Theological Perspectives on Free Will: Compatibility, Christology, and Community [1 ed.] 9781003306191, 9781032274423, 9781032306735

Table of contents :
Cover
Half Title
Series
Title
Copyright
Contents
List of contributors
1 Introduction: Free Will in Philosophical Theology
PART I Compatibility
2 Why Christians Should (Still) Be Compatibilists
3 Weighing Compatibilism and Libertarianism in Analytic Theology
4 Can an Unfree Person Have a “Deep Self?”
5 The Semantic Case Against Open Theism and Experimental Philosophy
6 Freedom, Even If God Decrees It
PART II Community
7 Free Together: On Christian Freedom and Group Ontology
8 Free Will, Cognitive Biases, and Theology
9 Bound Choice and the Mystery of Eternal Punishment – A Trilemma for Monergistic Accounts of Grace
10 A Theological Three-Body Problem: Why Lutherans Are Bound to Struggle With Free Will?
PART III Christology
11 Another Look at the Final Temptation of Christ
12 The Impeccable Freedom of Christ
Index

Citation preview

Theological Perspectives on Free Will

Free will is a perennial theological and philosophical topic. As a central dogmatic locus, it is implicated in discussions around core Christian doctrines such as grace, salvation, sin, providence, evil, and predestination. This book offers a state-of-the-art look at recent debates about free will in analytic and philosophical theology. The chapters revolve around three central themes: the debate between theological compatibilists and libertarians, the communal nature of Christian freedom, and the role of free will in Christology. With contributions by leading scholars, the volume provides a valuable overview of current arguments as well as novel openings and ideas for further discussion. Aku Visala is Research Fellow in the Faculty of Theology at the University of Helsinki, Finland. Olli-Pekka Vainio is University Lecturer of Systematic Theology at the University of Helsinki, Finland.

Routledge Studies in Analytic and Systematic Theology Series editors James Turner, Thomas McCall and Jordan Wessling

Impeccability and Temptation Understanding Christ’s Divine and Human Will Edited by Johannes Grössl and Klaus von Stosch Forgiveness and Atonement Christ’s Restorative Justice Jonathan C. Rutledge Theological Perspectives on Free Will Compatibility, Christology, and Community Edited by Aku Visala and Olli-Pekka Vainio The Church and the Problem of Divine Hiddenness Derek S. King Forsaking the Fall Original Sin and the Possibility of a Nonlapsarian Christianity Daniel H. Spencer Identity and Coherence in Christology One Person in Two Natures Paul S. Scott

For more information about this series, please visit: www.routledge.com/Routledge-Studiesin-Analytic-and-Systematic-Theology/book-series/RSAST

Theological Perspectives on Free Will Compatibility, Christology, and Community Edited by Aku Visala and Olli-Pekka Vainio

First published 2024 by Routledge 4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 605 Third Avenue, New York, NY 10158 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2024 selection and editorial matter, Aku Visala and Olli-Pekka Vainio; individual chapters, the contributors The right of Aku Visala and Olli-Pekka Vainio to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted 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. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library ISBN: 978-1-032-27442-3 (hbk) ISBN: 978-1-032-30673-5 (pbk) ISBN: 978-1-003-30619-1 (ebk) DOI: 10.4324/9781003306191 Typeset in Sabon by Apex CoVantage, LLC

Contents

List of contributors   1 Introduction: Free Will in Philosophical Theology

vii 1

AKU VISALA

PART I

Compatibility25   2 Why Christians Should (Still) Be Compatibilists

27

JESSE COUENHOVEN

  3 Weighing Compatibilism and Libertarianism in Analytic Theology

47

KEVIN TIMPE

  4 Can an Unfree Person Have a “Deep Self?”

75

LEIGH VICENS

  5 The Semantic Case Against Open Theism and Experimental Philosophy

87

FERHAT YÖNEY

  6 Freedom, Even If God Decrees It

104

FR. JAMES DOMINIC ROONEY, OP

PART II

Community113   7 Free Together: On Christian Freedom and Group Ontology D. T. EVERHART

115

vi  Contents   8 Free Will, Cognitive Biases, and Theology

136

SIMON KITTLE

  9 Bound Choice and the Mystery of Eternal Punishment – A Trilemma for Monergistic Accounts of Grace

152

AKU VISALA

10 A Theological Three-Body Problem: Why Lutherans Are Bound to Struggle With Free Will?

171

OLLI-PEKKA VAINIO

PART III

Christology185 11 Another Look at the Final Temptation of Christ

187

DAVID WORSLEY

12 The Impeccable Freedom of Christ

203

JOHANNES GRÖSSL

Index

214

Contributors

Aku Visala is University Researcher in the Faculty of Theology at the University of Helsinki, Finland. D. T. Everhart is Postdoctoral Reseach Fellow in Theology at the University of Leeds, UK. David Worsley is Lecturer in the Department of Philosophy at the University of York, UK. Ferhat Yöney is Assistant Professor at Istanbul Medeniyet University, Turkey. James Rooney (OP) is Assistant Professor at Hong Kong Baptist University, China. Jesse Couenhoven is Professor of Moral Theology at Villanova University, USA. Johannes Grössl is Assistant Professor of Fundamental Theology and Comparative Religious Studies at University of Würzburg, Germany. Kevin Timpe is the William H. Jellema Chair in Christian Philosophy at Calvin University, USA. Leigh Vicens is Professor of Philosophy at Augustana University, USA. Olli-Pekka Vainio is University Lecturer of Systematic Theology at the University of Helsinki, Finland. Simon Kittle is an independent scholar based in University of Leeds, UK.

1 Introduction Free Will in Philosophical Theology Aku Visala

Free will is a perennial theological and philosophical topic. As a central dogmatic locus, it is implicated in debates about core Christian doctrines. Due to its elusive and multifaceted nature, free will has become a divisive topic among Christian denominations and theological schools. The doctrines most shaped by assumptions about the will and its freedom, such as sin, grace, theological anthropology, and predestination, remain subjects of ecumenical disagreements even today. In the last two decades or so, free will has emerged as a central topic in Christian analytic philosophical theology. This is due to two reasons. First, while Christian theology has always debated topics related to free will, the core issue has remained unsolved: is human freedom compatible with thorough divine determination? The preferred solution to this compatibility problem has subsequent effects on how other theological doctrines are understood. Whatever account of free will (or the lack of it) we choose, it will have a significant impact of our overall theological system. The second reason for why free will is now at the foreground of philosophical theology is the maturation of the free will debate in analytic philosophy. For the past decades, analytic philosophers have rigorously debated the conditions, nature, and the possibility of human free will. As we will soon see, four families of views have emerged and the core disagreements have been identified.1 This whole debate has given philosophical theologians new tools and conceptual resources that they have now begun to apply to theology. The new tools will allow theologians to make progress in solving ancient theological problems. The volume at hand contains novel chapters that contribute to the debate about free will and moral responsibility in analytic philosophical theology. In the midst of the worldwide COVID-19 pandemic, we managed to organize the annual Helsinki Analytic Theology Workshop (HEAT) in Spring 2021. The chapters that constitute this volume are fruits of this workshop. The chapter at hand provides a general introduction to the topic of free will and moral responsibility in the context of philosophical theology. It consists of four sections. First, I will outline the main accounts of free will and moral responsibility in analytic philosophy. These accounts have emerged as DOI: 10.4324/9781003306191-1

2  Aku Visala possible solutions to the compatibility problem. The compatibility problem is often framed in terms of free will and determinism: if some form of determinism (divine, physical, causal, and psychological) turns out to be true, will this rule out the possibility of human free will and moral responsibility? I will then go on to suggest that Christian theologians also face multiple variations of the compatibility problem in their own context, especially because of the prevalence of divine determinism in the Christian tradition. Third, I  will briefly outline some accounts of free will and responsibility that attempt to solve the theological compatibility problem. Finally, I will offer a brief summary of chapters of the volume so that the reader will have an easier time following the twists and turns of the complex dialectic. Free Will and Moral Responsibility We should start with the concept of free will itself. The notion of free will, as it relates to many debates in philosophy and theology, is crucial for moral responsibility. We wonder what differentiates those actions of humans that make us apt subjects of responsibility attributions from those that do not. The prime candidate for drawing this difference is the notion that some actions are “up to us,” whereas some other actions are not. In other words, some actions depend on our agency in a way that other actions do not. If an action belongs to the first group, if it is really up to us, it seems appropriate to hold us responsible for that action. However, if an action is not really up to us, it does not seem fair to hold us responsible for it. In what follows, I will use the notion of control to denote this possible “up-to-usness” of our actions. Moral responsibility is at the core of the free will debate. Moral responsibility as a psychological phenomenon consists of various attitudes that we direct toward each other and ourselves. Such attitudes involve, for instance, blame, praise, punishment, asking for forgiveness and giving it, anger, gratefulness, and resentment. These attitudes also include a host of moral judgments about culpability, moral quality, and other morally relevant features. Our ordinary social life consists of navigating this network of attitudes and expectations and judgments. We habitually adopt responsibility attitudes toward one another and ourselves as we assess how we manage to fulfill our duties and moral expectations toward one another. The normative issue at the heart of the debate is about the aptness or appropriateness of these attitudes and judgments: under what conditions are moral agents deserving of blame or praise? This is the notion of basic desert. Derk Pereboom explains: For an agent to be morally responsible for an action in this sense is for it to be hers in such a way that she would be 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

Introduction 3 at issue 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.2 The existence of such appropriate conditions seems indubitable: it is inappropriate or unjust to blame a person for an action that the person did not control. Moreover, the desert is basic in the sense that moral judgments and responsibility attitudes are grounded in the fact that the person actually committed the action and it was hers in the appropriate sense. What it means for an action to belong to the agent, to “be hers,” can be understood in terms of the control, or the free will condition: for a person to be justly or aptly targeted with moral responsibility attitudes, like blame, she must be in control of her actions. While most philosophers agree about the necessity of the control condition, they disagree about the exact nature of it. What kind of control is required for moral responsibility? There are many candidates on table. Following Alfred Mele, we can distinguish existing accounts of control in terms of how demanding they are.3 Low grade accounts of control construe free action in terms of sourcehood: an action of the agent is free, when that action has its sources in the agent, rather than outside her. According to one popular account, the agent controls the action, if that action is a product of a mechanism sensitive to rational considerations and that mechanism properly belongs to the agent.4 Another account states that a free action is an action that has its roots in the agent’s deeply held values, beliefs, and goals. If the action flows from these deeply held views, it is appropriate to hold the person responsible for it.5 The core claim of high grade accounts of free will is that free actions are products of choice. If a person never had a choice about anything, she could not be said to control her actions. Choice, in turn, should be conceived in terms of having access to metaphysically possible alternatives. This is the famous alternative possibilities condition, or the power to do otherwise.6 Only those actions count as free and responsible that the agent could have refrained from committing, while all the background conditions remain the same. Rather than simply relying on the idea of sourcehood, this account requires what Kevin Timpe and others call leeway freedom.7 Another high grade account of free will does not require that the agent have leeway freedom with respect to all responsible actions. Instead, sourcehood incompatibilism holds that some actions can be free without leeway because they are properly grounded in the agent’s character.8 A person with an impeccable moral character, for instance, cannot really choose an evil option in some moral situation. In this sense, the impeccable moral agent has no choice, no leeway. This is no problem for free will, however, because good actions committed by the impeccable agent have their sources in the agent’s moral character. A person’s actions flow from her moral character, so in order for a person to be responsible for her actions, she must exercise control over

4  Aku Visala the “springs of her actions,” the moral character that issues them. In order for such a control over one’s character to be possible, the agent’s character cannot be completely determined by causes that are outside the control of the agent. The Compatibility Problem The debate about free will revolves around different versions of the compatibility problem. One way to frame the problem is as follows: How do we fit in? . . . We now have a reasonably well-established conception of the basic structure of the universe. We have plausible theories about the origin of the universe in the Big Bang, and we understand quite a number of things about the structure of the universe in atomic physics and chemistry. There is, however, an interesting tension. It is not easy at all to reconcile the basic facts with a certain conception we have of ourselves. . . . We have a conception of ourselves as conscious, intentionalistic, rational, social, institutional, political, speech-act performing, ethical and free will possessing agents. Now, the question is, How can we square this selfconception of ourselves as mindful, meaning-creating, free, rational, etc. agents with a universe that consists entirely of mindless, meaningless, unfree, nonrational, brute physical particles?9 The problem consists of reconciling our self-conception of free and rational agency with the naturalistic view of the world as ultimately non-rational and non-conscious. One central point where these two pictures clash is the assumption that the world is deterministic. Everything in the world has a cause. Given that there are no extra physical forces or influences acting upon the physical world, everything that happens seems to be the result of prior causes. The result is the doctrine of physical or causal determinism, according to which the totality of the past up to the present moment and laws of nature only allow for one unique future. In other words, given the past up to the present moment, only one future is possible. There are other ways of defining determinism. Logical determinism holds that a set of facts about the past combined with a set of facts about laws of nature entail all facts about the future as well. So, if there were a god-like intelligence that would know all truths about the past and laws of nature, that intelligence could deduce all the truths about the future as well. There are also other forms of determinism that might raise questions about human free will. Psychological determinism holds that the action of the agent is determined by all the prior psychological facts that pertain to the agent. So, if the agent’s reasons, desires, beliefs, and other relevant mental features were fixed, they would always produce the same action as an outcome. The compatibility problem will take different forms depending on the account of control we adopt. Consider the sourcehood account combined with

Introduction 5 low grade free will first. The worry there is that the truth of causal or physical determinism will result in a situation where the sequence of events that leads to the agent’s action bypasses the agent’s own rational decision-making process. In other words, the agent’s rational processes do not make a difference with respect to her actions. Alternatively, the problem could be set up like this: if determinism were true, the core self of the agent (her deeply held values and beliefs) could not be the source of the agent’s actions. In both cases, the agent would be an unfit target for moral evaluation and blame on the basis of her actions since those actions would not really have their sources in her. These kinds of problems are not insurmountable for a low grade sourcehood theorist. Defenders of low grade free will invoke the way in which our ordinary practices of moral evaluation work: we do not make metaphysical commitments to indeterminism when we judge people blameworthy; instead, we simply assess whether the person acted rationally in the light of her own beliefs and values and understood the moral nature of her action. The sourcehood theorist will offer an account of action, where the agent can act rationally and on the basis of her deeply held values and goals in a deterministic world. So, the truth of determinism should not matter for the existence of free will and moral responsibility at all. Furthermore, the low grade theorist will invoke various imaginary scenarios that are designed to show that alternative possibilities are not really needed for moral responsibility after all.10 What counts is the actual sequence of events that leads to the agent’s action, not what the agent could have done. So, a human agent can justifiably be held responsible for her actions even if she has only low grade free will. Sourcehood can be the case even given the truth of determinism. As pointed out earlier, the sourcehood account can also be combined with a high grade version of free will. Here, the compatibility problem takes a more difficult form. Remember that according to source incompatibilism, the agent controls those actions that are issued by those features of her character that she has had a role in shaping. The truth of determinism – the worry is – would undermine the agent’s ability to control those features that shape her character. The truth of determinism would entail that the ultimate control over her character would rest outside the control of the agent, thereby making any kind of basic desert blame or culpability judgment unjustified. A  family of arguments called the manipulation arguments is supposed to provide the reason why the agent requires ultimate control over her character in order to be held responsible.11 Manipulation argument begins with the claim that if the person A’s character – her tendencies, inclinations, values, and goals – were a product of covert manipulation by a person B, then person A would not be responsible for actions that flow from her character. Various science fiction scenarios have been presented to support this claim. Next, the argument states that there is no relevant difference between this kind of covert manipulation by another person and a world where determinism is true. The truth of determinism would be analogous to the covert manipulator so that it would make it impossible for the person to be a source of her character. Therefore, free will is incompatible with determinism.

6  Aku Visala Finally, it is not difficult to see how the compatibility problem emerges for the leeway theorist. According to the leeway view, exercising control over an action requires that the agent has access to metaphysically alternative possibilities. However, if causal or logical determinism were true, the access to such alternative possibilities would be ruled out almost by definition. If determinism means that given the past, only one possible future is possible, it seems to automatically rule out the existence of open futures. The reason the leeway theorist takes leeway control as a necessary feature is the consequence argument.12 In short, the consequence argument suggests that if the person is unable to control events in her distant past and that determinism is true, then the person is unable to control the present events (including her actions) as well. This is because the truth of determinism entails that the present is just a product of the totality of past events and the laws of nature. Since the person is unable to change past events and the laws of nature, the person is unable to control how the present and the future turn out, including her own actions. I have now briefly outlined a number of ways of how determinism could undermine the ability of humans to control their actions. Before I  go on to describe the four prevalent accounts of free will and moral responsibility, one additional issue needs to be raised. Notice that various arguments invoke everyday intuitions about free will and moral responsibility. The defender of low grade free will argues that our everyday views include no deep metaphysical assumptions about alternative possibilities. Against this, the leeway theorist appeals to the everyday notion of choice. Similarly, the sourcehood incompatibilist argues that our everyday view of manipulation and responsibility exposes our hidden assumptions about the conditions of moral responsibility. These appeals to our everyday views raise a crucial question: what is our everyday view of free will and how do we really justify our moral responsibility practices? We should distinguish the descriptive project of understanding our everyday view from the normative project of trying to offer an account of free will and moral responsibility.13 The descriptive project is about the nature of our everyday views of free will and moral responsibility. How do people understand the concept of free will? What kinds of moral responsibility practices they actually have? The normative project concerns the issue of what we are supposed to think about free will and moral responsibility. Finally, there is the issue of revision – the revisionist project. If it turns out to be the case that our everyday views and practices are in line with our normative account of free will and moral responsibility, there is no need for substantial revisions. However, it might turn out that some of our everyday practices and views are no longer appropriate after we have conducted our normative inquiry. In such a case, revisions are required. Four Views on Free Will We have finally mapped out the concepts and issues that allow us to distinguish four different families of views about free will and moral responsibility.14 First, there are libertarians, according to which human beings in

Introduction 7 general are capable of free will that renders us as apt targets of various basic desert responsibility attitudes and judgments. Libertarians consider themselves as defenders of the everyday view of free will that – according to them – includes the ideas of undetermined choice and (perhaps) control over one’s character. For them, the everyday view of free will closely corresponds to a high grade view of free will. It follows from these commitments that libertarians accept the incompatibility of free will and determinism. If determinism were true, human agents could not make choices between alternatives, nor could they control their character. However, humans do have free will so indeterminism must be true. The compatibilists agree with libertarians that human beings indeed have free will, but they disagree about its nature. According to the compatibilists, our everyday view of free will and moral responsibility correspond to low rather than high grade free will. Moreover, the compatibilists argue that humans indeed have free will of the low grade variety, and that high grade free will is not necessary for moral responsibility. Given their low grade view of free will, the compatibilists maintain that the truth of determinism is compatible with human free will. The third view, free will skepticism, agrees with libertarianism on many points; both are incompatibilist for instance. Like libertarians, free will skeptics hold that our everyday view of free will is of the high grade variety and that free will and determinism are incompatible. However, they argue that human agents cannot perform actions that could be considered free in the high grade sense. Derk Pereboom, for instance, argues that source incompatibilism is really a requirement for basic desert moral responsibility and develops his own manipulation argument to demonstrate this. However, Pereboom continues that regardless of the truth of determinism, human beings can never achieve this kind of control over their characters and actions. Pereboom mounts a two-pronged attack: no matter whether determinism is true or false, the “springs of the agent’s actions” are not under her control.15 The skeptic concludes that human agents do not control their actions in a way required for basic desert responsibility. In the end, no human agent really deserves to be the target of any kind of negative or positive responsibility attitude. Given the centrality of responsibility in our moral and religious life, the skeptic now needs to offer reconstructed account of our everyday moral responsibility practices without invoking basic desert. Many skeptics have argued that such an account can be provided in order to retain most of our moral responsibility practices (like praise, blame, and forgiveness). Some practices (like retributive punishment) should be jettisoned, but independent reasons can also be provided for their elimination. Invoking forward-looking considerations rather than backward-looking, like basic desert, can justify many of these practices. By forward-looking considerations, the skeptic refers to the practical usefulness and benefits of various responsibility practices. Some forms of holding a person responsible can be justified by invoking the good effects that it has. Not only does holding people responsible encourage

8  Aku Visala moral behavior in the future, but some practices like incarceration protect the public from harm.16 However, there are skeptics that deem most of our moral responsibility practices wholly unjust and demand radical reconceptualization or direct elimination of those practices.17 Finally, there is a fourth view that is a newcomer at the scene. Free will revisionists agree with libertarians and skeptics that our everyday view of free will corresponds closely to high grade accounts of free will.18 However, they hold that the actions of human agents do not fulfill the conditions of high grade free will. Instead, they correspond more closely to low grade accounts. Nevertheless, the revisionists hold onto basic desert responsibility. On their view, our basic desert concept can be reconstructed so as to make room for low grade free will. The revisionist offers a moderately reconstructed and revised account of our free will language and practices. She maintains that while our freedom is not exactly what we might think, we should not reject and jettison our basic desert language and moral practices, like the free will skeptics do. So, human agents can deserve blame and praise for their actions, but not on the grounds that people normally think. Divine Determinism and the Compatibility Problem So far, I have mapped out various accounts of free will and moral responsibility without reference to theology. I constructed the compatibility problem in terms of the naturalistic picture of the world, which seems to entail determinism of some form or another. One might initially think that a theistic view of the world was more hospitable to our self-conception as free, rational and conscious, and intentional agents. According to theism, the world does not consist of non-rational and unfree particles only, but includes a free, perfect, rational, and living mind, which is the source of everything else. However, a number of divine attributes might raise compatibility worries not wholly unlike those faced by the naturalists. For instance, if God’s knowledge of the future is construed as perfect and including what created agents freely do in Table 1 Four Views on Free Will 

Do we have free will? Is the everyday view correct? Is free will compatible with determinism? Do we need to revise our everyday view?

