur when we make torn decisions, when certain reasons make the choice of one option, A, attractive and certain other reasons make another option, B, attractive. He believes that when we make a decision in such situations our reasons cause our decision, but it’s not necessarily the case that they deterministically cause it. When such decisions are undetermined, they are not random because in these situations there are reasons for our decisions and the reasons cause our choice without determining it.

Alternative Models of Basic Free Actions

107

Furthermore, he says that when we make such undetermined choices, in choosing we assign greater weight to one set of reasons. In choosing, we establish which considerations are of more value to us. He also says that this prioritization of certain values over others oftentimes sets a precedent for how we will decide in future situations. In this way, these undetermined decisions we make contribute to shaping our hierarchy of values, making them similar in certain respects to Kane’s SFAs. Nozick’s view suffers from problems similar to the other event-causal views we’ve considered in Part One of this chapter. The problem with the view is that there is not enough agential involvement in the deliberation process in the moments leading up to choice to make sense of the agent’s control over and responsibility for which choice gets made in basic undetermined free choices. For instance, suppose an agent is considering doing A or doing B, and he makes an undetermined choice of A. On Nozick’s view if this is a free choice the agent will have reasons to choose A and these reasons will be the cause of his choice without determining his choice. But the problem here is that there’s nothing in the theory to explain why the choice of A isn’t something that happens to him as opposed to something he does. It is unclear how he makes the choice of A happen as opposed to having the conflict between his reasons for A and his reasons for B resolved by a neural coin toss over which he has no control. In contrast, on Kane’s view since the agent exerts an effort to choose A prior to the choice of A and since the choice is indeterministically caused by this effort, it is the agent that makes the choice of A happen. And, of course, since he simultaneously makes an effort to choose B, then had he chosen B he’d have been responsible for that choice instead. Despite this problem with Nozick’s view, I want to use his notion of assigning weight to reasons in developing a new way to think about basic, underivatively free actions. To make this idea useful for our purposes, we need to reconceive of the timing of such “weightings.” Nozick says that at the moment of decision, in choosing one course of action over another, we assign greater value to the reasons in favor of that action (Nozick 1981, 296–297). But suppose we move the assigning of weights to reasons back prior to choice, and suppose that the weight we assign to the reasons is not causally determined. Suppose as well that once the weights are assigned to the reasons this then typically dictates in a determined way which choice gets made. For instance, suppose I am torn between doing A or B and have good reasons to A and good reasons to B, and, thus, I’m deliberating about which to do. Suppose that during deliberation I do in a causally undetermined way assign greater weight to the reasons for A—that is, I assign greater value to those reasons than the reasons for B—and this then leads me to choose A instead of B. Notice what’s going on here. On this model, my choice is the product of a causally indeterministic process of deliberation in which I actively assign more value to one set of reasons (the reasons for A) than the other set of reasons (the reasons for B). Then due to this assignment of weights to the

108 Alternative Models of Basic Free Actions reasons, I choose to A. Here the particular choice I make—the choice of A—will result from a causally indeterministic process in which I’m actively involved through the assigning of weight to the competing sets of reasons. Further, how I assign these weights then dictates the choice that gets made— the choice of A. Here the choice doesn’t happen to me; rather, it is a direct consequence of something I do—assigning a certain value to the reasons under consideration. On this modified Nozickian view, the agent’s active assigning of weights to his reasons plays the role of the efforts of will in Kane’s view. By assigning a greater value to the reasons for A and lesser weight to the reasons for B, I actively do something which makes me then choose A and not B. On this view, the choice of A doesn’t just happen to the agent; rather, just as on Kane’s view the agent makes it happen by his actively and intentionally assigning a greater weight to the reasons for A. The view I’m describing here has some kinship with the modest libertarian view I described in Part One of this chapter, but at the same time it differs from it in significant respects. The similarity is that, like the modest libertarian view, this view suggests that our underivatively free choices wouldn’t be undetermined right up to the moment of choice. On this modified Nozickian view, there’s indeterminism in the weighting of the reasons, the assigning of different amounts of value to the reasons, but after that the agent’s choice will typically be a determined outcome based on those assigned weights. The key difference between this modified Nozickian view and the modest libertarian view is that the latter view doesn’t include the crucial agential involvement in the deliberation process to make sense of agent control of and responsibility for the choice. Recall that on the modest libertarian view there is to be indeterminism in the coming to mind of certain ideas or reasons and then once all the relevant ideas or reasons have come to mind the process leading up to choice is causally determined. The problem with this is that we don’t control what ideas or reasons come to mind in deliberation. This just happens to us. And, as my earlier example of Billy making his choice of a college shows, we will have no way of making sense of how agents have control over and responsibility for such choices. In contrast, on the modified Nozickian view the indeterminism occurs in the assigning of weights to the reasons. But this is something the agent actively does; it is not something that happens to him. Thus, this modified Nozickian view avoids this crucial flaw in the modest libertarian position. A critic may say that on this model of basic underivatively free decisions it is not actually the choice which is the underivatively free action here; it is the assigning of weights to the reasons. It is the placing of greater value on one set of reasons, say, the reasons for doing A than the reasons for doing B, or vice versa. This is the real undetermined free act at issue, not the choice of A or B. For the latter will typically be a determined choice grounded on the prior undetermined assignment of relative values to the competing reasons. Further, one may want to know what it is about the assigning of these relative values to the competing reasons which makes them something the

Alternative Models of Basic Free Actions

109

agent does as opposed to something which happens to him. We might understand this weighting of reasons on a Kanean model. For instance, suppose the agent exerts competing efforts to (1) weight the reasons for A over B and (2) weight the reasons for B over A; and then one of these competing efforts ultimately dictates which set is weighted more heavily. If so, then this will be an account of the weighting process which Kane could accept as a free willed weighting. As such, a critic might think this view just gives a confused vision of SFAs, which once we get clear about the actual SFA involved—the weighting of the reasons—may be fully coherent with a Kanean view of things. To adequately answer this objection, we’ll need to give an account of how the agent can freely and responsibly weight the reasons without falling back into a model which relies on the dual efforts view. To do this we should view the assigning of the relative weights to the reasons as basic in a way that the dual efforts are basic on Kane’s view. Recall how Clarke and Mele argued that since Kane’s assassin freely tried to shoot the prime minister, our dual efforts in SFAs must be free; and they pushed Kane for an account of the freedom of these efforts. In response Kane argued that in SFAs it suffices for the freedom of the agent’s efforts if the agent meets plausible compatibilist standards of free action. In other words, in making her efforts the agent is not forced, coerced, subject to covert controllers, acts on desires endorsed by her second-order desires. According to Kane, the libertarian freedom and ultimately responsibility we exhibit in our SFAs is grounded on the fact that we possess some kind of compatibilist freedom and responsibility for our dual efforts leading up to undetermined choice. What I propose is that just as Kane says we meet compatibilist requirements in our dual efforts making us responsible for them, so too on this modified Nozickian view we meet compatibilist requirements in the undetermined assigning of weights to the reasons. Thus, when, say, the agent assigns greater weight, more value, to the reasons for A than the reasons for B and he meets compatibilist requirements in doing so—that is, he’s not forced, coerced, manipulated—then he assigns the weight to those reasons freely in some compatibilist sense and is responsible in some compatibilist way in doing so. By meeting these compatibilist requirements while assigning the weights, this ensures that it is the agent who is responsible for assigning the weights; she is not subject to the controlling will of another through force, manipulation, or coercion, and she is not overcome by foreign emotions that do not reflect her higher-order desires. In this way, even though it will be causally undetermined which weights are given to her reasons, it will still be true that whichever way she weights the reasons she will have control over the outcome, in the sense that she did the weighting on her own without being forced, coerced, manipulated, subject to compulsive desires. Finally, when she makes her choice based on an undetermined weighting which she did, her choice will then be a free choice for which she is ultimately responsible.

110 Alternative Models of Basic Free Actions Here we have an event-causal libertarian view of basic undetermined free choice without dual efforts involved which nonetheless has an appropriate level of agential involvement in the deliberative process to make sense of the agent control over and responsibility for the choice made. Here the choice made does not merely happen to the agent, for she is actively involved in the undetermined assigning of the weights to the reasons which ultimately dictate which choice she makes. A Kanean critic might want to have us reconsider what’s going on in the assigning of weights. Just because an agent meets compatibilist criteria in the assigning of weights, it doesn’t mean that she’ll have the proper kind of control over the particular assignment made to make her ultimately responsible for it. Suppose the agent is making a basic undetermined free choice between A or B. In doing so, suppose in some causally undetermined way she gives greater weight to the reasons for A and so she chooses A. Without dual efforts involved in the assignment of those weights, it’s too hard to see why the assignment will be a reflection of her will and not just a random consequence of causally indeterministic neural processes. Further, since the agent’s choice would then be subject to how those weights are assigned, it remains unclear how the agent’s choice is a product of her will as opposed to the random working out of neural processes over which she has no control. However, the modified Nozickian view has no more of a problem here than the Kanean view does. Recall that on Kane’s view it will be causally undetermined which of the competing efforts of will wins out and determines which choice gets made. Nonetheless, whichever choice gets made it will be a reflection of the agent’s will, since either choice, A or B, will be something she tried to do. Analogously, on the modified Nozickian view, while it will be causally undetermined which set of reasons is given the greater weight, the reasons for A or the reasons for B, it will be a reflection of the agent’s own will through the fact that she assigns the weights, while uncoerced, unforced, unmanipulated, reasons-responsive, and so forth. As such, whatever assignment of weights is made it will be a reflection of the agent’s own will, just as much as the agent’s undetermined choices in SFAs are a reflection of the agent’s will on the Kanean view. But the Kanean may say the choice of A or B will just be a determined consequence of the undetermined weighting. Thus, again, the Kanean may say that the actual SFA here is not the choice of A or B, it is the weighting. In the weighting process, the agent can either value A more than B or value B more than A. To establish the agent’s responsibility for the causally undetermined weighting it will not suffice to note that the agent meets compatibilist criteria in assigning these weights. For even though such criteria are met the weighting may still be a consequence of neural processes over which she has no control. Whereas on Kane’s view of SFAs in making a choice either way the agent will succeed at something she was antecedently trying to do, and, thus, she will be responsible either way.

Alternative Models of Basic Free Actions

111

To answer this potential Kanean criticism, it is important to establish that on this modified Nozickian view the SFA is not the weighting but it is the choice that follows upon the weighting. The choice will indeed typically be a causally determined consequence of the direction the weighting process goes. As such, this modified Nozickian view is significantly different from Kane’s view. Kane wants the SFA to be something which is itself causally undetermined and not the determined result of something else which was causally undetermined. Despite this difference, I would note that on the modified Nozickian view of SFAs, even though the choice made in the SFA is determined: (1) the agent still could have done otherwise in the sense that had she weighted the reasons differently she would have chosen otherwise; and (2) regardless of which choice is made it will be a reflection of the agent’s will, since either way it will be an expression of a causally undetermined weighting which she performed. Insofar as the weighting of the reasons is done while meeting compatibilist criteria—while unforced, uncoerced, unmanipulated, and while reasons-responsive, and so forth—it is the agent that does the weighting and as such the weighting is not a product of overwhelming urges, external manipulation, but a reflection of her own will, her own deep or rational self. Suppose the choice of A is made. Since on this view no dual efforts are involved, the agent’s choice of A will not be an instance of succeeding at what she was antecedently trying to do. But she will still be responsible for this choice as it will still be a reflection of her will in the sense that it results from a causally undetermined weighting of her reasons in which she was actively involved and in which the weighting was uncoerced, not manipulated, reflective of both her first- and second-order desires, reasonsresponsive. And, while her choice may be a determined result of the way in which she weighted the reasons, there remains a rich sense in which she could have chosen otherwise. For, as noted, she could have done otherwise had she weighted the reasons differently. I describe this as a “rich sense” of could have done otherwise, because on this view the weighting of the reasons is causally undetermined. As such, while the choice made is determined she really could have done otherwise had the weighting taken a different path. Compatibilist accounts of the could have done otherwise condition on free choices cannot give you this, as they must account for this condition in a manner that is consistent with a fully determined universe. Before concluding, I should note that some may find Kane’s view preferable on the ground that his view allows that in our SFAs we may be said to act freely and responsibly while our choice is undetermined right up to the moment of choice. It may be thought this view allows the agent greater control over his choice at the very moment of choice. On the model I’ve presented here, the agent will typically be determined to choose as he does by the time he makes his choice. It seems to me that this is a price which must be paid if we are to give an account of basic undetermined free choices without making any appeal to dual efforts. But, if this modified Nozickian

112 Alternative Models of Basic Free Actions model is coherent, it can be said to make up for this price by avoiding the difficulties which come from embracing the Kanean dual efforts model.

Notes 1. I propose making such a modification to Kane’s view in Lemos (2011b). 2. For more on the differences between Kanean SFAs and Balaguer’s torn decisions, see Balaguer (2010, 74). 3. In Balaguer (2004) and (2010), he focuses on a different version of the luck argument—the replay argument—maintaining (for similar reasons to the ones I’ve noted) that even if in 100 replays of her decision the agent chose differently (say 50 times one way and 50 times the other way), this would not imply the existence of control-robbing luck. I have spoken in terms of similar agents deciding in different possible worlds in keeping with my example of BW and BW*, who face the same decisions in different possible worlds. 4. See Balaguer (2004, 389). There, in discussing a different but similar example of a torn decision, he makes the same kind of point about agent control.

6

Moral Responsibility Denial and the Problem of Punishment

In Chapter Three of this book, I argued that while libertarianism seems to be immune to manipulation arguments, compatibilism does suffer from serious problems regarding manipulation cases. Given the issues I raise in Chapter Three as well as the very insightful criticisms of compatibilism which have already been made by Peter Van Inwagen (1983, 2002), Derk Pereboom (2001, 2014), Robert Kane (1996), and Bruce Waller (2011), I think we have good reason to reject compatibilist approaches to understanding free will and moral responsibility. Admittedly, the debate about the merits of compatibilism is ongoing and the literature on the subject is vast. Thus, to fully demonstrate why compatibilism should be rejected would require doing a lot more than I have done in this book. Indeed, it seems that task would require a book-length discussion all by itself. However, I do think the arguments which have been made by Van Inwagen and Pereboom and others are on the right track, and I think the points I make in Chapter Three add some new reasons to think compatibilism should be rejected. Thus, I am inclined to think that if we have the kind of free will which supports moral responsibility and deserved praise and blame and reward and punishment, then it must be some kind of libertarian free will. In the first half of this book, I’ve discussed at length what I regard as the most plausible kind of libertarian view on offer, Robert Kane’s naturalistic, eventcausal libertarian view. I’ve argued that his view has the resources to answer many of the criticisms against it and that it overcomes weaknesses in various alternative libertarian views. I’ve also noted that there are some lingering problems with the Kanean view which need to be resolved to make it more plausible, and in Chapter Five I suggested some alternative ways to think about SFAs which I think might help remedy these problems. The defense of libertarianism which I’ve provided in the first half of this book has focused on the coherence of the view. That is, I’ve tried to show that there is a coherent naturalistic, event-causal libertarian view of free will which can make sense of all that we can reasonably want from our conceptions of free will and moral responsibility. In focusing on these issues, I have left aside the issue of whether or not such libertarian free will exists. In the end, like Kane, I do not think the existence of libertarian free will can

114

Responsibility Denial and Punishment

be proved through introspection or intuition. Nor do I think we have the empirical scientific evidence or the metaphysical arguments to show that it exists. At the same time, I don’t think there are sufficient evidentiary reasons to reject the claim that we lack libertarian free will. Some might think the scientific evidence weighs convincingly against the idea that any of our decisions are causally undetermined in such a way that allows for free will. But such is not the case. For an excellent discussion of this matter, see Balaguer (2010, Ch.4). Consider also the remarks by Robert Kane (2016a) on this matter. It is of course, an empirical and scientific question whether any indeterminism is there in the brain in ways appropriate for free will. No purely philosophical theory can settle the matter. It is interesting, however, that in the past decade there has been more openness and discussion on the part of some scientists and philosophers about this possibility. See, e.g. Bishop (2011), Baker and Gollub (1990), Hilborn (2001), Hobbs (1991), Kellert (1993), Balaguer (2010), Heisenberg (2013), Glimcher (2005), Maye et al. (2007), Hameroff and Penrose (1996), Shadlen (2014), Brembs (2010), Stapp (2007), Doyle (2011), Tse (2013), Jedlicka (2014), and Briegel and Mueller (2015). (Kane 2016a) Like Kane and Balaguer, my view is that it’s an open question whether or not we possess libertarian free will. Maybe we don’t have it. Who knows? Having said all of this, though, notice the situation here. I’ve said we should reject compatibilism. Thus, if we are to defend free will, it should be a free will of the libertarian sort. But, along with Kane and Balaguer, I think we must admit that there’s no sufficient scientific, metaphysical, introspective, or intuitive evidence for or against the existence of the kind of libertarian free will which is necessary to moral responsibility. This raises a kind of existential question—namely, how should we live in the face of this situation? When I or others do good things, should I think of myself or others as deserving of praise or reward? When I or others do bad things, should I think of myself or others as deserving of blame or punishment? Blame and punishment for wrongdoing are only appropriate when directed at targets who are morally responsible for their wrongdoing. It would seem that if I don’t know whether persons have the kind of free will which is necessary for moral responsibility, then I should not engage in blaming and punishing myself or others for their wrongdoing. The question that concerns me here might be put more generally as follows: should we abstain from living, believing, and acting as though people have the kind of free will which supports moral responsibility until we have sufficient scientific or metaphysical or some other theoretical evidence that it exists? Or might there be good moral or pragmatic reasons to continue living and acting as if or believing that people have such free will? It is my

Responsibility Denial and Punishment

115

view that despite the lack of theoretical—that is, scientific, metaphysical, introspective, or intuitive—evidence for the existence of libertarian free will, there are nonetheless good moral reasons to continue living and acting as if and believing that we have it. It is morally problematic to deny that we have the kind of libertarian free will which is the foundation of moral responsibility, as doing so undermines sufficient respect for human rights, human dignity, and the concept of moral obligation. While some kind of ethics could be maintained if one denies that we are morally responsible, it will be an impoverished ethics. In this and subsequent chapters, I will defend the view that the belief in libertarian free will and the moral responsibility it supports can be justified on moral grounds. I will start in this chapter by critically examining the views of some contemporary philosophers who believe that (a) we lack the kind of free will which makes us morally responsible and (b) we are morally better off rejecting the notion of moral responsibility in the just desert sense. I will argue that such a view runs into significant problems regarding criminal justice by posing a threat to the rights of innocent persons who have committed no crimes.

Moral Responsibility Skepticism and Criminal Detention/Quarantine Some contemporary philosophers claim that human beings lack the kind of free will required for moral responsibility in the basic desert sense. Derk Pereboom (2001, 2014) is one of these philosophers. In a recent book, he explains moral responsibility in the basic desert sense as follows: [F]or an agent to be morally responsible for an action in [the basic desert] sense is for it to be hers in such a way that she would deserve to be blamed if she understood that it was morally wrong, and she would deserve to be praised if she understood that it was morally exemplary. The desert here is basic in the sense that the agent would deserve to be blamed or praised just because she has performed the action, given an understanding of its moral status, and not, for example, merely by virtue of consequentialist or contractualist considerations. (Pereboom 2014, 127) Some other philosophers who say we lack this kind of responsibility are Gregg Caruso (2016; with Pereboom forthcoming) and Bruce Waller (2011, 2015).1 Holding such a view raises a number of puzzles and problems in the practical arena. For instance, suppose I often knowingly engage in many morally exemplary acts, sacrificing for the welfare of others. Should I take pride in my moral accomplishments? If I don’t have the kind of free will that makes me responsible for these actions, then I don’t deserve praise for these

116

Responsibility Denial and Punishment

actions. And if I don’t deserve praise for these actions, then it is unclear that I should take pride in the fact that I have done them. Rather, it would seem that I am just lucky to have been subjected to the genetic and environmental conditions that have made me the morally exemplary sort of person that I am. Or suppose that I am a very conscientious worker and more conscientious than my fellow employees and because of this I get more of my work done than my fellow employees. Then do I deserve better pay than my fellow workers? If no one is morally responsible in the basic desert sense, then one cannot justify giving me higher pay solely on the grounds of desert. But, perhaps, higher pay for the more hard working and more successful workers could still be justified on the grounds that it provides incentive for workers to work harder at their jobs. While the practical puzzles and problems raised by this view on responsibility are many, I want to focus here on the problems this view has with the treatment of those convicted of criminal offenses. If no one is morally responsible in the basic desert sense, then what should we do with those who have been convicted of committing crimes? It could be argued that the penal system in the United States should continue to exist and operate as it does today even if no one is morally responsible in such a way as to deserve blame and punishment for what they do. It could be argued that the state should continue to mete out harsh prison sentences and even death penalties in some cases, as providing these harsh penalties and detaining or killing offenders prevents the convicted from committing more crimes and discourages many others from committing crimes. However, Caruso, Pereboom, and Waller don’t believe this. Rather, they contend that insofar as no one is morally responsible in the basic desert sense, then there should be significant reform of the penal system. These philosophers don’t believe that the harsh treatment of convicted criminals can be justified on grounds of deterrence. They think we could reduce and prevent crime using much less harsh measures. On their view, harsh measures, such as those used in the United States, could be justified only if criminals deserved to suffer in the basic desert sense for their crimes. Since they don’t believe we have the relevant kind of moral responsibility, they don’t believe that the harsh treatment of criminals can be justified.2 What they recommend we do with those convicted of criminal offenses is very different from what we currently do with criminals in the United States. They are in agreement that many people in the United States who are in prison should not be in prison. Prison or detention centers should be reserved for only the most violent and dangerous criminals. There are drug users, drug dealers, thieves, prostitutes, and others who are not violent and who are doing time in prison. Caruso, Pereboom, and Waller believe such persons should not be incarcerated or detained for such crimes, unless it can be shown that they are a significant threat to the health and safety of others. Furthermore, on their view no one is morally responsible in the basic desert sense and so no one really deserves to be sent to prison or detained,