Libertarianism Compatibilism Skepticism

Revisionism

Yes (high grade) Yes (low grade) No

Yes (low grade)

Yes (high grade) Yes (low grade) No (high grade) No (high grade) No

Yes

No

Yes, but with revisions

No

No (or minor revisions)

Yes (major or radical revisions)

Yes (moderate revisions)

Introduction 9 the future, it seems that created agents have no choice about the future. Similarly, if God’s relationship to creation is considered such, that God is causing everything therein, we can wonder whether God would also be the cause of all the actions of human agents and whether those actions could be different from what they actually are. There are also worries that are more specific to Christian theology rather than generic theism. Some core ideas of Christian theology, like sin, grace, and salvation, suggest a kind of psychological determinism. Original sin, for instance, introduces significant psychological constraints as to how the human intellect and will work. The question is whether a person, whose moral cognition and will are impaired by sin, can be justifiably held responsible for her actions, especially if sin leads to the inability to act in ways pleasing to God. It would seem that God would be unjust to hold such a person responsible for sinful actions. Consider a parallel theological statement of the compatibility problem: We have a set of claims about God and the created world that comes through revelation and theological tradition. These beliefs are held by most Christians and belong to the core of the religion itself. So, we believe that God is the ultimate source of everything else that exists. Not only does God give being to everything, God is also in ultimate control of his creation. His divine plan underlies all that happens and his omnipotence partakes in all causal relations. God is also a righteous judge of human evil and sin. However, because of the fall, sinful humans lack the ability to perform actions that are pleasing to God. Finally, we understand God as perfect goodness, indeed, Goodness itself, in whom there is no evil, just superabundant love. God simply acts out of God’s perfect goodness, when he condemns sinful human beings to eternal torment. However, there is an interesting tension here. It is not at all easy to reconcile these theological basic facts with our self-conception. We see ourselves as conscious, free, intentional, social, and institutional agents. The question is how we can reconcile this view of ourselves as morally responsible agents with the basic theological facts that imply that God’s agency is ultimately the source of everything, including our free actions. God knows what we will do and has control over it. Sin constrains our control over our actions such that we cannot act in ways that would make God respond to us with good will. Furthermore, it seems plausible that if God is the source of the world, God is also responsible for the enormous evil and suffering in it, including the sinfulness of humans. Why is God not held accountable for evil? Christian theologians inherit two central motives from the Scriptures. On the one hand, Christian theologians want to hold onto human moral responsibility. In the Christian story, humans are accountable for their actions in

10  Aku Visala the eyes of God and their fellow humans. This is especially apparent in the Hebrew Scriptures, where the core narrative is that of God’s election of the people of Israel and Israel’s response to this election in various circumstances. God’s covenant with Israel calls out for a free response: Israel expresses her willingness to remain in the covenant by following God’s law (e.g., Deuteronomy 30:15–20). God promises to reward those who do his will and punish those who do not. God treats humans as autonomous, morally responsible agents. God issues commandments, expresses his will, anger, displeasure, and pleasure to people, and expects people to comply. If human beings were merely automata, such actions on God’s part would make very little sense. On the other hand, it is also widely acknowledged in the Scripture that God is in charitable providential control over his creation and is the ultimate source of salvation. Some facts about God and his actions seem to undermine human free will. In the Hebrew Scriptures, we can observe the theme of God’s unilateral and sovereign action toward his creation and his people. God sometimes elects individuals and groups for certain treatment. The most difficult cases are those where God elects some individuals or groups as target of his anger and wrath. The most famous is the hardening of the heart of the Pharaoh in Exodus (10:12). The theme also emerges in Paul as an explanation for why Jews refuse to believe in Christ (Romans 9). The main source of compatibility problems is the theological analogue of causal determinism, divine determinism. The exact mechanism of divine determinism can be understood in different ways, but the core idea is that all created events and facts are ultimately made necessary by God’s will, power, or knowledge, including the free acts of creatures.19 Many divine determinists invoke the Scriptures as evidence for its truth. Some passages imply that God is in full providential control over his creation (Psalms 103:19; Daniel 4:35; Matthew 10:29). Second, the Scriptures also seem to imply that God has perfect knowledge of the future, including human free actions (Isaiah 42:9; Acts 3:18). Because God is necessarily infallible, it follows that human future (and present and past) actions are, in fact, not contingent but necessary. Some theologians argue that the truth of divine determinism is entailed by the very nature of God as the omnipotent and sovereign creator of the universe. God is said to concur with creaturely causes (lat. concursus Dei). In contemporary debates, this divine property is often called God’s omnicausality. Some theological commitments also raise questions about the possible truth of psychological determinism. If free action requires that the agent controls her psychological states or her character, some accounts of theological anthropology, sin, and human psychology seem to produce compatibility problems. The Lutheran reformer Philip Melanchthon, for instance, argues in his early work that sinful humans have no free will because their actions are caused by their strongest desire. Moreover, the strongest desires are sinful and selfish desires, and there is nothing in the human mind that could resist or otherwise control them. As a consequence, humans are unable to perform selfless actions that would be ultimately pleasing to God.20 This problem

Introduction 11 easily generalizes if one assumes a strong or robust notion of original sin. The strong doctrine of original sin says that there are certain psychological defects that render sinful humans incapable of performing actions that are pleasing to God. While some actions can be good in human eyes, even the best human actions are sinful in God’s eyes because they can be traced back to selfish motives. This raises a compatibility problem about blameworthiness: how can humans be blamed for their sinful actions, if they lack control over their sinful nature from which those actions flow? In contemporary philosophical theology, there are Augustinian compatibilists who defend a low grade account of free will in order to preserve human moral responsibility.21 Guillaume Bignon, for instance, argues that human agents can be held responsible for those actions that have their roots in their deeply held values, beliefs, and goals.22 Despite the truth of divine determinism, sinful human beings still act voluntarily when they sin. While sin has severely impaired the workings of human moral cognition, sinful actions are traced back deeply held values and beliefs that happen to be selfish and against God’s will. Neither sinners nor saints exercise leeway control over their deeply held values and desires. Until God acts to free a person from sin, there is nothing the sinner can do to overcome her sinful nature. For Augustinian compatibilists, true freedom does not reside in making choices between morally relevant alternatives, or exercising control over one’s character, but rather being able to make the objectively right choice without external or internal constraints. Jesse Couenhoven calls this normative freedom.23 So, a saint might very well be unable to choose a sinful action, but that fact makes the saint freer than a person who is able to act in sinful ways as well. A sinful person is a slave to her selfish desires and cannot uninhibitedly act in moral ways, whereas the saint’s will is not constrained by sinful desires and can therefore do what is good unimpaired. So, when God heals the human soul from sin, God actually narrows down the range of possible human actions, but this is nevertheless freedom enhancing: it becomes psychologically impossible for the saint to love and do what is evil. Varieties of Libertarianism in Philosophical Theology The recent PhilSurvey from 2020, that records the views of analytic philosophers on various philosophical questions, concluded that compatibilism is the mainstream view among analytic philosophers. The results suggest that approximately 60% of respondents accept or lean toward compatibilism. Libertarianism is a minority position accepted or sympathized by 18% of responding philosophers. Finally, approximately 10% accepted or leaned toward free will skepticism.24 As the previous PhilSurvey from 2009 already noted, theistic philosophers tend to be much more favorable toward libertarianism than non-theists. Similarly, philosophers of religion tend to be libertarians whereas philosophers outside philosophy of religion tend not to be. While theism is a minority

12  Aku Visala view among analytic philosophers (approximately 18%), 50% of theistic philosophers favored libertarianism. This is surprising since the mainstream view among analytic philosophers is compatibilism. Why is libertarianism so relatively popular among theists and Christians in particular?25 Libertarian accounts of free will appear to have multiple theological benefits, mainly related to various aspects of the problem of evil. Kevin Timpe and Daniel Speak suggest that the main reason is the success of free will theodicies. Most famously, Alvin Plantinga formulated a version of the free will defense to the problem of evil, which entails libertarian free will. Similarly, John Hick’s popular soul-making theodicy also assumed the existence of libertarian free will.26 The popularity of libertarianism also stems from the fact that human libertarian free will offers a way to exculpate God from human evil and sin – as well as render God’s eternal judgment just. An account of moral responsibility in the Christian context has two functions: first, it needs to secure human moral responsibility in the face of various compatibility problems, and second, it must also exculpate God from involvement in human sin and evil. The second function is important but often goes unnoticed. While the compatibilist might succeed in providing a plausible account of human moral responsibility under divine determinism, the second task is much more difficult for the compatibilist compared to the libertarian. If human agents have low grade free will, God can exercise significant control over human free actions without thereby constraining or hindering human free will. In other words, if God’s control and human freedom are compatible, God will become morally responsible for what humans do. Not only would God be deeply involved with human evil and sin, but God’s status as the ultimate judge and punisher would also be undermined. If God makes human sin and evil necessary and they are ultimately under God’s control, it does not seem to be appropriate that God would stand in judgment of humans, blaming and punishing humans for their sins.27 The libertarian has a relatively easy solution to these problems. Assuming high grade human free will guarantees human moral responsibility and explains sin and evil in the world without implicating God. If humans had high grade freedom, not only would human agents be responsible for their actions and deserving of punishment, but also God could not control their behavior without overriding their free will. So, when God decided to create a world with high grade human free will, God also decided to create a world where evil, sin, and the eternal rejection of God by created agents became possible. God might want to create such a world because of the value high grade human free will adds to world: for instance, one could argue that without such free will, there could be no moral and spiritual development and loving and caring relationships between God and humans.28 Given the challenges of Augustinian compatibilism in accounting for the existence of evil and exculpating God from human evil, many theologians have sought to develop alternatives to divine determinism. If divine

Introduction 13 determinism were jettisoned, there would be more room for libertarian free will in the human case. Two such options have emerged in the recent debates: Molinism and Open Theism. Jesuit Luis de Molina and his followers developed an ingenious way of combining comprehensive divine providence and foreknowledge with human libertarian free will. They developed an account of providence and foreknowledge that would provide what divine determinists want. On this account, God has access to middle knowledge that pertains to truths about what would happen in different circumstances. These truths include the counterfactuals of freedom, namely, truths about what free agents would do in different circumstances. By having access to counterfactuals of freedom, God can choose to actualize that possible world where the created agents in that world act according to God’s will. God can guarantee a complete and comprehensive providence and foreknowledge without actually determining human actions as the divine determinist is forced to do.29 Another alternative to divine determinism is Open Theism. It is a family of views according to which the future of the world is radically open. Not even God knows how events, especially free actions of created agents, will turn out. On this view, humans have libertarian free will and it follows from this that God’s foreknowledge and providence cannot be as complete as Molinists and divine determinists have thought.30 God’s creation involves contingency and randomness, which allow for the free choices of created persons. God’s interaction with the world is not strictly unilateral, but instead dynamic and reactive. God has given the created world significant causal independence. Here the open theist is motivated by God’s love rather than power and sovereignty: God respects the autonomy and freedom of created persons and the created world. The open theist picture of God and the world gives space to robust free will. Richard Swinburne, for instance, argues that free will requires a special form of causation, agent-causation. For Swinburne, free will is the power of the agent to initiate actions that are not determined by prior causes. In other words, agents are capable of free will only when they are able to bring about actions that have no prior sufficient causes: free will is freedom from determinism.31 Ultimate sourcehood is guaranteed by the fact that the agent herself stands outside the flow of ordinary causation, as it were, and can generate new events (actions) without determining prior causes. Swinburne argues that this is possible, because human beings are more than simple physical substances. To be an agent is to be a non-physical substance that has the power to initiate brain events that lead to actions on the basis of prior mental states.32 Recently, a number of philosophers have developed a source incompatibilist account that is much indebted to Thomas Aquinas. Eleonore Stump and Kevin Timpe, among others, have defended what some call virtue libertarianism. On this view, free and responsible actions require no leeway, but rather they need to have their sources in the agent in a robust way. As other source

14  Aku Visala incompatibilists, Timpe locates freedom in the agent’s moral character: the agent is responsible for those actions that have their roots in the agent’s virtues and vices.33 Virtues and vices are relatively stable character traits that lead the person to respond to various moral situations in a predictable manner. Virtues, such as courage, wisdom, and justice, and vices, like viciousness and laziness, are products of many factors, but these factors must include the agency of the person herself. While the amount of control that the agent exercises over her actions at a given time is limited, over a longer period of time the person can shape the core aspects of her character and form virtues (or fail to do so and develop vices). If determinism were true, a person could not contribute to her own virtues and vices in any substantial way. To the extent that a person has contributed to the development of her character, she deserves praise for her virtues and blame for her vices and for those actions that flow from them. While most Christian theologians fall under compatibilism or libertarianism, there are some exceptions. Most famously, Martin Luther argued for a skeptical conclusion: at least with respect to salvation, human agents have no free will whatsoever. For Luther, only high grade free will deserves the title “free will.” Unfortunately, High grade free will, the ability to choose between morally significant options without being constrained by anything else, belongs to God only. Humans cannot have such free will because of comprehensive divine determinism and robust human sinfulness. Divine determinism makes leeway impossible, and sin makes it impossible for humans to perform meritorious acts in God’s eyes. While Luther rejects human free will, he holds onto moral responsibility. Humans are unable to control their actions in high grade sense, but they are nevertheless responsible for their actions. Invoking an Augustinian account of moral responsibility, Luther holds that a human agent can be held responsible for actions that flow from her deeply held beliefs, intentions, and values.34 In contemporary context, accounts that reject the compatibility of free will and determinism but affirm the compatibility of moral responsibility and determinism are called semicompatibilist.35 However, some aspects of Luther’s view could be interpreted as coming close to free will skepticism rather than compatibilism. First, Luther seems to think that there is something seriously wrong with our everyday views of free will. Our everyday view construes us as free and autonomous, whereas in reality, we are slaves to our sinful desires. So, a Christian account of human freedom would be a radically revisionist one. Second, Luther also seems to be ambivalent about basic desert, at least with respect to God’s treatment of humans. God can do whatever God wants: if humans deserve some kind of treatment on the grounds of their actions, God would be obligated to comply. God would “owe” humans either positive or negative treatment. However, since God is ultimately free and sovereign, God “owes” nothing to created beings, so God need not take into account what people deserve.36 Very few contemporary theologians or philosophers are willing to go this far.

Introduction 15 One possible exception might be the free will skeptic Derk Pereboom, who has highlighted the benefits of divine determinism and developed strongly revisionist account of moral responsibility practices.37 A Preview of Coming Attractions The volume consists of three parts: (1) compatibility, (2) community, and (3) Christology. The first part Compatibility contains chapters that directly address various aspects of the compatibility problem in theology. The main question is whether Christian theologians should adopt a libertarian or a compatibilist account of free will. The second part Community includes chapters that cover a variety of issues regarding the relationship between individual agents and the community. Theological anthropology of the twentieth century has taken what some call a relational turn. Rather than conceiving human agents as individual atoms, we should see them as parts of larger networks of community and culture. Similarly, some analytic philosophers have begun to rebel against what they see as unjustified individualism in the free will debate. However we conceptualize human freedom, it must take into account that humans are social and cultural creatures, and their actions and attitudes emerge in constant dialogue with other members of the community.38 Moreover, some chapters included in this section also address various challenges that the Lutheran theological tradition has with free will and moral responsibility. Finally, chapters in the third part examine free will in the context of Christology. Many aspects of the free will debate in theology pertain to the possible free will of human agents. According to Christian theology, however, there are other agents as well. The most relevant of such agents is God. Given the perfection of God, it makes sense to see God’s freedom as perfect as well. Divine free will is the most perfect free will possible. Unique issues and problems emerge when the second person of the Trinity is incarnated in the person of Jesus Christ. If the divine nature of Christ is to retain ultimate freedom, one must make sense of various scriptural depictions of Christ, where Christ seems to be less than perfectly free. Questions arise, for instance, about whether Christ could have failed to resist temptations. If Christ did not have the possibility of failure, as his divine perfection would suggest, how could Christ have high grade free will in the first place? Similarly, if Christ was unable to sin (i.e., could not have sinned), how can Christ be free in the high grade sense? Part I: Compatibility

First, Jesse Couenhoven’s chapter Why Christians Should Be Compatibilists outlines an overall vision of Christian theology from a compatibilist perspective. More specifically, Couenhoven offers an axiological examination whereby he assesses the costs and benefits of compatibilism about free will

16  Aku Visala and moral responsibility. Couenhoven argues that the theological “cost” of being a compatibilist is lower than that of being a libertarian. The main advantage of compatibilism is that it puts less pressure on various theological convictions than libertarianism. This allows the theological reflection of these topics to develop more naturally. Couenhoven develops his argument by analyzing the best libertarian and compatibilist approaches to theological core topics, like the freedom of God, Christ, the heavenly saints, and fallen humanity, and the role of grace in the Christian life. Couenhoven also discusses themes that are widely taken to favor libertarianism, like the justification of evil and eternal punishment. Even in these cases, Couenhoven concludes, compatibilism is the more preferable view. Kevin Timpe’s chapter Weighing Compatibilism and Libertarianism in Analytic Theology offers a direct libertarian response to Couenhoven’s chapter. While Timpe agrees that Christians have reasons to be compatibilists, he contends that those reasons are not decisive in favor of endorsing compatibilism. First, Timpe examines compatibilism and libertarianism as a family of views about free will and moral responsibility. Second, he suggests that while compatibilism might help to solve some particular theological problems better than others, one should consider the big picture. The most coherent big picture is provided by libertarianism, argues Timpe. So, the disagreement between Timpe and Couenhoven is at this level of holistic vision rather than individual theological doctrines. The third chapter deals with a very specific facet of the compatibility problem. It was already outlined earlier how both libertarians and free will skeptics share the claim that determinism and free will are incompatible. While libertarianism and compatibilism are both widespread among philosophical theologians, free will skepticism is a minority view. However, Leigh Vicens’ chapter Can an Unfree Person Have a “Deep Self?” examines some theological motivations for free will skepticism. The greatest challenge of free will skeptics is that its proponents are forced to give up a lot – or at least so it might seem. Much of our moral and social life is associated with the assumption of free will, and many religious practices and theological doctrines would seem to depend upon it as well. Recently, however, some free will skeptics have questioned these associations. Derk Pereboom has argued that with some modifications, much of our moral life can withstand free will skepticism. For even if, because we lack free will and so are not ultimately responsible for what we do, we do not deserve certain treatment in response to our behavior, treating each other in these ways can be justified by forwardlooking considerations. Vicens builds upon Pereboom’s account and argues that even if we have no free will, we can still deserve some moral responses. She suggests that even an unfree person could have a “deep self,” that is, could possess deep values, beliefs, and desires. Moreover, a person with a deep self could deserve certain kinds of moral attributions. The fourth chapter contributes to the debate about the compatibility problem by discussing some results of experimental philosophy. Ferhat Yöney’s

Introduction 17 chapter The Semantic Case against Open Theism and Experimental Philosophy responds to recent arguments against Open Theism. According to these arguments, Open Theism goes against the ways in which ordinary human beings talk about knowledge of the future. According to Alfred Freddoso, there is a prevalent use of future contingents in ordinary language as if they were true either retrospectively or prospectively. Moreover, Jonathan Kvanvig and William Lane Craig argue that in ordinary language, future contingents are used as objects of knowledge. Yöney responds to these arguments by invoking recent work in experimental philosophy. Experimental philosophy demonstrates that although there is a universal core for epistemic notions, epistemic intuitions of ordinary agents and the ways they use ordinary language are not reliable (though not totally misguided) in cases, which are not clear cases of knowledge and non-knowledge. On the basis of this, Yöney suggests that future contingents are not clear cases of knowledge. If future contingents are not clear cases of knowledge, the way they are spoken of in ordinary language does not pose a challenge against Open Theism. In the fifth chapter, Freedom, Even If God Decrees It, James Rooney examines Thomistic notions of God’s sovereignty and causality. If God is the primary cause of all human free actions, how could those actions remain free in the high grade sense? If God’s intention or decree is logically prior to a human free action, that action becomes determined and no longer free in the leeway sense. Recently, W. Matthews Grant has provided an account of how it might be possible to reconcile God’s universal causality with human leeway freedom. Grant argues that leeway freedom would indeed be lost if God determined human actions via God’s decrees or will. These factors would be intrinsic and unchanging features of God, thus always logically prior to human free actions. However, Grant proposes a model according to which God does not cause human free actions via anything intrinsic to God. He calls this the Extrinsic Model of divine action. Rooney responds to Grant’s account by arguing that two classical theories of grace, despite holding that God causes creaturely acts in virtue of a divine decree or intention, remain libertarian in just the same sense Grant’s account is. Rooney suggests, controversially, that this is true even of Banezian theories of grace, on which God causes free actions directly. Part II: Community

In his chapter Free Together: On Christian Freedom and Group Ontology, D. T. Everhart criticizes the individualist assumptions made in the free will debate. He argues that recent work in ontology and agency of groups can help theologians to address better the relationship between individual freedom and groups. Everhart examines the function of Christian freedom in the Pauline epistles, giving special attention to our belonging to others in the body of Christ. The purpose of human freedom for the Apostle Paul is made clear in his ecclesiological ethics. For Paul, freedom is something one person

18  Aku Visala either takes up or lays down for the sake of the other. Drawing on both theological considerations and analytic social ontology, Everhart offers an account of corporate freedom. On this account, Christian freedom pertains to both groups of Christians and individuals. The freedom of the individual resides in the power of the individual to either use or give up their freedoms for the sake of the whole. This account of freedom for the sake of others, Everhart argues, does not undermine the self-determination of human agents. Human agents can still be sources of their own actions. However, the account will situate the self-determination of agents in a network of dependency on other human agents for their individual agency. The flourishing of free, individual human agents occurs in belonging to one another in community; we are free when we are free together. In his chapter, Free Will, Cognitive Biases, and Theology, Simon Kittle examines a possible skeptical challenge emerging from social and cognitive psychology. First, Kittle outlines his account of libertarian leeway free will. Human free will requires that human agents have a conscious choice about what they do. Without such a choice, they cannot be held responsible for their actions. Any kind of determinism, be that causal or divine determinism, rules out free will. For such a free will, an agent needs additional cognitive capacities: accurate conscious access to one’s propositional attitudes and the ability to weigh and process the contents of those attitudes according to rational standards. Second, Kittle examines the results of social and cognitive psychology suggesting that the human cognition is widely affected by a-rational factors that operate outside our conscious access. This produces a significant skeptical problem: all these influences undermine the kind of conscious control over action that free will requires. Kittle argues that the scientific results undermine our ordinary notion of free will, and that the correct response is to move toward free will skepticism. The next two chapters in this section deal with a specific theological tradition, Lutheranism, and examine how the tension between various theological commitments therein shapes the Lutheran view of free will (or the lack of it). First, Aku Visala’s chapter Bound Choice and the Mystery of Eternal Punishment presents a problem for strongly monergistic accounts of grace. This problem is created by three seemingly incompatible commitments. First, it is widely assumed by Lutherans (and many other Christian theologians) that God is perfect love and wills the salvation of all human persons. Second, Augustinians, like Lutherans and Reformed theologians, hold that human beings have no leeway freedom with respect to their salvation. The human will is “bound” with respect to salvation; humans lack control over whether they come to salvific faith. It is God who exercises control over a person’s coming to faith, not the person herself. Third, Lutherans and most other theologians assume that not all persons are saved. The seeming incompatibility emerges in two ways. Accepting the first two commitments puts pressure on the third commitment. Similarly, if one accepts the second and the third, it seems that one is forced to give up the first. In his chapter, Visala outlines why

Introduction 19 Lutherans have strong reasons to accept all three commitments and identifies how the conflicts between them arise. He also examines a distinctively Lutheran solution – the appeal to mystery or paradox – and argues that the cost of invoking mystery or paradox in the face of the trilemma incurs a high theological cost. Olli-Pekka Vainio continues the examination of Lutheran theology and free will in his chapter Why Lutherans Are Bound to Struggle With Free Will. Since Martin Luther’s On the Bondage of the Will, Lutherans have had an ambivalent relationship to the issue of free will. According to Vainio, Luther ends up between a rock and a hard place: purely from a philosophical point of view, Luther wishes to maintain both individual responsibility and divine determinism. The later Lutheran generations try to remain loyal to Luther’s intentions, but remain uneasy about the consequences of his extreme formulations and their metaphysical underpinnings. Immanuel Kant famously uses Lutheran Confessions on this issue as an example of antinomy that follows when one tries to hold on to two contradictory ideals. Vainio argues that Lutherans often subscribe to determinism in theory while being libertarians in practice. This way the antinomy, or paradox, is dissolved. Vainio interprets the doctrine of bound choice as a tool of spiritual life and source of mediation of one’s sinfulness. However, this leads Lutherans to a conflict from which there seems to be no easy way out. In securing the monergistic doctrine of grace, they are left with two unsatisfactory options. First, the monergistic doctrine seems to lead to universalism (which they reject). Second, allowing a two-tier approach to theological truths seems to lead to sacrificium intellectus (which they reject). Part III: Christology

David Worsley’s chapter The Primal Sin and Christ’s Final Temptations begins the section on Christology and free will. Worsley follows Eleonore Stump’s account of Christ’s passion and identifies a similarity between Christ’s temptation in Gethsemane and the “temptation” facing the primal sinner. In the case of the primal sinner, the temptation was between beatific revelation, a revelation that would perfect their intellect and satisfy their will, and friendship with God, doing God’s will at the (possible) permanent cost of this revelation. Christ’s temptation, on this account, was between maintaining beatific revelation in his human nature, the best thing for him, and giving it up on the cross, in order to do the will of God the Father. Worsley argues that if the primal sinner’s action can be attributed to ‘sin’, we should be open to doing the same in Christ’s case, too, were he to succumb to this temptation. Similarly, if we follow Richard Swinburne and think that Christ could only be tempted to do less than the best, we should be open to seeing the primal sinner’s action as less than the best (but not sin), too. In any case, if this temptation was sufficient to motivate the primal sin, if there was going to be a temptation that might tempt Christ to sin, it would be this.

20  Aku Visala In the final chapter, The Impeccable Freedom of Christ: Supererogatory Choice or Mere Disposition, Johannes Grössl discusses the problem of Christ’s impeccability. Most Christians believe – in accordance with Chalcedon and the Third Council of Constantinople – that Jesus Christ was truly human, yet unable to sin. However, if being human, as libertarians understand it, implies the power to choose between morally significant options, it seems that Christ cannot be both truly human and divine, for the latter implies impeccability. This is why Christology is a common motivation to favor compatibilism. Grössl seeks to undermine this argument from Christology to compatibilism by presenting and contrasting two viable libertarian solutions to the problem. On the one side, the popular supererogation view by Richard Swinburne says that although morally significant libertarian free will is essential to human nature, the power to choose between good and evil is not. In some cases, a human being can only possess the power to choose between performing and omitting a supererogatory free act. On the other side, the dispositional view says that although the capacity for developing a power to choose between morally significant options is essential to human nature, certain external circumstances can prevent this ability from being actualized. In addition to laying out anthropological differences between these approaches, Grössl suggests that the dispositional view does not stand in sharp contrast to the supererogation view. Rather, it can fill a significant explanatory gap in Swinburne’s theory by answering the question of why libertarian free will is actualized in most people as a power to perform or omit evil, and in some (in Christ) merely as a power to perform or omit supererogatory good acts. Notes 1 For an informative overview, see John Martin Fischer, Robert Kane, Derk Pereboom & Manuel Vargas, Four Views on Free Will (Oxford: Blackwell, 2007). 2 Derk Pereboom, Free Will, Agency, and Meaning in Life (Oxford: Oxford University Press, 2014), 2. 3 Alfred Mele, Free: Why Science Has Not Disproved Free Will (Oxford: Oxford University Press, 2014). 4 This is the reason-responsiveness account: John Martin Fischer & Mark Ravizza, Responsibility and Control: A Theory of Moral Responsibility (Oxford: Oxford University Press, 1998) 5 This is sometimes called the attributionist account of moral responsibility. See, for example, Smith, Angela. “Control, Responsibility, and Moral Assessment.” Philosophical Studies 138 (2008): 367–92. 6 Peter van Inwagen, An Essay on Free Will (Oxford: Oxford University Press, 1986). 7 Kevin Timpe, “Leeway vs. Sourcehood Conceptions of Free Will.” in Kevin Timpe, Megan Griffiths  & Neil Levy (ed.), The Routledge Companion to Free Will (London: Routledge, 2017). 8 One influential sourcehood incompatibilist account is Robert Kane, The Significance of Free Will (Oxford: Oxford University Press 1996). 9 John Searle, Freedom and Neurobiology: Reflections of Free Will, Language, and Political Power (New York: Columbia University Press, 2004), 4–5.

Introduction 21 10 These include the Frankfurt cases. See essays in David Widerker  & Michael McKenna (ed.), Moral Responsibility and Alternative Possibilities: Essays on the Importance of Alternative Possibilities (Aldershot: Ashgate, 2006). 11 For an overview of manipulation arguments, see Kristin Mickelson, “The Manipulation Argument.” in Kevin Timpe, Megan Griffith & Neil Levy (ed.), The Routledge Companion to Free Will (London: Routledge, 2017). 12 Peter van Inwagen, An Essay on Free Will (Oxford: Oxford University Press, 1986). 13 These distinctions can be found, for instance, in Shaun Nichols, Bound: Essays on Free Will and Responsibility (Oxford: Oxford University Press, 2015), 4–11. 14 These views are outlined in Fischer, Kane, et al. Four Views on Free Will. 15 Derk Pereboom, Free Will, Agency, and Meaning in Life (Oxford: Oxford University Press, 2014). 16 For a recent skeptical take on punishment, see Gregg Caruso, Rejecting Retributivism: Free Will, Punishment, and Criminal Justice (Cambridge: Cambridge University Press, 2021). 17 Bruce Waller, The Stubborn System of Moral Responsibility (Cambridge: The MIT Press, 2015). 18 Manuel Vargas, Building Better Beings: A  Theory of Moral Responsibility (Oxford: Oxford University Press, 2013). 19 Peter Furlong, The Challenges of Divine Determinism: A Philosophical Analysis. (Cambridge: Cambridge University Press, 2019). 20 Markus Höfner, “The Affects of the Soul and the Effects of Grace: On Melanchthon’s Understanding of Faith and Christian Emotions.” in Michael Welker (ed.), The Depth of the Human Person: A Multidisciplinary Approach. (Grand Rapids: Eerdmans, 2014), 218–35. 21 Jesse Couenhoven, Stricken by Sin, Cured by Christ: Agency, Necessity and Culpability in Augustinian Theology (Oxford: Oxford University Press, 2013); Paul Helm, The Providence of God (Downers Grove: IVP Academic, 1994). 22 Guillaume Bignon, Excusing Sinners and Blaming God: A Calvinist Assessment of Determinism, Moral Responsibility and Divine Involvement in Evil (Eugene: Pickwick, 2018). 23 Couenhoven, Stricken by Sin, Cured by Christ. 24 https://survey2020.philpeople.org 25 The connection between Christian theism and libertarianism has not gone unnoticed among critics of libertarianism and theism. Indeed, some philosophers have suggested that the popularity of libertarianism is ultimately due to the influence of theism and Christian theology on philosophical background assumptions. Manuel Vargas, for instance, has suggested that theism is the “elephant in the room” when libertarianism is discussed. This would constitute a problematic case of “motivated reasoning” – a situation where a philosopher misconstrues the strength of an argument or a piece of evidence because of a-rational emotional or volitional factors, like desiring or preferring a specific outcome. See Manuel Vargas, “The Runeberg Problem: Theism, Libertarianism, and Motivated Reasoning.” in Kevin Timpe & Daniel Speak (ed.), Free Will and Theism (Oxford: Oxford University Press, 2016). 26 Kevin, Timpe & Daniel Speak, “Introduction.” in Kevin Timpe & Daniel Speak, (ed.), Free Will & Theism (Oxford: Oxford University Press, 2016), 16–17. 27 Patrick Todd, “Does God Have the Moral Standing to Blame?” Faith and Philosophy 35/1 (2018): 33–55. 28 Kevin Timpe, Free Will in Philosophical Theology (London: Bloomsbury, 2014). 29 See, for example, Ken Perszyk (ed.), Molinism: The Contemporary Debate (Oxford: Oxford University Press, 2012).

22  Aku Visala 30 See, for example, William Hasker, Providence, Evil, and the Openness of God (London: Routledge, 2004). 31 Richard Swinburne, Responsibility and Atonement (Oxford: Oxford University Press, 1989). 32 Richard Swinburne, Mind, Brain, and Free Will (Oxford: Oxford University Press, 2013). 33 Kevin Timpe, Free Will in Philosophical Theology (London: Bloomsbury, 2014). 34 Aku Visala & Olli-Pekka Vainio, “Erasmus versus Luther: A Contemporary Analysis of the Debate on Free Will.” Neue Zeitschrift fur Systematische Theologie und Religionsphilosophie 62(3) (2020). 35 The most famous representative of semicompatibilism is John Martin Fischer. 36 I am not claiming that Luther is a consistent free will skeptic. However, these themes can be found in his writings. See Visala  & Vainio, “Erasmus versus Luther.” 37 Derk Pereboom, “Libertarianism and Theological Determinism.” in Kevin Timpe & Daniel Speak (ed.), Free Will and Theism (Oxford: Oxford University Press, 2016), 112–31. Derk Pereboom, “Theological Determinism and the Relationship with God.” in Hugh McCann (ed.), Free Will and Classical Theism (Oxford: Oxford University Press, 2017), 201–20. 38 See Katrina Hutchison, Catrionna Mackenzie & Marina Oshana, Social Dimensions of Moral Responsibility (Oxford: Oxford University Press, 2018). For the theological issues involved, see Aku Visala, “Free Will and Relationality: Theological, Philosophical, and Scientific Perspectives.” Philosophy, Theology, and the Sciences 8/2 (2022): 184–208.

Bibliography Bignon, Guillaume. Excusing Sinners and Blaming God: A Calvinist Assessment of Determinism, Moral Responsibility, Divine Involvement in Evil. Eugene: Pickwick, 2018. Caruso, Gregg. Rejecting Retributivism: Free Will, Punishment, and Criminal Justice. Cambridge: Cambridge University Press, 2021. Couenhoven, Jesse. Stricken by Sin, Cured by Christ: Agency, Necessity, and Culpability in Augustinian Theology. Oxford: Oxford University Press, 2013. Fischer, John Martin  & Mark Ravizza. Responsibility and Control: A  Theory of Moral Responsibility. Oxford: Oxford University Press, 1998. Fischer, John Martin, Robert Kane, Derk Pereboom & Manuel Vargas. Four Views on Free Will. Oxford: Blackwell, 2007. Flint, Thomas. Divine Providence: The Molinist Account. Ithaca: Cornell, 1998. Frankfurt, Harry. “Freedom of the Will and the Concept of the Person.” The Journal of Philosophy 68(1) (1971): 5–20. Furlong, Peter. The Challenges of Divine Determinism. Oxford: Oxford University Press, 2019. Hasker, William. Providence, Evil, and the Openness of God. London: Routledge, 2004. Helm, Paul. The Providence of God. Downers Grove: IVP Academic, 1994. Höfner, Markus. “The Affects of the Soul and the Effects of Grace: On Melanchthon’s Understanding of Faith and Christian Emotions.” in Michael Welker (ed.), The Depth of the Human Person: A Multidisciplinary Approach. Grand Rapids: Eerdmans, pp. 218–35, 2014. Hutchison, Katarina, Catrionna Mackenzie & Marina Oshana (ed.). Social Dimensions of Moral Responsibility. Oxford: Oxford University Press, 2018.

Introduction 23 Kane, Robert. The Significance of Free Will. Oxford: Oxford University Press, 1996. Mele, Alfred. Free: Why Science Has Not Disproved Free Will. Oxford: Oxford University Press, 2014. Mickelson, Kristin. “The Manipulation Argument.” in Kevin Timpe, Megan Griffith & Neil Levy (ed.), The Routledge Companion to Free Will. London: Routledge, pp. 166–78, 2017. Nichols, Shaun. Bound: Essays on Free Will and Responsibility. Oxford: Oxford University Press, 2015. Pereboom, Derk. Free Will, Agency, and Meaning in Life. Oxford: Oxford University Press, 2014. Pereboom, Derk. “Libertarianism and Theological Determinism.” in Kevin Timpe & Daniel Speak (eds.), Free Will and Theism. Oxford: Oxford University Press, pp. 112–31, 2016. Pereboom, Derk. “Theological Determinism and the Relationship with God.” in Hugh McCann (ed.), Free Will and Classical Theism. Oxford: Oxford University Press, pp. 201–20, 2017. Perzyk, Ken (ed.). Molinism: The Contemporary Debate. Oxford: Oxford University Press, 2012. Searle, John. Freedom and Neurobiology: Reflections of Free Will, Language, and Political Power. New York: Columbia University Press, 2004. Smith, Angela. “Control, Responsibility, and Moral Assessment.” Philosophical Studies 138 (2008): 367–92. Swinburne, Richard. Mind, Brain, and Free Will. Oxford: Oxford University Press, 2013. Swinburne, Richard. Responsibility and Atonement. Oxford: Oxford University Press, 1989. Timpe, Kevin. Free Will in Philosophical Theology. London: Bloomsbury, 2014. Timpe, Kevin. “Leeway vs. Sourcehood Conceptions of Free Will.” in Kevin Timpe, Megan Griffiths & Neil Levy (eds.), The Routledge Companion to Free Will. London: Routledge, pp. 213–24, 2017. Timpe, Kevin & Daniel Speak. “Introduction.” in Kevin Timpe & Daniel Speak (ed.), Free Will & Theism. Oxford: Oxford University Press, pp. 1–26, 2016. Todd, Patrick. “Does God Have the Moral Standing to Blame?” Faith and Philosophy 35/1 (2018): 33–55. van Inwagen, Peter. An Essay on Free Will. Oxford: Oxford University Press, 1986. Vargas, Manuel. Building Better Beings: A Theory of Moral Responsibility. Oxford: Oxford University Press, 2013. Vargas, Manuel. “The Runeberg Problem: Theism, Libertarianism, and Motivated Reasoning.” in Kevin Timpe & Daniel Speak (ed.), Free Will and Theism. Oxford: Oxford University Press, pp. 27–47, 2016. Visala, Aku. “Free Will and Relationality: Theological, Philosophical, and Scientific Perspectives.” Philosophy, Theology, and the Sciences 8/2 (2022): 184–208. Visala, Aku & Olli-Pekka Vainio. “Erasmus versus Luther: A Contemporary Analysis of the Debate on Free Will.” Neue Zeitschrift für Systematische Theologie und Religionsphilosophie 62(3) (2020): 311–35. Waller, Bruce. The Stubborn System of Moral Responsibility. Cambridge: The MIT Press, 2015. Widerker, David  & Michael McKenna (ed.). Moral Responsibility and Alternative Possibilities: Essays on the Importance of Alternative Possibilities. Aldershot: Ashgate, 2006.

Part I

Compatibility

2 Why Christians Should (Still) Be Compatibilists Jesse Couenhoven

Nearly 20 years ago, Lynne Rudder Baker offered some theological reasons not to be libertarian in her “Why Christians Should Not be Libertarians.”1 Although it may still be true that “the prevailing view of Christian philosophers today seems to be that Christianity requires a libertarian conception of the will,” Baker’s argument (and others like it) has had an effect.2 Theologically driven compatibilism, which was once both under-developed and mainly the province of Calvinists, has become a broader undertaking, and conceptually more complex.3 At the same time, philosophers, who for a time rarely engaged theological matters, are now writing complex treatises on topics such as divine determinism and the impeccability of Christ.4 Christian libertarianism has also grown deeper and more complex. Not surprisingly, theologically astute libertarians have written replies to Baker’s essay (and my own work) that have changed the conversation in important ways. Thus, it is high time to consider her overarching question once again. My central contention in this essay is that the more things change, they more they stay the same. There continue to be significant theological reasons for Christians to prefer compatibilism. In general, that is because the theological “cost” of being a compatibilist is lower than that of being a libertarian. Compatibilism has a theoretical advantage: it puts less pressure on theological convictions than libertarianism, which allows theological reflection to develop more organically, on its own terms. In developing this thesis, I  endeavor to keep in mind a risk and an opportunity. Any wide-ranging discussion of a topic as complex as that of free will risks superficiality.5 That risk comes with a potential reward: an integrated perspective on topics that are often treated individually. Taking such a perspective, while being careful about the details, will allow significant overarching themes to emerge. My discussion strategy is as follows: I summarize and assess what I take to be the best libertarian and compatibilist approaches to challenging questions regarding God’s freedom, the freedom of Christ, and the freedom of the heavenly saints, considering along the way what they say about fallen human agency and the role of grace in the Christian life. I will conclude by reflecting on two areas of theology that have widely been taken to be strengths for DOI: 10.4324/9781003306191-3

28  Jesse Couenhoven libertarian approaches, the topic of hell and the problem of divine goodness (theodicy). In spite of the ambition of this discussion, it should be clear that many important and relevant topics will not be addressed. The goal of this essay is correspondingly limited. I seek to offer some reasons why Christians should be compatibilists. It is also worth clarifying that my inquiry concerns the relative costs and benefits of being libertarian or compatibilist for Christians who are attracted to more or less orthodox views of the topics just mentioned (this is obviously a vague parameter, but what counts as orthodoxy is not always easy to specify). My presumption is that whichever view of free will cuts off fewer possibilities of belief has conceptual advantages. Making Sense of Divine Freedom What some call “leeway libertarianism” is still taken for granted by many libertarians. That view takes having a choice between alternative possibilities as fundamental to the idea of freedom. As Kevin Timpe has put it, for leeway incompatibilists, “free will is primarily a function of being able to do otherwise than one in fact does.”6 Having such leeway is valued for two related reasons. First, if it is genuinely up to you to choose between significant alternatives, then your choice is the deciding factor, and you are in control. Second, having such control is necessary for being a moral agent who can do right or wrong and be praised or blamed for it. Many theological libertarians still find this view intuitive, and it has also been held by influential philosophers.7 When, in early publications of my own, I followed Baker in claiming that compatibilists can do better than libertarians in making sense of Christian doctrines, I, like her, had such approaches in mind.8 If significant choice between alternative possibilities is the sine qua non of freedom, and thus required for moral agency, God seems to be in a bad position. As a necessarily perfect being, God cannot choose evil, and to keep the argument short, that means God lacks significant moral choices. This makes God lack moral praiseworthiness, which in turn makes even imperfect human moral agents more praiseworthy than God – surely, an unfortunate result. I take it that this and related theological arguments against leeway libertarianism offer powerful reasons for Christians with an interest in orthodoxy (and likely other theists, too) to question the view. Timpe, at least, has agreed.9 Timpe, Eleonore Stump, Katherin Rogers, and others have responded, however, by arguing that there is a better way to be libertarian, one that does as good a job accounting for central Christian doctrines as compatibilist approaches – and is better on questions of theodicy.10 Source libertarians, for whom “free will is primarily a function of an agent being the source of her actions in a particular way,” shift the focus away from alternative possibilities, helpfully recasting the idea of being in control.11 My discussion here will focus on Timpe’s attractively named version

Why Christians Should (Still) Be Compatibilists 29 of this view, virtue libertarianism (which arguably can be traced back as far as Saint Anselm). Timpe develops the virtue aspect of this view by arguing that freedom has a kind of teleology. Having choices is good not intrinsically but instrumentally because it makes possible genuine moral virtue – habits developed by a lifetime of free choices. Having such virtue is compatible with having fewer and fewer choices, as one’s volitional history gradually rules certain things out via the formation of ingrained habits. Having choices between alternative possibilities still matters, then, but not as the intrinsic heart of freedom. Rather, freedom itself is a certain sort of self-determination, and choices are instrumental to that end. Accordingly, if one could have ultimate sourcehood without the ability to do otherwise, one would be free nevertheless. Virtue libertarianism is, from my Augustinian point of view, a significant improvement on leeway incompatibilism. It better accommodates Augustine’s claim that the highest freedom is a power for goodness.12 It is a better fit for traditional Christian doctrines about the need for grace in the moral life as well.13 But let us briefly consider the details, focusing for now on God’s exemplary freedom. The triune God is, of course, the most free of any being in Christian theology. The virtue libertarian take on why divine freedom is so great in spite of God’s inability to sin is straightforward but sophisticated. Because God is a necessary being, necessarily perfect in every way, God cannot sin. In a human being, this might undermine freedom since without the ability to sin, human agents cannot form their own moral character. However, because God is neither made to have God’s perfections by some other power nor has those perfections contingently, God does not need to take control of the divine character. Since there is no ultimate source of the divine perfections other than God, God does not have to make virtue God’s own over time by making repeated, undetermined, choices. We may simply credit God for the virtues on display in the life of the immanent Trinity. As Timpe writes, God is the source of His actions given that He acts on the basis of motivational reasons, cannot choose to do something for which He has no motivational reasons, and is not causally determined to do so by anything outside of Himself.14 It is hard to overstate how much better this view is than the “volitionalist” view that requires God to choose God’s own perfections in order to receive the sort of praise Christians customarily bestow on their divinity.15 However, it still seems to me that compatibilism offers some theological advantages as we puzzle over why God’s freedom is the best, freedom par excellence. Before I make that argument, it will be helpful to briefly sketch the kind of compatibilist view I have in mind. The Augustinian compatibilist account I offer here as a friendly competitor to virtue libertarianism is meant to be a big tent.16 There are any number of compatibilist approaches, but those that best

30  Jesse Couenhoven make sense of divine freedom have four main features. First, they offer an explanation of why a person can be free even when they lack the sort of undetermined sourcehood that libertarians deem essential to freedom. The basic claim is that we can properly consider persons morally responsible (and thus open to praise or blame) for their reason-responsive attitudes and behaviors. How a person came by the attitudes that lead to their behaviors is not irrelevant to ascriptions of responsibility, but a person could nevertheless be responsible for a behavior that God has determined them to do, as long as God works with their intellect and wills. When they act out of their beliefs and desires, they are free. Second, these views are non-volitionalist, meaning that they do not consider choosing among morally significant alternatives the core of what it means to be free.17 Third, and related, these views are open to the idea that freedom comes in degrees, and that the best kind of freedom is the ability to flourish as the kind of being you are. Freedom, on such views, is a kind of power, the power to live well. Finally, it is important to clarify, in the face of common misunderstandings, that being a compatibilist does not commit one to belief in determinism. As I understand it, compatibilism commits one to the belief that freedom is compatible with certain sorts of determination, but (1) one can be agnostic about whether those factors actually exist and (2) what those factors are thought to be will depend in part on other commitments a compatibilist may have. For example, I  am a compatibilist who believes in irresistible divine grace and predestination but I  am doubtful about the existence of physical determinism and agnostic about divine determinism.18 As we will see, I also think that God’s freedom is best described in compatibilist terms, even though determinism is not even a possibility in God’s case. From the compatibilist perspective just sketched, the virtue libertarian account of God’s freedom is impressive. Indeed, the two accounts overlap in being non-volitionalist. As I have suggested, volitionalists, compatibilist or incompatibilist, have in common the view that responsibility is tied to histories of self-defining choices.19 But since the God of traditional theology has no such history, non-volitionalist theories are in a better position to account for the nature of divine freedom. On such views, although God’s greatness creates certain necessities, that is not a problem for God because God is, by definition, not subject to other powers but is a se. God’s greatness consists, in part, in God’s inability to sin, but lacking the ability to be bad is not a weakness. Divine freedom is not potential but actuality, a perfect life in love. In spite of its strengths, I  nevertheless have a significant concern about the virtue libertarian take on divine freedom. It relates to the fact that God is generally taken to be free not only in God’s acts but also with regard to the divine character. We can present the worry in two ways. First, for virtue libertarians, as on most views, having free will means having control – free will is the control condition on moral responsibility.20 I wonder, though, if it makes sense to say that God controls the divine perfections. Talk of control makes it sound as though God stands over God’s own character, intending it to be a certain way, when in fact God is (partially) constituted by it.21

Why Christians Should (Still) Be Compatibilists 31 To put the concern another way, if undetermined sourcehood is a necessary condition of freedom, in what sense is God the source of the divine perfections? Perhaps the idea is that God is the source of the divine perfections because, being undetermined, nothing else controls God. However, not being controlled does not seem to be enough for sourcehood. In the human case, it is widely accepted that lack of determinism by itself does not put a person in control. The same applies to God. Virtue libertarians certainly can say that God is the source of God’s actions, because they flow from God’s motivations. That arguably offers a kind of control of God’s actions. The very being of God, however, seems like a different story. The doctrines of divine aseity and simplicity, which many Christian libertarians embrace, suggest that God is God’s attributes, not that God (or the divine will) is their source. Control seems to require having a grasp on something, or standing behind it, which is not how God and God’s necessary excellences are related. In conclusion, then, I doubt that God has the kind of control of divine character virtue libertarians require for ascriptions of moral praise.22 As an Augustinian compatibilist, my inclination is to deal with this problem by not requiring that God have the kind of control that comes from being the unconditioned source of something. Rather than claiming that God is the source of what God is, which threatens to create a problematic recursivity, it seems better to focus on the fact that God is the owner of the divine attributes, in the sense that they are rightly ascribed to God. God’s freedom is not backward looking, requiring us to inquire into the history of the divine being. It is excellent in its unparalleled power for life. Unlike other agents, God is fully able to do and be what God pleases to do and be. Notably, compatibilists can agree at this point that God is too free to be causally determined – but the rationale differs from the libertarian take. On a normative conception of freedom, the highest freedom is a kind of power, not potential but actuality, the ability to do and be good. That power is the greatest when nothing can possibly stop it. So although not being determined is not a necessary feature of more mundane forms of freedom, it is a necessary feature of God’s agency, which exhibits freedom par excellence. Making Sense of Heavenly Freedom The freedom of the heavenly saints has important similarities and differences to divine freedom. In a way, they are like God, settled in a power to love well, and therefore unable to sin. The difference, from a virtue libertarian perspective, is that they must have histories of self-constituting choices. Because human beings are created, and thus subject to other powers, they need the ability to do otherwise, at least at certain points in their moral careers, in order to be free agents. By making choices between morally significant alternatives, they are able to become the sources of their own moral identities, and thus come to deserve praise or blame for them. Timpe, writing with Timothy Pawl, has drawn on Aristotle’s insight that acts or omissions made

32  Jesse Couenhoven over a lifetime can become so ingrained as to become habits that are psychologically impossible to change. This “deposit” of self-constituting actions can produce good or bad character; the heavenly saints are those who are so settled in their good character that they find it impossible to do evil. They might continue to have choices about what goods to involve themselves in, and how, but their wills are fundamentally set, oriented toward the good.23 I think this offers an adequate explanation of saintly impeccability – even in this life we sometimes see people for whom certain virtues have become second nature. I wonder, however, how many of us non-saints it applies to. It is difficult to believe that most human beings reach anything like heavenly perfection in character via their earthly choices. Insofar as our wills are stably set on anything in this life, it is still often on the less than fully admirable. An obvious response is to consider the possibility that purgatory will intervene between death and heaven for all those who are not secure in their perfection, until they are. This undoubtedly helps. It is worth noting, however, that purgatory has traditionally been at least a somewhat controversial conjecture. If libertarianism makes purgatory a mandatory feature of Christian theology, it puts a kind of pressure on theologians at this point that compatibilism would not. One might also wonder how many sinners can work their way up from fractured virtue to wholehearted affirmation of the good, even with purgatorial assistance. This bottom-up approach may also run into trouble with traditional claims about the unity of the virtues. If it is hard to be fully just without being fully wise, and so on, it is hard to achieve the virtues one by one. While moral exercise can help us grow in virtue in this life, and perhaps the next as well, it may well be the case that a rounded perfection in virtue, unlike the piecemeal collection of habits we tend to develop to get us by, cannot be achieved without something like the gift of a “new heart” often spoken of in the Bible. Part of the attraction of Augustinian compatibilism, of course, is its claim that such gifts of grace do not undermine freedom, but rather, enhance it, by helping cure the moral and psychological diseases that keep us from living well. It is possible, however, that the libertarian can make a similar move, which could make purgatory theologically non-necessary. Eleonore Stump has argued that divine grace is necessary but not sufficient for developing the virtues the Christian needs in this life and the next. Her suggestion is that divine grace can be prevenient, in a way – God need not wait for us to seek divine assistance, but can take the initiative without violating our free will as long as grace is an offer we can refuse. She argues that human wills are able to remain quiescent in response to God’s outstretched hand, neither saying yes nor saying no. This omission to act does not take the lead, putting us in God’s place. Rather, it humbly permits God to heal and strengthen our intellects and wills. Thus, it is up to us whether we develop in virtue or take up the divine offer of salvation, but it is not simply up to us. Her libertarian take on the necessity of grace is anti-Pelagian, Stump argues, because it accepts the Augustinian claim that sinful human beings are not able to do or be good without the help of divine grace.24