Responsibility Denial and Punishment

117

but they acknowledge that for the sake of protecting society at large some violent criminals should nonetheless be held in detention centers. On their view, drug dealers, thieves, prostitutes, white collar criminals, should still be regarded as criminals who’ve engaged in illegal activities, but they should not be held in detention centers. Rather, they should be monitored, offered counseling, be expected to pay restitution, engage in community service, but the detention and isolation of such criminals is unwarranted if they pose no immediate threat to the health and safety of others. Pereboom and Caruso explicitly justify detention of violent criminals by analogy to the practice of quarantining people with dangerous communicable diseases, such as Ebola. Just as we think it is morally acceptable to quarantine certain individuals who through no fault of their own have contracted dangerous communicable diseases, so too it is morally acceptable to quarantine violent criminals who through no fault of their own pose a threat to the rest of society. While Waller doesn’t explicitly appeal to the quarantine analogy, he thinks violent criminals should be locked up for similar reasons. Further, all three of these philosophers believe there is no justification for making the conditions of detained criminals harsh. Rather, they think they should be detained and offered many opportunities to help them become better, more well-functioning, and law-abiding members of society. They view harsh prison conditions as in conflict with the goal of rehabilitation and indefensible on retributive grounds since no one is morally responsible in the basic desert sense for their deeds. In Pereboom (2001, 2014), he makes it clear that if no one is morally responsible in the basic desert sense then there is no way to provide a retributive justification for the imprisonment/detention of violent criminals. The idea here is that if no one is responsible in this sense, then no one deserves to be imprisoned or detained simply because they deserve it; and since retributive theories of punishment justify such treatment in terms of just desert, it follows that there can be no retributive justification for imprisonment/ detention if no one is morally responsible in the basic desert sense. Pereboom goes on to note that there are several different nonretributive ways the responsibility denier can defend the imprisonment/detention of violent criminals. He explains the difference between several different theories of punishment, such as the moral education theory, the utilitarian deterrence theory, and the right to self-defense deterrence theory; and he gives reasons to reject each of these approaches. Instead, he adopts the quarantine model. Like Pereboom and Waller, Gregg Caruso also believes human beings lack moral responsibility in the basic desert sense and he too embraces the quarantine model as the best way of dealing with violent criminals. In Caruso (2016), he develops the quarantine model in some different ways than Pereboom, linking it to a public health/medical ethics framework.3 Responsibility deniers, such as Pereboom, Caruso, and Waller, frequently claim that their view does not have as many practical problems as critics

118

Responsibility Denial and Punishment

suggest. Indeed, these three particular figures believe that by denying the existence of responsibility we actually will be able to provide a more humane and effective way of dealing with criminals and reducing crime. However, I will show that we have good reason to be very concerned about the ethics of any approach to criminal detention that does not rest upon some conception of moral responsibility in the basic desert sense. I will argue that the views of criminal detention expressed by Caruso, Pereboom, and Waller suffer from a problem of detaining those who have committed no crimes. In particular, I will argue that on their assumption that no one is ever responsible in the basic desert sense then they will have no moral basis for objecting to the use of a certain sort of policy and practice which would knowingly lead to the conviction and sentencing of more innocent people who have committed no crimes.4 In developing my argument, I will appeal to an argument that Saul Smilansky (1990) has made against utilitarian deterrence theories of punishment. As noted Derk Pereboom does himself reject such theories and instead endorses the quarantine model. But I will show that the quarantine model also suffers from the problem raised in Smilansky’s argument. In recent years Smilansky (2011) has offered direct criticism of Pereboom’s quarantine model by arguing that it leads to the problem of “funishment,” which he regards as a practical reductio of Pereboom’s view. The issue he raises is that since no one deserves criminal detention, those subjected to it would have to be made so comfortable as compensation for their detention that there would be little to no disincentive to engage in crime. While this is an interesting argument and worthy of serious discussion, it shall not be my concern in what follows.5

Smilansky and the Problem of Punishing the Innocent In Smilansky (1990) he argues that any utilitarian deterrence theory of punishment which bases the justification of punishment solely in term of its propensity to prevent crime and not on the desert of the criminal faces a serious problem regarding the punishment of the innocent. The problem is that if the only reason to punish is to prevent crime, then there is no reason why those punished must actually be guilty of having committed a crime. In some cases, the goal of preventing crime can be served just as well, if not better, by punishing an innocent person who has committed no crimes. As Smilansky notes, this kind of criticism is often defended with fanciful examples in which to prevent riots an innocent person is singled out and framed for a crime he did not commit and then punished. Friends of the utilitarian deterrence theory often counter this move with the assertion that the examples are indeed fanciful and in real life one would hardly ever encounter a scenario where framing and punishing an innocent person for a crime is actually defensible on utilitarian grounds.6 I would note here as well that in his classic treatment of the issue John Rawls (1969) makes a rule-utilitarian

Responsibility Denial and Punishment

119

argument that it would never be beneficial for a society to have rules allowing for the framing and punishment of innocent people, citing the general anxiety among the population that this would evoke if the citizens knew the state was authorized to act in this fashion. However, in defense of the problem of punishing the innocent Smilansky finds a way to build the argument without the use of fanciful examples. He argues that we would convict and punish more criminals if we lowered the standards used for evidence in criminal trials (1990, 258–259). For instance, in the United States to be convicted of a criminal offense the accused must be shown to be guilty beyond a reasonable doubt. This is a very high evidentiary standard. Due to the use of this standard many accused persons who are actually guilty escape conviction and punishment. At the same time due to the use of this high evidentiary standard fewer innocent people are punished. Smilansky argues that it would likely lead to a greater decrease in crime if the evidentiary standards were lowered to, say, a simple preponderance of the evidence as used in civil cases. According to Smilansky, the problem this suggestion presents for utilitarian deterrence theories is that such a relaxation in evidentiary standards would likely lead to the following consequences: 1. Significantly fewer criminals would escape conviction, and (assuming the utility of punishment) this would benefit the vast majority of people: directly and indirectly it would reduce crime, the fear of crime, insurance premiums, policing and security costs, etc. 2. There would not be a significant increase in the arbitrary power of people in positions of authority. 3. General respect for the law and the judicial system need not diminish (even if the details of the procedures would be realized): as the judicial system will be significantly more effective, possibly just the reverse. 4. Most innocent people would still not be at any sort of probable risk of being ‘punished’. 5. Most people’s chances of being wrongly ‘punished’ might be negligible, compared to their chance of being severely hurt by a criminal who got away because of the previous stringent procedures. 6. More innocent people, however, are very likely to be ‘punished’. (Smilansky 1990, 259) Here Smilansky shows in convincing fashion that if the only reason to punish is to improve the general welfare of society by reducing crime, then this would justify the state in adopting policies which it knows would lead to the conviction and punishment of more innocent people who have committed no crimes. Interestingly, the argument here does not involve an appeal to fanciful examples. It also gives a way around the problem posed by Rawls’s rule-utilitarian response to the problem of punishing the innocent, as there doesn’t seem to be any reason to think that slightly lowering the evidentiary

120

Responsibility Denial and Punishment

standards for criminal conviction would have the problematic utilitarian consequences which Rawls finds in allowing for the occasional framing and punishing of innocent people. Smilansky takes it as obvious that punishing innocent people who do not deserve it is inherently wrong. Thus, he thinks we have strong prima facie reasons to avoid enacting policies that increase the likelihood that innocent people shall be punished. But, as he demonstrates, utilitarian deterrence theories, which seek to justify punishment solely on the grounds of its benefit to society without attending to the demands of just desert, have no ground for acknowledging the inherent wrongness of punishing the innocent. Thus, such theories may all too easily justify actions which lead to the punishment of more innocent people.7

Extending Smilansky’s Argument to the Quarantine Model Now Smilansky’s argument is directed against utilitarian deterrence theories of punishment, but, as I note above, Pereboom himself rejects such theories of punishment, acknowledging their tendency to give way to the problem of punishing the innocent and the use of overly harsh punishments (2001, 166–167). Instead, he and Caruso adopt the quarantine model, and Waller seems to hold a similar view even if he doesn’t call it such. In what follows, I will show that the same basic argument which Smilansky makes against utilitarian deterrence theories of punishment can be extended to the quarantine model of criminal detention, presenting a serious problem for such theories. Pereboom and company would likely balk at such a suggestion. They might say that in matters of public health we detain only those who are known to carry dangerous infectious diseases. We hold them and treat them, and we release them back into society only after they are no longer infectious. So, too, with criminal detention it would only be acceptable to detain known criminals. However, such a defense of the quarantine model would be misguided. It is not true that in matters of public health we detain only those known to carry infectious disease. Rather, we also detain those who are likely carriers of the disease, as sometimes people may have a disease and spread it before showing symptoms. The U.S. Centers for Disease Control (CDC) actually distinguishes between isolation and quarantine. Isolation separates sick people with a contagious disease from people who are not sick. Quarantine separates and restricts the movement of people who were exposed to a contagious disease to see if they become sick. Those quarantined are not necessarily carriers of the disease; they are people who’ve been exposed and are likely carriers. Furthermore, federal and state laws allow for legally enforced quarantine of people who were merely exposed and who may not actually be carriers. In most states in the United States breaking a quarantine order is a criminal misdemeanor, and breaking federal quarantine orders is punishable by fines and imprisonment.8

Responsibility Denial and Punishment

121

Thus, the quarantine model does not actually discourage quarantine of the innocent, rather it allows for it as long as those detained are likely to be criminals. Notice that this plays right into the hands of the kind of argument strategy employed by Smilansky (1990). He argues that if no one is responsible in the basic desert sense and if the only point of punishment is to prevent crime then we might as well lower the evidentiary standards to get more criminals off the streets. This is akin to saying that we should detain not only those who are known to be violent criminals but also those who are likely to be violent criminals.9 To be clear, my point is that when responsibility deniers, like Caruso and Pereboom and Waller, endorse the detention of criminals on the model of the practice of public health quarantine they run right into the problem of punishing (or, perhaps, detaining) the innocent, which Smilansky has so forcefully raised against utilitarian deterrence theories of punishment. Quarantine is motivated primarily to serve the protection of the broader society. Those detained need not actually be carriers of illness—they only need to be persons who’ve been exposed and are likely carriers. If we detain criminals on similar grounds, then the point of their detention is not to give those detained their just deserts. Rather, the point is to protect the rest of the population. Thus, according to the quarantine model, there is no good reason why we should detain only known criminals; rather, it would make sense to relax the standards for criminal conviction so as to cast our net a bit more broadly convicting and detaining all those who most likely did commit violent crimes. Again, the problem here is that this would also lead to detaining more people who are innocent and have committed no crimes. The quarantine model does not exhibit enough respect for the rights of innocent people who’ve committed no crimes and who don’t deserve to be detained. Pereboom is aware of a problem very similar to the one I am raising. He writes: If justification of detention by this analogy is tenable, must it not then be legitimate to detain those who have not committed a violent crime, if by some means it has been ascertained that they are quite likely to do so? (2001, 176) In response to this problem, he states: To avoid this problem, it seems that invasive preventative measures should be restricted to those who have committed crimes. The right to liberty should count heavily here. This right would yield a strong reason not to detain someone even if there were some reason to believe that he is likely to commit a crime. (2001, 176–177)

122

Responsibility Denial and Punishment

Pereboom believes we need to be concerned about the right to liberty of persons. Infringement on such liberty is not to be taken lightly. Thus, he thinks detention for the sake of crime prevention should be restricted only to those who have committed crimes. It is likely that Pereboom would object to my argument that fans of the quarantine model have no recourse to block the use of lower evidentiary standards in determining criminal guilt. He would argue that concern for the right to liberty of persons supports the continued use of high evidentiary standards, such as evidence beyond a reasonable doubt. However, such a response to my argument would be problematic. First, while the right to liberty of persons should be protected, this does not necessarily mean that we should not lower the evidentiary standards for criminal conviction. As Smilansky argues, a lot more criminals would be taken off of the streets using a lower evidentiary standard. And, while more innocent people would be convicted of crimes and detained, it would still not be that many more. By taking more criminals off of the street we would actually be doing more to enhance and protect the liberty of persons, as a big part of our liberty involves the freedom from rape, murder, theft, fraud, and so forth. Thus, a responsibility denier who favors the quarantine model doesn’t really have good reason not to support lowering the evidentiary standards for criminal convictions. Second, recall that (1) the purpose of criminal detention on the quarantine model is to protect society from harm caused by criminals and, as I’ve argued, (2) we could serve this goal better by lowering the evidentiary standards for criminal convictions, allowing us to detain more criminals. Now, what I want to suggest is that were responsibility deniers, like Pereboom and Caruso and Waller, to say that we should not lower the evidentiary standards to protect the freedom of those who have not committed crimes, this would amount to a kind of unjustified discrimination against criminals. On the responsibility denier’s view, no one is responsible for what he does. On this view, the person who commits crimes is no more deserving of punishment than the person who never commits crimes. As I’ve argued, we could reduce crime more effectively by using lower evidentiary standards for criminal conviction and this would, of course, lead to the punishment of more people who have not committed crimes. However, if the responsibility denier is correct in saying that no one is responsible, then, given the purpose of quarantine style criminal detention, it is unjust to expect only those who have actually committed crimes to be detained. Why should the criminals be the only ones to carry the burden of criminal detention for the sake of crime prevention if they are no more deserving of the detention than innocent people? To refuse to lower the evidentiary bar for criminal conviction when (a) the goal is to protect society from crime and (b) no one really deserves detention is to unfairly single out the criminal population when they are not responsible for their criminal tendencies. In Pereboom and Caruso (forthcoming), they offer a couple of replies to my critique of their view. First, they contend that they do not justify the

Responsibility Denial and Punishment

123

use of the quarantine model on strict consequentialist grounds. Rather, they justify it on the grounds of the right of self-defense and the defense of others. On this view, we have the right to defend ourselves and others against those who pose a threat to our health and safety. Lowering the standards of evidence for criminal convictions would increase the frequency with which innocent people are convicted and punished. Innocent people are not a threat to our health and safety and, thus, we are not within our right to endorse lowering the evidentiary standards in this way. However, such a defense of their view is problematic. Recall that Pereboom and Caruso believe the quarantining of violent criminals is justified on the same grounds as quarantining to prevent the spread of dangerous infectious disease. Presumably, they believe the right to self-defense and the defense of others justifies quarantining those who pose a threat of spreading infectious disease. That seems right. However, notice, again, that this does not just allow for detaining those who actually are sick with the infectious disease; rather, it also allows for detaining those who are likely carriers of the disease. Since the right to self-defense justifies quarantining those who are not actually a threat to spreading disease but are mere likely threats, the emphasis on this right of self-defense model does not establish why the standards of evidence for criminal conviction should not be lowered. Second, Pereboom and Caruso also argue that the quarantine of violent criminals is not motivated simply to protect society from them. Rather, it also serves the goal of the moral reform of those detained and it allows for the possibility of reconciliation with those they have harmed. Since innocent people who have committed no crimes do not need this moral education and do not need to be reconciled with those who’ve been harmed, we should not lower the standards of evidence for criminal conviction. This response is also problematic. While it’s true that Pereboom and Caruso believe that those criminals subject to quarantine should be offered moral education and therapy and opportunities for reconciliation, this is not the primary motivation for their being quarantined. Recall that they believe it is only the violent criminals who pose a threat to others who should be quarantined. On their view, nonviolent criminals should not be quarantined; they should just be offered moral education and therapies so as to help them change their ways. However, as I’ve just argued, the right of self-protection and the protection of others justifies the detention not only of those who are actual threats to others but also those who are likely threats to others. Thus, it would still make sense to lower the evidentiary bar for criminal convictions, even though the moral education and therapies offered would be wasted on those innocent people who don’t need it. Now it might be thought that this waste of time with moral education and therapy for those who don’t need it is to be avoided and so we should not lower the bar for criminal convictions so as to avoid this waste. But this argument is problematic too, when we keep in mind that Pereboom and Caruso invoke a public health model in defending their approach. After all it is commonplace for doctors to give

124

Responsibility Denial and Punishment

medicines as a preventative measure in many cases. That is, sometimes we don’t know if persons are sick with a contagious disease but we give them medicines anyway to protect them and others from getting sick. So where has the argument taken us here? I have tried to show that when responsibility deniers, like Pereboom, Caruso, and Waller, endorse a quarantine model for the detention of dangerous criminals, they are adopting a position which has no basis for resisting the use of lower standards of evidence for conviction on criminal charges. Thus, they hold a position that has no basis for resisting the adoption of a policy which would lead to the detention of more innocent people who have committed no crimes. This is a disturbing consequence of their view, as it seems obvious that there is an inherent wrongness to punishing innocent people that makes doing so prima facie wrong, yet the quarantine model does not have the resources to acknowledge this and, thus, too easily justifies the use of policies that would increase the frequency with which innocent persons are detained. I would also note that Pereboom and Caruso believe that when it comes to dealing with dangerous criminals the quarantine model is the best theory available to responsibility deniers. Waller, who does not explicitly endorse the quarantine model, does nonetheless articulate a view that is very similar, if not identical, to it, and he too seems to think this is the best approach for dealing with dangerous criminals. Pereboom rightly notes that retributivism is incompatible with responsibility denial. And, while he acknowledges that there are other theories of punishment open to responsibility deniers, such as utilitarian deterrence theories and moral education theories, he contends that these other views are inferior to the quarantine model. Assuming he is correct in claiming that the quarantine model is the best approach for the responsibility denier, then my argument shows that responsibility deniers are saddled with a view about which we should have serious moral reservations.

Waller on Responsibility and the Punishing of Innocent People I’ve argued that responsibility deniers are saddled with a view that has no rational means for blocking the use of policies that would lead to the detention of more innocent people. Thus, such a position is likely to lead to the detention of more innocent people for crimes that they did not commit. However, in recent publications Bruce Waller (2011, 2015) confronts this suggestion head on, arguing that it is false. In doing so, he defends both a relevant weak thesis as well as a relevant strong thesis. The weak thesis states that belief in moral responsibility does not discourage the punishment of the innocent; the strong thesis states that belief in moral responsibility actually contributes to the punishment of innocent people. In Waller (2015), he argues that the more a culture is committed to individual moral responsibility the more obsessed with punishing it becomes and the less concerned with protection of the innocent. Indeed, he contends

Responsibility Denial and Punishment

125

that belief in moral responsibility is the wrong tool to use in protecting the innocent, “It’s not like using a sledgehammer to drive a nail; it’s more like using gasoline to put out a fire” (2015, 223). Here Waller makes his strong thesis manifest; he regards the belief in individual moral responsibility as actively contributing to the punishment of innocent people. In Waller (2015), he notes that in the United States various legal and police tactics are commonly used which are known to be unreliable. Due to the use of these tactics more innocent people are convicted and punished for crimes they did not commit, yet despite this the government does nothing to stop the use of these tactics. Among these tactics are the use of the testimony of jailhouse informants; the use of faulty crime lab techniques; the use of often overworked and sometimes incompetent public defenders; the use of plea bargains; and the suspension of the right of habeas corpus in dealing with suspected terrorists in fighting the “war on terror” (2015, 224–228). Waller notes that the United States is a country in which it is widely believed that people are morally responsible for their actions and yet in the United States these practices continue to be used while it is known that they lead to the conviction and punishment of many innocent people. Thus, he concludes that belief in individual moral responsibility does not encourage protection of the innocent from punishment; rather, it actually promotes the punishment of the innocent. In further support of his position, Waller talks about how the belief in individual moral responsibility makes us believe that criminals deserve to be punished and it makes us eager to see them punished. When a crime is committed, there is a cry for “justice”; that is, a cry for the punishment of the person who committed the crime, and who clearly—because morally responsible—justly deserves to pay the price. The more heinous the crime, the stronger the demand for just deserts. That powerful demand overwhelms concern about whether the person charged is guilty. (Waller 2015, 228) According to Waller, the belief in moral responsibility leads to belief in just desert and this makes us eager to see criminals punished. However, he also believes this eagerness often “overwhelms concern about whether the person charged is guilty.” In this way, he thinks the belief in moral responsibility actually promotes punishing the innocent. As noted above, Waller is defending both the strong thesis that belief in individual responsibility actually contributes to the punishment of the innocent and the weaker thesis that the belief in moral responsibility does not discourage punishment of the innocent. In what follows, I will show that neither of these theses is well supported by Waller. Concerning the stronger thesis that the belief in moral responsibility actually contributes to the punishment of the innocent, I would note that Waller

126

Responsibility Denial and Punishment

provides no significant evidence for this. He is quite correct that in the United States there is widespread belief in individual responsibility; and, sadly, in the United States the problematic use of jailhouse informants is widespread, as is the use of faulty crime lab techniques and the use of overworked and sometimes incompetent public defenders. It is also true that plea bargains are commonly used and habeas corpus rights have been suspended in dealing with suspected terrorists. It is true as well that these practices are known to contribute to the conviction and punishment of innocent people and the U.S. government has made little, or no, effort to change these things. However, this simply is not evidence that the belief in individual moral responsibility contributes to the punishment of the innocent. These points do nothing to establish a causal connection between belief in moral responsibility and the use of these practices which lead to the conviction and punishment of many innocent people. All that we really have here is evidence that in the United States where belief in individual moral responsibility is common there is also widespread use of legal practices and crime detection techniques which increase the frequency with which innocent people are convicted and punished. Such a correlation does not establish causation. The closest Waller comes to making the case that in the United States belief in moral responsibility contributes to punishing of the innocent is when he says, “When a crime is committed there is a cry for ‘justice’; that is, a cry for the punishment of the person who committed the crime, and who clearly—because morally responsible—justly deserves to pay the price” (2015, 228). Here, instead of establishing the mere correlation between the prevalence of belief in moral responsibility and the punishing of the innocent, Waller suggests how he thinks the belief in moral responsibility actually contributes to the punishing of the innocent. The basic idea is that the belief in moral responsibility makes us so strongly desire to see that justice be done by punishing the guilty that we fail to pay sufficient attention to whether those punished are in fact guilty. The belief in moral responsibility grounds the belief that those who commit crimes should be made to suffer punishment and the latter belief fuels the desire to see that the guilty be punished. Waller adds the “powerful demand [for just deserts] overwhelms concern about whether the person charged is guilty” (2015, 228). Here, indeed, there is a causal story being told and not mere clarification of disturbing correlations. And Waller is clearly correct to assert that belief in moral responsibility leads to the belief that those guilty of crimes should be made to suffer punishment and this leads to the desire to see the guilty punished. However, this by itself does not establish that the belief in moral responsibility contributes to the punishment of innocent people. To get there one has to add in another factor as Waller does when he asserts that the demand for just deserts overwhelms concern that the charged is guilty. Now it is true that sometimes people are so eager to see that someone get punished for a crime that they no longer sufficiently attend to or care whether the person punished is in fact guilty. But when this happens people