Why Christians Should (Still) Be Compatibilists 33 Given this account, virtue libertarians could affirm that God is able to graciously strengthen the character of the saints, or confirm them in their good choices, and in so doing to fulfill their prayers for salvation. That would make it possible for them to achieve heavenly perfection with the help of infused grace. In such cases, the saints would not (from a libertarian perspective) be entirely free with respect to every aspect of the perfected character created by such divine grace – which would have been given to them unilaterally – but they would nevertheless retain responsibility for their previous, undetermined choices, including their prayers for divine assistance. They would deserve praise for their heavenly virtues in a limited sense, since asking for divine help, or not opposing it, makes them a source of their newly formed character, but only in a modest respect (and their heavenly character would be derivatively free to a limited degree).25 This nuanced libertarian account of divine grace enabling heavenly freedom is theologically rich and elegant. Yet compatibilists still have minor advantages on this topic, in my view. First, theological compatibilism creates no pressure to rule purgatory in or out, leaving room for other considerations to drive accounts of the afterlife. Second, Augustinian compatibilism easily explains how unproblematic it would be for those who have died to quickly receive their hearts’ true desire, being transformed by union with Christ without any loss to their freedom or praiseworthiness. Their reception of divine grace need not be limited to cases where they fail to reject divine aid, but can include cases where God helps us without our having asked. Correspondingly, compatibilism allows the heavenly saints to receive more credit for their perfected goodness because the view embraces responsibility for merits they did not choose to have. A further consideration deserves mention at this point, though it will be further developed in the following section. Libertarianism requires human freedom to be more unlike God’s than compatibilism does. For libertarians, human freedom is volitionalist, dependent on undetermined choices. It is therefore a different kind of freedom than God’s. For compatibilists, by contrast, although human freedom lacks the ultimate perfection of God’s freedom, this is a difference in degree, not in kind. Participation in God’s life of freedom thus seems more straightforward on a compatibilist view. The heavenly saints share in God’s non-volitionalist freedom, which is the sort of freedom they too may have. This point will be highlighted by the discussion of Christ’s freedom in the following section. Making Sense of Incarnate Freedom Having discussed divine and human freedom, we are now in a position to consider an even more difficult topic, the freedom of the God-man. The challenge for libertarians is fairly easy to articulate: traditionally, Christ has been thought to have two natures, human and divine, each with its own will. Given the fact that God is necessarily without sin, the two natures can remain

34  Jesse Couenhoven in union only if the human will makes no sinful choices. If Jesus is fully human, however, he would seem to need to reach impeccability in the same manner as the heavenly saints, through a lifetime of choices that could have been otherwise. In turn, that leaves open the possibility that he could sin, thereby also leaving open the possibility of a failure in the union between the divine and human natures constituting him as a person. Requiring the human nature of Christ to have libertarian freedom thus seems problematic for the incarnation. As we survey attempts to respond to these challenges, it is clear that nothing approaching a consensus has been reached. One point of agreement among libertarian philosophers and theologians is that Jesus, qua human being, could have sinned. Attempts to develop the implications of that claim have diverged along three major paths. First, one might reject Christian orthodoxy and accept that because Jesus, the person, is human, he not only could have sinned; he actually did sin. Even if Christ was imperfect, he could learn from his mistakes and become perfected in the manner and degree necessary for him to do the salvific work God planned. On this view, Jesus’ moral agency is like that of the heavenly saints, although he may receive more grace that allows him to become virtuous more quickly, given the uniquely close association of divine and human in his case. How exactly that association is maintained during the course of Jesus’ history, given the need for the necessarily impeccable to be united with the sinful, is unclear.26 Second, what some call the “probation model” affirms that although Jesus could have sinned, he did not. According to this view, Jesus’ human nature was united with the second person of the Trinity on something like a trial basis.27 Had the human will been about to make a sinful choice, the unity of the two natures would have been dissolved in order to preserve divine impeccability. Perhaps there were, indeed, cases where this occurred (and, therefore, multiple would-be incarnations). In Jesus, however, all went as planned. One implication of this view is that the unification of natures in Jesus was not a unique commitment on God’s behalf, but a contingent affair. A further implication is that Jesus was not fully the Christ, the Messiah, until he died and, having proved himself, was resurrected. Third, what we might call the “Molinist view” develops a less radically contingent version of the previous view. Since divine foreknowledge gives God the ability to know what will happen, God can ensure that everything in this one particular union of the divine and human goes as planned by virtue of divine foreknowledge of what Jesus, qua human, will do throughout his life.28 On this view, Jesus is selected to be the God-man because God knows he will, qua human, choose virtuously. He is therefore selected because his free and undetermined choices ensure the success of the incarnational project. In spite of the creativity of these suggestions, it will be obvious that each comes with significant costs. The first option, which makes Jesus a sinner like us, rejects one of the central beliefs of Christianity. The Molinist view requires adherence to highly controversial ideas about divine “middle knowledge,”

Why Christians Should (Still) Be Compatibilists 35 and creates puzzles about whether Jesus could have a counterfactual identity as a merely human person who is separable from his identity as Christ in the incarnation. Finally, the probation view both turns the incarnation into a throw of the dice and raises questions about other aspects of Christian doctrine – especially Mary’s role in Christian piety, since she too would be “on probation” for her role as theotokos. In a way, however, the particular costs of any particular accommodation to libertarianism are not the point here; the overarching issue is the demanding nature of libertarian requirements for freedom, which rather than making room for a wide variety of possible theological views puts pressure on theologians to revise their views in order to fit its strictures. It seems safe to conclude that the traditional doctrine of the free impeccability of Christ’s human nature continues to present significant difficulties for libertarians. The contrast between libertarian and compatibilist approaches on this topic is clear enough that Pawl, in a recent discussion of the freedom of Christ, noted that defending claims about Jesus’ freedom is hardly a challenge for compatibilists: “If compatibilism were true, then one would be able to reconcile the freedom of Christ’s human will with the divine determination of that will very easily.”29 Following the suggestion of patristic authors such as Maximus the Confessor, compatibilists can argue that there need be no opposition between Christ’s human and divine wills because Christ’s human will, having been healed by grace, is oriented to follow the divine. Given the centrality of Christ in Christian theology, I  want to suggest that this problem of compatibility between human and divine agency is not an issue that can be restricted merely to Christology as a special case. Rather, the challenge libertarians face in making sense of Christ’s impeccable freedom highlights a broader theological challenge of making sense of the Christian story, in which participation in the divine is the telos of human agency. It is partly for that reason that Christians have traditionally taken Christ to be an exemplary figure. The union of divine and human natures in his person is, of course, usually considered sui generis, but there has nonetheless been a sense that as the firstborn of all creation his life with God offers a pattern to which the rest of us aspire. Notably, that pattern is one of radical divine prevenience. Jesus’ human nature is given a gift it has no opportunity to refuse, a union with the divine that constitutes him as the person he is, and his human nature as the sort of thing it is. Like Israel’s relationship with God more generally, this is not a status that was bestowed because it was merited. It is simply given, not merely as a gift but also as a task that presents special opportunities and challenges. The natural interpretation of Jesus’ freedom in this union is not that participation in the divine life is too much like being a puppet, or that it would be better had the human nature been given a choice in the matter. Rather, the natural reading of the “Christ event” is that being given a particular identity by one’s maker is quite good indeed, and leads not to the undermining of one’s agency or freedom but provides the basis for it.

36  Jesse Couenhoven Making Sense of Graced Freedom Christian doctrines of the fall and redemption are famously both complex and controverted. In this essay, there is not room to consider libertarian appropriations of the obviously compatibilist-friendly ideas of original sin or predestination.30 Regarding those topics, I  will simply reiterate that being a compatibilist does not require commitment to any strong claims about bondage to sin, election, or divine determinism. Compatibilism allows for a wide range of traditional soteriological views. This flexibility, I have suggested, is a significant attraction of the approach. Libertarianism, by contrast, exerts a kind of gravitational pull on Christian doctrine, bending it to its commitments. Rather than taking up specific soteriological doctrines, I want to return to a topic discussed briefly earlier, the virtue libertarian view of divine grace and its role in the Christian life. Expanding on Stump’s approach, Timpe offers the following account of saving grace: [I]ndividuals have no control over whether grace is given, but since they can resist grace, they do have indirect control over whether that grace causes its intended result. Agents thus have control over their own salvation, not in virtue of causing it, but in virtue of an omission that they indirectly control. The conjunction of God’s grace and the individual’s refraining from resisting will be jointly sufficient for the individual’s coming to saving faith.31 This account is subtle and intriguing, but it raises concerns for Augustinians like myself because it insists that whatever is morally praiseworthy about us must ultimately be up to us, at least to refuse. As I suggested earlier, this is theologically worrisome because of its tension with the idea of divine prevenience. It is also psychologically demanding since a libertarian view requires accepting divine grace to be the kind of intentional choice required for any act to count as freely self-determining. Let’s explore this psychological worry first, which will return us to the theological worry. Virtue libertarianism presumably requires conscious awareness of the operations of grace in us so that we have some realistic opportunity to resist that grace, if we so choose. If we are ignorant of the activity of God’s grace, our response or lack of response to it would not be under our control. The problem is that it seems psychologically unrealistic to say that we are or should always be aware of the work of grace in our lives. There are actually two problems here. One is that we do not always know what is going on inside ourselves, why we feel certain ways, or do various things. So grace might influence us without our knowing (indeed, that is mostly how it works, as far as I can tell). Second, it is hard to know what counts as grace, and thus as something that is a divine act, as opposed to all the things that are up to chance or other agents. This is one reason why

Why Christians Should (Still) Be Compatibilists 37 stories about grace are typically retrospective, realizations that more was going on than one thought at first. Consider, also, cases where we do have such awareness. If we are aware of divine grace as such, and know that it will inevitably work its results if it is not intentionally resisted, and we then fail to resist it, it seems as though we are actually consenting to it. In turn, that undermines Stump’s claim that the will is simply quiescent in these cases. This is a version of an old problem about whether acts and omissions are really so different, morally speaking. If grace must intentionally be permitted in order for it to help us, it seems to me that it makes little difference whether that happens via an act or an omission. The libertarian approach to soteriology is still, basically, that it is up to us whether to accept or reject the divine offer. A compatibilist doctrine of grace, by contrast, recognizes that divine grace may work in subtle ways. It does not undermine human agency, however, because it works with our intellects and wills, supporting not circumventing them. This leads to a question that connects with the earlier discussion of heavenly freedom. The typical libertarian picture of the moral life, even with divine grace taken into account, is of a sanctification that takes place slowly, in many small steps. As we noted earlier, however, many conversion stories tell of far more radical shifts, operations on the heart significant enough that Christians have been led to use the language of “new creation.”32 Is the libertarian approach able to accommodate such transformative grace? It would be a significant loss for a Christian doctrine of grace not to be able to do so, but I suspect that accommodating the possibility of significant and sudden shifts in moral and spiritual identity might be difficult even for source incompatibilists. Operative grace, as Augustine understood it, can make major changes to the beliefs and desires of God’s beloved. Fallen persons might have libertarian freedom with respect to their consent to receiving such grace, on an account like Stump’s, but what about the results of the changes themselves? Consider an analogy. In a science fiction future, I might consent to brain therapies that make me smarter.33 Even so, if they are successful I am not sure how much libertarian credit I deserve for my heightened intellectual abilities and the new me that results. Strangely enough, therefore, the saints who receive grace end up being second class, on a libertarian view. From the perspective of merit, it would be better to tough it out in purgatory than to accept divine assistance. Libertarian views about freedom thus seem to significantly restrict both what divine grace can do and the goodness of that grace. This creates tension with traditional Christian understandings of the role grace plays in our lives. In summary, as Heath White has noted, in spite of the genuine rapprochement between virtue libertarianism and Augustinian compatibilism, their theologies of grace differ greatly.34 Compatibilism’s openness to the idea that our agency is substantially from sources not of our own making leads to a more deeply relational way of thinking about our moral and spiritual lives. That not only raises natural questions about fairness but also fits the

38  Jesse Couenhoven Christian idea that participation is a key part of all personal reality, from the Trinity on down. Making Sense of Divine Goodness Even if libertarians have the theological weaknesses just mentioned, they have a ready reply. Compatibilism has its own problems. In particular, compatibilists have trouble making sense of evil. All theists have some difficulty with this topic, of course, but libertarianism offers the theist an option the compatibilist lacks – the free will defense. Timpe put the point this way, responding to Baker: [L]ibertarianism fares better in the face of the problem of evil than does compatibilism, since it has more resources to muster in its defense. If humans have libertarian free will, then God cannot create a world containing such agents and unilaterally guarantee that that world contains no evil. . . . Compatibilism, on the other hand, cannot so easily make the claim that God cannot create a world containing such agents and unilaterally guarantee that that world contains no evil.35 Timpe also argued that libertarianism makes it easier to respond to questions about why God would permit some created persons to go to hell.36 The compatibilist must therefore respond to two challenges. These challenges are related – hell, after all, is one of the evils that theists can be called upon to account for. Thus, it makes sense to begin by discussing theodicy generally.37 My argument will be that compatibilists are, at least, no worse off than libertarians when it comes to explaining why a good God would permit evil and hell. Indeed, as we will see, compatibilists may have an advantage in responding to the problem of horrendous evils. It is often said that compatibilists cannot make use of the free will defense. However, that claim is misleading, in three ways. First, it is not actually impossible for compatibilists to affirm the value of undetermined choices. As I noted in the first section of this essay, compatibilists do not have to be committed to determinism. They might, therefore, hold that some things happen because of luck. They might also hold that libertarian self-determination is good, when you can have it, but perhaps also rare, and not necessary for free agency or moral responsibility.38 Second, it is appropriate to have reservations about the highly demanding and blame-shifting appeals to human agency that have often accompanied free will theodicies. Leeway libertarian free will defenses have tended to operate on the assumption that once God creates persons, their free choices have such value that they must largely be left to their own devices, even if that means permitting grave harms that we would never permit one another to commit. In a systematic theological context, however, such notions are questionable, not least because they are in tension with the doctrine of grace discussed earlier.

Why Christians Should (Still) Be Compatibilists 39 Virtue libertarianism should be seen as an ally of at least some of these compatibilist concerns because it helpfully creates strictures that should guide the way Christian libertarians employ the free will defense. Free choice, we have seen, is not good in itself, but a means to the end of development in virtue. It makes sense to honor undetermined choices insofar as they serve that end. Moreover, the importance of grace in the Christian life means that it is not good for God to simply leave our moral and other development up to us. Thus, even from a libertarian perspective, the good of allowing created persons to determine their character for themselves only goes so far. Often, it is good for them to be given assistance and guidance in developing virtue and in making it possible for others to do the same. Virtue libertarians and Augustinian compatibilists both have reason, I am suggesting, to back away from some of the strong claims made on behalf of free choice by leeway libertarian theodicies. That said, it is important to highlight a third point, the fact that compatibilists can offer analogues to central aspects of well-known free will defenses. We can, for instance, recognize that independent choice may at times be valuable, for instrumental reasons. Divine permission to fail, morally and otherwise, makes it possible for human agents to make decisions and develop commitments of sorts that would not be possible in a world where we had no occasion to fear evil, or to be tempted by it. Facing the possibility of natural or moral evils may offer human agents unique opportunities to attain depth of character and moral and spiritual insight.39 There is thus some good in letting us make mistakes, so long as these mistakes are not overly destructive. Compatibilists can also argue that God permits some evils because a world in which God regularly intervened to limit the effects of evil would be one in which certain decisions would become unthinkable.40 You would not threaten harm if everyone knew it was impossible to follow through. Even on a compatibilist view God would have to carefully limit and shape divine influence on human agency, in order to honor the basic conditions for human action. Our agency requires lawlike continuity and regularity in the world so that we can act for the reasons that make sense to us, expressing our personalities in response to past events and in expectation of future outcomes.41 One might think of the trade-offs between the goods God seeks and the evils God permits as making possible certain sorts of narrative arcs (individual and communal). These narrative arcs place intrinsic constraints on even God’s ability to take away suffering. If God wants a certain sort of narrative, say of development in appreciation for the depths of the meaning of self-giving and redemptive love, or of the goodness of forgiveness, God must permit the sin and suffering that make such virtues possible.42 The sort of agency-development approach to theodicy just briefly summarized can be an appropriate element of attempts to understand why a good God might permit evils – so long as they do not seek to justify evils so awful that permitting them makes God seem intrinsically unloving, or so devastating that referring to agential development in their context becomes implausible. Such theodicies are humane when they do not seek to shift responsibility

40  Jesse Couenhoven for evils from God to humans, but rather show respect for the fragile and limited nature of human agency. They are also of limited value. As Marilyn Adams has argued, our world contains horrendous evils that make it difficult to see how those who have suffered them could have lives that are good to them. In the face of such evils, it is not helpful to refer to the theoretical good of making autonomous choices, nor is it fair to talk about the possibility of developing one’s agency. The problem with horrors is precisely that evils like severe physical suffering, significant biologically based improper function, and personality destroying or distorting abuse and trauma, extinguish such hope. The shadow of such evils should force us to shift our attention from the question of whether theists can justify the existence of any evil at all, to the question of whether God can be trusted and admired given the excessive horrors that fill our world. As Christians have sought to respond to the problem of divine goodness, two particularly compelling (if still incomplete) responses have been offered. Both are easier to defend and develop on compatibilist assumptions. First, Marilyn Adams has argued that God can defeat horrors by guaranteeing that each created person will participate in an incommensurable good – enjoyment of and participation in the supreme beauty of the divine life itself.43 Participation in such a great good ensures that horrors are marginalized in each person’s life so that they are not decisively meaning-defining for anyone. As Adams makes clear, this defense of divine goodness requires commitment to universal salvation. In turn, universalism fits best with compatibilist conceptions of human agency. The only sure way to ensure that all say “I do” at the divine wedding feast is for God to shape human agency, by bringing frail human agents into participation with God’s own life. This universalism is undoubtedly controversial, historically speaking, but it is hard to deny its theodical appeal. Second, Vince Vitale has recently developed a theodicy of identity that also is most easily appropriated by compatibilists.44 Drawing on Robert M. Adams’ suggestion that God may permit evil because doing so is the best way for God to ensure that we exist as the concrete, particular persons we are, Vitale argues that God may not be able to give us life without also allowing us to participate in the evils that have shaped our identity and contributed to making us who we are. This is not the same as the claim, discussed earlier, that suffering evil can have a pedagogical purpose. Rather, the suggestion is that if, when God created, he wanted certain concrete particular individuals to exist, a sure way to get them would be to ensure the existence of whatever is necessary for those individuals to exist, including the sorrows (and joys) that make them who they are. Given the ways in which our identities are tied up with a history of horrors, divine love for us might, ironically enough, give God a compelling reason to make room for evil. This line of thought does not explain or justify horrors, in the traditional sense of offering a reasonably satisfactory explanation for why they happen or how it is in certain respects better that they did. Nevertheless, it does

Why Christians Should (Still) Be Compatibilists 41 re-contextualize the ugliness of evil while offering us a way to see the God who permits such evil as loving. Electing to create a world like ours, and people like us, seems permissible if God cannot have us any other way, and if God elects and destines us for love, and thus for lives that are a blessing on the whole. At the same time, this approach requires that God determine at least certain features of our histories – our existence, and the factors that go into making us the individuals we are, to ensure that we happen. Compatibilism is therefore a better fit for this theodicy than libertarianism.45 I hasten to reiterate that neither of these approaches “solves” the problem of divine goodness in the face of horrendous evil. We cannot justify all the suffering we see throughout human (and animal) history. There are nevertheless some helpful things to be said, and my suggestion here has been that those things are easier to say if one is a compatibilist. Perhaps surprisingly, therefore, compatibilism is at least as helpful when it comes to theodicy as libertarianism, and perhaps more. Finally, let us briefly turn to the problem of hell. As I argued in the final chapter of my book on predestination, compatibilists are no worse off on this topic than libertarians.46 Both parties can offer explanations of why God would permit creatures to reject the offer of participating in the divine life. Doing so might, for instance, be necessary to build character of a certain sort. But the idea that God wants to give us a chance to make certain decisions on our own falls far short of the idea of hell. Hell – at least, as it has traditionally been conceived, as a place of eternal punishment – is not a necessary concomitant of what we might call non-salvation. Libertarian views do make somewhat better sense of why a person might repeatedly be allowed to refuse divine grace than compatibilist views. Our topic, however, is hell. And there is a huge gap between personal responsibility for saying no to God, and deserving eternal punishment. As we have noted, the free will defense is often employed in such a way that it presents a picture of a God who takes radical risks to permit the possibility of autonomous self-determination, even to the point of allowing us to eternally ruin our lives by making a variety of (typically, poorly informed) choices. It is natural to wonder why a God of love would allow us to make such mistakes. If hell is more like C. S. Lewis’s picture in The Great Divorce, where there are postmortem opportunities to turn hell into purgatory, we might still wonder why God should forever take no for an answer.47 We, after all, do not only take consenting persons to drug treatment programs. At a minimum, it seems better for the suffering not to be unlimited. Free choice does not, therefore, provide a compelling rationale for hell, understood as a place of eternal punishment. As a result, libertarians and compatibilists are in a similar position regarding hell – it remains a mystery.48 In summary, my suggestion is that compatibilism either helps with the problem of divine goodness, when it is paired with an affirmation of universal salvation, or, when paired with a doctrine of hell, makes the problem not noticeably worse than it is on libertarian views.

42  Jesse Couenhoven Conclusion: The Dialectical Situation In spite of its venerable history in Christian thought, compatibilism is, evidently, not intuitively attractive to many modern Christian philosophers and theologians. Although it is controversial to claim that freedom is compatible with determination, I have argued in this essay that this view has significant conceptual advantages for the theologically inclined. That is partly because compatibilism is a relatively lightweight and flexible view, adaptable to a variety of doctrinal stances. Unlike incompatibilism, which states what freedom must be, compatibilism takes a minimal stance, indicating what freedom can be. The Augustinian compatibilism I favor is more determinate, given its normative theological commitments. I have suggested, however, that it fits the Christian story better than its virtue libertarian competitor. Because of its emphasis on the continuity between divine and human freedom, and its insistence that our greatest freedom comes with participation in God’s life, it comes as little surprise that Augustinian compatibilism is better able to make sense of the liberating nature of the union of human and divine agency in Christ. More surprising might be my argument that compatibilism has certain advantages as we attempt to make sense of the problem of evil. But perhaps compatibilism’s Christological strengths should have alerted us to this possibility. After all, in both instances, we are concerned not with the fair sort of life to which libertarianism aspires, a morality tale about receiving deserved rewards, but with the kind of hard luck story for which compatibilism is prepared, a tragic yet beautiful tale about suffering, redeemed by shared love and grace.49 Notes 1 Lynn Rudder Baker, “Why Christians Should Not Be Libertarians: An Augustinian Challenge.” Faith and Philosophy 20(4) (2003): 260–78. 2 Baker, 460. 3 See, for example, Katherine Sonderegger, “The Doctrine of Providence.” in Francesca Aran Murphy & Philip G. Ziegler (eds.), The Providence of God: Deus Habet Consilium (New York: T&T Clark, 2009), 144–57; Jesse Couenhoven, “The Necessities of Perfect Freedom.” International Journal of Systematic Theology 14(4) (2012): 396–419; Michael Patrick Preciado, A Reformed View of Freedom: The Compatibility of Guidance Control and Reformed Theology (Eugene, Oregon: Pickwick Publications, 2019). 4 Leigh C. Vicens, Divine Determinism: A Critical Consideration (Madison: University of Wisconsin-Madison, 2012); Peter Furlong, The Challenges of Divine Determinism: A Philosophical Analysis (New York: Cambridge University Press, 2019); Heath White, Fate and Free Will: A Defense of Theological Determinism (Notre Dame, Indiana: University of Notre Dame Press, 2019); Timothy Pawl, In Defense of Extended Conciliar Christology: A Philosophical Essay (New York: Oxford University Press, 2019); Johannes Grössl  & Klaus von Stosch (eds.), Impeccability and Temptation: Understanding Christ’s Divine and Human Will (New York: Routledge, 2021). 5 I will accept as non-controversial the widely held claims that freedom is necessary for moral responsibility, and that moral responsibility, in turn, is what makes us proper subjects of moral praise and blame.

Why Christians Should (Still) Be Compatibilists 43 6 Kevin Timpe, Free Will in Philosophical Theology (New York: Bloomsbury, 2013), 8. 7 For a recent example of the former, see Friedrich Lohmann, “God’s Freedom: Free to Be Bound.” Modern Theology 34(3) (2018): 368–85. For the latter, see Alvin Plantinga, God, Freedom, and Evil (Grand Rapids: Eerdmans, 1989); Richard Swinburne, The Existence of God, 2nd edition (Oxford: Clarendon Press, 2004). 8 See Couenhoven, “The Necessities of Perfect Freedom.” The same seems to apply to Wes Morriston, “What Is so Good about Moral Freedom?” The Philosophical Quarterly, 50(200) (2000): 344–58; Wes Morriston, “Are Omnipotence and Necessary Moral Perfection Compatible? Reply to Mawson.” Religious Studies 39 (2005): 441–49. 9 Kevin Timpe, Free Will in Philosophical Theology (New York: Bloomsbury, 2013), 8. 10 Katherin Rogers, Anselm on Freedom (New York: Oxford University Press, 2008); Eleonore Stump, Aquinas (New York: Routledge, 2005). 11 Timpe, Free Will, 8. 12 For more on this topic, see Jesse Couenhoven, Stricken by Sin, Cured by Christ: Agency, Necessity, and Culpability in Augustinian Theology (chap. 3) (New York: Oxford University Press, 2013). 13 As a side note, the improved fit between source incompatibilism and traditional Christian doctrine also suggests that although modern source incompatibilism has typically been motivated by Frankfurt style counterexamples, it can also be motivated by doing theology. Bringing theology and philosophy into conversation thus enriches the conceptual possibilities available in both fields, a boon even to non-theists. 14 Timpe, Free Will, 112. 15 For example, Thomas V. Morris, Anselmian Explorations (Notre Dame: University of Notre Dame Press, 1987); Bruce McCormack, “Seek God Where He May Be Found: A Response to Edwin Chr. van Driel.” Scottish Journal of Theology 60(1) (2007): 62–79. 16 Thus, I will ignore the more controversial aspects of my own views, developed in Couenhoven, Stricken by Sin and elsewhere. Examples of the sort of compatibilist views I have in mind include Nomy Arpaly, Merit, Meaning, and Human Bondage: An Essay on Free Will (Princeton: Princeton University Press, 2006); Michael McKenna, Conversation and Responsibility (New York: Oxford University Press, 2012); Thomas Michael Scanlon, “Responsibility and the Power of Choice.” Think 12(33) (2013): 9–16; Angela M. Smith, “Responsibility as Answerability.” Inquiry: An Interdisciplinary Journal of Philosophy 58(2) (2015): 99–126. 17 See Neil Levy, “Luck and History-Sensitive Compatibilism.” The Philosophical Quarterly 59(235) (2009): 237–51; Jason Benchimol, “The Moral Significance of Unintentional Omission: Comparing Will-Centered and Non-Will-Centered Accounts of Moral Responsibility.” in Nicole A. Vincent (ed.), Moral Responsibility: Beyond Free Will and Determinism (New York: Springer, 2011), 101–20. I first appropriated their argument in Jesse Couenhoven, “The Problem of God’s Immutable Freedom.” in Kevin Timpe & Daniel Speak (eds.), Free Will and Theism: Connections, Contingencies, and Concerns (New York: Oxford University Press, 2016), 294–311. 18 As a result, I think part of Timpe’s reply to Baker is misleading. Even if compatibilism is true, that does not necessarily mean that God determines all actions. Kevin Timpe, “Why Christians Might Be Libertarians: A Reply to Lynne Rudder Baker.” Philosophia Christi 6(2) (2004): 282–83, https://doi.org/10.5840/pc20046230. 19 Harry Frankfurt, The Importance of What We Care About (New York: Cambridge University Press, 1988), is perhaps the best known example of compatibilist volitionalist. 20 Timpe, Free Will, 7.

44  Jesse Couenhoven 21 For helpful reflections on the question of what “control’ means, see Aku Olavi Visala, “Consciousness and Moral Responsibility: Skeptical Challenges and Theological Reflections.” Zygon 56(3) (2021): 641–65. 22 Patrick Todd and Peter Furlong have suggested that libertarians could respond by accepting that God is not praiseworthy (in the basic desert sense) for the divine perfections. I find this a costly concession because Christians have traditionally taken for granted that they can and should praise and worship God for those qualities. 23 Timothy Pawl and Kevin Timpe, “Heavenly Freedom: A Reply to Cowan.” Faith and Philosophy, 30(2) (2013): 188–97. 24 Stump, Aquinas (chap. 13); cf. Timpe, Free Will, 13. 25 For development of the point that divine cooperation permits greater heavenly libertarian freedom than unilateral grace, see Robert J. Hartman, “Heavenly Freedom and Two Models of Character Perfection.” Faith and Philosophy 38(1) (2021): 45–64. 26 This view naturally raises questions about the unity of the human and divine in the historical Christ, and whether sin should be ascribed to God. Johannes Grössl’s version of this theory addresses these questions by positing that the divine nature acts like a subconscious mind on the human consciousness of Jesus, pulling him toward perfection but not irresistibly so (see Grössl & Stosch, Impeccability and Temptation (chap. 11)). How exactly this differs from the role of the Spirit in the life of any human being is unclear. His view implies that the unity of natures in Jesus is loose enough that the sin of the human nature should not be ascribed to the divine nature. 27 Grössl & Stosch, 189. 28 Grössl & Stosch, 230. 29 Pawl, Conciliar Christology, 125. 30 See, however, Couenhoven, Stricken by Sin; Jesse Couenhoven, Predestination: A Guide for the Perplexed (New York: T&T Clark, 2018). 31 Timpe, Free Will, 64. 32 For a recent example of this kind of story, see Paul Kingsnorth, “The Cross and the Machine.” First Things, June, 2021, www.firstthings.com/article/2021/06/ the-cross-and-the-machine. 33 Think, for example, of Daniel Keyes, Flowers for Algernon (Orlando, FL: Mariner Books, 2005). 34 White, Fate and Free Will, 245–47. 35 Kevin Timpe, “Why Christians Might Be Libertarians: A Reply to Lynne Rudder Baker.” Philosophia Christi 6(2) (2004): 89–98. 36 Timpe, Free Will, 285. 37 For the sake of simplicity, I ignore the common philosophical distinction between a “defense” and a “theodicy” in the following discussion. 38 I develop this argument at somewhat greater length in the final chapter of Couenhoven, Predestination. I should clarify that although I am reluctant to affirm divine determinism, as a compatibilist I am also open to the possibility. 39 Cf. John Hick, Evil and the God of Love, 2nd edition (London: Macmillan, 1977). 40 It is worth noting that objecting to a highly interventionist picture of divine agency does not require the rejection of divine determinism, but specification of the particular form the latter view might take. My thanks to Simon Kittle for pressing me to clarify this point. 41 Cf. Clive Staples Lewis, The Problem of Pain (New York: Macmillan, 1962). 42 Cf. Alvin Plantinga, “Supralapsarianism, or ‘O Felix Culpa.’ ” in Peter van Inwagon (ed.), Christian Faith and the Problem of Evil (Grand Rapids: William B. Eerdmans, 2004), 1–25; Eleonore Stump, Wandering in Darkness: Narrative and the Problem of Suffering (New York: Oxford University Press, 2010).

Why Christians Should (Still) Be Compatibilists 45 43 Marilyn McCord Adams, Horrendous Evils and the Goodness of God (Ithaca: Cornell University Press, 1999). 44 Vince R. Vitale, Non-Identity Theodicy: A Grace-Based Response to the Problem of Evil (New York: Oxford University Press, 2021). 45 A version of Vitale’s theodicy can undoubtedly be appropriated by libertarians. It is clear, however, that compatibilism is a more natural fit for his view, especially given its implicit universalism. 46 Couenhoven, Predestination. 47 Clive Staples Lewis, The Great Divorce (New York: HarperOne, 2009). 48 Cf. White, Fate and Free Will (chap. 10). 49 I want to offer my thanks to Kevin Timpe for his generosity and insight as a conversation partner on many of the topics covered in this essay, and for many instructive comments from Johannes Grössl, Robert Harmon, Peter Furlong, Simon Kittle, Patrick Todd, Leigh Vicens, and Aku Visala.

Bibliography Adams, Marilyn McCord. Horrendous Evils and the Goodness of God. Ithaca: Cornell University Press, 1999. Arpaly, Nomy. Merit, Meaning, and Human Bondage: An Essay on Free Will. Princeton: Princeton University Press, 2006. Baker, Lynn Rudder. “Why Christians Should Not Be Libertarians: An Augustinian Challenge.” Faith and Philosophy 20(4) (2003): 260–78. Benchimol, Jason. “The Moral Significance of Unintentional Omission: Comparing Will-Centered and Non-Will-Centered Accounts of Moral Responsibility.” in Nicole A. Vincent (ed.), Moral Responsibility: Beyond Free Will and Determinism. New York: Springer, pp. 101–20, 2011. Couenhoven, Jesse. Predestination: A  Guide for the Perplexed. New York: T&T Clark, 2018. ———. Stricken by Sin, Cured by Christ: Agency, Necessity, and Culpability in Augustinian Theology. New York: Oxford University Press, 2013. ———. “The Necessities of Perfect Freedom.” International Journal of Systematic Theology 14(4) (2012): 396–419. ———. “The Problem of God’s Immutable Freedom.” in Kevin Timpe  & Daniel Speak (eds.), Free Will and Theism: Connections, Contingencies, and Concerns. New York: Oxford University Press, pp. 294–311, 2016. Furlong, Peter. The Challenges of Divine Determinism: A  Philosophical Analysis. New York: Cambridge University Press, 2019. Grössl, Johannes & Klaus von Stosch. (eds.). Impeccability and Temptation: Understanding Christ’s Divine and Human Will. New York: Routledge, 2021. Harry Frankfurt. The Importance of What We Care About. New York: Cambridge University Press, 1988. Hartman, Robert J. “Heavenly Freedom and Two Models of Character Perfection.” Faith and Philosophy 38(1) (2021): 45–64. Hick, John. Evil and the God of Love (2nd edition). London: Macmillan, 1977. Keyes, Daniel. Flowers for Algernon. Orlando: Mariner Books, 2005. Kingsnorth, Paul. “The Cross and the Machine.” First Things, June, 2021. www.firstthings.com/article/2021/06/the-cross-and-the-machine. Levy, Neil. “Luck and History-Sensitive Compatibilism.” The Philosophical Quarterly 59(235) (2009): 237–51. Lewis, Clive Staples. The Great Divorce. New York: HarperOne, 2009.

46  Jesse Couenhoven ———. The Problem of Pain. New York: Macmillan, 1962. Lohmann, Friedrich. “God’s Freedom: Free to Be Bound.” Modern Theology 34(3) (2018): 368–85. McCormack, Bruce. “Seek God Where He May Be Found: A Response to Edwin Chr. van Driel.” Scottish Journal of Theology 60(1) (2007): 62–79. McKenna, Michael. Conversation and Responsibility. New York: Oxford University Press, 2012. Morris, Thomas V. Anselmian Explorations. Notre Dame: University of Notre Dame Press, 1987. Morriston, Wes. “Are Omnipotence and Necessary Moral Perfection Compatible? Reply to Mawson.” Religious Studies 39 (2005): 441–49. ———. “What Is so Good about Moral Freedom?” The Philosophical Quarterly 50(200) (2000): 344–58. Pawl, Timothy. In Defense of Extended Conciliar Christology: A Philosophical Essay. New York: Oxford University Press, 2019. Pawl, Timothy & Kevin Timpe. “Heavenly Freedom: A Reply to Cowan.” Faith and Philosophy (kane) 30(2) (2013): 188–97. Plantinga, Alvin. God, Freedom, and Evil. Grand Rapids: Eerdmans, 1989. ———. “Supralapsarianism, or ‘O Felix Culpa.’ ” in Peter van Inwagen (ed.), Christian Faith and the Problem of Evil. Grand Rapids: William B. Eerdmans, pp. 1–25, 2004. Preciado, Michael Patrick. A Reformed View of Freedom: The Compatibility of Guidance Control and Reformed Theology. Eugene: Pickwick Publications, 2019. Rogers, Katherin. Anselm on Freedom. New York: Oxford University Press, 2008. Scanlon, Thomas Michael. “Responsibility and the Power of Choice.” Think 12(33) (2013): 9–16. Smith, Angela Marie. “Responsibility as Answerability.” Inquiry: An Interdisciplinary Journal of Philosophy 58(2) (2015): 99–126. Sonderegger, Katherine. “The Doctrine of Providence.” in Francesca Aran Murphy & Philip G. Ziegler (eds.), The Providence of God: Deus Habet Consilium. New York: T&T Clark, pp. 144–57, 2009. Stump, Eleonore. Aquinas. New York: Routledge, 2005. ———. Wandering in Darkness: Narrative and the Problem of Suffering. New York: Oxford University Press, 2010. Swinburne, Richard. The Existence of God (2nd edition). Oxford: Clarendon Press, 2004. Timpe, Kevin. Free Will in Philosophical Theology. New York: Bloomsbury, 2013. ———. “Why Christians Might Be Libertarians: A Reply to Lynne Rudder Baker.” Philosophia Christi 6(2) (2004): 89–98. Vicens, Leigh C. Divine Determinism: A Critical Consideration. Madison: University of Wisconsin-Madison, 2012. Visala, Aku Olavi. “Consciousness and Moral Responsibility: Skeptical Challenges and Theological Reflections.” Zygon 56(3) (2021): 641–65. Vitale, Vince R. Non-Identity Theodicy: A Grace-Based Response to the Problem of Evil. New York: Oxford University Press, 2021. White, Heath. Fate and Free Will: A  Defense of Theological Determinism. Notre Dame: University of Notre Dame Press, 2019.

3 Weighing Compatibilism and Libertarianism in Analytic Theology Kevin Timpe

In “Why Christians Should (Still) Be Compatibilists,” Jesse Couenhoven makes the case that compatibilism about free will and determinism fits better with key aspects of Christian doctrine than does libertarianism.1 As a result, he thinks that Christians have reasons to be compatibilists. Perhaps surprisingly, I agree. But while compatibilism does put “less pressure on [certain] theological convictions” (27) than libertarianism, that doesn’t mean that Christians should be compatibilists. In this essay, I respond to Couenhoven, both in terms of evaluating the particular theological doctrines that may favor one view of free will over the other, and in terms of overarching methodological issues about how we ought to evaluate philosophical positions with regard to our theological commitments. While I agree that there are theological reasons for Christians to be compatibilists, I don’t think that those reasons are decisive in favor of endorsing compatibilism. I proceed in three sections. In Section 1, I look at compatibilism and libertarianism as two (non-exhaustive) families of views regarding free will. Section 2 looks at how those families of view do with respect to the particular theological convictions that Couenhoven addresses in his chapter. While I  agree that there are a number that favor compatibilism, there are others where I  think that libertarianism has an advantage. How we rank the relevant overall merits of the two positions depends not only on how we look at individual issues but also on how to weigh and seek to integrate those issues into a cohesive whole. Couenhoven and I disagree primarily on the wholistic level. So, in Section 3, I try to give a partial account of how we should weigh the overarching philosophical and theological merits of two competing packages of views. Here, I’m less directly interested in what we should believe (though, it’ll come as no surprise, I have a view on that) than on how we weigh theological evidence as we go about figuring that out. Compatibilism and Incompatibilism Couenhoven frames his chapter as a reappraisal, 20 years later, of some of the issues raised by Lynne Rudder Baker’s 2003 article, “Why Christians Should Not Be Libertarians.”2 Baker’s paper is one to which, as Couenhoven DOI: 10.4324/9781003306191-4

48  Kevin Timpe notes, I responded with a not very originally titled “Why Christians Might Be Libertarians.”3 Couenhoven aims to give a more recent reassessment as another round in the contemporary conversation about how we ought to best understand the nature of free will in theology. Like Couenhoven, I’m going to limit myself to a comparison between compatibilism and libertarianism, realizing that there are other views that one would need to consider in order to proclaim that either compatibilism or libertarianism is the best view for the Christian to endorse. One would need to consider, for instance, skeptical views which deny the existence of free will.4 Without intending this as an argument against such views, let me just note that it’s not clear to me how to reconcile free will skepticism with Christianity given my own understanding of free will as the control condition on moral responsibility and what I think is a pretty clear commitment of Christianity to the existence of moral responsibility.5 Couenhoven himself takes on the first of these as a “widely held claim” (42, note 5). If we take the existence of free will as a given for the present context, we’re thus left comparing compatibilist views, particularly compatibilist views that affirm the existence of free will, with libertarian views.6 While I think that what Robert Kane calls the Compatibility Question sometimes receives more than its fair share of the philosophical discussions surrounding free will, in the present section I’m going to contribute to that tendency. Let me also note that both compatibilism and libertarianism so understood are best thought of as a cluster or family of views whose individual members differ on a number of other important issues (e.g., how often we’re free, event versus agent causation, sourcehood versus leeway approaches, time-slice versus historicist approaches). We should thus compare what we think are the best instances of each of those views, while humbly and fallibly admitting that we might not have the right view of the relative strength within a family, much less across families of views. It’s only by comparing what we think are the best compatibilist and libertarian views that we’ll have the resources to weigh which family of view has the advantage over the other.7 Couenhoven, unlike others, rightly doesn’t think that libertarian views of freedom require the ability to choose evil.8 Contrast this with, for instance, theologian John E. McKinley: “In the case of Jesus, the theory of libertarian freedom views his choices in the face of temptations as only free if he was able to choose otherwise, that is, to sin.”9 Since Couenhoven thinks source libertarian views are more promising than leeway libertarian views,10 and I myself have argued for the same, in what follows I limit my discussion to source-based libertarian views unless I  specifically specify otherwise. Similarly for compatibilism, where I’ll have in mind source-based compatibilist views that affirm the existence of free will.11 Referring back to an earlier article of his,12 Couenhoven says, When I  followed Baker in claiming that compatibilists can do better than libertarians in making sense of Christian doctrines, I, like her, had

Weighing Compatibilism and Libertarianism in Analytic Theology 49 such approaches in mind. If significant choice between alternative possibilities is the sine qua non of freedom, and thus required for moral agency, God seems to be in a bad position. (28, endnote omitted) Theologically driven compatibilism, according to Couenhoven, has been under-developed, especially by those who are not Calvinists, and many theologians have seemed to think that libertarianism is the only game in town. I’m not in a position to know the degree to which his characterization is true of theologians in general. In a recent handbook of analytic theology, for instance, David Fergusson defines “the classical model of providence” in such a way that it includes theological determinism, thereby requiring compatibilism if proponents of the classical model of providence also affirm the existence of human free will.13 And I  don’t think that Fergusson is alone among theologians in endorsing compatibilism. But my disagreement with Couenhoven isn’t primarily sociological. I agree with Couenhoven that the past two decades have seen an increase in the sophistication and broader influence of Christian compatibilist views. As someone who teaches at Calvin University, I  can’t help but be struck by Al Plantinga’s evaluation of compatibilism, penned while he himself worked at Calvin during the 1970s, from God, Freedom, and Evil. To give some context, Plantinga is considering compatibilism as an objection to his free will defense to the logical problem of evil. If compatibilism is true, then the good of free will, and likely the other goods that human freedom make possible, could be secured without the existence of moral evil. Plantinga responds: This objection to the Free Will Defense seems utterly implausible. One might as well claim that being in jail doesn’t really limit one’s freedom on the grounds that if one were not in jail, he’d be free to come and go as he pleased. So I shall say no more about this objection here.14 Two things to note briefly about this passage, despite having to admit that it makes me chuckle nearly every time I read it. First, Plantinga’s response captures only a particular kind of compatibilism, one that I’ve elsewhere referred to as “classical compatibilism.”15 There are, of course, other kinds of compatibilism that don’t require conditional analyses of the ability to do otherwise. Second, while I agree with Plantinga that compatibilism is false (and, I take it, necessarily false if false at all), one needs more of an argument against compatibilism than Plantinga gives; this is for two reasons. First, it’s too quick to dismiss compatibilism as “implausible” or “paradoxical” particularly if one seeks to take seriously those aspects of the Christian tradition that historically are compatibilist. Second, as Couenhoven notes, there are more recent and more compelling compatibilist accounts that don’t fall prey to the criticisms of the conditional analysis.

50  Kevin Timpe Elsewhere, I’ve argued that, contrary to what a number of libertarians claim, compatibilism should be our initial default since it makes fewer claims on the world than does libertarianism.16 Compatibilism, at its core, is a claim about logical possibility: it is logically possible to be both free and determined. Incompatibilism, on the other hand, is a claim about impossibility (it is impossible to be both free and determined) or, if one prefers, necessity (necessarily, every world in which determinism is true is a world in which there is no human free will and every world in which there is human free will is a world in which determinism is false). Here’s how William Lycan puts the point, which strikes me as correct: Compatibilism, not just about free will but generally, on any topic, is the default. For any modal claim to the effect that some statement is a necessary truth, I  would say that the burden of proof is on the claim’s proponent. A theorist who maintains of something that is not obviously impossible that nonetheless that thing is impossible owes us an argument. And since entailment claims are claims of necessity and impossibility, the same applies to them. Anyone who insists that a sentence S1 entails another sentence S2 must defend that thesis if it is controversial. If I tell you that “Pigs have wings” entails “It snows every night in Chapel Hill,” you need not scramble to show how there might be a world in which the first was true but the second false; rather, you would rightly demand that I display the alleged modal connection. And of course the same goes for claims of impossibility. The point is underscored, I think, if we understand necessity as truth in all possible worlds. The proponent of a necessity, impossibility, entailment or incompatibility claim is saying that in no possible world whatever does it occur that so-and-so. That is a universal quantification. Given the richness and incredible variety of the pluriverse, such a statement cannot be accepted without argument save for the case of basic logical intuitions that virtually everyone shares.17 Libertarianism requires more of the actual world than does incompatibilism since it is the conjunction of incompatibilism (which, as we’ve seen, is a claim about impossibility or necessity) and a contingent claim (namely, that there is free will). Compatibilism, as Couenhoven puts it, “cuts off fewer possibilities” (28) than does libertarianism. Or, as he puts it at the end of his paper: “Unlike incompatibilism, which states what freedom must be, compatibilism takes a minimal stance, indicating what freedom can be” (42). That said, I  think there are good philosophical arguments for incompatibilism over compatibilism, despite where the dialectical default lies. Fortunately for the reader, I  won’t rehash those arguments here.18 But if we’re to compare the reasons analytic theologians and philosophers of religion have for endorsing either compatibilism or libertarianism, we need to evaluate not only their theological benefits but also the philosophical

Weighing Compatibilism and Libertarianism in Analytic Theology 51 merits that each view has. It may well be that Couenhoven thinks that there are philosophical and not just theological reasons that favor compatibilism over libertarianism. But, if that’s the case, then an engagement with the philosophical arguments would need to be part of the comparative case made. If one doesn’t, then one view’s squaring with theological claims better than another won’t entail the overall superiority of that first view over the second. Suppose, for instance, that one thought there were good philosophical arguments for the existence of abstract objects, necessarily existing entities, such as properties or attributes, that necessarily exist since their existence depends on nothing.19 Would that mean that the philosopher would need to believe in the existence of abstract objects, despite thinking there are good reasons for believing in their existence, because Nicene Creed claims that God is “the maker of heaven and earth, of all things visible and invisible?” Not necessarily. Such a person could perhaps think in the following manner: The Nicene Creed is attempting to make a claim about the relation of God qua creator to creation. But abstract objects are the kinds of things that can’t enter into causal relations, and thus can’t enter into the causal relation that is creation. So we should understand the “all” in “all things” in the creed as a tacitly restricted quantifier, ranging over all created things. But we should already understand the “all” to be restricted since God is already excluded from the range of the generalization, despite being an invisibile.20 Here, it’s not that the creed isn’t normatively binding. But neither is it the case that our philosophical views are irrelevant and the creedal claim alone settles the metaphysical issues. The point is that the theological cost of having to interpret the quantifier as implicitly restricted might be outweighed by the philosophical reasons supporting the existence of abstract objects. We thus need to figure out how to weigh the overall philosophical and theological reasons for views, particularly if the relevant theological constraints don’t themselves settle the issue. I  return to this process of weighing in section “Holistic Weighing.” Specific Doctrines Before doing that, however, in this section I  want to briefly look at some of the specific doctrines that Couenhoven examines in order to weigh how compatibilism and libertarianism fair with respect to Christian theology as a whole, since these doctrines will figure into the overall weighing. For some of these doctrines, I agree with him that they, taken by themselves, might give some reason for preferring compatibilism. For others, unsurprisingly, I disagree with him regarding how we should evaluate them.