Responsibility Denial and Punishment

127

are not led by the desire that the person who committed the crime be punished; rather, as stated, they are led by the desire to see that someone be punished. Strictly speaking, it is not the belief in moral responsibility and the consequent desire that those who commit crimes get their just deserts which leads to the punishment of the innocent. Rather, it is the irrational urge to punish someone whether guilty or not that leads to this. Nonetheless it still might be thought that what triggers this irrational urge to see that someone whether guilty or not be punished is ultimately the belief in moral responsibility. The latter belief leads to the belief that someone deserves to pay for his crime. In this way it could be said that the belief in moral responsibility causally contributes to the punishment of the innocent. However, to argue in this fashion is highly problematic. Perhaps it’s true that were it not for the belief in moral responsibility then no one would feel that someone deserves to pay for his crime. But to then say that this belief encourages the punishment of the innocent doesn’t follow. After all, one can believe in moral responsibility and remain steadfast in his commitment to seeing to it that only those who are clearly guilty of committing crimes should be punished. So, the belief in moral responsibility is clearly not sufficient for the urge to punish the innocent. At best the belief in moral responsibility may be one of the background beliefs that leads some people to end up punishing innocent people. But this is a very weak claim. Suppose a crime, say a murder, has been committed in my community. Members of the community and I believe that people are morally responsible for their actions and so we believe that the guilty party should be brought to justice. Also suppose that I am the sheriff in this community and I am in charge of finding the guilty party. But also suppose that I am horrible at examining evidence and making proper inferences from evidence. Consequently, I mistakenly charge an innocent man with the crime. Also, suppose that the jury is racist and lazy and due to these factors they convict and sentence this man to life in prison. Here the belief that people are morally responsible does play some role in the conviction and sentencing of an innocent person, but its causal role in this is very minimal. For these other factors, such as my incompetence as a sheriff and the jury’s laziness and racism, are crucial elements that must be appealed to in providing a sufficient causal explanation for why this innocent man is punished. It is not at all clear that the belief in moral responsibility triggers the irrational desire that someone—whether guilty or innocent—must be punished. Interestingly, in his book (2015), Waller himself puts his own finger on a relevant rival account of how this irrational urge might arise. In Chapter Three, “The Strike-Back Roots of Moral Responsibility,” he talks about how human beings have an urge to strike at others when they have been wronged or threatened in some way; and he says there is also empirical evidence of this in other animals. He notes how oftentimes in human beings this urge to strike back when wronged or threatened is directed at those who’ve wronged us or threatened us, but just as often it is directed at others who

128

Responsibility Denial and Punishment

pose no threat or who’ve done no wrong. He notes how this tendency has been identified in nonhuman animals as well, and he talks about how being wronged or threatened creates damaging stress for us that we alleviate by lashing out at others whether they deserve it or not. Contra Waller, what I want to suggest here is that the tendency of human beings to punish the innocent is more likely a product of this irrational urge to strike back which we share with nonhuman animals than it is a product of belief in moral responsibility. When someone in our community has been harmed through crime, this creates stress in us as it does among other communal animals. Suppose we cannot know for sure who has committed the wrong. Even so, the stress is there along with the urge to strike back. The urge is strong, and we want to satisfy it so strongly that we are willing to convict and punish someone without sufficient evidence of their guilt. Notice that here the belief in moral responsibility doesn’t drive the punishment of the innocent. Rather, it is the animal urge to strike back whether those we strike are guilty or not. As noted Waller’s strong thesis is that the belief in moral responsibility contributes to the punishment of the innocent. But much of what he says just shows that in the United States where belief in moral responsibility is common the innocent are frequently punished and in the United States many legal and police tactics known to contribute to the punishment of the innocent persist. None of this shows that the belief in moral responsibility contributes to punishing the innocent. When Waller does get around to forwarding a causal hypothesis on the issue, he suggests that the strong desire to see that someone be made to suffer for his wrongdoing overwhelms the careful consideration of whether the accused is guilty. In this way, he thinks the belief in moral responsibility contributes to the punishment of the innocent. However, as I’ve shown not even this makes a convincing case that the belief in moral responsibility contributes to punishing the innocent. Rather, as Waller himself argues, there is a common and powerful urge in human beings and even in many nonhuman species to strike back or lash out when one is wronged or threatened. It may not be the belief in moral responsibility that leads to the punishment of the innocent, rather it may be this irrational urge which we share with nonhuman animal species. This rival hypothesis is just as plausible as Waller’s hypothesis. Thus, Waller gives us no good reason to think that belief in moral responsibility contributes to the punishment of the innocent. Interestingly, I would note that if my rival hypothesis is correct then instead of contributing to the punishing of the innocent the belief in moral responsibility actually helps discourage the punishing of the innocent. For suppose we do have this common and strong urge to strike back or lash out when harmed or threatened; and we are inclined to do this as it helps relieve our stress. As Waller himself notes, many humans as well as many animals will lash out at the innocent to relieve their stress. What I want to suggest is that if all of this is correct, then, rather than contributing to the

Responsibility Denial and Punishment

129

punishing of the innocent, the belief in moral responsibility and recognizing it as the ground for just desert and justified punishment actually discourages the punishing of the innocent. The belief in moral responsibility and the recognition that it is the basis for just desert may actually be what keeps many of us from lashing out at others in inappropriate ways. Imagine that I’ve been treated badly at work by my boss. I am stressed and I’d like to relieve this stress by striking back at him, but I know that would only make matters worse. I come home stressed and my wife and children annoy me, and I want to lash out at them. Perhaps what keeps me from doing so is my recognition that they are not morally responsible for any wrongdoing and, thus, deserve no harsh treatment. Or imagine that I am a member of a jury dealing with a trial in the case of a terrible rape or murder, and I know that I and members of the jury and the broader community would love to see someone punished for this. But remembering that people are morally responsible for their deeds and that only those morally responsible for bad deeds deserve to be punished, I am hesitant to convict the accused as the evidence of his guilt is insufficient. Contra Waller, there is some pretty good reason to think that the belief in moral responsibility actually plays a significant role in discouraging the punishment of the innocent. Let me remind the reader that Waller claims not only (1) that belief in moral responsibility does not discourage punishment of the innocent, but also (2) that belief in moral responsibility actually encourages the punishment of the innocent. Much of what I have said in this section has been intended to undermine his case for (2), the stronger thesis; but the preceding paragraph contributes to the case against (1)—his weaker thesis. In the preceding paragraph I suggest that belief in moral responsibility and recognizing its connection to justified punishment may actually be what keeps many of us from lashing out at innocent people in unjustified ways. Waller will most likely reject this, as in his book he states, “One might logically suppose that those most insistent on moral responsibility and justified punishment would be most careful about finding the guilty party and being certain of that guilt; empirically, the opposite is the case” (2015, 228). He thinks there’s too much empirical evidence against the hypothesis that belief in moral responsibility discourages the punishing of the innocent. Again, if it did discourage this then why is it that in the United States where there is such a widespread and strong commitment to individual moral responsibility it is so common to find innocent people in prison and such a willingness to allow faulty legal and police tactics to persist when we know they often result in the punishment of the innocent? According to Waller, if the belief in moral responsibility discouraged punishing the innocent, then this would not happen. However, such a retort to my argument would be problematic. Above I explain how it is that belief in moral responsibility and recognition of its connection to justified punishment serves as a corrective that keeps many people from lashing out at the innocent. The mere fact that in the United

130

Responsibility Denial and Punishment

States where belief in individual moral responsibility is common and yet many innocent people are punished and the government does nothing to stop the use of faulty tactics which lead to this is no evidence that the belief in moral responsibility does not discourage punishing the innocent. Perhaps there would be even more punishment of the innocent if we didn’t believe that people are responsible. Additionally, it is certainly possible that the belief in moral responsibility discourages the punishment of the innocent in the way I’ve suggested, while in the United States there are other values and beliefs at play that explains why it is nonetheless the case that many faulty legal and police tactics continue to be used and many innocent people get punished. What I want to suggest is that while belief in moral responsibility is a corrective that discourages punishment of the innocent, other factors are at work in our society which leads us to engage in the use of legal and police tactics which are faulty and, thus, contribute to the punishment of the innocent. While we value justice and only want to see the guilty punished for crimes, we also want to be secure in our persons and property. This desire for safety and security against criminal threats can make us blind to the dangers posed by certain legal and police tactics that can lead to the punishment of innocent people. Contributing to this blindness are systemic racism and classism. Large segments of the population are not likely to be one of the accused innocent; the latter are more likely to be poor or nonwhite. Thus, we are disinclined to be as concerned for those who will be unjustly punished. Consider how Waller cites the U.S. government’s suspension of the right to habeas corpus in dealing with terror suspects as part of the war on terror. He considers this as part of the evidence that belief in moral responsibility does not contribute to protection of the innocent. But, as I have argued, the belief in moral responsibility may well contribute to protection of the innocent, while other causal factors may override this concern. In the case of the war on terror, we are so concerned with our own security and safety and so little concerned with those who don’t look like us or speak our language or share our values that we are willing to overlook our concern for protecting the innocent.

Conclusion In this chapter, I’ve tried to show that when philosophers say that we lack moral responsibility in the basic desert sense and then they endorse nonpunitive detention of violent criminals, that is, quarantine, for the protection of society, they are forwarding a view that is morally problematic in significant respects. In particular, I have argued that such a view will have a very hard time explaining why we should not enact policies and procedures, such as lowering the evidentiary standards for criminal conviction, which would lead to the detention of more innocent people who have committed

Responsibility Denial and Punishment

131

no crimes. To enact policies and endorse procedures which we know will lead to more occasions in which innocent people are punished is to show insufficient concern for the dignity and worth of individual human beings. As such, we should have deep moral reservations about the position of responsibility deniers who endorse the quarantine model for dealing with violent criminals. A central part of my argument is that when one rejects belief in moral responsibility then one loses the resources to explain the inherent wrongness of punishing innocent persons and this is what makes responsibility deniers vulnerable to the problem of punishing the innocent. As such, my argument suggests that the belief in moral responsibility acts as a safeguard against punishing the innocent. Bruce Waller directly challenges this assumption in his writings, but as I have shown his challenge can be met.

Notes 1. Pereboom and Caruso are led to deny the existence of moral responsibility on the grounds of their hard incompatibilism. This is the view that whether all human behavior is determined or not there is still no free will and thus no moral responsibility. See Pereboom and Caruso (forthcoming). Waller denies moral responsibility on the grounds of determinism, as such he is a hard determinist about moral responsibility (2011, 2015). 2. While Pereboom, Caruso, and Waller base their case for significant reform of the U.S. penal system on their moral responsibility denial, this should not be taken to suggest that retributivists cannot also support many similar significant reforms which these philosophers embrace. 3. Regarding the quarantine model, it might be wondered: What if we have a category of persons who commit a violent crime but we know from empirical evidence, that it is very unlikely that they would ever be violent again? On the quarantine model endorsed by Pereboom and Caruso, there would be no grounds for detaining such a person. [See for instance Pereboom (2001, 175).] Such a person might be expected to see a counselor and be psychologically examined periodically, but there would be no need to hold this person in isolating captivity if he or she poses no threat to the rest of society. The quarantine model only allows for isolating detention to keep those who pose a threat to the rest of society from harming them. 4. David Hodgson has also argued that innocent people are more likely to be punished on the isolation model relative to its retributivist rival. See his Rationality + Consciousness = Free Will (2012), esp. Ch.11. 5. For a critical reply to Smilansky’s funishment argument, see Neil Levy (2012). Also, see Pereboom (2014, 172–173). Pereboom expresses agreement with the criticisms made by Levy. 6. Smilansky cites R.M. Hare and J.J.C. Smart as making these kinds of response to the problem of punishing the innocent. See Hare (1989a, 1989b) and Smart (1973). 7. Note the argument is that utilitarianism cannot acknowledge the inherent wrong of punishing the innocent. This makes punishing the innocent prima facie wrong. This does not mean that we cannot allow for some legal practices that might contribute to the punishment of some innocent people, for instance the use of eyewitness identifications in criminal trials. Given the unreliability of eyewitness identifications, an outright ban on them would help in reducing the number of

132

Responsibility Denial and Punishment

innocent people convicted of crimes. But without eyewitness identifications too many guilty parties would escape punishment, so despite the inherent wrongness of punishing the innocent we may nonetheless have sufficient reason to preserve a practice that contributes to the punishing of the innocent. 8. See www.cdc.gov/quarantine/aboutlawsregulationsquarantineisolation.html 9. It might be thought that the quarantine model would seem not only to allow for punishing people using a lower standard of evidence, but also it would allow for punishing people who we know have not yet committed a crime but who we know are likely to commit a crime in the future. This would be another troubling implication of the quarantine model. However, I will not develop this issue here, so as to keep my focus on the issue of lowering evidentiary standards. Here I just want to show how the insights of Smilansky’s argument about evidentiary standards can be brought to bear on the quarantine model for criminal detention invoked by Pereboom and Caruso.

7

Libertarian Free Will and Moral Obligation

Chapter Six makes the case for the view that to say we lack the kind of free will which makes us morally responsible in the just deserts sense leads to significant problems in the realm of criminal justice. These problems threaten the rights and dignity of persons. Chapter Six is part of my argument that we have compelling moral reasons to live and act as if we have the kind of libertarian free will which I have explained and defended in the first half of this book. In the present chapter, I seek to contribute in another way to my moral/ pragmatic defense of libertarianism by examining the very grounds of moral obligation. It is my view that human beings are subject to moral obligations; there are numerous things which we morally ought to do. In this chapter, I will argue that in thinking about moral obligation we should preserve the conceptual link between having the obligation to do something and having the kind of free will which grounds just deserts responsibility. As we will see, the notion of moral obligation can be preserved even if we believe human beings lack desert-grounding free will. But I will argue that in such a situation, the motivational force of the moral ‘ought’ is weakened. Such a view has been challenged in the literature. Bruce Waller (2011) provides an especially detailed challenge to such a view. I will begin by developing my own view and then go on to critically examine Waller’s view in hopes of clarifying why, despite his claims to the contrary, maintaining a belief in libertarian free will and the just deserts responsibility which it supports is so important to the concept of moral obligation and the strength of our moral system. The discussion which follows will center on the principle that “ought” implies “can.” Waller accepts a very weak version of this principle. I will defend a stronger version of it and argue, contra Waller, such a stronger version of this principle does not impoverish our moral system, rather it strengthens it.

“Ought,” Failure, and Free Will: Towards a Proper Conception of “Ought” Implies “Can” Suppose I see a large airliner—a 747—plummeting out of control towards the earth and I know it will crash leading to the deaths of all the passengers

134

Free Will and Moral Obligation

and crew. It would be good if I could stop this from happening. Additionally, if I could with little effort stop this from happening and if I had no more pressing obligations, then I would clearly be morally obligated to prevent the plane from crashing. However, I am not free to stop this from happening because I haven’t the power to stop it from happening. Thus, I have no moral obligation to stop it from happening and it would be ridiculous to say, “I ought to stop it from happening.” In contrast, were I Superman, gifted with the power of flight and superhuman strength, then I could easily prevent the crash from happening. And, if I had no more pressing obligations, I’d be morally obligated to stop the plane from crashing. Such considerations lend credence to the idea that “ought” implies “can.” Because I cannot stop the plane from crashing, it seems that I have no moral obligation to stop it from crashing. Thus, it seems that “ought” implies “can.” However, let’s consider another case. Suppose I am an older man who is very selfish and suppose my character is such that I will lie whenever I see that it is in my self-interest to do so. Suppose that you know this is a fixed aspect of my character such that I am incorrigible regarding this—I cannot stop or make myself stop lying out of self-interest. In this situation, would it follow that I have no moral obligation not to lie? I’m inclined to think that even if I cannot stop myself from lying it may be that I am nonetheless morally obligated to do so. Thus, there’s some sense in which I think the principle of “ought” implies “can” is false. Before saying more on this let’s consider one more case. Suppose Henry was taken as a very young child and had his brain altered by a neuroscientist such that Henry will always make decisions in accordance with his rational self-interest. Thus, due to this fixed programming Henry always lies whenever it is in his rational self-interest. Suppose I know adult Henry and I know how he came to be the selfish and dishonest person that he is, and I also know that he is incorrigible. Should I say of him that he ought not engage in these selfish lies? I don’t think so. For to say that a person ought not lie suggests that when he or she lies it reflects a failing on his or her part to live up to moral expectations. But in Henry’s case his lying reflects no failure on his part, because he never could have done otherwise. He is the selfish liar that he is through factors beyond his control. As such to assert that he is morally obligated to not lie is misplaced. In contrast as the incorrigible self-serving liar that I am, when I lie it may nonetheless be appropriate to say I ought not lie. This is because I am a normal human being and in normal human beings their character is in significant respects a product of many, many free choices and decisions they have made in the past. I am an incorrigible self-serving liar today in large measure due to prior free choices I’ve made in the past which have contributed to the shaping of my character. Thus, while I cannot help but act as a self-serving liar today, it is still appropriate to say I ought not act this way. For when I lie it reflects a failing on my part to live up to moral expectations. When I

Free Will and Moral Obligation

135

lie today, I cannot help but do so because of the man I’ve become. But I’ve become this way in large measure because of the free willed choices I’ve made in the past. Thus, when I engage in selfish lies today it reflects a failure on my part and it makes sense to say that I do not act as I ought to act. In comparison, Henry never could have been anything other than the calculating lying egoist he is. Thus, when he lies it does not reflect a failing on his part. For this reason, to say he ought not lie is misplaced. What these comments suggest is that when we say that one does not act as he ought this implies that one fails to meet moral expectations. But to fail at this implies that it was within one’s power to live up to the expectations. Since Henry never could have been anything other than a self-serving liar, his lies reflect no failing on his part. As such, it’s false that he’s morally obligated not to lie. In contrast, while I’m an incorrigible self-serving liar now, I could have made different choices in the past instead of the ones I made which have led me to be the way I am now. Thus, when I lie now it reflects a failure on my part and it makes sense to say that I do not act as I ought. In this way, as well, we can see that there is a way in which “ought” does imply “can.” Those who incorrigibly and inevitably act badly as adults fail to do as they ought insofar as it was within their power to avoid being the bad people they are now. Henry lacked this power. Thus, it’s false to say of him that he ought not act as he does. For, again, in acting as he does he exhibits no failing on his part, as it was never in his power for him to do otherwise. If I am right about all of this, then what this means as well is that to say that one morally ought to act in certain ways entails that one must have the kind of free will which can make one responsible for one’s actions. Suppose (a) I am an incorrigible self-serving liar and (b) I ought not act like this. If so then I must have had the free will to make different choices in the past which would have allowed me to be an honest person instead. Because I did not make those choices, I am responsible for my character. Thus, when I don’t meet moral expectations now, it reflects a failure on my part and it is then appropriate to say that I do not do as I morally ought to do. Henry is also an incorrigible self-serving liar. But he did not come to be this way through any free choices he made. Thus, when he doesn’t meet our moral expectations, it doesn’t reflect any failure on his part. So, it’s not appropriate to say that he does not do as he ought. It may be correct to say that Henry is dishonest and that he is selfish. It may be true as well that he acts badly and that he is a bad man. But when he lies, he doesn’t fail to do as he ought. Due to the interventions of the neural programmer in his early development, he was never able to act honestly. Thus, since it’s been impossible for him to ever act honestly, he has no moral obligation to do so. Furthermore, since he has not failed in any moral obligation to be honest, he’s not at fault for his dishonesty. When he lies he acts wrongly and it’s bad of him to be dishonest, but he isn’t morally bad since it was never in his power to do anything other than to act dishonestly.

136

Free Will and Moral Obligation

A Rival Account of the Moral “Ought” Some philosophers would disagree with the position I’ve articulated here. Some think that even if all human beings lack the kind of free will which grounds just deserts responsibility, it would still make sense to say that there are certain things which we morally ought or ought not to do. Bruce Waller is one such philosopher. He even contends that tying the moral “ought” to free will and moral responsibility as I do impoverishes morality (Waller 2011, 184–185). In what follows, I will critically examine this rival view in hopes of exhibiting the relative strength of my own preferred view. According to Waller, no one has the kind of free will which makes them morally responsible in the just desert sense, but even so making “ought” claims is still reasonable. He distinguishes between the admonishment sense of “ought” and the judgment sense of “ought.” To say “You ought not X” in the admonishment sense is to say this in the rebuking/warning sense. It is to say this in the sense of “X is wrong and shame on you for doing X.” Whereas to say “You ought not X” in the judgment sense is merely to acknowledge that doing X is wrong or bad (Waller 2011, 187–188). Bearing this distinction in mind, Waller says that when you know someone does bad acts and she is utterly incapable of changing her behavior in the future, then the admonishment use of “ought” is inappropriate. For instance, if you know I’m incapable of ceasing from engaging in selfish lies, then there is no point in saying to me “You ought not lie” in the sense of “It’s wrong for you to lie and shame on you for doing so.” He says to engage in the use of such an admonishing “ought” in this case may be “rude and sometimes cruel,” as saying this won’t change me for the better and it will just make me feel bad (2011, 187). But, to say in the presence of others that I ought not lie in the judgment sense is reasonable. For it reminds them that such behavior is bad and it may give them incentive to avoid such behavior themselves. In addition, Waller believes that the use of the admonishing “ought” can be reasonable in other circumstances. Suppose that you know I am a selfish liar but not an incorrigible one. In this case, saying to me in the admonishing sense “You ought not lie” may give me incentive to change my behavior in the future. According to Waller, the rationality of the use of “You ought/ ought not X” is grounded on whether it will help encourage better behavior. Regarding the admonishing “ought,” if you know the person to whom you direct it cannot change and given that it may hurt her, then using it makes no sense. But when the person to whom you direct it can change, then the admonishing “ought” may be appropriate as well as the judgment “ought.” On Waller’s view, when it comes to the use of “ought,” it doesn’t matter how one came to be the sort who engages in a bad course of action. All that matters is whether the use of the “ought” language can play a useful role in getting listeners to act better in the future.