52  Kevin Timpe Divine Freedom

The first of these issues that Couenhoven addresses is God’s freedom. Regarding divine freedom, he thinks that “compatibilism offers some theological advantages [over libertarianism] as we puzzle over why God’s freedom is the best, freedom par excellence” (29). I agree with the last part of this quotation. It is divine freedom that grounds and gives meaning to human freedom. It’s not clear to me, however, that he’s right that compatibilism helps us have a better understanding of divine freedom. Elsewhere, I’ve tried to show why I think a source libertarian view of divine freedom can satisfy the relevant desiderata. Couenhoven apparently isn’t convinced. So here let me advance a different argument for a libertarian understanding of divine freedom. Remember that libertarianism is the conjunction of two claims, one about the existence of free will and the other about the claim that incompatibilism is true.21 Like Couenhoven, I  take the Christian tradition to include the claim that God is free. In fact, as Couenhoven also agrees, God’s freedom is the ultimate kind of freedom. So we have the first of the two claims that constitutes libertarianism. What then of the claim that it is possible for God to be both free and determined? This strikes me as an impossibility, for given the divine nature, it is not possible for anything external to God to act on the divine nature in such a way as to determine it. That is, there is no possible world in which God is determined, and thus no possible world in which God is determined and free. Couenhoven admits as much: “determinism is not even a possibility in God’s case” (30).22 But such an impossibility supports incompatibilism not compatibilism. And so this may be a way, despite what Couenhoven says that human freedom is more like divine freedom than on the compatibilist picture, since on my libertarian view both human freedom and divine freedom are “dependent on undetermined choices” (33). Other comments suggest that the reason Couenhoven thinks compatibilism fairs better than incompatibilism regarding divine freedom involves how we think about control: As an Augustinian compatibilist, my inclination is to deal with this problem by not requiring that God have the kind of control that comes from being the unconditioned source of something. Rather than claiming that God is the source of what God is, which threatens to create a problematic recursivity, it seems better to focus on the fact that God is owner of the divine attributes, in the sense that they are rightly ascribed to God. (7 in manuscript) There are important ontological differences between human freedom and its divine exemplar, as I’ve admitted before. But those differences don’t rule out a libertarian understanding of divine freedom. In fact, as I suggested a few

Weighing Compatibilism and Libertarianism in Analytic Theology 53 paragraphs back, I find it puzzling to think that God’s exercise of freedom could be determined by anything outside of the divine nature. Describing divine freedom, Couenhoven also writes that God’s freedom is not backward looking, requiring us to inquire into the history of the divine being. It is excellent in its unparalleled power for life. Unlike other agents, God is fully able to do and be what God pleases to do and be. . . . On a normative conception of freedom, the highest freedom is a kind of power, not of potential but actuality, the ability to do and be good. That power is the greatest when nothing can possibly stop it. (31) But there’s nothing in this description of divine freedom that’s ruled out by my view, and I’ve argued extensively (though perhaps not persuasively) that considerations of divine freedom and God’s essential moral perfection don’t give us reason to prefer compatibilism over libertarian views.23 So I don’t see how these considerations, which I agree with, favor compatibilism over libertarianism. Heavenly Freedom

Turning then to perfected human freedom, here especially if we are to take seriously the idea that part of the Christian tradition that holds that heaven is a place where it is no longer possible to sin, we need to find a way to reconcile the absence of relevant possibilities for sin. But I think that can be done, as I’ve sought to show in a series of papers with Tim Pawl.24 Apparently Couenhoven agrees that we’ve succeeded, given that he says we have “an adequate explanation of saintly impeccability” (32). Couenhoven is right that our view requires that the redeemed’s moral character is formed in a particular way, and thus puts a demand on agent’s history that a compatibilist account of heavenly freedom doesn’t necessarily have.25 A compatibilist approach, he thinks, still has minor advantages over libertarian approaches (33). If he’s referring to not needing a historical condition that requires the world be indeterministic, then I agree. And for some folks, our explanation may require a purgatorial period for the justified-butnot-yet-sanctified’s moral character to be unified around the good. I  don’t think that we rule out “the zap” view, whereby God unilaterally brings about moral progress once the agent has consented to it, though I do think a more participative model makes more sense, even if it does require purgatory. If redeemed agents consent to the zapping, then it too could “explains how unproblematic it would be for those who have died to quickly receive their hearts’ true desire, being transformed by union with Christ without any loss to their freedom or praiseworthiness” (33). Again, the view that I’ve developed with Pawl isn’t intended to rule out such a possibility.

54  Kevin Timpe But even if there are two ways to be perfected, both are graced – either the grace of zapping or cooperating with grace in the moral transformation of this life (and in purgatory). When we are perfected, we do become “a new creation.” No matter how we become what we are called to be, it’s through grace and divine assistance, so I don’t understand why Couenhoven thinks that strangely enough . . . the saints who receive grace [and are “zapped” into perfection] end up being second class, on a libertarian view. From the perspective of merit, it would be better to tough it out in purgatory than to accept divine assistance. (37) This claim only makes sense on the assumption that the process of perfection in purgatory doesn’t involve divine assistance, which is a possibility that my libertarian view denies. In fact, I wrote an entire paper26 precisely to show the key role that participation plays in the cooperative project that is creaturely perfection – what Dockter refers to as “the idea of synergeia as the cooperation of human and divine”27 in salvation. Though it’s a more minor point, before leaving this section, let me briefly address one other issue. Couenhoven suggests that my and Pawl’s view “seems to ignore traditional claims about the unity of the virtues” (32). I don’t know exactly what he means by “the unity of the virtues” since that claim is interpreted in a number of different ways. Here are three different ways that the view has been interpreted: 1) The identity of the virtues: All of the apparently different virtues are really just one single overarching virtue. 2) The reciprocity of the virtues: While they are multiple virtues, they come as a necessary package such that if you have one virtue you have every virtue. 3) The interconnection of the virtues: If you have any virtue, you will have some sensitivity for considerations relevant to the other virtues (one way of understanding this sensitivity is in terms of correlations: having one virtue makes it more likely that you’ll have other similar virtues).28 It would certainly be harder for the view I’ve developed with Pawl to endorse the first of these, the claim about the identity of the virtues, given that it’s the strongest claim and, in general, the stronger the metaphysical claim, the less it will be compatible with. I  confess, though, that it’s not clear to me that our view about the process of perfecting our freedom in alignment with the good couldn’t be reconciled with the identity of the virtues. But suppose that there’s an argument that it couldn’t be. In one sense, that’s fine, since I already think that the identity of the virtues thesis is false for independent reasons. And the same is true regarding the reciprocity of the virtues thesis. Our view of the process of character formation aligns with the weakest of

Weighing Compatibilism and Libertarianism in Analytic Theology 55 the three claims about the unity of the virtues, that is interconnection of the virtues, and so I don’t see any cost here that I’m not already willing to pay. (In fact, not being able to consistently endorse that which I already think is false doesn’t strike me as much of a cost at all.) Incarnate Freedom

I turn then to free will of the incarnate second person of the Trinity: The challenge for libertarians is fairly easy to articulate: traditionally, Christ has been thought to have two natures, human and divine, each with its own will. Given the fact that God is necessarily without sin, the two natures can remain in union only if the human will makes no sinful choices. (33f) Tim Pawl and I have attempted to show that a libertarian understanding of incarnate freedom is defensible.29 Couenhoven thinks that “as we survey attempts to respond to these challenges, it is clear that nothing approaching a consensus has been reached” (34). I take it that this means, among other things, that Pawls’ and my view hasn’t yet received consensus status. I guess that’s perhaps somewhat disappointing, given that I think our view is true and that people should endorse the truth. It, however, certainly isn’t surprising that our view hasn’t yet received consensus. Given the contentious nature of philosophical and theological disagreement, I wouldn’t expect it to. But I agree with Couenhoven that there’s a sense in which compatibilism here has an easier time explaining how the human nature hypostatically united to the divine nature in the incarnation could be ensured to be sinless while still being free. The compatibilist could reason in the following manner, for instance: 1) Necessarily, a human nature’s freedom is compatible with that human nature being determined. 2) Necessarily, God would only hypostatically unite a particular human nature to the second person of the Trinity if God would ensure (via determinism if needed) that that human nature did not ever sin. Therefore: 3) Necessarily, if God hypostatically unites a human nature with the second person of the Trinity, then God would ensure (via determinism if needed) that that human nature did not ever sin.30 I agree with Couenhoven that the challenge of making sense of Christ’s impeccable freedom highlights a broader theological challenge of making sense of the Christian story, in which

56  Kevin Timpe participation in the divine is the telos of human agency. It is partly for that reason that Christians have traditionally taken Christ to be an exemplary figure. The union of divine and human natures in his person is, of course, usually considered sui generis, but there has nonetheless been a sense that as the firstborn of all creation his life with God offers a pattern to which the rest of us aspire. (35) And so thinking about incarnate freedom is closely connected with how we think about the perfected freedom of the redeemed. I don’t have much else to say here, instead pointing to what I think is the best work on this topic, notably Tim Pawl’s In Defense of Extended Conciliar Christology, especially Chapters 5 and 6. Even here, I must confess, a certain degree of reluctance given my co-author’s subsequent evaluation of our work together on the Incarnation: I thank Kevin Timpe for allowing me to use our co-authored article, ‘Freedom and the Incarnation.’ Since it is co-authored, it would be unfair of me to claim full credit for the content of that chapter. In light of the demands of justice, I will only claim credit for the insights, leaving the errors to my dear friend.31 If libertarianism can’t be reconciled with Conciliar Christology, then the former rather than the latter should be rejected. It’s just not obvious, at least not to me at any rate, that we’re forced to pick one at the expense of the other. Graced Freedom

In speaking of graced but not yet perfected freedom, Couenhoven writes that compatibilism allows for a wide range of traditional soteriological views. This flexibility, I have been suggesting, is a significant attraction of the approach. Libertarianism, by contrast, exerts a kind of gravitational pull on Christian doctrine, bending it to its commitments. (36) I think that both compatibilism and libertarianism exert a “pull” on other parts of the package of one’s theological and philosophical beliefs, though perhaps not equally so. As I’ve already indicated, compatibilism carries less metaphysical commitment regarding the way the world is than does libertarianism. And so, if we think of a metaphysical commitment as something that the rest of our philosophical and theological views need to be consistent with, I can see how perhaps we could think of libertarianism as exerting a greater of “pull” on other parts of our views than does compatibilism.

Weighing Compatibilism and Libertarianism in Analytic Theology 57 But I don’t see this as involving any more of a commitment than what one already has in virtue of endorsing libertarianism. My first paper on the subject was an attempt to show the compatibility of two claims: Claim 1: Divine grace is the efficient cause of saving faith. Claim 2: Humans control whether or not they come to saving faith.32 The motivation for Claim 1 was to avoid Pelagianism, which is a constraint I  think Christian philosophers and theologians have good reason to try to satisfy. The motivation for Claim 2 was to be able to preserve the compatibility of Claim 1 with a  libertarian account of free will, thereby avoiding commitment to theological determinism. Reflecting back on this earlier paper in a later one, I wrote as follows: In that paper, I focused on just those acts of will involved in coming to saving faith. I confess that I’ve always been somewhat uncomfortable with that earlier view given the metaphysical cost it required – namely the truth of certain controversial (but, I  still think, plausible) claims about causation (e.g., that omissions can’t be causes but instead are merely quasi-causes). If there’s another account that can preserve both Claim 1 and Claim 2, I’d welcome it.33 While I would still like there to be a lower metaphysical cost, I don’t know how to secure that. Given the rest of my philosophical and theological commitments, it’s a cost that, at present, I’m willing to pay (or, to use Couenhoven’s imagery, a gravitational pull I’m willing to be caught in). Divine Goodness and Evil

So far, I’ve focused on those specific theological issues where Couenhoven thinks the libertarian is at a disadvantage. With respect to those issues, Couenhoven thinks that compatibilism fits key aspects of Christian doctrine better than libertarian views: “the theological ‘cost’ of being a compatibilist is lower than that of being a libertarian. Compatibilism’s theoretical advantage is that it puts less pressure on theological convictions than libertarianism” (27). However, there are also other theological issues where, it is sometimes claimed, the compatibilist is at disadvantage. This is especially true of the reconciling of the existence of evil with God’s goodness. And one of these evils that are especially troubling is hell, as traditionally conceived.34 We might think of hell as a limit case for divine goodness. Or, as David Lewis once put it, the suffering involved in the traditional Christian understanding of hell “dwarfs the kind of suffering and sin to which the standard versions [of the problem of evil] allude.”35 This isn’t the place to try and canvas the voluminous work that has been done on responding to the problem(s) of evil in general or the problem of

58  Kevin Timpe hell in particular. The present issue, rather, is whether compatibilism or libertarianism has more resources in appealing to the problems. On this point, I appreciate Couenhoven’s honesty: “Compatibilism has its own problems. In particular, compatibilists have trouble making sense of evil” (38).36 While I don’t think that the truth of libertarianism would, by itself, assuage all concerns about evil, I do think that it makes the problem at least somewhat less intractable. Here’s how I put it in my earlier article: I do not pretend that either libertarians or compatibilists have fully answered the problem of evil in all its various forms. Nevertheless, I think that libertarianism fares better in the face of the problem of evil than does compatibilism, since it has more resources to muster in its defense. If humans have libertarian free will, then God cannot create a world containing such agents and unilaterally guarantee that that world contains no evil. Libertarians can therefore maintain a distinction between possible worlds and feasible worlds. Compatibilism, on the other hand, cannot so easily make the claim that God cannot create a world containing such agents and unilaterally guarantee that that world contains no evil. If God is the ultimate cause of all human actions (either via determining the state of the physical universe at a time in the distant past or not), then He bears some direct responsibility for every action that occurs.37 In subsequent years, I’ve come to realize that there are ways for the compatibilist to appeal to the free will defense too,38 and not just the libertarian, even though I don’t myself find those compatibilist free will defenses compelling. But the central point still strikes me as correct. And Laura Ekstrom agrees in her recent book: If we could act freely even when causally determined at every moment or divinely determined at every moment to act as we do, then there could be created beings with the power for free action yet no evil in the world.39 Notice that Ekstrom’s point merely requires the possibility of determined freedom, not it’s actuality. As Couenhoven points out in his chapter, compatibilism doesn’t commit one to the truth of determinism – just that the truth of determinism wouldn’t rule out freedom. But the compatibilist is forced with a dilemma. The compatibilist could endorse the truth of theological determinism.40 But doing so has immense repercussions for how we think about both God’s goodness and the existence of certain evils, such as the evil of hell.41 The view that God determines some humans to an eternity in hell is, in my view, a non-starter. Like Couenhoven and Baker (and Pereboom), if I thought that theological determinism were true, I’d endorse universalism. Speaking of

Weighing Compatibilism and Libertarianism in Analytic Theology 59 what he takes to be the most plausible determinist views regarding human suffering and God’s goodness, Couenhoven says that this defense of divine goodness requires commitment to universal salvation. In turn, universalism fits best with compatibilist conceptions of human agency. The only sure way to ensure that all say “I do” at the divine wedding feast is for God to shape human agency, by bringing frail human agents into participation with God’s own life. This universalism is undoubtedly controversial, historically speaking, but it is hard to deny its theodical appeal. (40) I come back to how we should weigh universalism in the final section. But if I were to become convinced of the truth of theological determinism, I’d likely endorse it precisely for that appeal. What then of the compatibilist who doesn’t also endorse theological determinism? Such a compatibilist can certainly appeal to an indeterministic understanding of human freedom. But note there that they face two issues. First, the compatibilist would need there to be an acceptable account of indeterministic human freedom. In this respect, the sort of compatibilist we have in mind here ought to hope that libertarians can give both an adequate free will defense and an adequate account of human freedom on the assumption of indeterminism; for if it ends up being the case that compatibilism is true but determinism is false, then they could make use of these views. That is, in such a case, the compatibilist could then deploy the indeterministic free will defense and indeterministic account of freedom developed by the libertarian.42 (They just wouldn’t, unlike the libertarian, think that these two views required the falsity of determinism.) In this way, compatibilists who don’t want to be committed to the truth of determinism should hope that the various arguments raised that indeterminism undermines free will (such as the Mind Argument) are unsuccessful. But if this is the direction that the compatibilist goes, and this is the second thing to note about the second disjunct of the dilemma, then the positive account of freedom the compatibilist would be using wouldn’t be better than the libertarian view.43 Where they would differ is that, as compatibilists, they could think that God could have determined human freedom, perhaps to get the very same goods that libertarian views seek to secure. However if God didn’t, then presumably it’s because there’s a good reason for God to have chosen the indeterministic rather than the deterministic route. But whatever those goods are will make it harder for them to argue that libertarian free will views are problematic. So any purported advantage for the compatibilist on the basis of indeterminism undermining free will evaporates. Consider, by way of example, the possibility Couenhoven considers that indeterministic freedom is “not a good in itself, but [as] a means to the end

60  Kevin Timpe of development in virtue” (39). If the compatibilist thinks that virtue couldn’t be developed without indeterministic freedom, then their view is just as much “held hostage” to the relevant empirical findings as the libertarian’s.44 But if the compatibilist thinks that the virtues in question could be developed through either deterministic or indeterministic freedom, and then if they deny that we are determined to develop virtue, there has to be a good reason for why God hasn’t chosen to realize this particular good through this particular means. God could determine us to develop virtue but, it certainly appears, hasn’t actually done so. Presumably God would have a justifying reason for not doing so. But what might it be? It’s not clear to me what such a reason would be, especially since (on the assumption of compatibilism) it couldn’t be a reason that depended upon the existence of free will. Unless we can come up with a plausible reason, this gives us some reason to reject the line of thought on which the virtues could be achieved through the means of some deterministic process. But then the compatibilist who holds that the virtues are developed through the means of indeterministic free choices isn’t in a better position than the libertarian. The purported comparative advantage disappears.45 Holistic Weighing Couenhoven admits that it matters what parameters one is holding fixed in the comparison; as he puts it, “my inquiry concerns the relative costs and benefits of being libertarian or compatibilist for Christians who are attracted to more or less orthodox views” (28).46 Much depends on how we interpret the “more or less orthodox,” and understandably but unfortunately Couenhoven gives us little guidance by way of how we delineate the relevant orthodoxy. As R. Lucas Stamps notes in an important paper on the authoritative role of tradition in analytic theology, [O]ne of the common complaints leveled against the emerging discipline of analytic theology (AT) is the allegedly tenuous relationship it enjoys with the history of scriptural interpretation and the history of Christian doctrine. Both critics and proponents of AT have highlighted the ahistorical character of some of the projects being carried out under the analytic banner.47 Indeed, it’s a charge that I myself have raised elsewhere. William Wood’s way of putting the concern is a particularly pithy way, and his evaluation of it strikes me as particularly true: Many academic theologians regard theology primarily as a historical discipline, and they regard analytic theology as perniciously ahistorical. . . . The objection is that analytic theology does not take history or

Weighing Compatibilism and Libertarianism in Analytic Theology 61 historical contingency seriously enough. Sometimes, this objection takes an even more direct form: analytic theologians are simply ignorant of the history of doctrine, and of historical sources more generally. .  .  . I agree that some analytic theologians and philosophers of religion do not take historical sources seriously enough. . . . But in such cases, the fault lies with the individual thinkers who make these mistakes, not with the analytic method itself.48 I agree with Wood. Every theologian, no matter their tradition, has to figure out how they’re going to relate to the history of Christian theology, and “no one can deny that the development of Christian Orthodoxy seems messy and historically contingent.”49 But for all that messiness and contingency, I think it’s worth taking seriously. Stamps questions whether it makes sense to attempt to speak of the Christian tradition given the multiplicity one finds in Christian theology: Some have questions whether or not we can meaningfully speak of the tradition of Christian doctrine, given the multiplicity of perspectives on offer in the history of Christian reflection. On certain doctrines, this warning is apt. What, we might ask, is the traditional understanding of Christ’s atonement? In this case, it is perhaps better to speak of traditions, plural, that have sought to understand Christ’s reconciling work of atonement, rather than the tradition, singular, that provides a definitive account. On other doctrines, however, we can detect a consensus of judgements that evince a profound consistency over time.50 I agree with Stamps that there is a range of doctrines where there isn’t a singular agreed-upon view; on such topics, while admitting the multiplicity of answers, I think it’s acceptable to say that the tradition is underdetermined. But, as Stamps notes, other doctrines aren’t so open. Demarcating those in the first group from those in the second will itself be contentious. Stamps himself suggests that, alongside the Christian Scriptures, there are three hierarchical levels: At the tops of this hierarchy, underneath the ultimate authority of Scriptures, are, first, the ecumenical creeds and councils. The three creeds recognized as ecumenical – the Apostles’, the Nicene, and the so-called Athanasian Creeds – summarize the consensus tradition of the Christian church on the chief doctrines of the Trinity and the Incarnation of the Son of God for the salvation of the world. The seven ecumenical councils give fuller shape to this consensual tradition and represent the essential of the faith of the undivided church bequeathed from the patristic era. Embedded within these Trinitarian and Christological guidelines are, second, certain classical doctrines that are implied by them. Revisionist analytic theologians may balk at this suggestion, but I believe that the doctrines associated with classical Christian theism are either necessary

62  Kevin Timpe consequences or else presuppositions of the creedal/conciliar pronouncements. .  .  . Third, on a more local level, church and denominational confessions of faith further delimit the parameters within which analytic theologians ought to exercise their craft.51 Tim Pawl refers to the conjunction of the Christological teachings of the first seven ecumenical councils as “Conciliar Christology.”52 We might, in a similar way, think of conciliar Christian theology as the conjunction of the authoritative teachings of the first seven ecumenical councils and the classical doctrines that are entailed by them. Such teachings will be roughly what Strands refers to as the second level of hierarchy.53 I’m not interested, here, in denominational theological constraints, though I think in other contexts they may have their place. So it’s only the first two levels of traditional hierarchy that I shall take as normative in what follows. But, as I shall argue, there’s another level between Stamps’ category of “classical doctrines” implied by the creeds and councils and denominational confessions. This additional level may not be as normatively binding as the classical doctrines, but it should carry some degree of admittedly defeasible prescriptiveness for all Christians, and not just those of a particular denomination. But before that argument, not every Christian philosopher or analytic theologian thinks that the councils are binding as understood. In a paper dedicated to the acceptability of monothelitism (the doctrine that the Incarnate Christ had only one will), Jordan Wessling differentiates two different principles that could govern how a philosopher or theologian, analytic or otherwise, interacts with Conciliar councils such as the Third Council of Constantinople (which rejected monothelitism as heresy, promoting instead dyothelitism, the doctrine that the Incarnate Christ had two distinct wills, one human and one divine): Conciliar Undercutting Principle (CUP):

The (evangelical) Christian is free to reject a conciliar pronouncement if this pronouncement is not taught or implied by Scripture. Conciliar Abrogation Thesis (CAT):

The (evangelical) Christian should reject a conciliar statement if and only if it either: (1) contradicts that which is taught in or implied by Scripture, or (2) is incoherent.54 Wessling states that proponents of both CUP and CAT can have “a deep commitment to the primary of Scripture,” though as compared to those who endorse CUP the proponent of CAT “differs . . . in that the epistemic weight he places on the councils as an accurate representation of Scripture.”55 The proponent of CAT, though not the proponent of CUP, can hold that councils are authoritative even when they go beyond the teachings of Scripture.

Weighing Compatibilism and Libertarianism in Analytic Theology 63 Wessling also differentiates two different reasons why one might think that the councils are theologically binding: P1. God ensures that the majority of the church does not formally err on those doctrines that are central to the Christian faith. P2. God has ensured that the teachings of [a particular ecumenical council] are true and binding on the (evangelical) church.56 Wessling says that P1 is “very plausible” and appears willing to endorse its truth in addition to its plausibility but rejects P2.57 He nevertheless thinks that the teachings of some of the ecumenical councils can be binding on the Christian, either by grounding that ecumenical teaching directly in Scripture or by arguing that (a) “God would not allow a state of affairs as bad as the erring of an ecumenical council on matters that are central to the faith”58 and (b) arguing that the teaching in question is “central to the faith.” But in virtue of which might one argue that a particular teaching is in fact central to the faith? Presumably, as Wessling’s treatment already indicates, by directly rooting it in Scripture. But that, as I understand church history, wouldn’t suffice for establishing homoousianism over Arianism, since both sides of that debate in the fourth century appealed to Scripture.59 Something similar holds for the condemnation of Pelagianism at the Council of Orange. CUP doesn’t seem sufficiently strong for establishing the contours of acceptable Christian theology since CUP cannot rule out Arianism or Pelagianism.60 So much the worse for CUP. Returning then to Stamps’ three hierarchical levels of tradition’s authority, do either of the two levels help us settle the libertarian/compatibilism debate? I don’t think they do. So far as I can tell, there’s nothing in the ecumenical creeds and councils that commit one to a particular understanding of the modal relationship between freedom and determinism. Nor does the totality of conciliar Christian theology explicitly endorse or entail either libertarianism or compatibilism. I thus agree with Baker, and with Couenhoven, that “there is a lot of room for the denial of libertarian accounts in the Christian tradition.”61 Similarly, neither do I think that the totality of conciliar Christian theology explicitly endorses or entails compatibilism.62 But there’s still a role for tradition to play that’s not merely Stamp’s third level of hierarchy, church and denominational confessions of faith, even if it’s also not as strong as the first two levels of his hierarchy. We might think it’s important to take seriously what the history of Christian theological reflection says about a topic even if that history doesn’t rise to the level of what’s contained in or entailed by the creeds and councils. We might endorse something along the lines of the following principle: Traditional Default Principle (TDP):

If the vast majority of Church tradition has taught some theological position X, that history gives us good reason to endorse X even if conciliar Christian theology endorse or entail X. But this reason is defeasible if

64  Kevin Timpe there are sufficiently good reasons (beyond the classical doctrines that are implied by conciliar Christian theology) for the falsity of X.63 I wish that I knew how to make TDP more precise, but I confess that I don’t.64 How strong do reasons have to be “sufficiently good reasons?” How do we demarcate what is allowed to count as part of “Church tradition?” Does a doctrine’s being widely endorsed more recently give us a higher epistemic bar that must be cleared to reject that doctrine as compared to more temporally distant endorsement? These questions, none of which I know how to systematically answer, all indicate the need for further analytic theological work regarding tradition. But to illustrate how this principle might be used, even in broad strokes, I want to return to the issue of universalism. There are a number of different forms of universalism.65 We can distinguish, for instance, at least the following three forms: Hopeful universalism:

the view that it is appropriate for us to hope that all will be redeemed eventually and that no one will be assigned to hell for eternity. Contingent universalism:

the view that as a matter of contingent fact all will be redeemed eventually and that no one will be assigned to hell for all eternity. Necessary universalism:

the view that it is necessarily true that all will be redeemed eventually and that no one will be assigned to hell for all eternity.66 The first form of universalism is a claim about our evaluative stance toward the salvation of all, rather than about their actual salvation. The second and third are claims about eschatological reality. My particular interest is with contingent universalism, the weaker of these two theses – though I think that the following general argument would also hold against the stronger claim as well. So in what follows, “universalism” should be understood to refer to contingent universalism so defined. Is universalism true? According to the seventeenth-century Jesuit theologian Denis Pétau, Nothing is more firmly rooted in the minds of Christians, both learned and uneducated, than that the torments of demons, and of damned [humans], since these too are immortal, will be eternal and will never end. This question we are addressing [that is, whether Christianity is

Weighing Compatibilism and Libertarianism in Analytic Theology 65 consistent with universalism] will therefore seem to some to be superfluous and even ridiculous.67 Pétau’s claim strikes me as too strong. While there are statements about the existence of hell and that it will be eternal for those who die in mortal sin, so far as I can tell there’s nothing in conciliar Christian theology that explicitly requires or entails that any humans will be in hell.68 That is to say, so far as I can tell, universalism is a live option in conciliar Christian theology. But in the history of the Church, even though universalism is found as early as the second century, it’s certainly a minority view.69 Michael J. McClymond’s evaluation seems more apt than Pétau’s: The overwhelming majority of Christian believers through the centuries have been particularists [that is, those who think not every person is ultimately redeemed]. They believe that certain persons – or a particular group of persons – will finally be saved and dwell forever with God, while others will finally be lost and irrevocably separated from God in hell.70 McClymond understands the “curb appeal” of universalism, but his twovolume work argues that “the universalist house proves to be not a very livable place. The longer one looks at this house and examines the plumbing, wiring, and crawl space beneath, the less attractive it becomes.”71 His overarching conclusion is that “generally speaking, universalism relies on nonliteral interpretations of Scripture and a substantial rejection of church tradition, which in its official doctrinal declarations (rather than in the sphere of private opinions) is consistently particularist on the question of final salvation).”72 From this, we can distill that while there doesn’t appear to be anything in the conciliar Christian theology that explicitly teaches or entails universalism or its denial, the vast majority of Christian tradition has held a rejection of the claim that all will be redeemed. Applying TDP then, given that the vast majority of Church tradition has rejected contingent universalism, even if it is not ruled out by either conciliar Christian theology or classical doctrines, I think that we have good reason to reject contingent universalism (even if we are hopeful universalists). And given, further, that I think that if compatibilism were true then universalism would be true, this gives us theological reason to reject compatibilism even if the philosophical arguments regarding the compatibility question aren’t clearly decisive. Couenhoven writes: [I]n summary, my suggestion is that compatibilism either helps with the problem of divine goodness, especially when it is paired with an affirmation of universal salvation, or, when paired with a doctrine of hell, makes the problem not noticeably worse than it is on libertarian views. (41)

66  Kevin Timpe It should be clear that I don’t share his evaluation. Especially since I think we have good philosophical reasons for thinking compatibilism is false and incompatibilism is true, this strengthens my reasons for rejecting Couenhoven’s view.73 Christians should not, in an epistemic sense of “should,” be compatibilists even if there is not conclusive theological or philosophical evidence that they cannot, in a normative sense, be.74 Notes 1 Subsequent references to Couenhoven’s chapter in this will be made parenthetically. 2 Lynn Rudder Baker, “Why Christians Should Not Be Libertarians: An Augustinian Challenge.” Faith and Philosophy 20(4) (2003): 260–78. 3 Kevin. Timpe, “Why Christians Might Be Libertarians: A Reply to Lynne Rudder Baker.” Philosophia Christi 6(2) (2004): 89–98. I confess I’m pleased to see, looking back, that my title captures a degree of humility that I’ve come to think is even more appropriate for our philosophical reflection on God than I did when I wrote my reply. In fact, there I refrained from arguing that “Christians must, or even should, be libertarians” (280). 4 See, for instance, Derk Pereboom, “Libertarianism and Theological Determinism.” in Kevin Timpe & Daniel Speak (eds.), Free Will and Theism: Connections, Contingencies, and Concerns (Oxford: Oxford University Press, 2016), 112–31. 5 See Kevin Timpe, Free Will in Philosophical Theology (London: Bloomsbury, 2014). Obviously, Pereboom would disagree. 6 As I understand it, either compatibilism or incompatibilism is true, and necessarily so; see Kevin Timpe, Free Will: Sourcehood and Its Alternatives, 2nd edition. (London: Bloomsbury, 2013), particularly Chapter 3, and Kevin Timpe, “Attitudinism, the Compatibility Question, and Ballung Concepts.” in Gregg D. Caruso (ed.), Ted Honderich on Consciousness, Determinism, and Humanity (Cham: Palgrave MacMillan, 2018), 181–94. Following van Inwagen, I take libertarianism to be the conjunction of incompatibilism with the free will thesis; see Peter van Inwagen, “How to Think about the Problem of Free Will.” Ethics 12 (2008): 327–41. 7 So in making these comparisons, we’re making “best of each group” comparisons not average or cherry-picked comparisons, which strikes me as the right way to do it. It also means that proponents of each group have an interest then in seeing the best developed instances of the other family so that we can make the comparisons in this way. 8 For historical grounding for this claim, see Aquinas, De Malo (trans. Richard J. Regan) (Oxford: Oxford University Press, 2003), q.16, a.5, response. For a more recent discussion, see Timothy O’Connor, “Freedom with a Human Face,” Midwest Studies in Philosophy 29 (2005): 207–27. 9 John McKinley, “Seven questions ingredient to Jesus Christ’s temptations.” in Johannes Grössl & Klaus von Stosch (eds.), Impeccability and Temptation: Understanding Christ’s Divine and Human Will (New York: Routledge, 2021), 123. This is, unfortunately, only one of numerous confusions regarding free will that McKinley makes in his chapter. 10 Couenhoven seems to take source libertarian and virtue libertarian views to be equivalent (29), though as I see it what I called virtue libertarianism is one kind of source libertarianism, not equivalent with it; see Kevin Timpe, Free Will in Philosophical Theology, chapter 1. 11 See Timpe, Free Will (chap. 8).

Weighing Compatibilism and Libertarianism in Analytic Theology 67 12 Jesse Couenhoven, “The Necessities of Perfect Freedom.” International Journal of Systematic Theology 14(4) (2012): 396–419. 13 David Fergusson, “Providence.” in James M. Arcadi & James T. Turner, Jr (eds.), T&T Clark Handbook of Analytic Theology (London, T&T Clark Ltd, 2021), 156–58. 14 Alvin Plantinga, God, Freedom, and Evil (Grand Rapids: Eerdmans, 1977), 32. Interestingly, compatibilism doesn’t even make an appearance in the longer The Nature of Necessity (Oxford: Oxford University Press, 1974), despite its covering much of the same territory as God, Freedom, and Evil. Instead, Plantinga defines free will in a way that precludes compatibilism; see 166. In his earlier God and Other Minds (Ithaca: Cornell University Press, 1967), Plantinga writes that “it seems to me altogether paradoxical to say of anyone all of whose actions are causally determined that on some occasions he acts freely” (134). 15 Timpe, Free Will (chap. 5). 16 Timpe, Free Will (chap. 2). 17 William Lycan, “Free Will and the Burden of Proof.” in Anthony O’Hear (ed.), Minds and Persons (Cambridge: Cambridge University Press, 2003), 109. See also David Chalmers, The Conscious Mind: In Search of a Fundamental Theory (New York: Oxford University Press, 1996), 96: “In general, a certain burden of proof lies on those who claim that a certain description is logically impossible. . . . If no reasonable analysis of the terms in question points towards a contradiction, or even makes the existence of a contradiction plausible, then there is a natural assumption in favor of logical possibility.” 18 See Timpe, Free Will. My discussion there is admittedly now a decade dated and would need to be updated to address further compatibilist developments. 19 What more could be said about the nature of abstract objects by someone thinking along these lines? See van Inwagen, “A Theory of Properties.” in Dean Zimmerman (ed.), Oxford Studies in Metaphysics (volume 1) (Oxford: Clarendon Press, 2004), 107–38. 20 One might think in this manner because Peter van Inwagen does think this way in “God and Other Uncreated Things.” in Kevin Timpe (ed.), Metaphysics and God (New York: Routledge, 2009), 3–20. 21 See Peter van Inwagen, “How to Think about the Problem of Free Will.” Ethics 12 (2008): 327–41. 22 Mark Murphy describes divine freedom this way: “God would be fully free – God would not have God’s choosing and acting in any way constrained by nonrational impulses or external coercion” (God’s Own Ethics (Oxford: Oxford University Press, 2019), 162). 23 See Kevin Timpe, “The Best Thing in Life is Free: The Compatibility of Divine Freedom and God’s Essential Moral Perfection.” in Hugh McCann (ed.), Free Will and Classical Theism: The Significance of Freedom in Perfect Being Theology (Oxford: Oxford University Press, 2016). 24 Timothy Pawl & Kevin Timpe, “Incompatibilism, Sin, and Free Will in Heaven.” Faith and Philosophy 26(4) (2009), 396–417; Timothy Pawl  & Kevin Timpe, “Heavenly Freedom: A  Reply to Cowan.” Faith and Philosophy 30(2) (2013), 188–97; Kevin Timpe  & Timothy Pawl, “Paradise and Growing in Virtue.” in Ryan Byerly & Eric Silverman (eds.), Paradise Understood: New Philosophical Essays about Heaven (Oxford: Oxford University Press, 2017), 97–109. 25 Though here, it would depend on if the source-compatibilist’s view was closer to John Martin Fischer’s view or Harry Frankfurt’s; see Timpe, Free Will, chapter 8 for relevant discussion. 26 Kevin Timpe, “Cooperative Grace, Cooperative Agency.” European Journal for Philosophy of Religion 7(3) (2015): 223–45.

68  Kevin Timpe 27 Cornelia Dockter, “God’s Work and Human’s Contribution: Jesus’ Sinlessness in Theodor of Mopsuestia’s Christology.” in Johannes Grössl & Klaus von Stosch (eds.), Impeccability and Temptation: Understanding Christ’s Divine and Human Will (New York: Routledge, 2021), 81. 28 See Kevin Timpe  & Craig Boyd, “Introduction.” in Virtues and Their Vices (Oxford: Oxford University Press, 2014), 9–11 for a longer discussion. 29 See Timothy Pawl & Kevin Timpe, “Freedom and the Incarnation.” Philosophy Compass 11(11) (2016): 743–56; Timothy Pawl, In Defense of Extended Conciliar Christology: A Philosophical Essay (Oxford: Oxford University Press, 2019). 30 Johannes Grössl also thinks that compatibilism does better on this score, but his treatment is perplexing because he seems to hold that source incompatibilism (and thus also source libertarianism) is a variety of compatibilism; see his “Christ’s Impeccability.” in James M. Arcadi & James T. Turner Jr (eds.), T&T Clark Handbook of Analytic Theology (London: T&T Clark Ltd, 2021), 215–29. 31 Pawl, In Defense of Extended Conciliar Christology, viii. For those who do not know the two of us well, this brief insertion, like Pawl’s acknowledgement, is said in Timpe-depricating jest. In working with Tim over the years, I’ve aligned myself with Kurt Cobain’s lyrics from “All Apologies”: “find my nest of salt/everything’s my fault/I’ll take all the blame.” 32 Kevin Timpe, “Grace and Controlling What We Do Not Cause.” Faith and Philosophy 24(3) (2007): 284–99. 33 Timpe, “Cooperative Grace, Cooperative Agency,” 226. 34 For a discussion of what I mean by that, see Timpe, Free Will in Philosophical Theology, 59–71. 35 David Lewis, “Divine Evil.” in Kevin Timpe (ed.), Arguing About Religion (New York: Routledge, 2009), 472. 36 Though Couenhoven also says, “no worse off than libertarians when it comes to explaining why a good God would permit evil and hell” (38). I, unsurprisingly, find this claim less compelling. 37 Timpe, “Why Christians Might Be Libertarians,” 283–4. 38 See, for instance, Jason Turner, “Compatibilism and the Free Will Defense.” Faith and Philosophy 30(2) (2013): 125–37. 39 Laura Ekstrom, God, Suffering, and the Value of Free Will (Oxford: Oxford University Press, 2021), 43. 40 While compatibilists aren’t the only ones who can affirm the truth of determinism, an incompatibilist who did would thereby be denying human freedom and moral responsibility. Since, I take it, God would not subject someone to hell who didn’t deserve it, I’ll set aside determinisitic incompatibilism here for the moment. 41 As I understand it, the problem of hell is a version – a particularly virulent version – of the problem of evil. Jonathan Kvanvig refers to it as “the worst instance of the problem of evil” (The Problem of Hell (New York: Oxford University Press, 1993), 4. 42 I seem to remember that John Fischer initially raised this line of thinking to me in conversation. 43 They would also need to hope that, contra Ekstrom’s God, Suffering, and the Value of Free Will, such freedom is worth the cost. More on this below. 44 I adopt the language of libertarianism being “held hostage” from John Martin Fischer, though apply it here to a slightly different issue than he does. See, for instance, his “Problems with Actual-Sequence Incompatibilism.” The Journal of Ethics 4 (2000): 323–8. 45 Both the libertarian and the compatibilist who thinks that we have non-determined freedom would have to address the question, raised by David Lewis: “God could have settled for a world with compatibilist freedom and that he could have set things up so as to keep his creatures out of trouble. So to escape the

Weighing Compatibilism and Libertarianism in Analytic Theology 69 problem, theists will have to explain why the value of incompatibilist freedom is so great that it outweighs the extraordinary torment by those who forever resist” (Lewis, “Divine Evil,” 474). This line of reasoning is explored at length by Laura Ekstrom, if the existence of such freedom is worth its cost; see Ekstrom, God, Suffering, and the Value of Free Will, particularly chapters 2 and 5. 46 Stamps 2021 differentiates “maximalist” from “minimalist” approaches to tradition in analytic theology, based on how strict of a confessional confession the proponent of those approaches have. I don’t want to require confessional buy-in, but would like for there to be at least Conciliar buy-in, though I realize that some will surely find this an arbitrary preference. 47 R. Lucas Stamps, “Norma Normata: The Role of Tradition in Analytic Theology.” in James M. Arcadi & James T. Turner Jr (eds.), T&T Clark Handbook of Analytic Theology (London: T&T Clark Ltd, 2021), 45. 48 William Wood, Analytic Theology and the Analytic Study of Religion (Oxford: Oxford University Press, 2021), 14f. Throughout his book, Wood argues that a number of the major criticisms leveled against analytic theology may have purchase against particular instances of analytic theology, but don’t hold as objections to analytic theology per se. 49 Wood, Analytic Theology, 14. As Stamps notes, “questioning the role that tradition plays in theology is not unique to the discipline of AT [i.e., analytic theology]. All theologizing from a Christian perspective must give some account of the place of tradition in the theological task” (Stamps, “Norma Normata,” 45). 50 Stamps, “Norma Normata,” 47. 51 Stamps, “Norma Normata,” 49. Stamps draws on both Thomas H. McCall, An Invitation to Analytic Christian Theology (Downers Grove: IVP Academic, 2015) and Oliver D. Crisp, God Incarnate: Explorations in Christology (London: T&T Clark Ltd, 2009) but makes this second category of the hierarchy explicit in a way that they do not. 52 Timothy Pawl, In Defense of Conciliar Christology: A Philosophical Essay (Oxford: Oxford University Press, 2016). 53 This set of teachings may be slightly narrower than what Strand means by the second level since I want to include only the council’s teachings and what is entailed by them, whereas Strands refers to what is implied by the councils’ teachings. It’s not clear to me, though, that he means implication as opposed to entailment. 54 Jordan Wessling, “Christology and Conciliar Authority: On the Viability of Monothelitism for Protestant Theology.” in Oliver D. Crisp and Fred Sanders (eds.), Christology: Ancient & Modern (Grand Rapids: Zondervan, 2013), 153 and 158. I confess it’s not clear to me why Wessling focuses on the evangelical Christian in these principles, rather than the Christian in general. I suspect as a Protestant, he’s not as interested in what, say, Catholic or Orthodox Christians think, but even that doesn’t fully explain the restriction to evangelicals. In more recent work (“Crisp on Conciliar Authority: A  Response to Analyzing Doctrine.” Philosophia Christi 23(1) (2021)), Wessling focuses more on Protestantism in general than evangelical Protestantism. It seems to me to be more important to focus in this context on what Christians should do, and not just what a particular subset of Christians should do. (In other contexts, it makes sense to me to take on additional assumptions that only some Christian traditions endorse.) I take it that Catholics would reject CUP, even when not restricted simply to evangelicals; but I don’t see this as begging the question in the present dialectical question. For if CUP is true, then Catholicism is false. That is, at most one of Catholicism or CUP is true. Furthermore, I take it that the Catholic could endorse CAT, again without the restriction just to evangelicals; they’d simply hold that the right-hand disjunctive side of the biconditional is never satisfied. 55 Wessling, “Christology and Conciliar Authority,” 158f.

70  Kevin Timpe 56 Wessling, “Christology and Conciliar Authority,” 159 and 160. As with CUP and CAT, it’s unclear to me why Wessling restricts P2 to evangelical Christians; see note 55. 57 Wessling, “Christology and Conciliar Authority,” 160. He explicitly rejects P2 as applied to the Third Council of Constantinople; he may not reject it as applying to all teachings of the councils that go beyond what is taught by Scripture. He’s thus an example of what Stamps characterizes as a “minimalist” approach to Christian tradition. I think that Stamp is correct to be concerned about such an approach: In some cases, perhaps especially among Protestant analytic theologians, who may attribute less binding authority to creeds and councils in favor of some form of biblicism, the minimalist’s methodology may be an example of cherry-picking. So, for example, analytic theologians who reject dyothelitism seem to privilege the first four ecumenical councils ending in Chalcedon, and thus sense a lower burden to defend the Sixth Ecumenical Council [i.e., the Second Council of Nicaea in 787]. (Stamps, “Norma Normata,” 52f) Wessling also writes that if one endorses theological determinism, then we should be deeply skeptical about claiming to know what God would or would not allow when it comes to theological error; see 164. However, since both Couenhoven and I  reject such determinism, I  won’t explore this issue further at present. 58 Wessling, “Christology and Conciliar Authority,” 163. More recently, Wessling has indicated that “my traditionalism has only increased since the construction of my [2013] paper” (Wessling, “Crisp on Conciliar Authority,” 45). 59 It’s possible that Wessling does think that Scripture does establish homoousianism despite Arius’ appeal to Scripture. If so, we have different readings both of what the Scriptures establish and Church history. For another discussion of Wessling’s “Christology and Conciliar Authority,” see Oliver D. Crisp, “The Divine and Human Will of Christ.” in Grössl & Klaus von Stosch (eds.), Impeccability and Temptation: Understanding Christ’s Divine and Human Will Johannes (New York: Routledge, 2021), 199–215; see also Oliver D. Crisp, Divinity and Humanity (Cambridge: Cambridge University Press, 2007), 35–6. For Tim Pawl’s view of the dialectic with regard to principles like CUP and CAT in relation to Conciliar Christology, see Pawl, In Defense of Extended Conciliar Christology, 6f. 60 Another concern with CUP is that it seems it would require a particularly strong view of the consistency of the Scriptures that I think isn’t justified. Parts of the Old Testament, for instance, are henotheistic rather than monotheistic. Given that the two views are inconsistent, how do we decide which to take as normative? Conciliar Christian theology gives us an easy response. 61 Baker, “Why Christians Should Not Be Libertarians,” 462; I indicated my agreement with this as early as 2004 with my Timpe, “Why Christians Might Be Libertarians.” 62 It does seem to me that conciliar Christian theology entails the existence of moral responsibility, so would seem to rule out views, compatibilist or incompatibilist, that deny free will. See also the discussion in Mawson 2016. 63 TDP is similar to be not identical with what Wessling calls the doxastic preservation principle, or DPP: “it is extremely implausible that God would allow the vast majority of the church to be led into error on a matter central to the faith” (Wessling, “Crisp on Conciliar Authority,” 44; quoting Crisp, Analyzing Doctrine, 191). 64 In a recent book on the intellectual virtues, and with intellectual carefulness specifically in mind, Nathan King develops the point that some topics of belief are more important than others. “So, which objects deserve our careful thought? And of those objects that do deserve care, how much do they deserve? Unfortunately – but appropriately given our topic – these questions don’t admit precise answers. There’s no algorithm for intellectual carefulness” (King, The Excellent Mind:

Weighing Compatibilism and Libertarianism in Analytic Theology 71 Intellectual Virtues for Everyday Life (New York: Oxford University Press, 2021), 72. But King does give a couple of guidelines for what we need to be extra careful in thinking about. “When our beliefs about a topic X have logical implications for a large number of other beliefs, we should consider X carefully. . . . [And] other things being equal, topics central to human flourishing tend to deserve more care than those that are peripheral” (ibid.). This is good advice, and I have sought to follow it in developing this section. 65 Insofar as universalism is consistent with all being saved through Christ, we ought not think that universalism implies a denial of Christian exclusivism. 66 If it ends up being true that God’s decision to create is itself the result of a contingent choice, we could rewrite the view here as “necessarily, if God freely chooses to create moral agents, all those agents will be redeemed eventually and that no one will be assigned to hell for all eternity.” Insofar as nothing of substance for the larger part depends on this precision, I’ll ignore it other than this footnote. 67 Denis Pétau, Theologicorum dogmatum tomus tertius (Paris: n.p., 1644), T.3.199, quoted in Daniel Pickering Walker, The Decline of Hell: Seventeenth-Century Discussions of Eternal Torment (Chicago: University of Chicago Press, 1964), 21. David Lewis also asks “Is universalism really a Christian option?” (Lewis, “Divine Evil,” 476). He argues that it is not given that redemption is central to Christianity. But this isn’t right. If it’s only contingent that there’s a creation at all, there being the need for redemption of that creation would also be contingent. 68 Something similar is found in Catholic theology; see, for instance, Catechism of the Catholic Church §1033–1035 which claims those who die in mortal sin will spend an eternity in hell; it does not explicitly claim that there are any individuals for whom this is true. 69 Ilaria L. E. Ramelli’s book A Larger Hope? Universal Salvation from Christian Beginnings to Julian of Norwich (Eugene: Cascade Books, 2021) is a worthwhile read on the history of Christian universalism. But even at its best, it only establishes universalism as a thread running throughout church history. While I  am neither a biblical scholar nor a historical theologian, I find his book often fails to provide sufficient evidence for his interpretations. I do find his argument that what the Second Council of Constantinople (553CE) condemned was apocatastasis embedded within a larger metaphysic of the transmigration of the soul and not universalism per se to be plausible; see Ramelli, chapter 9. 70 Michael J. McClymond, The Devil’s Redemption: A New History and Interpretation of Christian Universalism (Grand Rapids: Baker Academic, 2018), xxi. See also Richard Bauckham, “Universalism: An Historical Survey.” Themelios 4 (1978): 47–54. 71 McClymond, The Devil’s Redemption, xxi. McClymond argues throughout that “Christian universalism is not like traditional Christian theology with salvation for all superadded” (17), but that all major aspects of systematic theology are shaped by universalistic commitments. 72 McClymond, The Devil’s Redemption, 21. 73 One could, I guess, argue that any true view is part of “theology,” and thus there’s no philosophical versus theological distinction that I’m assuming here. If that’s the case, then I take it the overall theological case would favor incompatibilism. But I don’t want to argue for that way of laying out the boundaries of the disciplines here – for one, it would potentially make every truth claim a theological truth. And that just seems to get the terrain wrong if we’re trying to relate theological to another discipline (since the entire language of “relates to another discipline” presupposes a difference between them). 74 I’m very grateful for helpful comments from John Martin Fischer and Michael DeVito on earlier drafts. Despite our disagreements, Jesse Couenhoven is a wonderful and gracious interlocutor; I’ve learned much from his work and our interactions over the years.