Free Will and Moral Obligation

137

With these points in mind, I should note that there is a sense in which Waller thinks “ought” implies “can.” He and I are in agreement about the case of the 747 plummeting to earth. No one standing on the ground has an obligation to stop the plane from crashing. Waller contends that there is nothing that any human could possibly do to prevent the plane from crashing. Thus, he maintains that there is no meaningful sense in which it can be said that anyone ought to stop the plane from crashing. Waller and I also agree that it can be appropriate to say of particular individuals that they “ought” to do certain things that they are currently incapable of doing. However, Waller believes the admonishing “ought” should only be used in such cases when it is possible that in the future the person may become capable of doing as he ought. For instance, on Waller’s view while I might not at present be able to stop engaging in selfish lies, I might nonetheless be capable of slowly changing my character such that down the road I will be able to avoid telling such lies. According to Waller, in this case the admonishing use of “ought” would be appropriate as it may encourage me to change myself so that in the future I will stop telling such lies. Additionally, regarding Henry who has been programmed to always engage in selfish lies and is incorrigible, the use of the admonishing “ought” would not be appropriate. However, telling others in the judgmental sense that Henry ought not tell such lies would, on Waller’s view, be appropriate. In contrast, on my view it does matter how one came to be the sort who engages in bad actions. Regarding the case of Henry, I don’t think there is any legitimate sense in saying he ought not tell these lies. Waller thinks it would be acceptable to use the judgmental ought here, as it may encourage others to avoid acting as he does. However, to say that he ought not act this way suggests a failing on his part when he tells selfish lies and it suggests that he is morally responsible for these lies. On my view, to say in any sense, whether admonishing or judgmental, that one ought to X entails that they have the kind of freedom with regard to doing X which makes them responsible for whether they perform X or not. Thus, it would be inappropriate to say, even in the judgmental sense, that Henry ought not lie. Furthermore, we can encourage others not to lie without suggesting that Henry ought not lie. For instance, we might simply note that he acts badly in telling such lies. Also, consider the case in which I am an incorrigible selfish liar, and suppose that I came to be this way through free choices I’ve made in the past. In this case, it would make sense to say that I ought not act this way. For to say this acknowledges that my behavior is bad and that it is bad as a result of my own failing. In this case, Waller would object to admonishing me by saying “You ought not lie,” because it may hurt my feelings without any hope of getting me to change. At the same time, he would endorse the judgment use of “ought” in this case, especially if expressed to others. Telling others “John ought not to lie like that” might encourage them to be honest.

138

Free Will and Moral Obligation

Assessing Waller’s View As noted, Waller believes that it makes sense to say that people ought to act in certain ways even though no one has the kind of free will which makes them morally responsible in the just desert sense. He believes this because he thinks that making such claims about the obligations of persons will encourage people to act rightly even though they lack the kind of free will which makes them responsible in the just desert sense. Additionally, while Waller accepts a very weak version of “ought” implies “can,” he believes that accepting any stronger version would weaken or impoverish one’s moral system. In what follows, I will critically examine Waller’s position regarding these matters.

How Effective Will the Moral “Ought” Be in a World Without Just Deserts Responsibility? On Waller’s view, no one has the kind of free will which makes them morally responsible in a just deserts sense. That is, no one has the kind of free will which makes them deserving of praise/blame or reward/punishment. Despite this, he still thinks it makes sense to say that there are certain things which we ought and ought not do. If I am constantly engaging in self-serving lies, Waller believes it makes sense to say that I ought not do this. If one reasonably thinks it would help me change my conduct by telling me I ought not do this, then one would be justified in admonishing me with a “You really ought not tell such lies.” And even if one thinks I’m incorrigible, then one would still be justified in telling me and others that I ought not act this way in the purely judgmental sense in which it is merely acknowledged that I act wrongly. One problem with Waller’s view concerns the fact that he rejects practices of blame and punishment. Since he thinks none of us has the kind of free will which makes us morally responsible in the just deserts sense, he is led to conclude that blaming and punishing for wrongdoing are unjust. This creates a problem for the efficaciousness of the use of “ought.” Suppose we lived in a culture which accepts Waller’s assertions that no one is morally responsible in the just desert sense and that no one deserves blame or punishment. If so, then how effective would telling me I ought to X and admonishing me to X be if I am part of such a culture? If I am part of such a “no blame” culture and if I believe I am not responsible for my conduct in a way that merits blame, then when you tell me I ought to X I may well feel little incentive to do X. After all, I won’t be blamed or punished for my not doing X. I may know that some people may frown upon me and my conduct and some may not want to associate with me, but as long as I have other friends whom I care about and who accept me and I don’t mind being rejected by some folks, then I might not feel much incentive to conform to the claim that I ought to X.

Free Will and Moral Obligation

139

It might be thought that when you are admonishingly told or it is simply judgmentally asserted that you ought not X and then you engage in X anyway, you are still likely to feel guilty for doing X even if you are not blamed or punished for it. This by itself should make the use of “ought” judgments fairly effective in getting people to act rightly. But it is hard to see how a person can harbor feelings of guilt in a culture where no one believes in just deserts type responsibility. If I believe I am not responsible in the way that grounds just desert, then it’s hard to see why I should feel guilty for what I’ve done. Here it might be thought that if enough people tell you not to X and you X anyway and if you see that others are hurt through your conduct, then while not feeling guilt for what you’ve done you may still feel bad for what you’ve done. This by itself might make the use of “ought” language widely effective in encouraging right conduct. For even if folks do not see themselves as guilty, they still don’t want to be the sources of wrongdoing and harm to others. Along these lines, Waller writes: It is very disturbing to discover that one’s character has deep flaws. Imagine you are called before the House UnAmerican Affairs Committee, and under intensive authoritarian pressure you “name names” of your friends. After leaving the hearings, you recognize your vile behavior and the character flaw behind it. This would be a profoundly disturbing revelation, and a source of deep regret, even self-disgust, and would be so even if you firmly believed that you were not morally responsible for the character flaw or the resulting behavior: “I realize that I’m not morally responsible for this flaw in my character; after all, I can see the powerful environmental forces that shaped me to be weak and acquiescent when confronted by figures of authority. But I am very sad to recognize this deep moral failing in my own character.” Alice—as a moral responsibility abolitionist—will not blame herself for her bad character, and if “self-reproach” is synonymous with “self-blame”, then Alice will not self-reproach or self-blame. But Alice might feel profound and sincere regret at the character and behavior that she acknowledges as her own. (Waller 2011, 192–193) The latter points give some reason to think that even in a society which embraces Waller’s skepticism about moral responsibility, the use of “ought” judgments would still be effective in promoting right action. According to Waller, in such a society people will still feel bad about themselves and feel bad for what they’ve done when they fail to do as they ought. However, there will likely be a significant portion of the population which will need a stronger incentive to obey moral norms. Not everyone will feel so bad upon the recognition that they are the sources of wrongful conduct or the injury of others. A significant portion of the population may need the threat of

140

Free Will and Moral Obligation

shaming and punishment to be dissuaded from wrong action. Without this in our motivational tool-kit we may find the persuasive force of “ought” language significantly undermined. Additionally, we should worry about whether there would be as many people who feel bad about harming others if they are raised in a culture that rejects moral responsibility, blaming, shaming, and punishment. Consider Waller’s example in the passage I cite above. This is a plausible example, but I wonder if one would feel so bad about their wrongdoing or harmful conduct if one was raised in a culture where moral responsibility, blame, shame, and punishment are rejected. It is said in the passage that as a moral responsibility abolitionist Alice doesn’t believe she’s morally responsible for her actions, but she still feels regret for what she does when she harms others or otherwise acts badly. But it may be that what enables her to feel such regret is the fact that she was raised in a culture where there is belief in moral responsibility and there are practices of blaming, shaming, and punishment. If one grows up in a culture which embraces moral responsibility as well as the practices of blaming, shaming, and punishment and then in adulthood one comes to believe no one is morally responsible in the desert bearing sense, then it may be that one’s feelings of regret are the lingering effects of having been raised in a culture that embraces the moral responsibility system. For these reasons, it’s not clear how widespread these guilt-like feelings would be among wrongdoers in a culture in which moral responsibility skepticism and blaming, shaming, and punishment are absent. Do Stronger Versions of “Ought” Implies “Can” Weaken Our Moral System of Thought? Bruce Waller cites David Copp (2003) as an advocate of the “ought” implies “can” principle, quoting the following passage from Copp: But if the kleptomaniac lacks the ability to prevent himself from stealing, then I think he is not morally required not to steal. If he does steal, he would have performed an action of a kind that is wrong even though his stealing was not wrong in my view since he did not violate a moral requirement. The requirement he otherwise would have faced was defeated by his inability. (Copp 2003, 282 cited in Waller 2011, 184) Assuming that, as is typically the case, the kleptomaniac made no free choices which led to his condition, I would agree that he is not obligated not to steal. Again, my view is that if we have character flaws which determine us to act wrongly and the character flaws are not the product of our own free choices, then in committing wrongs as a result of these flaws we do not fail in our moral obligations. In such cases, since we never could have done otherwise, we have no moral obligation to avoid the wrongful conduct in

Free Will and Moral Obligation

141

question. At the same time, I do believe that if I am now necessitated to steal due to a character flaw that is the result of prior free choices I’ve made, then even though at this point I couldn’t do otherwise than steal I fail to do as I ought. Judging from the passage of Copp’s which Waller cites, it is unclear whether Copp would agree with my opinion on such a case. Waller does not tease out all the different ways in which the “ought” implies “can” principle might be understood. Thus, it is not entirely clear what he would make of my own version of the principle. Regardless, he is in general quite critical of stronger versions of the principle such as mine. In response to Copp’s point, he writes: [W]hy should we narrow our ethics in this manner? It would severely weaken our system of moral thought, and ethics could not function as well—if at all—for some of the most important things that moral thought does. (Waller 2011, 184) Later in the same paragraph, Waller goes on to talk about Robert Harris who in real life was subjected to a horrific upbringing and then in later life committed horrible crimes. It is reasonable to think that Robert Harris was causally determined to be the vicious person that he was such that he couldn’t help but commit the horrible crimes he committed. Waller writes: Robert Harris is genuinely bad, and his brutal acts are morally bad, and it’s important to note these things so that we can make efforts to prevent him from doing such acts in the future, attempt to reform his character, and—above all—examine the conditions that shaped his vicious character and try to change them so that fewer people like Robert Harris emerge. Saying that Robert doesn’t deserve blame for his genuine moral faults does not impair the effectiveness of our system of moral thought; in contrast, we do severely limit our moral system if we must say that (because he could not act virtuously) it was not wrong for him to act viciously and that he was under no moral obligation to refrain from murder. (184–185) Waller clearly believes that given his upbringing Robert Harris could not have been otherwise than the vicious person he was and that he was thus determined to engage in the horrible crimes he committed. Despite his lack of responsibility for engaging in these crimes, Waller thinks it is vitally important that we continue to believe that he was morally obligated not to commit these crimes. Even though we see that he could not have done otherwise, Waller believes it is still true that Harris ought not to have committed those horrible crimes. And, he thinks that to deny this is to embrace a moral system that is impoverished.

142

Free Will and Moral Obligation

Waller’s assessment of the Robert Harris case clearly differs from my own. Assuming that Robert Harris was necessitated by his upbringing to be the vicious person he is and that given this he was determined to commit the crimes that he did, then on my view we should not believe that he was morally obligated to avoid committing those crimes. Again, we can say that what he did was wrong and bad and we can say that he is a bad person; but on my view, it just doesn’t make sense to say that he ought to have done otherwise. Because to say he ought not to have acted in these ways suggests that when he doesn’t do as he ought, then he has failed in some way. And, crucially, the thing is Robert Harris hasn’t failed, because he never could have done otherwise. His actions were not up to him, they were the consequence of his horrible upbringing. Notice, as well, that Waller wants to say that Harris is morally bad and that what he does is morally wrong, but I am uncomfortable with this. Since he is utterly devoid of responsibility for his actions and since he thereby hasn’t failed in his obligations in doing these bad things, I don’t think that he is morally bad. Rather, it seems to me he is dangerous and a bad human being and he does bad things; but in doing these bad things he does not act as a moral agent who is subject to obligation and so he is not morally bad when he acts in these ways. Think of the difference our reactions are to the wrongdoing of a Robert Harris and the wrongdoing of someone else who, unlike Harris, made himself through his own free choices into the bad person he is. When we understand the different life circumstances of these people, seeing that Harris is not responsible for being as he is and seeing that the other person has freely made himself into a bad person, we are angry with the latter person but with Harris we are just dismayed about the circumstances which have given rise to his being the bad person he is. This reflects the fact that Harris is not morally bad, whereas the latter is. Harris is a mere unfortunate product of his environment, whereas the latter having made himself into a bad person who harms others makes us justifiably angry with him. In the passage I quote above, Waller claims it is important that we note that Harris’s acts are “morally bad” and that he was under a moral obligation to refrain from murder if we are to “make efforts to prevent him from doing such acts in the future, attempt to reform his character, and— above all—examine the conditions that shaped his vicious character and try to change them so that fewer people like Robert Harris emerge.” But this doesn’t follow logically. It suffices to recognize that he is dangerous, a man who has brought great suffering upon others and will do so again unless he is detained. Additionally, seeing that his actions are a result of his character, a character he has had no hand in shaping, we have reasons to help in changing his character so that he won’t continue to act in this way in the future. It’s not at all clear that we need to view him as someone who is morally bad and who fails in his obligations in order to promote the ends of preventing him and others from acting in such ways in the future.

Free Will and Moral Obligation

143

In a later passage, Waller writes: Consider another case: Sam ought to stop being jealous. Suppose we learn that Sam (because of his conditioning or genetics or other similar reasons) cannot presently stop being jealous. It may still make sense to say that Sam ought to stop being jealous: it makes sense, and it may serve several useful functions. First, even if Sam cannot presently stop being jealous, if he believes that he ought to do so (perhaps as the result of being admonished to that effect), he might take steps to make it possible to stop being jealous in the future (for example, he might seek the services of a good psychotherapist). Second, “You ought not be so jealous” may be useful, even if we believe that Sam cannot presently exercise control over his jealousy, and indeed even if there are not steps available to Sam that would lead to gaining such control; even if he cannot stop being jealous, he may come to see it as a character flaw and not something to be acted upon, rather than a virtue that is a good guide to action. Furthermore, he may work to prevent his children from following in his own flawed path. Finally, even if we think that Sam has no chance of reform and that he will never even be capable of seeing his jealousy as a flaw, it may be useful to say that Sam ought not be jealous; such an admonition may help shape others who are currently more malleable to avoid such a character flaw. Of course, if Sam has no resources for reform, then it may be useless to tell Sam that he ought not be jealous, but that fact certainly does not make the statement false or incoherent. (Waller 2011, 185–186) Waller’s point is that we weaken our moral system if we are unwilling to say that those who cannot currently act rightly are not obligated to act rightly. This weakens the moral system because it takes away a key tool for encouraging people to become right acting. My own view allows that even if Sam cannot avoid acting jealously it may nonetheless be true that he ought not do so. Presuming that human beings have the kind of free will which makes them responsible, it is reasonable to think that most people who are necessitated to act jealously as adults are this way in part because of free choices they’ve made in the past which have led them to be the jealous sorts of persons they are now. So, on my view, most jealous adults are partly to blame for their jealousy and when they act jealously they do not do as they ought. Thus, on my view, which embraces a stronger version of “ought” implies “can” than does Waller, the great majority of adults who act jealously and do so of necessity from their character can rightly be said to fail in their obligation to avoid acting jealously. Additionally, on my view, this means that if Sam came to be the jealous sort in the way that most adults do, then even though he cannot currently help but act jealously, it’s still true that he ought to avoid doing so.

144

Free Will and Moral Obligation

At the same time, on my view if we find out that due to genetic or environmental factors he could never have been other than the jealous man he is today, then even if he is capable of being reformed, we should not say of him in his present condition that he ought not act jealously. Again, on my view, if we say he ought not act jealously, then when he does this would imply some sort of failing on his part. But if he never could have been anything other than the jealous sort he is, then his being this way indicates no failing on his part; rather, his being this way is just an unfolding of the given. Waller will note that if Sam is capable of reform, then denying he ought not act jealously will take away a key tool for getting him to change. Thus, in a case like this, even my own version of the “ought” implies “can” principle weakens the moral system. But, here, we might wonder how essential this tool is. If there are other ways that are just as effective at getting someone like Sam to change his ways, then we shouldn’t be worried about this. Are there equally effective ways? Yes! Suppose I never tell Sam he is morally obligated to stop acting jealously. Instead, I show him how his jealousy is damaging himself and his relations with others. Suppose as well that I show him how it hurts others whom he cares about and suppose I show him that he cares about his reputation and how his jealousy is damaging his reputation. All these things may well be demonstrable to him and, if so, I strongly suspect they would be more effective in getting Sam to change than asserting “You have a moral obligation to change.” The latter has a sort of righteous emptiness to it that is more likely to fall on deaf ears than are these other tactics I’ve described. In defense of Waller, it might be said that the tactics I propose are just ways of getting Sam to see what he morally ought to do. Until he sees that not acting jealously is what he ought to do, then he won’t be moved to change. Perhaps, in this way preserving the use of “ought” even when one cannot do as he ought is still fundamental to motivating right action. However, such a response would be misguided. First, it’s not at all clear that people must conceptualize their conduct as what “ought” to be done in being moved to help themselves or others that they care about. Sam may simply want to help himself and others by changing his behavior and be moved straightaway to doing so, acting on such desires without thinking about what he “ought” to do. Second, even if he did need to make the connection between these reasons and the conclusion that he ought to change his behavior, it’s not at all clear that this would be a moral “ought.” Wanting to help oneself and the people one cares about and seeing what one ought to do to achieve these ends doesn’t seem sufficient for establishing a moral “ought.” After all what one sees as serving these ends may actually be unjust to others while benefitting oneself and those whom one cares about. Perhaps a better defense of Waller’s view would simply note that the strategy I’ve proposed for encouraging Sam to change may fail. He simply may not care enough about himself and his relations and the welfare of others to change. In this case, having the moral ought in one’s toolbox to motivate a change in Sam may do the trick. Since my view rules out the use of such an

Free Will and Moral Obligation

145

“ought” in Sam’s case, then my view has fewer resources to motivate right action. However, even here the argument seems weak. It seems to me that if Sam won’t change for his own good, the good of his relations, or for the sake of others, then he is hardly likely to be moved by being told that morality requires it. Regarding the latter point, some of my earlier comments are relevant. If we all come to believe no one is morally responsible in a just deserts sense, then the tendency to feel guilt or regret for one’s wrongdoing may be undermined in significant respects. In which case, being told what morality requires would likely have less motivational force for us.

Tragic Conflicts of Obligation and the Importance of Excusing Conditions In his book Waller talks about how persons can be subject to situations in which they have obligations to do two different actions but they can only do one of them. [I]n Greek tragedy, Antigone is obligated both to bury her brother and to follow the king’s law prohibiting burial. Thus she has conflicting obligations: she ought to do both, but obviously she cannot. To the Greeks, this seemed an unfortunate situation, though certainly not impossible. (Waller 2011, 182) It could be argued that the version of “ought” implies “can” which I endorse is inconsistent with the fact that one can have such conflicting obligations which cannot both be satisfied. Thus, the version of “ought” implies “can” which I endorse should be rejected. My view actually allows that in some circumstances one might face genuinely conflicting obligations. If I have made free decisions in the past which have structured my character in such a way that I now find myself obligated to do A and obligated to do B and I can only do one of them, I may nonetheless genuinely have such conflicting obligations and now be doomed to fail in satisfying one of them. However, not all conflicts of obligation arise in this way. Sometimes they arise not because of any free character-shaping choices one has made in the past but through chance occurrences in one’s life circumstances. For instance, to prevent my entire city from being flooded I may have to let my first child drown. It could be said that I am obligated to protect my city and I am obligated to protect my child. However, due to chance circumstances I am now faced with a situation where satisfying the one excludes the possibility of satisfying the other. Should we say that in circumstances such as this that I am genuinely obligated to both protect my city and protect my child, when I can only do one of these? Even here I am inclined to think that one may have the obligation to protect the city and the obligation to protect one’s child, but that one is excused from the requirements of one of them because of the presence of a conflicting obligation. Thus, even here it makes sense to say that one genuinely has conflicting obligations.

146

Free Will and Moral Obligation

Ishtiyaque Haji (2000a, b), who believes in the principle of “ought” implies “can,” thinks that there can be no such conflicting obligations. If your central claim is that on a particular occasion, perhaps due to unlucky personal history, it is possible that some persons have both an overall moral obligation to do A and an overall moral obligation to refrain from doing A . . . then I’d say that according to the theories about the concept of moral obligation that I favor, such situations are impossible: they can’t occur. (Haji 2000b, 369) According to Haji, since “ought” implies “can,” one cannot be obligated to do A and at the same time obligated to do B if one is incapable of doing both. However, while, like Haji, I endorse a strong version of “ought” implies “can,” I don’t see why we must claim that there can be no such genuinely conflicting obligations. Rather, it seems to me that this is an area where the notion of excuse becomes important. As I noted above, it seems to me that sometimes our obligations do genuinely come into conflict such that we cannot fulfill all of them. But this does not imply that in not fulfilling one of two conflicting obligations we exhibit moral failure; rather, in such circumstances we are excused from satisfying one of them. Waller seems to think that the presence of such conflicting obligations presents a challenge for those, like Haji and I, who endorse stronger versions of “ought” implies “can.” He talks about how the belief that there can be no such conflicting obligations arises from a belief in a just world. He says the ancient Greek “gods were spiteful, arbitrary, and cruel, and the best human efforts could be thwarted by cosmic caprice” (Waller 2011, 182). Thus, for the Greeks it was not hard to imagine that we could face genuine conflicts in our obligations. But on a Christian world view, “obligations and capacities must coincide: a just God would give no obligations beyond our abilities to fulfill them” (182). Waller suggests that the belief that there can be no genuinely conflicting obligations and that “ought” implies “can” is a survival of a Christian world view. He thinks that once we embrace a naturalistic world view we shall see that there is no reason to think the world is just and that we will be able to satisfy all our genuine obligations. [I]n the natural world, devoid of divine order, there is no such assurance [that our obligations and capacities will coincide]. Having evolved in this world it is hardly surprising that it accommodates us reasonably well, but fitting our moral obligations to our powers is well beyond what natural selection is likely to provide. Whatever one believes to be the source of one’s obligations, there is no reason to suppose that the natural world is designed to help us meet them . . . When we look closely, “ought implies can” has little to recommend it other than the exalted stature of Kant, the antinaturalist desire to transcend the natural

Free Will and Moral Obligation

147

world and set humans apart from other animals, and the belief that some divine force imposes a moral order on the world so that we will never face the misfortune of having a moral obligation we are incapable of meeting. (182–183) Contra Waller, the presence of conflicting obligations is no threat to stronger versions of “ought” implies “can” such as the one I endorse. It is indeed the case that one can have genuinely conflicting obligations which cannot both be satisfied. But, as I’ve argued, this has no negative implication for the principle that “ought” implies “can,” as this is an area in which the concept of excusing conditions becomes important. In these situations, one may have a genuine conflict in one’s obligations but be excused from satisfying at least one of them due to the presence of the conflict. If I have an obligation to A and an obligation to B and I cannot do both, then it seems that I ought to do whichever is the weightier of the two. If they are of equal weight—which I admit is entirely possible—then I must arbitrarily choose from among them and satisfy only one of these obligations. Should I satisfy A, I should not be blamed or seen as failing in my obligation to do B. Rather, the presence of the conflict excuses me from the obligation to do one of them.

More on Obligation and Failure: The Distinctive Nature of the Moral “Ought” Earlier in this chapter, I argued that there is a conceptual link between having a moral obligation and failure. When a person, P, is morally obligated to X, then, in the absence of an excusing condition, P’s nonperformance of X must involve some failing on the part of P. This was intended to help explain why Henry, who was programmed at birth to engage in self-serving lies, has no obligation to not tell such lies. For when he lies it exhibits no failing on his part, since he never could have done otherwise. In contrast, even though my formed character today may necessitate that I tell lies, it can still be coherently maintained that I ought not lie. For when I lie it is due to my failure to have formed my character differently. Since it was within my power to have formed a different, more honest character and I freely chose not to, I am now subject to an obligation which I am at present incapable of fulfilling. It might be thought that this line of argument is problematic because of the conception of failure which it employs. Couldn’t it be said that when Henry lies he fails to be honest, fails to do as he ought, even though he never could have done otherwise? If so, then my defense of a strong “ought” implies “can” principle is problematic. In support of this criticism, suppose that when I turn the key in the ignition of my brand-new car it fails to start the engine. If this happens, it is likely the determined consequence of some mistake in the construction of

148

Free Will and Moral Obligation

the car’s ignition system. Suppose this is the case. If so, the car couldn’t do otherwise than fail to start. Thus, the fact it couldn’t do otherwise does not establish its inability to fail at doing as it ought. So, too, when Henry tells his lies he may legitimately be said to fail at doing as he ought even though he never could have done otherwise. It must be conceded that when the key is turned and the engine doesn’t start, this is a failure in the functioning of the car. It must also be conceded that this failure is a determined consequence of the way in which the car was manufactured. It is also true that given the norms established by proper car functioning, the car ought to start when the key is placed in the ignition and turned. Notice here we are talking about a man-made product, an artifact, a car. But similar claims could be made about natural objects. Suppose a child is born with eyes that don’t function properly and the child is blind. As a determined consequence of processes occurring during fetal development, the child’s eyes don’t develop as they ought and due to this she is left blind. Teleological systems, such as cars and human bodies, have certain norms for proper functioning which establish how their parts ought to perform. When certain tasks are not performed in these systems, these amount to a kind of failing. In not allowing the child to see, the child’s eyes do not work as they ought and this is a failing in the functioning of her eyes. While all of these points can be made and are true, they don’t have application in the context of moral “oughts,” moral obligation, and moral failure. For one is morally good when one does of one’s own free will what one ought to do and morally bad when one does of one’s own free will what one ought not to do. As such, to exhibit moral failure one’s noncompliance with the moral norms must be the result of one’s own free will. If one is determined to tell lies, as Henry is, by forces outside of his control, then one does not comply with the moral rules. But, this does not constitute moral failure, because in this case one’s noncompliance is not free willed. Moral failure is conceptually linked to free willed noncompliance with the moral norms. Considered as a person, a moral agent, Henry doesn’t fail to do as he ought. Since he cannot do otherwise than lie, the moral rule against lying has no application to him. Thus, his lying exhibits no moral failure on his part. When one is subject to a moral obligation it is presumed that one has the power to shape his own self in such a way that he will comply with the moral rules. With respect to lying, Henry lacks this power and so it’s not true that he has a moral obligation not to lie. Thus, when he lies this exhibits no moral failure on his part. I think these considerations about the moral status of Henry are supported as well by how the reflective person would respond to someone like Henry and his lying when she knows how he came to be the lying sort he is. If the reflective person sees that his lying is the result of factors beyond his control such that he never could have done otherwise, then she will not be angry with him nor blame him for his lying. Rather, the reflective person, seeing that Henry never could have done otherwise, will see his harmful

Free Will and Moral Obligation

149

lies as akin to the harm caused by a rattlesnake when it bites a child that steps on it. The snake acts from a preprogrammed instinct when it bites and Henry does so as well when he lies. The reflective person won’t be angry at nor blame the snake. So, too, if she knows how he came to be the lying sort of person he is, the reflective person won’t be angry at nor blame Henry for his lies. The reflective person may be angry or sad that harm has been caused by the snake, but that is different from being angry at or blaming the snake. Like when Henry tells his lies, the snake does not act as a moral agent when it causes harm. Thus, it is an unfitting object of our anger or blame.

Conclusion Waller embraces only the weakest version of the “ought” implies “can” principle. He thinks “ought” implies “can” only when no human could possibly do what it is said that one ought to do. Thus, in the case of the falling 747, he says it is nonsensical to say that one ought to prevent it from crashing, as no human being could possibly stop this from happening. I embrace a stronger version of the “ought” implies “can” principle, which entails that one is obligated to X only if either (a) he is currently capable of doing X or (b) his inability to do X is a result of free choices he has made in the past which have shaped his character such that he is now incapable of doing X. Waller argues that stronger versions of the principle, such as the one I endorse, would weaken the moral system by robbing us of tools to motivate right conduct. However, I have argued here that such is not the case. I’ve also argued that it’s not a view like mine that would weaken our moral system, but, rather, a view like Waller’s. For my view presupposes that for human beings to have moral obligations they must have the kind of free will which can make us morally responsible in the just deserts sense. To be obligated to do X I must currently have the power to do X or if I am not currently able to X it must have been in my power to have formed a different character from the one I have now which makes it impossible for me to X. This demands that we have the kind of free will over our actions which supports just deserts responsibility. But Waller says that we lack this kind of free will. Because he thinks we lack such free will and just deserts responsibility, he rejects the practices of praise/blame and reward/ punishment. As I’ve argued, the problem with this is that without these practices in our motivational tool-kit it is no longer clear that use of “ought” claims, whether of admonition or of mere judgment, will have much motivational force. In the end, it seems to me that if there’s a problem with the assertion that we have the kind of free will which makes us morally responsible in the just deserts sense, it’s not because of its conceptual ties to the “ought” implies “can” principle and the practices of praise/blame and reward/punishment. Rather, the problem is with the lack of sufficient empirical, metaphysical, or introspective evidence of its existence. But, as I’ve argued in earlier chapters,

150

Free Will and Moral Obligation

despite the lack of this kind of evidence, we have sound moral reasons to act as if we have such free will or to believe we have it. Indeed, this chapter contributes to this case by showing why an “ought” implies “can” principle grounded on the belief in such free will does not weaken our moral system but strengthens it in precisely those areas where Waller thinks the moral responsibility denier’s view is stronger.

8

Hardheartedness and Libertarianism

In previous chapters, I argued that denying that human beings possess the kind of free will which makes them morally responsible is problematic. In Chapter Six I noted how this leads to problems in the realm of criminal justice. In particular, moral responsibility denial makes it too easy to justify the adoption of policies which would lead to the detention of innocent persons who’ve committed no crimes. As such, moral responsibility denial could easily lead to the adoption of policies which do not exhibit sufficient respect for human rights and the dignity of persons. In Chapter Seven I argued that denying we have the kind of free will which grounds just desert also runs into problems making proper sense of moral obligation and it may lead to the weakening of our moral system. As I’ve said in earlier chapters, I don’t think that at this point in time it can be proved that we have the kind of libertarian free will which is necessary to moral responsibility. Despite this, I’ve claimed that there are good moral reasons to maintain the belief in such free will, as it helps support moral policies which show sufficient respect for human rights and dignity and it helps provide for a more coherent and effective notion of moral obligation. The two previous chapters are intended as support of this thesis. In the present chapter, I want to explain and respond to an important challenge that has been raised against the kind of position I’m defending. This challenge has been made most clearly and forcefully by Richard Double (2002), who argues that it is “hardhearted” of libertarians, morally inappropriate of them, to treat people as morally responsible for their actions while also believing they lack epistemic justification for the belief in free will which supports moral responsibility. As he puts it: All libertarians who confess that they lack epistemic justification that anybody chooses indeterministically should exonerate everyone of moral responsibility. (2002, 229) Again, such a view stands as a direct challenge to my argument, and other philosophers have either made similar arguments or expressed sympathy with the position Double takes on this matter.1

152

Hardheartedness and Libertarianism

In “The Moral Hardness of Libertarianism” (2002), Richard Double rightly notes that most libertarians do confess they lack epistemic justification that anybody chooses indeterministically, that is, while believing libertarian free will exists, most libertarians admit that there is scant epistemic justification for belief in its existence.2 Thus, his argument implies that most libertarians should exonerate everyone of moral responsibility. Double also contends that when libertarians who believe they lack justification for their belief in free will refuse to exonerate everyone of moral responsibility and, instead, continue to do so, thereby endorsing practices of blame and punishment, such libertarians are being hardhearted. He states, “most libertarians are hardhearted (unsympathetic, not morally conscientious)” (p. 226). In what follows, I will present Double’s argument for this view and critically respond to it. In developing my response, I will defend a kind of objection which he considers and tries to answer in his essay. According to this objection, “the pragmatic argument,” even if the libertarian lacks epistemic justification for belief in libertarian free will and moral responsibility, the belief in these is justified on moral grounds. Double contends that Immanuel Kant and William James have defended this approach to the issue, and he rejects this approach arguing that it is grounded on problematic consequentialist moral considerations. I will show that the pragmatic argument need not be grounded on consequentialist moral considerations. I will provide and defend a nonconsequentialist, pragmatic response to Double’s argument.

Double’s Argument Double argues as follows: 1) Libertarians believe that we should hold persons morally responsible only if they exercise libertarian free will. 2) Libertarians believe that we should hold persons morally responsible. 3) Most libertarians believe that we have scant epistemic justification that persons have libertarian free will. So, 4) most libertarians believe we have scant epistemic justification for believing that persons meet one of the necessary conditions for being morally responsible, while still believing we should hold persons morally responsible. 5) Sympathetic or morally conscientious persons do not hold people morally responsible for their actions unless they have epistemic justification for doing so. To do this is to be hardhearted. So, 6) most libertarians are not sympathetic and morally conscientious, i.e. they are hardhearted. (pp. 226–229)

Hardheartedness and Libertarianism

153

In his article it is clear that Double is primarily worried about the willingness of libertarians to hold people morally responsible for wrong actions, for doing so leads to negative reactive attitudes, blame, and sometimes punishment. He states: Holding persons morally responsible includes a wide range of positive and negative behaviors: expressed reactive attitudes, verbal recrimination, praise and blame, retributive punishment and just-deserts rewards, all the way to eternal torment in Hell and bliss in Heaven. Libertarians disagree among themselves over how much of that range moral responsibility includes. But because even the mildest of these adverse behaviors harms persons, libertarians use the assignment of moral responsibility as a justificatory mantra that turns otherwise immoral treatment into just-deserts goods. (pp. 227–228) In holding people morally responsible for their actions we legitimize blame, scorn, resentment, and sometimes punishment directed towards those we deem morally responsible for wrongdoing. These reactions harm those to whom it is directed. Thus, we are obligated to withhold judgment of the responsibility of persons for wrongdoing until we have adequate evidence of their responsibility. Double notes that most libertarians don’t believe there is adequate epistemic justification for belief in the existence of libertarian free will; rather, they regard its existence as an open question. At the same time they believe its existence is necessary for being morally responsible and they think we should hold people morally responsible for their actions. Thus, Double concludes that most libertarians unjustifiably endorse practices which bring harm upon persons who may not deserve to be harmed. As such, Double believes most libertarians are hardhearted. Much of Double’s essay involves constructing possible objections to his argument and responding to them. He refers to one of these objections as “a pragmatic argument” for belief in libertarianism. The idea behind the objection is that a libertarian might say that while he lacks epistemic justification for belief in libertarian free will, there are still good moral reasons to believe such free will exists. In other words, some libertarians might say that while there’s no good scientific, logical, or metaphysical reason to believe in libertarian free will, there are, nonetheless, good moral reasons to do so. Double briefly characterizes this approach and quickly moves on to criticize it. He writes: A third reply [to Double’s argument] might defend the claim that persons make libertarian choices, despite its lack of epistemic credentials, by claiming that our believing in libertarian free will and moral responsibility has salutary moral effects, a claim Kant shares with James. Of course, this claim is debatable armchair psychology. To the best of my

154

Hardheartedness and Libertarianism knowledge, psychologists have not studied the results of the belief and non-belief in moral responsibility, so the whole issue seems to me conjectural . . . But let us concede libertarians the dubious psychological premise and see if it helps to rebut my objection. Any libertarian who stresses the pragmatic value of believing in libertarianism is committed to consequentialist cost-benefit analysis. So, even if the psychological speculation were true, the benefit of inculcating better moral behavior among the public at large must be weighed against the possible cost of the unfairness of assigning moral responsibility to persons who do not deserve it. Instead of reveling in the hope that we might have libertarian freedom, a compassionate incompatibilist would consider the injustice we do to persons when we retributively punish them under the banner of libertarian free will and then consider the probability that we actually have such freedom . . . [L]ibertarians who endorse this pragmatic argument by their own premises, find themselves in an unenviable position. They are committed to believing that an improbable moral good (encouraging better moral behavior at large) outweighs the probable moral evil of inflicting negative reactive attitudes, blame, and retributive punishment on undeserving victims. (p. 230)

Here Double notes that libertarians might defend themselves by arguing that while there is no adequate epistemic justification for belief in libertarian free will and moral responsibility we should still believe in it for moral reasons. After noting that Immanuel Kant and William James have made this kind of argument, he quickly moves on to make two central criticisms of this approach. First, he says there are no empirical psychological studies supporting the view that belief in libertarian freedom and responsibility encourages moral conduct. Second, he states that even if this premise is granted this defense rests on a problematic “consequentialist cost-benefit” analysis.

A Defense of the Pragmatic Argument Double’s reply to the pragmatic argument is problematic on both fronts. First, while relevant psychological studies had not been conducted at the time Double wrote his essay, they have since been conducted. Some empirical psychological studies suggest that encouraging disbelief in free will has the effect of promoting antisocial behavior. A study by Vohs and Schooler (2008) suggests that encouraging disbelief in free will and responsibility leads to cheating and stealing. A study by Baumeister et al. (2009) suggests it leads to aggression and reduced helpfulness. It is unclear just how relevant these studies are to the issues at hand. Just because a disbelief in free will leads to these does not mean that encouraging belief in libertarian free will and responsibility is essential to encouraging moral conduct. It could be that promoting compatibilist views would be equally effective. It is an interesting

Hardheartedness and Libertarianism

155

question as to whether these empirical studies are trustworthy and whether they are especially relevant to pragmatic arguments for belief in libertarian free will. At the same time, it is likely that further studies of different kinds would be needed to determine more clearly the nature and extent of the morally beneficial consequences of belief in libertarian free will.3 I will not pursue these matters here. Double himself is willing to grant that belief in libertarian free will and moral responsibility may have some salutary moral effects. He says that even if this point is granted the pragmatic justification is still problematic because it is based on consequentialist reasoning that will not stand up to critical scrutiny. I think Double makes a mistake in developing his response to the pragmatic argument. In responding he only envisions the pragmatic argument in consequentialist terms. However, the pragmatic argument can also be developed along nonconsequentialist, deontological lines. It is especially odd that Double envisions the pragmatic argument solely on consequentialist grounds while he acknowledges Immanuel Kant as being one of the advocates of the pragmatic approach. Kant is famous for rejecting consequentialist approaches to moral reasoning. In what follows, I will defend the pragmatic argument against Double’s criticism by showing how it can be developed along nonconsequentialist lines. I will give further defense of this approach by considering various possible objections to it and responding to them. My argument will be Kantian in spirit, but I am not to be understood as claiming that Kant held the views that I will be developing here. I am not engaged in Kant scholarship; rather, my intent is to appeal to some Kantian principles in developing a response to Double’s argument. Kant’s principle of ends states that one should always treat humanity, whether in one’s own person or in that of others, as an end and never as mere means. This is a nonconsequentialist moral principle. The very content of the principle manifests its nonconsequentialist nature, as it states that we should always treat humanity as an end and never as a mere means. Were the principle consequentialist it would not say this, as there are clearly cases where the greater good may be served by treating a person as a mere means. To treat a person as an end is to show respect for the person’s autonomy; it is to show respect for that person’s ability to make choices for himself and to act in accord with them. To treat someone as mere means is to disrespect this capacity of persons. Rape, murder, theft, slavery all involve treating others without regard for their own choices. The person raped did not choose to have sexual relations with the rapist—the sexual relations are forced upon him or her. The person robbed did not choose to give up his or her property—it is taken against his or her will. According to the principle of ends, these acts are wrong simply because they involve treating persons as mere means and not as ends in themselves who possess a capacity of choice that deserves respect. According to this principle, rape, murder, and theft are not wrong due to their bad consequences. Indeed, one of the cardinal virtues of this Kantian principle is that it captures the widespread intuition that acts

156

Hardheartedness and Libertarianism

like murder and theft can be wrong even when the consequences of these actions are good on the whole. Now, as noted, what makes persons deserving of respect is their capacity for choice and their ability to live their lives in accordance with their choices. This leads to questions about the nature of choice and this takes us to the link between competing theories of free will and the pragmatic reply to Double’s charge that, “All libertarians who confess that they lack epistemic justification that anybody chooses indeterministically should exonerate everyone of moral responsibility.” Choice can be understood along either hard determinist, hard incompatibilist, compatibilist, or libertarian lines. On all of these views, choice is to be understood as the end result of deliberation. There is no doubt that all human beings do frequently engage in deliberations about what they shall do and in doing so they assume that what they will do is up to them. Now on the hard determinist view of things, our choices are never freely performed and we are never responsible for them because what we choose is just a necessary consequence of prior factors which were in turn necessitated by even earlier events and so on going back in time. Hard incompatibilists agree that we never engage in free choice and that we are never responsible for our choices. They believe that all of our actions are either determined or, perhaps, some of them are undetermined, but either way we do not make free choices and we are not responsible for them. If our choices are determined, then they are not free for the reasons indicated by the hard determinists. Furthermore, if they are undetermined, then they are random occurrences, meaning that we lack the kind of control over them for them to be products of free will. On the libertarian view, choices can be freely made and we can be morally responsible for them. For this to be the case at least some of them must be causally undetermined. The libertarian does not have to view all free choices as undetermined. As Robert Kane has noted, the libertarian can view determined choices as free in a derivative sense if they are the consequence of a character formed by prior undetermined free choices. (See, for instance, Kane 1996, 2007a, b, 2011.) Compatibilists also believe choices can be free and that we can be morally responsible for them. However, unlike the libertarians, they believe that even if all events, including all of our choices, are determined, then we can still make free choices and be responsible for them. For the most part libertarians do not think compatibilist accounts of freedom and responsibility make sense. Hence, the famous quips of William James and Immanuel Kant, two historically famous defenders of libertarianism; James called compatibilism a “quagmire of evasion” (1884, 149) and Kant called it a “wretched subterfuge” (1788, 95–96). Further, there are some pretty good reasons to think compatibilism is deeply flawed—consider Peter Van Inwagen’s consequence argument (1983) or Derk Pereboom’s four-case argument (2001) or Robert Kane’s argument from ultimate responsibility (1996), as well as the argument I make in Chapter Three of this book.

Hardheartedness and Libertarianism

157

A libertarian may reasonably come to believe that there is no plausible compatibilist account of freedom and moral responsibility. If so, he will be led to think there is either libertarian free will or no free will and no moral responsibility at all. This point is very significant in developing a nonconsequentialist pragmatic response to Double’s argument. For when choice is viewed on the hard determinist or hard incompatibilist models it is not perceived as free choice for which the agent is responsible. On these models no human beings ever make free choices for which they are also morally responsible. If choice is understood in these terms, it is hard to see how the human capacity for choice gives us the special dignity and worth that entails we should always be treated as ends and never as mere means. Indeed, for Kant the capacity for free choice was the grounds for thinking of human beings as autonomous beings deserving of respect. Thus, in order to account for this autonomy he was led to conceive of humans as possessing a transcendental (noumenal) self that stood outside the realm of deterministic causal law. While I have no interest here in defending the notion of a Kantian transcendental (noumenal) self, I think it is correct to believe that we cannot make sense of the Kantian principle of ends unless human choices are perceived as free choices for which we are morally responsible. Furthermore, assuming libertarians are correct in their rejection of compatibilist models, it follows then that we can only make sense of the principle of ends on the assumption that human beings have libertarian free will. Before giving further defense of these points let’s pause to review the thread of the argument so far. Recall that the pragmatic objection to Double’s argument states that there can be good moral reasons to regard persons as possessing libertarian free will and being morally responsible even if we don’t have adequate epistemic justification for this. In reply, Double says that such an objection must be grounded on problematic consequentialist reasoning. What I am suggesting is that a libertarian who finds compatibilism implausible and who accepts the nonconsequentialist principle of ends on independent moral grounds may be rationally led to posit the existence of libertarian free will and moral responsibility without sufficient epistemic support, as it is the only coherent way to make sense of what he knows to be true in the realm of morality. One might come to believe that the principle of ends is true for independent moral reasons. For instance, one might for independent moral reasons come to believe that the Kantian principle of universalizability is true and then deduce the principle of ends from it. That is, one might consider treating persons as mere means and apply the principle of universalizability in assessing the rightness of acting in such a way. In doing so, he might see that one cannot reasonably will that everyone treat persons as mere means and, then, conclude that we should always treat persons as ends in themselves. Or, one might notice that many problems with utilitarianism involve instances of unjustified treatment of persons as tools in the pursuit

158

Hardheartedness and Libertarianism

of the greater good, and then one might be led to see that each of these is a case involving the violation of a general rule that persons should always be treated as ends. Now, having come to accept the principle of ends on independent moral grounds, one might then come to the realization that rational acceptance of it requires that we believe humans have free will. Further, if one has good reason to believe compatibilist accounts of free will are implausible but that there are plausible libertarian accounts, then one might rationally be led to posit the existence of libertarian free will and assume that people are morally responsible. That is, one might rationally assume this without epistemic justification, as this is the only way to make sense of the principle of ends, which one has already come to accept for independent moral reasons.

Interpretive Issues Regarding Epistemic Justification and Pragmatic Justification In later sections of this chapter, I will further develop and defend this pragmatic response to Double’s argument by raising possible objections and responding to them. Before moving on to these objections and replies to them, I want to confront an interpretive issue that may be of significance in understanding and assessing my response to Double’s argument. Recall that Double says most libertarians believe there is scant epistemic justification for the existence of libertarian free will and yet they believe that libertarian free will is necessary for moral responsibility and that we should hold people responsible for their actions. This leads Double to say that libertarians are hardhearted. In his essay Double does not clarify what he means by “epistemic justification.” He says: Immanuel Kant proclaims that we can have no epistemic justification for believing that persons make libertarian choices, but recommends that we postulate on faith alone the existence of trans-empirical selves “in” a noumenal world who (that?) make such choices. (p. 227) He also says Kant defends belief in libertarian free will on pragmatic grounds in arguing that there are good moral reasons to believe in libertarian free will. Further, in explaining the Kantian “pragmatic argument,” Double says libertarians, like Kant, “impute to persons libertarian free will (without epistemic justification) in order to make them act morally better” (p. 230). A few points are in order here. First of all, I think it is a mistake to interpret Kant as arguing that we should believe in libertarian free will so as to encourage people to act morally better. Rather, according to Kant, as practical beings living in community with others we understand ourselves as being bound by certain moral obligations—obligations which reflection tells

Hardheartedness and Libertarianism

159

us we can only possess if we have free will. According to Kant, reflection on our moral understanding leads to the conclusion that we must assume we have free will, a free will which Kant believes can only be rightly comprehended in libertarian terms. As Double notes, William James believes we should assume the existence of libertarian free will to encourage moral behavior, but this is not the nature of the Kantian moral argument for belief in libertarian free will. Having said this, it becomes clear that the Kantian moral argument for belief in libertarian free will is grounded on the attempt to make sense of our moral understanding. We understand ourselves as having certain obligations, such as the obligation to always treat persons as ends, and noting this Kant assumes (or posits) that we must have libertarian free will if these obligations really exist. For Kant the belief in libertarian free will is not justified on the grounds of promoting moral conduct; rather, it is justified as the only way to make rational, coherent sense of what we take to be true as practical beings who understand ourselves as subject to moral obligations. As such, there is a sense in which the Kantian case for belief in libertarian free will might be better understood as an epistemic justification. However, as I note above, Double asserts that according to Kant there is no epistemic justification for such a belief. In a certain sense this is also correct. According to Kant, in the realm of theoretical understanding—that is, in the realm of mathematics, empirical science, and pure metaphysics—there is no good reason to suppose that we have libertarian free will. According to Kant, it is only when we examine our moral understanding that we find justification for belief in libertarian free will. The nonconsequentialist moral justification for libertarian free will which I explicate and endorse is more in line with the way in which Kant actually argues. As such there is a sense in which my argument might be understood as epistemic and not pragmatic. Double himself never explicates the concept of “epistemic justification.” So, it is hard to know whether he would take my nonconsequentialist moral argument as an epistemic justification or not. In what follows, I will refer to my nonconsequentialist moral justification as a pragmatic argument—as it is grounded on what we must believe to make sense of our experience as practical beings who understand themselves as having certain moral obligations. Furthermore I, like Kant and many other libertarians, hold that there is not sufficient epistemic justification for belief in libertarian free will, as there is nothing in the realm of theoretical understanding which gives sufficient proof of its existence. Double argues that libertarians who believe there’s no epistemic justification for belief in libertarian free will and who hold people morally responsible are hardhearted. If he agrees that my nonsequentialist moral argument is a kind of nonepistemic, pragmatic argument and if I can show that this is a rationally defensible alternative, then Double’s argument is problematic. However, if Double takes my nonconsequentialist moral argument as an epistemic argument and if I’m right that it is very close to the spirit of

160

Hardheartedness and Libertarianism

Kant’s argument, then Double is mistaken in saying Kant believes there is no epistemic justification for belief in libertarian free will. Furthermore, in the conclusion of Double’s paper he asserts: Libertarians can answer this paper’s objection if they 1) reject incompatibilsm, or 2) quit holding persons morally responsible or 3) provide epistemic justification that persons make enough libertarian choices to support the practices of expressing reactive attitudes, blaming, and punishing retributively. Because libertarians cannot do either of the first two and remain libertarians, and seem unable to do the third, the three strategies are unpromising. (p. 233) Keeping this in mind, if Double takes my nonconsequentialist argument as an epistemic argument and if I show that this is a rationally defensible argument, I will have gone some way to answering Double’s argument via method (3) cited above.

Initial Objections and Replies Is This Kantian Proposal Still Fundamentally (and Problematically) Consequentialist? In reply to my nonconsequentialist pragmatic argument Double still might insist that it involves a problematic consequentialist calculation. He might say that one is still choosing to believe in the existence of libertarian free will and moral responsibility not due to epistemically sound reasons but solely because doing so has the consequences of propping up the Kantian principle of ends. To achieve the goal of respect for this principle in the community and believing society is better off embracing this principle one encourages belief in libertarian free will and moral responsibility. Additionally, Double might argue that this utilitarian calculation is morally problematic as it is unlikely that pursuit of the possible benefits, if any, which come from endorsing libertarian free will and moral responsibility is justified given the significant probability that this may lead to blaming, resenting, and punishing people who because of their lack of free will are not morally responsible for their conduct and, thus, undeserving of such treatment. However, such a response would miss the point of what I have been arguing. The kind of libertarianism I am describing does not endorse belief in libertarian free will and moral responsibility on the grounds of its good consequences for society. I am not arguing that society is better off embracing the principle of ends. It may be or it may not be, but that is not the point. The point is that persons have a certain dignity that makes them deserving of a certain kind of respect which makes it immoral to treat them as mere means. If one comes to accept this principle of ends for independent moral

Hardheartedness and Libertarianism

161

reasons and then comes to the realization that persons can only have this dignity if they have libertarian free will, then one will have a pragmatic justification for belief in free will without possessing epistemic justification for it. Does Rational Acceptance of the Principle of Ends Require Belief in Libertarian Free Will? Double might also challenge the assumption that we can only make sense of the principle of ends on the assumption that libertarian free will exists. He might argue that this is a controversial claim and without a good argument for it my critique might rightly be regarded as question begging. In response to this criticism, a couple of points are in order. First, it is Double who argues that adoption of the pragmatic argument which says we are morally better off by believing in libertarian free will is essentially a problematic consequentialist mode of thinking. I have shown how a libertarian can make the pragmatic argument on nonconsequentialist grounds. It could be said that I am not obligated to provide a full defense of this nonconsequentialist approach in making my case for the problematic nature of Double’s argument. Second, I have already suggested some of the reasoning that would support the idea that rational acceptance of the principle of ends does require a belief in libertarian free will. As I have already argued, to treat someone as an end is to respect their capacity to live and act in accord with their freely formed choices. As such, the adoption of this principle presupposes that people have the capacity to make free choices. Now, Double might say, “Yes, but this does not require a belief in libertarian free will.” To which my reply would be: Correct. If compatibilism could be shown to make coherent sense of the nature of human freedom and responsibility, then belief in libertarian free will would not be required to make sense of this principle. But, as I’ve noted, most libertarians do not find compatibilism to be plausible and there are some pretty good reasons to think this. Again, consider Van Inwagen’s consequence argument, Pereboom’s four-case argument, Kane’s argument from ultimate responsibility, as well as my own. Furthermore, while there are objections to these arguments and the debate about the merit of these arguments is still ongoing, one can still be epistemically warranted in rejecting compatibilism on the basis of such arguments. Surely, one need not have a reply to every possible objection to one’s argument to be warranted in accepting it. At this point, instead of claiming that I have begged the question, Double might argue that in point of fact a belief in free will is not required to make sense of the Kantian principle of ends. He might say that the reason why people should be treated as ends is that they have the rational capacities that make them able to understand the world and the consequences of their actions and they are able to deliberate and make choices in accordance with such knowledge. It could be argued that even if such choices are not freely

162

Hardheartedness and Libertarianism

made by persons we should still respect the capacities people have to make such choices by treating them as ends.4 Such a response would be misguided. To see this, consider that someone, call him “the Puppetmaster,” has the power and knowledge to take every young child that is not yet of the age to reason and deliberate and impose a set of beliefs and values and reasoning skills upon them such that they would then deliberate, choose, and act in accord with these. Further, imagine that the Puppetmaster is kind and wise and that he endows every child with good values, true beliefs, and sound reasoning abilities. Consequently, when the children begin to think and reason and choose they are always led to make the right decisions. Finally, imagine that once the Puppetmaster has given these children their beliefs, values, and reasoning abilities he does not interfere with them in later life; rather, he lets them think, reason, and choose in accord with the mental programming he has provided for them. Now, would the Puppetmaster have violated the principle of ends in doing this to every child? It seems fairly obvious that he would have. However, if you think we can make sense of the principle of ends without believing in free will but that we can make sense of it just in terms of the human capacity for choice, then you will have a hard time explaining how the Puppetmaster has violated this principle. For, according to the example the Puppetmaster has in no way intruded upon the ability of persons to make choices. He has simply given them the beliefs, values, and reasoning abilities that will dictate how they will deliberate and choose in later life. Had he not done this, then on the deterministic model genetics and environment would have provided the mental programming; and done a worse job of it, I might add, since the Puppetmaster provides programming that always leads to right action. In contrast, if we understand the principle of ends as involving a belief in free will, then we can make sense of the wrongness of the Puppetmaster’s action. That is, if we think that persons are to be treated as ends because of their capacity for free choices, then the Puppetmaster has clearly violated the principle of ends. The Puppetmaster makes everyone such that they choose in accord with the mental programming he has provided for them. While he does not interfere with their capacity to choose, he does interfere with the freedom of their choices in the sense that the agents subjected to his programming do not have the freedom to shape their own beliefs, values, and decision-making style. It might be objected that I am setting the requirements of human freedom at ridiculous heights—that I am assuming human freedom involves a capacity to create one’s own character and that this makes the standard of human freedom unattainable since we must all start with the given of genetic and environmental input. However, such a retort is misguided. I’m only advocating that the kind of freedom needed to make sense of the principle of ends includes the ability to have a role in shaping one’s own beliefs, values, and decision-making strategies. This does not require an ability to create oneself ex nihilo. Of course, we have to start with what is given to us from the

Hardheartedness and Libertarianism

163

lottery of genetics and environment, but from there we must have the freedom to critically evaluate our inherited system of beliefs and values and to accept or reject what we are initially given. Compatibilists think we can get the requisite kind of freedom to do this on a deterministic model, but most libertarians don’t think this will suffice. Nonetheless, whether we conceive of this freedom in compatibilist or libertarian terms, my point is that without this limited freedom to self-create (not to self-create ex nihilo) which the Puppetmaster would deny us, then we don’t have the freedom required by the concept of personhood which is invoked in the principle of ends. Thus, if one does not believe we have such freedom, then one will not be able to make sense of the fact that were the Puppetmaster to do this to every child he would be violating the principle of ends. Does Appealing to the Principle of Ends in Supporting the Pragmatic Argument Lead to Contradiction? Another objection to my argument might note that when we hold people responsible for wrongdoing and blame and punish them while believing that we have insufficient epistemic justification for belief in their guilt, then we treat them as mere means, violating the principle of ends. Thus, if one concedes that there is no epistemic justification for belief in libertarian free will and one believes libertarian free will is necessary for moral responsibility and one still holds people morally responsible for wrongdoing, then one violates the principle of ends. In this way, to make appeal to the principle of ends in supporting the pragmatic argument for holding people responsible without epistemic justification involves one in a contradiction. In response to this, I suggest that in the arena of practical reason—where, among other things, we assess the moral value of our own actions and those of others and we try to determine what is right to do—if we are to rationally employ the principle of ends, then we must assume there is free will. Further, the assumption that there is free will legitimizes attributions of moral responsibility and blaming and praising and punishing and rewarding. If in the sphere of practical reason I am led to wonder what it is right and wrong to do and these quandaries lead me to adopt the principle of ends on the basis of good reasons, then I am warranted in my acceptance of this principle and acting in accord with it even though it commits me to a belief in free will that I cannot epistemically justify in the domain of theoretical reason. If I am right about this, then I can be warranted in my attributions of moral responsibility and in my praising and blaming and rewarding and punishing, and doing so involves me in no contradiction. Here Double might say that there surely is a contradiction, because we justly say that a person violates the principle of ends when he blames and punishes another person while he knowingly lacks sufficient evidence of his guilt. Thus, if we say that it’s acceptable to blame and punish when we know there is insufficient epistemic justification for belief in the existence of free

164

Hardheartedness and Libertarianism

will, then there is a contradiction. For if the principle of ends is violated in the former case, then it must be violated in the latter case. However, such an argument is grounded on confusion. The reason we take care not to blame and punish persons based on insufficient evidence of their guilt is because we already take them to be ends in themselves with the power of free choice that makes them deserving of the respect which is commanded by the principle of ends. If there’s a rattlesnake in the road where my children are playing, then I may kill it or, at least, forcibly remove it from the road to protect them. If there’s a suspicious looking person walking in the road where my children are playing, I’m not entitled to kill him nor forcibly remove him from the road. And why not? Because as a human being the principle of ends applies to him and I should not bring harm upon him unless he does through his own free will commit certain acts which merit a response that may be harmful to him. If the principle of ends is the central element of my moral outlook, then I will hold to the principle that people should be regarded as innocent until there is sufficient evidence of their guilt. But a proper understanding of the principle of ends is grounded on the presumption of free will. Consequently, if rational moral considerations lead me to the adoption of the principle of ends, then I can without contradiction adhere to the principle of ends and the doctrine of innocent until proven guilty, even though I admit that from the perspective of theoretical reason it is an open question as to whether free will exists. Shouldn’t the Dictates of Theoretical Reason Be Given Greater Weight in Our Thought and Action Than the Dictates of Practical Reason? Another objection might note how I am suggesting that practical reason might lead us to the acceptance of a moral principle—the principle of ends—which, as I argue, presumes the existence of free will. The critic might also note that I concede that in the domain of theoretical reason there is no sufficient epistemic justification for this belief in free will; rather, it is an open question whether it exists. Here the critic might assert that theoretical reason—what reason dictates regarding science, math, logic, metaphysics, and so forth—should have primacy of place in our thinking and how we live our lives. Thus, if a moral principle entails adoption of a belief that theoretical reason cannot support, such as a belief in free will, then we should not embrace the moral principle. In response, I want to first note that when a moral principle conflicts with something that we clearly know to be true in the realm of theoretical reason then we should reject the moral principle. But such is not the case regarding the issue before us. Rather, my view, like that of most libertarians, is that it is an open question whether free will exists; there’s not a whole lot of evidence for its existence or for its nonexistence. As I’ve argued, rational acceptance

Hardheartedness and Libertarianism

165

of the principle of ends involves a presumption of the existence of free will. If I have good moral reasons to adopt the principle of ends, then I don’t see why I should withhold from adopting it just because theoretical reason provides no sufficient evidence of free will. To suggest that I should wait until there is theoretical proof of free will is to give theoretical reason an exalted status without good reason for doing so.

Double’s Response Considered In a recent publication (2017), Richard Double provides a reply to my argument. He notes how my argument hinges upon my claim that the principle of ends presupposes that persons have some variety of free will, either compatibilist or libertarian. He goes on to reject this claim, stating, “I can acknowledge the intuitional pull of the principle of ends without saying that the principle of ends presupposes either compatibilist or libertarian free will” (2017). Noting that this claim is not unsupported he goes on to critically consider the argument I make for it—my “Puppetmaster Argument”— in which I have the reader imagine a scenario in which every young child that is not yet of the age to reason and deliberate has a set of beliefs and values and reasoning skills imposed upon them such that in living their lives they will deliberate, choose, and act in accord with these values imposed upon them by the Puppetmaster. Recall that (1) the Puppetmaster imposes good values, true beliefs, and sound reasoning abilities upon all of these children and (2) I go on to suggest that were the Puppetmaster to do this to all of these children it would amount to a violation of the Kantian principle of ends. I argue that if you think we can make sense of the principle of ends without believing in free will then you will have a hard time explaining how the Puppetmaster has violated this principle. In response to this argument, Double says that he is of two minds. On the one hand, he is inclined to (a) reject the idea that the Puppetmaster treats these children as mere means, violating the principle of ends, “by taking away [their] free will.” In support of this he says that my Puppetmaster plays a role that is “an irrelevant addition to an otherwise deterministic picture. Lemos has located the Puppetmaster so gingerly in the scenario that his presence is irrelevant to whether or not the children are treated as ends” (2017). On the other hand, he also feels that (b) perhaps the Puppetmaster does treat the children as means, but he thinks this can be explained without assuming that the children have any free will. I will have more to say about this shortly. Before doing so, I would note that Double concludes his essay saying that he does not have a settled view on which of these two responses he favors, but he feels that either strategy is “adequate to undo the force of the Puppetmaster thought-experiment” (2017). Now, regarding strategy (a), here Double suggests that for someone with compatibilist intuitions there aren’t any grounds to say that the Puppetmaster takes away the free will of these children and, thus, that they are

166

Hardheartedness and Libertarianism

treated as mere means. He notes that he has some “compatibilist-like intuitions” which move him to make this point. Here a couple of points are in order. First, it is true that on some compatibilist models there won’t be any reason to believe that these children have been robbed of their own free will. For instance, on Walter Stace’s view (1952), where one’s act is free as long as it is caused by psychological states internal to the agent, the interventions of the Puppetmaster would pose no threat to their free will. But, some other compatibilist models might not be consistent with such a conclusion, as some compatibilist theories might not regard choices as free if made from such manipulated mental states. (See, for instance, the historicist externalist models of Mele (1995), Fischer and Ravizza (1998), and Haji (1998.) It is my suspicion that the latter kinds of compatibilist theories will be more plausible than the former. Second, and more important, my argument is intended to show how libertarians can reasonably be led to embrace the view that we have libertarian free will and engage in blame and punishment without epistemic justification for their belief in free will. I argue that if libertarians have independent moral reasons to think that the principle of ends is true and independent reasons to think that compatibilism is false, then they can reasonably be led to think that we must have libertarian free will. The idea here is that if the principle of ends is true and the basis for it is the fact that human beings have a capacity for autonomy that requires free will and if there are good reasons to reject compatibilism, then we have good reasons to believe we have libertarian free will. I mention these points because, while it is true that compatibilist intuitions may lead one to think that the children in my example may still be free despite the manipulations of the Puppetmaster, these points carry no weight against my argument. This is because my argument is premised on the idea that one can have good reason to reject such compatibilist intuitions. Again, Peter Van Inwagen’s consequence argument (1983) or Derk Pereboom’s four-case argument (2001) or Robert Kane’s argument from ultimate responsibility (1996) might serve as the rational basis for rejecting compatibilist intuitions. Double’s strategy (b) suggests perhaps the Puppetmaster does treat the children as mere means but we can make sense of this fact without supposing that he does so by violating their free will. He writes: I think we can explain the Puppetmaster’s violation of the principle of ends by simply citing the unusual causal etiology of the children’s choices. After all, we would be shocked to discover that we are subject to the Puppetmaster’s machinations ourselves. Imagine the distress of thinking: “But for the Puppetmaster’s agency, I would not have lived my life the way I did.” Putting myself in the children’s position, I would feel embarrassed to learn of the Puppetmaster’s role in my life. (2017)

Hardheartedness and Libertarianism

167

The point made here doesn’t clarify how we can make sense of the Puppetmaster’s treating these children as mere means without an appeal to his intrusion on their free will. It is true that were the Puppetmaster to do this to children, it would make the causal etiology of their choices unusual. But consider that a coach might condition his players in unusual ways to respond to certain game situations; however, this would not by itself imply that he treats them as mere means. Additionally, while it is true that when the children become adults they may be shocked to learn that they have been programmed to believe, think, and act as they do by the Puppetmaster, this by itself does not mean that they have been treated as mere means. A psychologist might make it clear to me that many of my thoughts and choices and much of my reasoning is a consequence of the way my parents raised me and I might be shocked and embarrassed to learn this (perhaps, because I now see that I took misplaced pride in my accomplishments) but this would not mean that my parents had treated me as mere means. Rather, they might well have treated me with love and respect while passing on certain values and beliefs and virtues that have made me think and reason and decide as I do. Perhaps, Double’s point is that we should not do things to others that will lead them to be shocked or embarrassed. When we do this to others without their consent, then we violate the principle of ends, treating others as mere means. Since the Puppetmaster programs the decision-making strategies of these children without their consent and since they would be shocked or embarrassed to find about this later, he violates the principle of ends. But, this too seems problematic. Suppose he programmed their decision-making strategies for them and made them so that they could never find out that he has programmed them. It seems fairly obvious that he would still have violated their autonomy. Suppose that some evil genius has the ways and means to and does gradually manipulate my consciousness over several years so that I no longer hold a certain set of values and decision-making procedures but instead I hold the beliefs and values he wants me to hold, and suppose he makes it impossible for me to find out about this manipulation. This still seems to be a violation of the principle of ends. Double also claims that, “To say that the only way the Puppetmaster can violate the children’s status is by denying them compatibilist or libertarian free will is tantamount to saying that free will trumps all the other qualities that make persons worthy of respect—things like intelligence, creativity, fortitude and sympathy—which do not require free will” (2017). However, my argument does not commit me to such a view. I am not saying this is the only way the Puppetmaster can violate the children’s status as ends; rather, I am saying that given the scenario laid out in the thought experiment this is the way he violates their status as ends. Notice the Puppetmaster is not portrayed as violating their intelligence, creativity, fortitude, or sympathy. They may retain all of these other powers and abilities; but they lose their ability to freely construct their own moral characters because the Puppetmaster

168

Hardheartedness and Libertarianism

intervenes early on in their lives in such a way that they lose this capacity. It is consistent with my view to hold that infringements on persons’ intelligence, creativity, and so forth may also constitute violations of the principle of ends.

Conclusion I want to note that Richard Double is to be commended for putting his challenge to libertarians in such clear and logical terms. The case he has made presents a worthy challenge to libertarians and it demands a thoughtful response. In providing a response, I have chosen to focus on the pragmatic objection to his argument. Double foresees this kind of objection and he responds to it. In doing so, however, he mistakenly assumes that any such response must be grounded on consequentialist considerations. Contra Double, I have tried to show that there is a plausible nonconsequentialist way of developing and defending the pragmatic response to his argument. In so doing, I hope to have shown how it is that libertarians can be pragmatically justified in holding persons responsible for their actions while admitting they lack epistemic justification for their beliefs in libertarian free will. In concluding I would note that the libertarian perspective might sometimes play a role in leading some people to be overly harsh in their blaming and punishing of persons. When we view persons as possessing a libertarian free will that gives them ultimate responsibility for their character and actions, it can easily lead us to think wrongdoing merits equal levels of blame and punishment directed towards anyone who has committed the same offense. But here we have to be careful. Just because two persons have libertarian free will, it does not mean their life circumstances and the pressures and temptations they face are the same. It is unjust not to take these matters into consideration when levying blame and punishment upon persons. It may well be that one thief or drug dealer deserves less punishment than another even if they’ve committed the same crimes and both have libertarian free will. This is because we should in blaming and punishing acknowledge that one might have faced greater pressures and temptations, making it more difficult for him to act rightly. These considerations are perfectly consistent with a libertarian perspective and they must be kept in mind by libertarians so as to avoid actually being or becoming hardhearted.

Notes 1. For works by other authors who’ve either made similar arguments to Double or expressed agreement with his “hardheartedness” argument against libertarianism, see Derk Pereboom (2001, 198–199, 2014, 177–178), Tammler Sommers (2012, 151–152), Manuel Vargas (2007, 140), and Bruce Waller (2011, 40). 2. Double cites Immanuel Kant and William James as two key historical figures who accept a libertarian view and who believe there is no sufficient epistemic justification for the view. He cites the following recent and contemporary figures

Hardheartedness and Libertarianism

169

as examples of libertarians who believe there is no sufficient epistemic justification for belief in the existence of libertarian free will: Roderick Chisholm (1976); Richard Taylor (1966); Peter Van Inwagen (1983); William Rowe (1995); Timothy O’Connor (1995a); Randolph Clarke (1995); Robert Kane (1996); and Mark Balaguer (1999). The libertarian views of the latter four thinkers are developed further in O’Connor (2000); Clarke (2003); Kane (2007a, b, 2011); and Balaguer (2010). 3. For critical replies to these empirical studies see James Miles (2013) and Bruce Waller (2011, 279–282). For a defense of these studies in response to Miles’s criticisms, see Vonasch and Baumeister (2013). 4. For a defense of this approach see Derk Pereboom (2001, 150–152).

9

Conclusion

What I have tried to do in this book is to show that a basically Kanean event-causal libertarian view of free will and moral responsibility is plausible and coherent. While I do think the Kanean view runs into problems regarding its conception of dual efforts of will, I have in Chapter Five proposed ways of modifying the view so as to avoid those problems. Recall that I argued the phenomenological problem and the problem of irrationality continue to plague Kane’s view, and I suggested two new ways of conceptualizing the nature of self-forming acts (SFAs) so as to avoid these problems. One of these proposals was more Kanean in nature, suggesting that we preserve the idea of dual efforts of will but that we see them not as efforts to choose A and to choose B but as efforts to keep the option of choice A alive and the option of choice B alive. The other proposal involved a more radical departure from Kane’s view of SFAs, suggesting that we see SFAs as undetermined choices which result from the way in which agents actively weight the competing possible courses of action considered in deliberation. Modifying some ideas from Nozick (1981), it was suggested that perhaps in SFAs the agent engages in a causally undetermined assigning of weights to competing reasons during deliberation and this then dictates which choice gets made in our SFAs. As I said in Chapter Five, I am unsure which of these models is preferable. It is my hope that they will draw the attention of others so that we might in time discover an even more coherent and defensible conception of libertarian free will. It is important that there be continued research and study of this topic, as the stakes in this debate are high. As I’ve argued in the later chapters of this book, it is morally important whether or not we conceive of human beings as having the kind of free will which grounds claims of just desert. I argued that when it is denied that we have such free will this leads into a conception of human beings that poses a threat to human dignity and the rights of the innocent. It also weakens the notion of moral obligation in significant respects. Some smart and resourceful philosophers, such as Derk Pereboom, Gregg Caruso, and Bruce Waller, as well as others, would disagree with these claims. So, as is common in philosophy, my arguments on this front are controversial. But, I would note that I am not alone in

Conclusion

171

thinking that there are serious moral threats that come with denying we have the kind of free will which grounds just deserts. Immanuel Kant is on my side here, and more recently Saul Smilansky (2000, 2005, 2011, 2017), who rejects libertarianism, has nonetheless done much to draw attention to the moral problems which come with skepticism about free will and moral responsibility. There will, of course, also be compatibilist critics of the views I defend in this book. As I’ve said several times, there are powerful criticisms of compatibilism which have been made by Peter Van Inwagen, Derk Pereboom, Robert Kane, and Bruce Waller. I’ve always been sympathetic with their arguments, and I’ve tried to make my own contribution to the case against compatibilism in Chapter Three of this book. Compatibilist responses to my critique and the critiques made by others should be welcomed and carefully considered. I say this because there is one great advantage compatibilism has over libertarianism—namely, that were there a plausible and coherent compatibilist account of free will then the question of the existence of free will would be solved. Indeed, I wish I could say with intellectual honesty that compatibilism about free will is true, but I cannot. Thus, in championing free will I’ve been led to endorse the libertarian view, and as this book would suggest to do this well is not easy! For instance, even if I’ve shown that the view is coherent (something which is itself very difficult to do), I’ve done little, if anything, to show that libertarian free will exists. Ultimately, to know whether it exists will require further discoveries and theorizing in neuroscience and psychology. As Kane has said, There are empirical aspects of the free will issue that mere philosophical speculation cannot co-opt. If free will of a nondeterminist kind should exist in nature, then the atoms must swerve to make room for it, and they must swerve in places where it matters—in the brain, for example. But neither can philosophical speculation be indifferent to the empirical questions. (Kane 1996, 17) Empirical psychology and neuroscience are relatively new disciplines. There is much that we still don’t know about how the mind operates. As Alfred Mele (2006, Ch.2, 2013) and Mark Balaguer (2010, Ch.4) show, it is an open empirical question whether or not science can prove or disprove the truth of libertarian free will. Again, it would be nice if we could prove the truth of compatibilism. For then we would not be left with the lingering worry about whether in fact anyone has desert bearing free will. In the end, though, I just don’t think compatibilism has been shown to be true and I believe there are good reasons to think it is false. And, despite worries about the existence of free will I believe for reasons already given that we should live and act as though we have it. But we should do so while being sensitive to the psychological and

172

Conclusion

social realities of ourselves and others, so as not to be overly harsh in our assessments of ourselves and others when we and others fall short of moral norms and fail in our obligations. If we have libertarian free will, it does not mean that everyone has an equal level of control over their actions and the path their lives take and the formation of their character. If we have libertarian free will, it does not mean that everyone is equally responsible for their actions and their character. One can be a libertarian without being overly punitive and judgmental regarding moral failings or overly complacent in one’s advantages, thinking of them all as justly deserved. It is here where the morally conscientious libertarian must learn to dance in his dealings with himself and others.

Bibliography

Allen, R. F. (2005). Free will and indeterminism: Robert Kane’s libertarianism. Journal of Philosophical Research 30: 341–355. Ayer, A. J. (1954). Freedom and necessity. In Philosophical Essays, 3–20. New York: St. Martin’s Press. Baker, G. and Gollub, J. (1990). Chaotic Dynamics: An Introduction. Cambridge: Cambridge University Press. Balaguer, M. (1999). Libertarianism as a scientifically reputable view. Philosophical Studies 93: 189–211. ———. (2004). A coherent, naturalistic, and plausible formulation of libertarian free will. Nous 38: 379–406. ———. (2010). Free Will as an Open Scientific Problem. Cambridge, MA: MIT Press. ———. (2014). Replies to McKenna, Pereboom, and Kane. Philosophical Studies 169: 71–92. Baumeister, R., Masicampo, E., and DeWall, C. (2009). Prosocial benefits of feeling free: Disbelief in free will increases aggression and reduces helpfulness. Personality and Social Psychology Bulletin 35: 260–268. Berkeley, G. ([1710] 1998). A Treatise Concerning the Principle of Human Knowledge, ed. J. Dancy. Oxford: Oxford University Press. Bishop, R. C. (2011). Chaos, indeterminism, and free will. In The Oxford Handbook of Free Will, 2nd ed., ed. R. Kane, 84–100. New York: Oxford University Press. Brembs, B. (2010). Towards a scientific concept of free will as a biological trait. Proceedings of the Royal Society: Biological Sciences 298: 930–939. doi:10.1098/ rspb.2010.2325 Briegel, H. J. and Mueller, T. (2015). A chance for attributable agency. Minds and Machines 25: 261–279. Caruso, G. (2012). Free Will and Consciousness: A Determinist Account of the Illusion of Free Will. Lanham, MD: Lexington Books. ———. (2016). Free will skepticism and criminal behavior: A public health quarantine model. Southwest Philosophy Review 32: 25–48. Caruso, G. and Pereboom, D. (forthcoming). Hard-incompatibilist existentialism: Neuroscience, punishment, and meaning in life. In Neuroexistentialism: Meaning, Morals, and Purpose in the Age of Neuroscience, ed. G. Caruso and O. Flanagan. New York: Oxford University Press. Chisholm, R. M. ([1964] 2002). Human freedom and the self. The Lindley Lecture. Department of Philosophy, University of Kansas. In Free Will, ed. R. Kane, 47–58. Oxford: Blackwell.

174

Bibliography

———. (1976). Person and Object: A Metaphysical Study. La Salle: Open Court. Clarke, R. (1995). Toward a credible agent-causal account of free will. In Agents, Causes, and Events, ed. T. O’Connor, 201–215. New York: Oxford University Press. ———. (2002). Libertarian views: Critical survey of noncausal and event-causal accounts of free agency. In The Oxford Handbook of Free Will, ed. R. Kane, 356–385. New York: Oxford University Press. ———. (2003). Libertarian Accounts of Free Will. Oxford: Oxford University Press. ———. (2012). How to manipulate a compatibilistically free agent. American Philosophical Quarterly 49: 139–149. Copp, D. (2003). ‘Ought’ implies ‘can’, blameworthiness, and the principle of alternative possibilities. In Moral Responsibility and Alternative Possibilities: Essays on the Importance of Alternative Possibilities, ed. M. McKenna and D. Widerker, 265–299. Burlington, VT: Ashgate. Dennett, D. (1978). On giving libertarians what they say they want. In Brainstorms, 286–299. Cambridge, MA: MIT Press. ———. (1984a). Elbow Room. Cambridge, MA: MIT Press. ———. (1984b). I could not have done otherwise—so what? The Journal of Philosophy 81: 553–567. ———. (2003). Freedom Evolves. New York: Viking Press. Doyle, R. (2011). Free Will: The Scandal in Philosophy. Cambridge, MA: I-Phi Press. Double, R. (1991). The Non-Reality of Free Will. New York: Oxford University Press. ———. (2002). The moral hardness of libertarianism. Philo 5: 226–234. ———. (2017). The hard-heartedness of some libertarians: A reply to John Lemos. The Journal of Philosophical Research 42. doi:10.5840/jpr2017511103 Edwards, P. (1958). Hard and soft determinism. Originally in Determinism and Freedom in the Age of Modern Science, ed. Sidney Hook. New York: New York University Press. Reprinted in Free Will, ed. R. Kane, 59–67. Oxford: Blackwell. Ekstrom, L. (2000). Free Will: A Philosophical Study. Boulder: Westview Press. ———. (2002). Libertarianism and Frankfurt-style cases. In The Oxford Handbook of Free Will, ed. R. Kane, 309–322. New York: Oxford University Press. ———. (2003). Free will, chance, and mystery. Philosophical Studies 113: 153–180. Fischer, J. M. (1982). Responsibility and control. The Journal of Philosophy 89: 24–40. ———. (1994). The Metaphysics of Free Will. Oxford: Blackwell. ———. (1999). Recent work on moral responsibility. Ethics 110: 93–139. ———. (2000). Excerpts from John Martin Fischer’s discussion with members of the audience. Journal of Ethics 4: 408–417. ———. (2006). My Way: Essays on Moral Responsibility. New York: Oxford University Press. Fischer, J. M. and Ravizza, D. (1998). Responsibility and Control: A Theory of Moral Responsibility. New York: Cambridge University Press. Flanagan, O. (2002). The Problem of the Soul: Two Visions of the Mind and How to Reconcile Them. New York: Basic Books. Frankfurt, H. (1969). Alternate possibilities and moral responsibilities. The Journal of Philosophy 66: 829–839. ———. (1971). Freedom of the will and the concept of a person. The Journal of Philosophy 68: 5–20.

Bibliography

175

———. (2002). Reply to John Martin Fischer. In Contours of Agency: Essays on Themes From Harry Frankfurt, ed. S. Buss and L. Overton. Cambridge: MIT Press. Ginet, C. (1990). On Action. Cambridge: Cambridge University Press. ———. (1996). In defense of alternative possibilities: Why I don’t find Frankfurt’s argument convincing. Philosophical Perspectives 10: 403–417. ———. (1997). Freedom, responsibility, and agency. Journal of Ethics 1: 85–98. ———. (2002). Reasons explanations of actions: Causalist versus noncausalist accounts. In The Oxford Handbook of Free Will, ed. Robert Kane, 386–405. New York: Oxford University Press. Glimcher, P. W. (2005). Indeterminacy in brain behavior. Annual Review of Psychology 56: 25–56. Goetz, S. (1988). A noncausal theory of agency. Philosophy and Phenomenological Research 49: 303–316. ———. (1997). Libertarian choice. Faith Philosophy 14: 195–211. Griffith, M. (2010). Why agent caused actions are not lucky. American Philosophical Quarterly 47: 43–56. Haji, I. (1998). Moral Appraisability. New York: Oxford University Press. ———. (1999). Indeterminism and Frankfurt-style examples. Philosophical Explorations 2: 42–58. ———. (2000a). Control requirement for moral appraisability: An asymmetry. Journal of Ethics 4: 351–356. ———. (2000b). Excerpts from Ishtiyaque Haji’s discussion with members of the audience. Journal of Ethics 4: 368–381. ———. (2005). Libertarianism, luck and action explanation. Journal of Philosophical Research 30: 321–340. Haji, I. and Cuypers, S. (2001). Libertarian free will and CNC manipulation. Dialectica 55: 221–238. Hameroff, S. and Penrose, R. (1996). Conscious events as orchestrated space-time selections. Journal of Consciousness Studies 3: 36–53. Hare, R. M. (1989a). Punishment and retributive justice. In Essays on Political Morality. Oxford: Oxford University Press. ———. (1989b). Some reasoning about preferences. In Essays in Ethical Theory. Oxford: Oxford University Press. Heisenberg, M. (2013). The origins of freedom in animal behavior. In Is Science Compatible With Free Will?, ed. A. Suarez and P. Adams, 95–103. New York: Springer. Hilborn, R. (2001). Chaos and Nonlinear Dynamics: An Introduction, 2nd ed. Oxford: Oxford University Press. Hobbes, T. ([1651] 1958). Leviathan. Indianapolis: Bobbs-Merrill. Hobbs, J. (1991). Chaos and indeterminism. Canadian Journal of Philosophy 21: 141–164. Hodgson, D. (2012). Rationality + Consciousness = Free Will. New York: Oxford University Press. Holbach, P. ([1770] 1999). Systeme de la Nature. Trans. H. D. Robinson. In Reason and Responsibility, 10th ed., ed. J. Feinberg and R. Shafer-Landau, 416–421. Belmont: Wadsworth. Hume, D. ([1748] 1974). An Enquiry Concerning Human Understanding. Locke, Berkeley, and Hume: The Empiricists. New York: Anchor Books. Huxley, A. (2006). Brave New World. New York: Harper Collins.

176

Bibliography

James, W. ([1884] 1956). The dilemma of determinism. In The Will to Believe: Human Immortality, 145–183. New York: Dover. Jedlicka, P. (2014). Quantum stochasticity and (the end of) neurodeterminism. In Quantum Physics Meets the Philosophy of Mind, ed. A. Corradini and U. Meixner, 183–197. Berlin: DeGruyter. Kane, R. (1985). Free Will and Values. Albany, NY: SUNY Press. ———. (1996). The Significance of Free Will. Oxford: Oxford University Press. ———. (1999a). Responsibility, luck, and chance: Reflections on free will and indeterminism. Journal of Philosophy 96: 217–240. ———. (1999b). On free will, responsibility, and determinism: Response to Clarke, Haji, and Mele. Philosophical Explorations 2: 105–121. ———. (2002a). Some neglected pathways in the free will labyrinth. In The Oxford Handbook of Free Will, ed. R. Kane, 406–437. New York: Oxford University Press. ———. (2002b). Free Will. Oxford: Blackwell. ———. (2003). Responsibility, indeterminism, and Frankfurt-style cases. In Moral Responsibility and Alternative Possibilities, ed. M. McKenna and D. Widerker, 91–106. Aldershot, UK: Ashgate. ———. (2005). A Contemporary Introduction to Free Will. Oxford: Oxford University Press. ———. (2007a). Libertarianism. In Four Views on Free Will, ed. J. M. Fischer, R. Kane, D. Pereboom and M. Vargas, 5–43. Malden, MA: Blackwell. ———. (2007b). Response to Fischer, Pereboom, and Vargas. In Four Views on Free Will, ed. J. M. Fischer, R. Kane, D. Pereboom and M. Vargas, 166–183. Malden, MA: Blackwell. ———. (2011). Rethinking free will: New perspectives on an ancient problem. In The Oxford Handbook of Free Will, ed. R. Kane, 381–404. New York: Oxford University Press. ———. (2013). Frankfurt-style examples and self-forming actions. In Free Will and Moral Responsibility, ed. I. Haji and J. Caouette, 58–73. Cambridge, UK: Cambridge Scholars Press. ———. (2014a). Torn decisions, luck, and libertarian free will: Comments on Balaguer’s free will as an open scientific problem. Philosophical Studies 169: 51–58. ———. (2014b). New arguments in debates on libertarian free will: Responses to contributors. In Libertarian Free Will: Contemporary Debates, ed. D. Palmer. New York: Oxford University Press. ———. (2016a). The complex tapestry of free will: Striving will, indeterminism, and volitional streams. Synthese. doi:10.1007/s11229-016-1046-8 ———. (2016b). Moral responsibility, reactive attitudes, and freedom of will. Journal of Ethics 1–3: 229–246. Kant, I. ([1785] 1981). The Groundwork of the Metaphysics of Morals. Trans. James W. Ellington. Indianapolis: Hackett. ———. ([1788] 1927). The Critique of Practical Reason. Trans. T. K. Abbott. New York: Longman, Green. Kellert, S. (1993). In the Wake of Chaos. Chicago: University of Chicago Press. Lemos, J. (2007). Kanian freedom and the problem of luck. The Southern Journal of Philosophy 45: 515–532. ———. (2008). Commonsense Darwinism: Evolution, Morality, and the Human Condition. LaSalle: Open Court.

Bibliography

177

———. (2011a). Kane’s libertarian theory and luck: A reply to Griffith. Philosophia 39: 357–367. ———. (2011b). Wanting, willing, trying, and Kane’s theory of free will. Dialectica 65: 31–48. ———. (2013). Hard-heartedness and libertarianism. Philo 16: 180–195. ———. (2014). Libertarianism and free determined decisions. Metaphilosophy 45: 675–688. ———. (2015). Self-forming acts and the grounds of responsibility. Philosophia 43: 135–146. ———. (2016). Moral concerns about responsibility denial and the quarantine of violent criminals. Law and Philosophy 35: 461–483. ———. (2017a). Hard-heartedness and libertarianism again: A rejoinder to double. Journal of Philosophical Research 42. doi:10.5840/jpr2017511104 ———. (2017b). A libertarian response to Dennett and Harris on free will. Rivista Internazionale di Filosofia e Psicologia 8: 231–246. ———. (forthcoming). Kane, Balaguer, libertarianism, and luck. In Free Will and Action: Historical and Contemporary Perspectives, ed. F. Grgic. Dordrecht: Springer. Levy, N. (2005). Contrastive explanations: A dilemma for libertarians. Dialectica 59: 51–61. ———. (2012). Skepticism and sanction: The benefits of rejecting moral responsibility. Law and Philosophy 31: 477–493. Maye, A., Hsieh, C. H., Sugahara, G., and Brembs, B. (2007). Order in spontaneous behavior. PLoS ONE 2: e443. doi:10.1371/journal.pone.00066443 McCall, S. (1984). A Model of the Universe: Space, Time, Probability, and Decision. Oxford: Clarendon Press. McCann, H. (1998). The Works of Agency: On Human Action, Will, and Freedom. Ithaca: Cornell University Press. McKenna, M. (2004). Responsibility and globally manipulated agents. Philosophical Topics 32: 169–192. ———. (2012). Moral responsibility, manipulation arguments, and history: Assessing the resilience of nonhistorical compatibilism. Journal of Ethics 16: 145–174. Mele, A. (1995). Autonomous Agents: From Self-Control to Autonomy. New York: Oxford University Press. ———. (1998). Review of Robert Kane’s The Significance of Free Will. Journal of Philosophy 95: 581–584. ———. (1999a). Kane, luck, and the significance of free will. Philosophical Explorations 2: 96–104. ———. (1999b). Ultimate responsibility and dumb luck. Social Philosophy and Policy 16: 274–293. ———. (2006). Free Will and Luck. New York: Oxford University Press. ———. (2013). Free will and neuroscience. Philosophic Exchange 43: 1–17. ———. (2017). Two libertarian theories: Or why event-causal libertarians should prefer my daring libertarianism to Robert Kane’s view. Royal Institute of Philosophy Supplement 80: 49–68. Mele, A. and Robb, D. (1998). Rescuing Frankfurt-style cases. Philosophical Review 107: 97–112. ———. (2003). Bbs, magnets, and seesaws: The metaphysics of Frankfurt-style cases. In Moral Responsibility and Alternative Possibilities, ed. M. McKenna and D. Widerker, 127–137. Aldershot, UK: Ashgate.

178

Bibliography

Miles, J. (2013). Irresponsible and a disservice: The integrity of social psychology turns on the free will dilemma. British Journal of Social Psychology 52: 205–218. Mill, J. S. (1874). A System of Logic. New York: Harper and Row. Moore, G. E. (1912). Free will. In Ethics, 84–95. Oxford: Oxford University Press. Nielsen, K. (1971). The compatibility of freedom and determinism. In Reason and Practice. New York: Harper and Row. Nozick, R. (1981). Philosophical Explanations. Cambridge, MA: Harvard University Press. O’Connor, T. (1995a). Agent causation. In Agents, Causes and Events, 173–200. New York: Oxford University Press. ———. (1995b). Agents, Causes, and Events. New York: Oxford University Press. ———. (1996). Why agent-causation? Philosophical Topics 24: 143–158. ———. (2000). Persons and Causes: The Metaphysics of Free Will. New York: Oxford University Press. ———. (2009). Agent-causal power. In Dispositions and Causes, ed. T. Hanfield, 189–214. Oxford: Oxford University Press. ———. (2011). Agent-causal theories of freedom. In The Oxford Handbook of Free Will, ed. R. Kane, 309–328. New York: Oxford University Press. Pereboom, D. (2001). Living Without Free Will. Cambridge, UK: Cambridge University Press. ———. (2007a). Hard incompatibilism. In Four Views on Free Will, ed. J. M. Fischer, R. Kane, D. Pereboom and M. Vargas, 85–125. Oxford: Blackwell. ———. (2007b). Response to Kane, Fischer, and Vargas. In Four Views on Free Will, ed. J. M. Fischer, R. Kane, D. Pereboom and M. Vargas, 191–203. Oxford: Blackwell. ———. (2014). Free Will, Agency, and Meaning in Life. New York: Oxford University Press. Pereboom, D. and Caruso, G. (forthcoming). Hard-incompatibilist existentialism: Neuroscience, punishment, and meaning in life. In Neuroexistentialism: Meaning, Morals, and Purpose in the Age of Neuroscience, ed. G. Caruso and O. Flanagan. New York: Oxford University Press. Rawls, J. (1969). Two concepts of rules. In The Philosophy of Punishment, ed. H. B. Acton. London: Macmillan. Reid, T. ([1788] 1969). Essays on the Active Powers of the Human Mind. Cambridge, MA: MIT Press. Rowe, W. (1991). Thomas Reid on Freedom and Morality. Ithaca: Cornell University Press. ———. (1995). Two concepts of freedom. In Agents, Causes, and Events, ed. T. O’Connor, 151–171. New York: Oxford University Press. ———. (2000). The metaphysics of freedom: Reid’s theory of agent causation. American Catholic Philosophical Quarterly 74: 425–446. ———. (2006). Free will, moral responsibility, and the problem of ‘oomph’. Journal of Ethics 10: 295–313. Shadlen, M. (2014). Comments on Adina Roskies, ‘Can neuroscience resolve issues about free will?’. In Moral Psychology, Vol. 4, ed. W. Sinnott-Armstrong, 139–150. Cambridge, MA: MIT Press. Skinner, B. F. (1948). Walden Two. New York: Macmillan. Smart, J. J. C. (1963). Free will, praise, and blame. Mind 70: 291–306. ———. (1973). An outline of a system of utilitarian ethics. In Utilitarianism: For and Against, ed. J. J. C. Smart and B. Williams. Cambridge: Cambridge University Press.

Bibliography

179

Smilansky, S. (1990). Utilitarianism and the ‘punishment’ of the innocent: The general problem. Analysis 50: 256–261. ———. (2000). Free Will and Illusion. Oxford, UK: Clarendon Press. ———. (2005). Free will and respect for persons. Midwest Studies in Philosophy 29: 248–261. ———. (2011). Hard determinism and punishment: A practical reductio. Law and Philosophy 30: 353–367. ———. (2017). Pereboom on punishment: Funishment, innocence, motivation, and other difficulties. Criminal Law and Philosophy 11: 591–603. Sommers, T. (2012). Relative Justice: Cultural Diversity, Free Will, and Moral Responsibility. Princeton: Princeton University Press. Spinoza, B. ([1677] 1992). The Ethics. Trans. S. Shirley. Cambridge, MA: Hackett. Stace, W. ([1952] 1993). The problem of free will. In Reason and Responsibility, 10th ed., ed. Joel Feinberg and R. Shafer-Landau, 435–441. Belmont, CA: Wadsworth. Originally in W. Stace, Religion and the Modern Mind. London: J.B. Lippincott Co. Stapp, H. (2007). The Mindful Universe. Berlin: Springer. Strawson, G. (1986). Freedom and Belief. New York: Oxford University Press. ———. (2000). The unhelpfulness of indeterminism. Philosophy and Phenomenological Research 60: 149–155. Strawson, P. F. (1962). Freedom and resentment. In Freedom and Resentment. London: Methuen. Taylor, R. (1966). Action and Purpose. Englewood-Cliffs: Prentice-Hall. ———. (1992). Metaphysics, 4th ed. Englewood-Cliffs: Prentice-Hall. Tse, P. (2013). The Neural Basis of Free Will: Criterial Causation. Cambridge, MA: MIT Press. Usher, M. (2006). Control, choice, and the convergence/divergence dynamics: A compatibilist probabilistic theory of free will. The Journal of Philosophy 304: 188–213. Van Inwagen, P. (1983). An Essay on Free Will. Oxford: Clarendon Press. ———. (2002). Free will remains a mystery. In The Oxford Handbook of Free Will, ed. R. Kane, 158–177. New York: Oxford University Press. Vargas, M. (2007). Revisionism. In Four Views on Free Will, ed. J. M. Fischer, R. Kane, D. Pereboom and M. Vargas, 126–165. Malden, MA: Blackwell. Vohs, K. and Schooler, J. (2008). The value of believing in free will: Encouraging a belief in determinism increases cheating. Psychological Science 19: 49–54. Vonasch, A. and Baumeister, R. (2013). Implications for free will beliefs for basic theory and societal benefit: Critique and implications for social psychology. British Journal of Social Psychology 52: 219–227. Waller, B. (1988). Free will gone out of control. Behaviorism 16: 149–162. ———. (1990). Freedom Without Moral Responsibility. Philadelphia: Temple University Press. ———. (1998). The Natural Selection of Autonomy. Albany, NY: SUNY Press. ———. (2011). Against Moral Responsibility. Cambridge, MA: MIT Press. ———. (2015). The Stubborn System of Moral Responsibility. Cambridge, MA: MIT Press. Watson, G. (1975). Free agency. Journal of Philosophy 72: 205–220. ———. (1987). Responsibility and the Limits of Evil. In Responsibility, Character, and the Emotions: New Essays in Moral Psychology, ed. F. Schoeman, 256–286. Cambridge: Cambridge University Press.

180

Bibliography

Widerker, D. (1995). Libertarianism and Frankfurt’s attack on the principle of alternative possibilities. Philosophical Review 104: 247–261. Wolf, S. (1987). Sanity and the metaphysics of responsibility. In Responsibility, Character, and Emotions, ed. Schoeman, 45–64. Cambridge: Cambridge University Press. ———. (1990). Freedom Within Reason. Oxford: Oxford University Press.

Index

agent-causal libertarianism 7n4, 8, 10, 11, 28n2 alternative possibilities 40–43; and selfforming acts 11–15 antecedent determining control 19–20, 64–65 Antigone 145 autonomy 155, 157, 166, 167 Baker, G. 114 Balaguer, M. 7n4, 25–27, 28n6, 87–91, 92, 93, 112nn2–4, 114, 169n2, 171 basic desert see desert basic free acts 9; Balaguer’s conception of 87–91; different conceptions of 83–112; Mele’s daring libertarian model of 91–98; a modest libertarian model of 85–87; a modiefied Nozickian model of 106–112; a non-Kanean dual efforts model of 99–106; a simple dual wantings model of 84–85; see also self-forming acts Baumeister, R. 154, 169n3 Bishop, R. 114 blame: and control over self-forming acts 20; and dual efforts canceling out responsibility 60–62; and the effectiveness of the moral “ought” 138–149; and the hard-line response to manipulation cases 48–51; and moral responsibility skepticism 113–116; the pragmatic defense of 163–169; and the problem of hardheartedness 152–172; and relation to free will 1–7 Brembs, B. 114 Briegel, H.J. 114

Caruso, G. 6n2, 115–118, 120–124, 131nn1–3, 132n9, 170 Center for Disease Control (CDC) 120, 132n8 chaos: in neural functioning 12–13 Clarke, Randolph 7n4, 18, 28n2, 28nn6–7, 66–73, 77–81, 81n1, 83–85, 109, 169n2 Clarke, Roger 39–46 classism: and punishing the innocent 130 compatibilism 3–4, 5, 29–30, 113, 114, 156, 157, 161, 166, 171; and CNC manipulation 30–33; and hard-line responses to manipulation cases 46–51; and soft-line responses to manipulation cases 52–56 constraining control (CC) 31 contrastive explanation 19–20, 60–62 Copp, D. 140–141 covert nonconstraining control (CNC) 31–32, 34, 46, 51, 52 daring libertarianism 22, 91–98, 105 Dennett, D. 7nn3–4, 11, 22–23, 28nn4–6, 31, 33, 37, 56n6, 85–87 desert i, 6, 50–51, 115–131, 133, 136, 138–140, 145, 149, 151, 153, 170–171; basic 115–121, 130 determinism 2–4, 6nn1–2, 7n3, 8, 11, 27, 31, 55, 86, 92, 131n1; and compatibilism 3–4, 11; defined 2; and hard determinism 2–3; and libertarianism 4, 8; and semicompatibilism 31; see also hard determinism dignity 6, 115, 131, 133, 151, 157, 160, 161, 170 Double, R. 151–169

182

Index

Doyle, R. 114 dual wantings model: of libertarian free will 84–85

and punishment 113–132; and selfforming acts (SFAs) 12–15 funishment: problem of 118, 131n5

efforts of will 9, 10, 22–28, 35–37, 48, 65, 73–81, 83–112, 170; and plural voluntary control 17–19; and self-forming acts 11–15; see also selfforming acts Ekstrom, L. 7n4, 23–25, 28nn5–7, 83–85, 106; and the irrationality problem with dual efforts 77–81; and the phenomenological problem with dual efforts 73–77 epistemic justification: vs. pragmatic justification 158–160 event-causal libertarianism 7n4, 9, 10, 27, 29, 37, 62, 65, 83, 85, 92, 106, 113, 170; alternative versions of 22–27, 83–112 evidentiary standards: for criminal conviction 118–124, 130, 132n9

Ginet, C. 7n4, 28nn3–4, 54 Glimcher, P.W. 114 Gollub, J. 114 Griffith, M. 62–66 guidance control 21, 31, 82n9

Fischer, J.M. 7n3, 28n4, 31, 32, 33, 37, 47, 52, 53, 54, 56n5, 56n8, 166 Frankfurt, H. 7n3, 28n4, 30, 32–35, 37, 40–46, 47, 48, 52–53, 56n3 Frankfurt-style case 40–46 free will: agent-causal libertarian conception of 8–9; alternative eventcausal libertarian views of 22–27, 84–112; and alternative possibilities 11–15; Balaguer’s model of 87–91; compatibilism about 3–4; as a condition of moral responsibility 1–2; a daring libertarian model of 91–98; a defense of the pragmatic approach to libertarian views of 154–169; denial of 2–3; a dual wantings model of 84–85; and efforts of will 12–15; the existence of 27–28; and hard determinism 2–3; and hard incompatibilism 3; Kane’s libertarian theory of 8–28, 57–60; libertarianism about 4; and manipulation arguments 29–56; a modest libertarian model of 85–87; a modified Nozickian model of 106–112; and moral obligation 133–150; noncausal libertarian conception of 8–9; a non-Kanean dual efforts view of 99–106; and plural voluntary control (PVC) 17–19; the problem of hardheartedness and libertarian views of 151–169;

irrationality problem: and dual efforts of will 77–81, 84, 99–100, 105, 170 isolation: of criminals 117, 131n4, 132n8; vs. quarantine 120

habeas corpus 125, 126, 130 Haji, I. 28n4, 34–35, 37, 47, 48, 81n2, 146, 166 Hameroff, S. 114 hard determinism 2–3, 6n1, 156 hardheartedness 6, 151–168 hard incompatibilism 3, 6n2, 131n1, 156–157 Harris, R. 141–142 Heisenberg, M. 114 Hobbs, J. 114 Huxley, A. 32, 46

James, W. 152–154, 156, 159, 168n2 Jedlicka, P. 114 just desert see desert Kane, R. viii, 5, 7n4, 113–114, 156, 161, 166, 169, 170–171; critiques of his libertarianism considered 60–81; his CNC manipulation argument explained 30–33; his libertarian theory explained 8–28, 57–60; his view contrasted with alternative models of basic free actions 83–112; manipulation arguments against his view 34–46; overly speculative nature of his view 99, 105 Kant, I. 146, 152–168, 171 Keller, S. 114 Levy, N. 60–61, 131n5 libertarianism 3, 5–6, 7n4, 171; alternative event-causal versions of 83–112; Balaguer’s 87–91; critiques of Kane’s version of 60–81; daring 91–98; and the existence problem 4, 27–28; Kanean vs. agent-causal varieties of 8–11; Kanean vs. other event-causal versions of 22–27;

Index Kane’s version of 8–28, 57–60; manipulation arguments against 34–46; modest 85–87; modified Nozickian version of 106–112; non-Kanean dual efforts version of 99–106; pragmatic defense of 113–115; and the problem of hardheartedness 151–168; and selfforming acts (SFAs) 11–22; a simple dual wantings model of 84–85 luck: problem of 19–22, 23, 27, 28n5, 58–60, 62, 81n3, 85–88, 90, 97–98, 112n3, 116, 146 manipulation arguments 5, 26, 29–56, 57, 113; against compatibilism 30–33, 46–55; against Kane’s libertarianism 34–46 Maye, A. 114 McKenna, M. 30, 32–33, 52, 54, 56n3, 56n5, 56n8; and hard-line reply to manipulation argument 46–51 Mele, A. 7n4, 22–23, 28nn4–6, 37, 47, 52–54, 56n5, 56n8, 59, 66–73, 81n5, 104–105, 109, 166, 171; and daring libertarianism 91–98; and modest libertarianism 85–87 modest libertarianism 22, 85–87, 92, 107 modified Nozickian view: of free will 106–112 moral failure: and moral obligation 133–150; and link to free will 148 moral obligation 6, 115, 133–150, 151, 158–159, 170, 172; conflicts of 145–147 moral responsibility: alternative event-causal views of 22–27; and alternative possibilities 11–15; and basic free actions 83–112; and CNC manipulation 30–33; critiques of compatibilist views of 29–33, 46–56; denial of 113–132; existence of 27–28; Kane’s conception of 8–28; and manipulation arguments 29–56; and moral obligation 133–150; and plural voluntary control 17–19; pragmatic defense of belief in 154–169; and problem of hardheartedness 151–169; and problem of punishment 113–132; and relation to free will 1–7 Mueller, T. 114 neural coin toss 26–27, 89–92, 107 neural networks 13

183

noncausal libertarianism 8–11, 22 non-Kanean dual efforts view: of free will 99–106 Nozick, R. 106–112, 170 obligation see moral obligation “ought”: of admonishment vs. judgment 136–140, 149; see also moral obligation “ought” implies “can” 133–135, 137, 138, 140–150; see also moral obligation parallel processing 14, 73–74, 102 Penrose, R. 114 Pereboom, D. 6n2, 18, 29, 35–39, 113, 131nn1–3, 131n5, 132n9, 156, 161, 166, 168n1, 169n4, 170, 171; and the quarantine model of criminal justice 115–124 phenomenological problem: and dual efforts of will 73–77, 99, 101, 105–106, 170 plural voluntary control 17–20, 34, 43–44, 48, 62–65, 95, 99 praise: and dual efforts canceling out responsibility 60–62; and the effectiveness of the moral “ought” 138–149; and the existence question concerning free will 114–115; and hard-line responses to manipulation cases 48–51; and moral responsibility skepticism 113–116; and relation to free will 1–7 principle of ends 155–158, 160–168 principle of universalizability 157 proper function 148 punishment: and connection to blame 49–51; connection to free will 1; and the effectiveness of the moral “ought” 138–149; moral responsibility denial and the problem of 113–132; the pragmatic defense of 163–169; and the problem of hardheartedness 152–172; and problem of punishing the innocent 118–132; and wise parenting 82n10 Puppetmaster argument 162–163, 165–167 quantum indeterminacy 20–21, 38, 90, 98 quarantine model: of criminal justice 115–118, 120–124, 130–131, 131n3, 132nn8–9

184

Index

racism: and punishing the innocent 127, 130 Ravizza, M. 7n3, 31, 33, 37, 47, 52, 53, 54, 56n5, 166 Rawls, J. 118–120 regret 101, 139–140, 145 resentment 49, 51, 153 rights: human rights 115, 121, 126, 133, 151, 170

teleological guidance control 21–22 torn decisions 26–27, 87–91, 97, 106, 112n2, 112n4 Tse, P. 114

Schooler, J. 154 self-forming acts (SFAs) 9, 55, 58, 83, 170; and alternative possibilities 11–15; arbitrariness of 15–17; and efforts of will 11–15; and plural voluntary control 17–22 semi-compatibilism 31 Shadlen, M. 114 Skinner, B.F. 32, 46, 56n3 Smilansky, S. 6n2, 118–122, 131nn5–6, 132n9, 171 Stace, W. 30, 32, 166 Stapp, H. 114 Strawson, P.F. 49–50 strike-back urge 127–128

Van Inwagen, P. 28n4, 29, 37, 113, 156, 161, 166, 169n2, 171 Vohs, K. 154 volitional streams 14, 21, 94

ultimate responsibility 12, 15, 19, 71–73, 77, 87, 103, 156, 161, 166, 168 Usher, M. 21

Waller, B. 6n2, 29, 81n2, 113, 115, 116–118, 120–124, 131, 131nn1–2, 168n1, 169n3, 170, 171; and moral obligation 133–150; and problem of punishing the innocent 124–130 Watson, G. 7n3, 49, 50 weighting of reasons 106–112, 170 Widerker, D. 40, 42, 43, 45 will-settled choices 78