72  Kevin Timpe Bibliography Aquinas, Thomas. De Malo (trans. Richard J. Regan). Oxford: Oxford University Press, 2003. Baker, Lynn Rudder. “Why Christians Should Not Be Libertarians: An Augustinian Challenge.” Faith and Philosophy 20(4) (2003): 260–78. Bauckham, Richard. “Universalism: An Historical Survey.” Themelios 4 (1978): 47–54. Chalmers, David. The Conscious Mind: In Search of a Fundamental Theory. New York: Oxford University Press, 1996. Couenhoven, Jesse. “The Necessities of Perfect Freedom.” International Journal of Systematic Theology 14(4) (2012): 396–419. Crisp, Oliver D. Divinity and Humanity. Cambridge: Cambridge University Press, 2007. Crisp, Oliver D. God Incarnate: Explorations in Christology. London: T&T Clark Ltd, 2009. ———. “The Divine and Human Will of Christ.” in Johannes Grössl & Klaus von Stosch (eds.), Impeccability and Temptation: Understanding Christ’s Divine and Human Will. New York: Routledge, pp. 199–215, 2021. Dockter, Cornelia. “God’s Work and Human’s Contribution: Jesus’ Sinlessness in Theodore of Mopsuestia’s Christology.” in Johannes Grössl & Klaus von Stosch (eds.), Impeccability and Temptation: Understanding Christ’s Divine and Human Will. New York: Routledge, 2021. Ekstrom, Laura. God, Suffering, and the Value of Free Will. Oxford: Oxford University Press, 2021. Fergusson, David. “Providence.” in James M. Arcadi  & James T. Turner, Jr (eds.), T&T Clark Handbook of Analytic Theology. London: T&T Clark Ltd, pp. 156–58, 2021. Fischer, John Martin. “Problems with Actual-Sequence Incompatibilism.” The Journal of Ethics 4 (2000): 323–28. Grössl, Johannes. “Christ’s Impeccability.” in James M. Arcadi & James T. Turner Jr (eds.), T&T Clark Handbook of Analytic Theology. London: T&T Clark Ltd, pp. 215–29, 2021. King, Nathan. The Excellent Mind: Intellectual Virtues for Everyday Life. New York: Oxford University Press, 2021. Kvanvig, Jonathan. The Problem of Hell. New York: Oxford University Press, 1993. Lewis, David. “Divine Evil.” in Kevin Timpe (ed.), Arguing about Religion. New York: Routledge, 2009. Lycan, William. “Free Will and the Burden of Proof.” in Anthony O’Hear (ed.), Minds and Persons. Cambridge: Cambridge University Press, 2003. McCall, Thomas H. An Invitation to Analytic Christian Theology. Downers Grove: IVP Academic, 2015. McClymond, Michael J. The Devil’s Redemption: A New History and Interpretation of Christian Universalism. Grand Rapids: Baker Academic, 2018. McKinley, John. “Seven Questions Ingredient to Jesus Christ’s Temptations.” in Johannes Grössl & Klaus von Stosch (eds.), Impeccability and Temptation: Understanding Christ’s Divine and Human Will. New York: Routledge, 2021. Murphy, Mark. God’s Own Ethics. Oxford: Oxford University Press, 2019. O’Connor, Timothy. “Freedom with a Human Face.” Midwest Studies in Philosophy 29 (2005): 207–27.

Weighing Compatibilism and Libertarianism in Analytic Theology 73 Pawl, Timothy. In Defense of Conciliar Christology: A Philosophical Essay. Oxford: Oxford University Press, 2016. Pawl, Timothy. In Defense of Extended Conciliar Christology: A Philosophical Essay. Oxford: Oxford University Press, 2019. Pawl, Timothy & Kevin Timpe. “Freedom and the Incarnation.” Philosophy Compass 11(11) (2016): 743–56. ———. “Heavenly Freedom: A Reply to Cowan.” Faith and Philosophy 30(2) (2013): 188–97. ———. “Incompatibilism, Sin, and Free Will in Heaven.” Faith and Philosophy 26(4) (2009): 396–417. Pereboom, Derk. God and Other Minds. Ithaca: Cornell University Press, 1967. ———. God, Freedom, and Evil. Grand Rapids: Eerdmans, 1977. ———. “Libertarianism and Theological Determinism.” in Kevin Timpe  & Daniel Speak (eds.), Free Will and Theism: Connections, Contingencies, and Concerns. Oxford: Oxford University Press, pp. 112–31, 2016. ———. The Nature of Necessity. Oxford: Oxford University Press, 1974. Ramelli, Ilaria L. E. A Larger Hope? Universal Salvation from Christian Beginnings to Julian of Norwich. Eugene: Cascade Books, 2021. Stamps, Lucas R. “Norma Normata: The Role of Tradition in Analytic Theology.” in James M. Arcadi & James T. Turner Jr (eds.), T&T Clark Handbook of Analytic Theology. London: T&T Clark Ltd, 2021. Timpe, Kevin. “Attitudinism, the Compatibility Question, and Ballung Concepts.” in Gregg D. Caruso (ed.), Ted Honderich on Consciousness, Determinism, and Humanity. Cham: Palgrave MacMillan, pp. 181–94, 2018. ———. “Cooperative Grace, Cooperative Agency.” European Journal for Philosophy of Religion 7(3) (2015): 223–45. ———. “Grace and Controlling What We Do Not Cause.” Faith and Philosophy 24(3) (2007): 284–99. ———. Free Will in Philosophical Theology. London: Bloomsbury, 2014. ———. Free Will: Sourcehood and Its Alternatives (2nd edition). London: Bloomsbury, 2013. ———. “The Best Thing in Life is Free: The Compatibility of Divine Freedom and God’s Essential Moral Perfection.” in Hugh McCann (ed.), Free Will and Classical Theism: The Significance of Freedom in Perfect Being Theology. Oxford: Oxford University Press, 2016. ———. “Why Christians Might Be Libertarians: A Reply to Lynne Rudder Baker.” Philosophia Christi 6(2) (2004): 89–98. Timpe, Kevin  & Craig Boyd. Virtues and Their Vices. Oxford: Oxford University Press, 2014. Timpe, Kevin & Timothy Pawl. “Paradise and Growing in Virtue.” in Ryan Byerly & Eric Silverman (eds.), Paradise Understood: New Philosophical Essays about Heaven. Oxford: Oxford University Press, pp. 97–109, 2017. Turner, Jason. “Compatibilism and the Free Will Defense.” Faith and Philosophy 30(2) (2013): 125–37. van Inwagen, Peter. “A  Theory of Properties.” in Dean Zimmerman (ed.), Oxford Studies in Metaphysics (volume 1). Oxford: Clarendon Press, pp. 107–38, 2004. ———. “God and Other Uncreated Things.” in Kevin Timpe (ed.), Metaphysics and God. New York: Routledge, pp. 3–20, 2009. ———. “How to Think about the Problem of Free Will.” Ethics 12 (2008): 327–41.

74  Kevin Timpe Walker, Daniel Pickering. The Decline of Hell: Seventeenth-Century Discussions of Eternal Torment. Chicago: University of Chicago Press, 1964. Wessling, Jordan. “Christology and Conciliar Authority: On the Viability of Monothelitism for Protestant Theology.” in Oliver D. Crisp & Fred Sanders (eds.), Christology: Ancient & Modern. Grand Rapids: Zondervan, 2013. ———. Jordan. “Crisp on Conciliar Authority: A Response to Analyzing Doctrine.” Philosophia Christi 23(1) (2021). Wood, William. Analytic Theology and the Analytic Study of Religion. Oxford: Oxford University Press, 2021.

4 Can an Unfree Person Have a “Deep Self?” Leigh Vicens

What else must one give up if one gives up belief in human freedom? Historically most philosophers have answered, “a lot.” Much of our moral and social life is associated with the assumption of free will, and many religious practices and theological doctrines would seem to depend upon it as well. Recently, however, some free will skeptics have questioned these associations. Derk Pereboom has argued that with some modifications, much of our moral life can withstand free will skepticism. For even if, because we lack free will and so are not ultimately responsible for what we do, we do not deserve certain treatment in response to our behavior, treating each other in these ways can be justified by “forward-looking” considerations. I have gone further to maintain that some responses to our conduct might still be deserved without free will. In this chapter, I summarize the debate as it stands so far before discussing an objection to my view and offering an answer to it. Ultimately, I suggest that the solution offered is compatible at least with certain central Christian notions. Free Will Skepticism and Basic Desert: Summarizing the Debate I begin by considering why anyone would give up belief in human freedom in the first place. There are, broadly, three sorts of reasons that are normally offered for free will skepticism: philosophical, empirical, and theological. Philosophical considerations most often raise the specter of determinism or indeterminism. Either view of the world may seem to preclude the possibility of our controlling how certain events – including our own choices and actions – turn out: determinism, because it would entail that our choices are the consequences of events of the distant past over which we have no control1; and indeterminism, since a lack of determination would seem to leave our choices not determined by anything, or anyone – including ourselves – and so random.2 Empirical considerations involve the role of consciousness in self-governance, including the automaticity of much of our behavior and (lack of) awareness of our own motivations for our choices.3 Theological considerations may differ significantly between religions. Within Christianity, theological determinism has been an enduring if now minority view of divine DOI: 10.4324/9781003306191-5

76  Leigh Vicens providence, for reasons relating to divine sovereignty4; and considerations of the universality and inevitability of human sin may also push one in the direction of at least postlapsarian determinism.5 What aspects of our moral and social life are traditionally thought to require free will? Free will is considered necessary – and, indeed, often defined as the “control condition” – for moral responsibility. And the assumption of moral responsibility is thought necessary for the appropriateness of our practices of holding each other morally responsible – blame and praise, as well as punishment and reward. Many philosophers maintain that we assume moral responsibility not only when we overtly praise and blame each other but also when we hold “reactive attitudes” such as gratitude, resentment, guilt, and forgiveness.6 For instance (so the story goes), when one feels guilty, one takes oneself to be responsible for wrongdoing, and when one experiences resentment, one takes another to be responsible for wronging one. While the exact nature of forgiveness is a debated topic, most parties agree that to forgive another, one has to take the other to be responsible for wrongdoing. One popular view defines forgiveness as the overcoming or forswearing of resentment against another7; in this case, if resentment involves the assumption of moral responsibility, then forgiveness does as well. If the absence of moral responsibility would undermine much of our moral and social lives, it would seem to completely annihilate our religious life – at least, in traditions such as Christianity, in which concepts such as sin, guilt, condemnation, and forgiveness are paramount. If we are not responsible for what we do, then we should not feel guilty for our sins, nor do we deserve condemnation. And where there is no possibility of condemnation, it would seem, there can be no possibility of divine forgiveness – a possibility at the heart of the Gospel. And yet some free will skeptics have questioned the association between free will and at least some of these important attitudes and practices. Most famously, Derk Pereboom has argued that, while that without free will we would lack “basic desert” moral responsibility, or the sort of responsibility that would make us fundamentally deserving of responses to our (im)moral behavior, we may still be justified in maintaining these responses on the basis of consequentialist or “forward-looking” justifications. Recently, Pereboom has argued that protesting or opposing the immoral conduct of someone who is not ultimately responsible for that conduct can be justified insofar as the protest aims at “moral formation of character, reconciliation in relationships, retention of integrity of a victim, and protection from harm.”8 Similarly, renouncing one’s former opposition to a wrongdoer, when the wrongdoer has expressed contrition,9 can be justified, presumably since the forward-looking aims of protest, such as moral formation of character and protection from harm, have been met. Such protest and renunciation of protest are the freewill skeptic’s versions of blame and forgiveness. I have gone further10 to suggest that some responses to immoral conduct might even be deserved without free will. The sense of “desert” I mean is the “basic” one at issue in the free will debate. Justifications for treatment that

Can an Unfree Person Have a “Deep Self?” 77 appeal to basic desert can be characterized in part negatively: for one thing, such justifications do not appeal to consequentialist or “forward-looking” considerations.11 So isolating a criminal on the grounds that he is a danger to society, or praising a student merely in order to increase her confidence on a test, would not be justified in terms of basic desert – nor, of course, would protesting someone’s behavior simply to bring about moral reform. Basic desert justifications are also not contractualist: if a Monopoly player “deserves” to go to jail, or a lottery winner deserves a million dollars, this desert is also not the “basic” sort at issue in the free will literature.12 In addition to ruling out consequentialist and contractualist considerations, basic desert justifications do not appeal to relational features of an agent and her situation. If a justification for treating someone in a certain way (or not treating her that way) appeals, for instance, to others’ standing to blame, this justification would not be a basic desert one. What is left when consequentialist, contractualist, and relational considerations are ruled out? Put simply, a justification for treating someone a certain way is a basic desert justification when it appeals only to the fittingness between the intrinsic features of an agent and her (im)moral action, and the deserved treatment. I have argued13 that protest, in particular, may be basically deserved by a wrongdoer even if free will skeptics are right that no one ever freely engages in wrongdoing. While this may seem like simply a statement of semi-compatibilism – that is, that someone can be (and be held) morally responsible for her behavior in the absence of free will – I have maintained that it is not; my position is that even if, because a person lacks free will, a central form of moral responsibility is ruled out, the person may still be (basically) deserving of certain treatment. One way to see that this is not a standard form of semi-compatibilism is to note that I hold that a lack of free will would rule out some treatment as basically deserved: any treatment that involves the intention to cause suffering or harm as payback for wrongdoing – that is, any “retributivist” response – would be ruled out. But this does not rule out all basically deserved treatment for wrongdoing; for protest (among other types of blaming responses to immoral behavior) need not involve a retributivist intention. I have suggested elsewhere that the possibility of divine protest – especially if it is deserved – might go a considerable way to shoring up Christian doctrine against the threat of free will skepticism since it could account for much of the New Testament language of divine wrath and related concepts.14 I would now add that divine forgiveness construed as Pereboom proposes – as the renunciation of protest – would fit with my view as well, and further the project of showing that free will skepticism is compatible with central Christian doctrine. Deserved Protest and the Deep Self: An Objection (or Two) But objections might be (and have been) raised to the idea that a free will skeptic could maintain the coherence of deserved protest – and so, the possibility of forgiveness. The main objection I will consider here is that if a person

78  Leigh Vicens does not act freely, then we can only reasonably direct our protest toward the person’s behavior and not toward the person herself. When Pereboom discusses free will skepticism, he says it would rule out moral responsibility “in the basic desert sense,” a sense he cashes out in terms of the agent’s action “belonging to” her15 or being hers “in such a way that she would deserve blame if she understood that it was morally wrong” and praise if morally exemplary.16 Drawing on Pereboom’s terminology, the objection may thus be put as follows: if an agent’s action does not belong to her in this sense, then while it may make sense to oppose her action, or the message it conveys, it does not make sense to oppose her – because there is not the right sort of self at which to direct the protest.17 Why is this a problem for me? Well, it seems that the concept of basic desert is necessarily tied to that of the self. While actions (or even inanimate objects) might in some sense deserve things – for example, that performance deserves the gold medal; that paper deserves a C – a performance or a paper cannot basically deserve anything. And in particular we might think that if divine judgment is to involve protest, then that protest must be directed at a person; for it is persons (“the living and the dead”) and not simply actions that will be condemned or forgiven. Moreover, Justin Capes has raised a further objection to Pereboom’s view of forgiveness as the renouncing of protest which would cause trouble for my position as well. Capes argues that if Pereboom’s view of forgiveness is correct, [Y]ou can’t forgive those who have wronged you while also continuing to engage in deliberate protest of (the meaning of) their bad behavior. But this implication is problematic. To see why, consider. . . . Pereboom’s example of a friend who mistreats you in some way. Later, let us imagine, the friend apologizes, you accept his apology, and you forgive him. But now why couldn’t your friend join with you in protesting the claim made by his mistreatment of you and support your right to be treated well? If you have accepted your friend’s apology, and the meaning of (or the claim made by) the past offense is the real object of your joint protest, it seems that you and your friend may both stand up for morality and your self-worth by protesting his past mistreatment of you in appropriate contexts and ways. It seems that it is both possible and often entirely fitting to forgive a wrongdoer and yet also to continue to protest (the meaning of) his offense. But if so, then forgiving can’t be the renunciation of moral protest, nor can it require such renunciation.18 This second problem seems to build on the first: it is because the free will skeptic must convey protest as targeting a person’s behavior (and the meaning or message it conveys), rather than the person herself, that one cannot forgive (on Pereboom’s account of forgiveness) while continuing to stand in opposition to the wrongdoing. One cannot, as it were, forgive the sinner but still oppose the sin, if one was only opposed to the sin and not the sinner to begin with. Thus, both problems would seem to depend on the claim that the

Can an Unfree Person Have a “Deep Self?” 79 free will skeptic cannot protest a person herself, but only her behavior.19 And so, if the skeptic can give an account of a “deep self” that a person who lacks free will might retain, and that might be the target of (basically deserved) protest, both problems will be solved. Being a Self Without Free Will: Having the Will One Wants to Have I think a person who is unfree might still have (or, perhaps more accurately, can still be) a self that (basically) deserves certain treatment. To show this, I will point to some (compatibilist) conditions commonly thought sufficient for free will, and note that while these conditions may obtain in the absence of other conditions that free will skeptics think are necessary for free will, if these former conditions are met, then at least an action can “belong” to an agent in a deep enough sense that the agent herself can deserve protest for immoral behavior. I begin with Harry Frankfurt’s hierarchical view of free will and moral responsibility. Frankfurt has asserted that a person has a free will when she has the will that she wants to have, which involves the satisfaction of a secondorder desire that a certain first-order desire be effective, or actually move one to act.20 Imagine, for instance, that a person wants to please her husband, and also wants to have a career, hobbies, and friends. And suppose that she is aware that these desires are in conflict (because her controlling husband is displeased with the idea of his wife having a life of her own) and that reflecting on this conflict, she decides to please her husband rather than pursue these other things. Then, on Frankfurt’s account, she is acting freely, and is responsible for her behavior since she has the will (i.e., effective first-order desire) that she wants to have. Frankfurt’s view has been criticized on the now-familiar grounds that it fails to take into account how a person has come to have the desires she does, including her second-order desires.21 We might suppose that, in the case under consideration, the controlling husband has manipulated his wife for years to diminish her self-confidence and sense of agency – or, instead, we might suppose that she grew up in a home and community where women were always submissive to their husbands so that she cannot even imagine a different way – much less desire it. Either way, it is possible that the wife has the will she wants to have, and yet many would say she is not really free or fully responsible for her own submissive behavior. Imagine, though, that the wife has internalized her husband’s, or family’s or community’s, values, some of which are deeply immoral. And now imagine that the wife acts out of these values in an immoral way – say, she insists that her own daughter submit to the domineering men in her life. It seems to me that because she has the will that she wants to have, she would be an appropriate target of protest. And it would not simply be her behavior that was protest-worthy, but the person herself, and the internal features of the person – her misogyny, her disrespect of the agency of her own daughter, her

80  Leigh Vicens failure of empathy and hardness of heart in response to the obvious suffering she is causing, and so on. Frankfurt emphasizes the importance of being “critically aware” of one’s own will22; on his view, it seems, second-order desires involve such awareness. In later writing, he characterizes a second-order volition as involving the identification of oneself with a certain second-order desire23 – an identification which also seems to require some conscious awareness. And so it might seem that I am saying that, even if because one is manipulated or socially engineered to behave in a certain way, one is not free, so long as one is consciously aware of, or reflectively identifying with, the desires on which one acts, one can (basically) deserve certain treatment in response. But while conscious self-reflection may be sufficient for having a self that could basically deserve certain treatment, I am not sure it is necessary. To see why not, I turn to a second (compatibilist) account of moral responsibility: attributionism. Being a Self Without Free Will: Revealing One’s Cares Attributionists such as Angela Smith and Michael Brownstein hold that a person is morally responsible for what she does if it is attributable to her, in the sense that it reveals “an agent’s stable and identity-grounding attitudes.”24 Different attributionists identify what is attributable to the self in different ways, such as one’s “evaluative judgments or appraisals”25 or one’s “cares” in the sense of what matters to one, or what one is “emotionally tethered” to.26 Importantly, attributionists reject what Smith calls “voluntariness as a precondition for moral responsibility,”27 the requirement that morally responsible action be under (or traceable to an instance of) the agent’s voluntary control, as well as what might be called “consciousness conditions,” that is, that consciousness of certain facts about one’s action is required for moral responsibility.28 But just as many reject Frankfurt’s view on the grounds that it fails to take into account the causal history of a person’s (higher order) desires – and so, fails to rule out cases of manipulation or social engineering as undermining free will – so many reject attributionist accounts of moral responsibility on the grounds that if an action is not under (or at least traceable to an instance of) a person’s conscious voluntary control, then it cannot be considered free and responsible.29 I agree with those who reject attributionism on these grounds. It would not be fair, for instance, to inflict retributive punishment on someone whose bad behavior was never under his conscious voluntary control: he could never deserve such treatment. Still, suppose that it turns out that the voluntariness or consciousness conditions are rarely, if ever, met in the real world. And imagine again someone who meets the attributionist conditions for moral responsibility, but does not meet the voluntariness and/or the consciousness conditions. For instance, Brownstein notes that while inadvertently stepping

Can an Unfree Person Have a “Deep Self?” 81 on a stranger’s toes on a crowded subway is not something we normally think is reflective of a person’s deeper self, if I step on someone’s toes because I am aggressively pushing my way toward my favorite spot near the window, then there is something about me – something about who I am as a moral agent – that is expressed through my behavior, and I now appear to be open to evaluation for this act. This is to say that the action appears to be attributable to me, even if a number of putatively “exculpating” conditions obtain. I might not know that I  have stepped on anyone’s toes, and might not have intended to do so, and I might even have tried hard to avoid everyone’s toes while I raced to my favorite spot. Regardless whether my action is non-conscious and non-volitional in this way, or whether I disavow “New York Style” subway riding, what I’ve done expresses something morally important about me. I’m not just klutzy, which is a kind of “shallow” or “grading” evaluation (Smart, 1961). Rather, a bystander would be quite right to think “what a jerk!” as I push by.30 Now imagine that you end up riding the 5:00 AM train home with such a passenger every day, and at least once a week, he tramples on your toes. It would seem entirely appropriate for you to voice some protest31 – and the protest would not simply be directed at the behavior of toe-trampling; it would be directed at the person, and his (perhaps non-conscious) attitudes and values: he cares more about getting a seat on the train than he does about respecting others’ personal space or physical well-being! Such a person deserves to be confronted. In, I have argued that even if a person lacks free will in acting immorally, the person may still basically deserve protest. The person herself, and not simply her action, could be deserving of protest in virtue of her actions being connected in a certain way to herself: being the manifestation of the will she wants to have or a reflection of her “identity-grounding attitudes.” Moreover, if forgiveness is understood as the renunciation of protest, then it is possible to forgive a person who has acted wrongly without free will, and one may do so while still opposing the (meaning of) her immoral conduct. While certain empirical considerations mentioned earlier – regarding the extent to which we are aware of our motives in acting and identify with the desires that move us to act – might call into question whether we meet the first set of conditions for basic desert, suggested by Frankfurt’s hierarchical view of moral responsibility, the attributionist conditions seem more immune from such concerns. They also seem compatible with the philosophical and theological considerations discussed earlier that have led some to question the existence of human freedom, such as (theological) determinism or the inevitability of sin. Indeed, Jesse Couenhoven has suggested (personal communication)

82  Leigh Vicens that Martin Luther – who (in)famously insisted on the active divine determination of all our (im)moral actions, as well as a lack of (libertarian) human freedom – was a kind of proto-attributionist about human moral responsibility. Interpreting Luther’s remark that saints and sinners “do what is in them,” Couenhoven writes, The most fundamental human motivations are aspects of personality that are too deeply imbedded within human hearts and minds to be objects of choice. Saints and sinners make their choices because of who they are; they are not, Luther thought, able to choose who they are.32 Basic Desert Redux I will now consider one final objection to my view that, even without free will, protest can be basically deserved. The objection is that my view of basic desert is too thin, and not really the sense at issue in the free will debate. Pereboom himself acknowledges that certain negative, backward-looking attitudes can be fitting, and yet he does not consider these to be basically deserved. For instance, he acknowledges that “feeling pain on account of a recognition that one has not lived up to one’s moral standards” is fitting, but that this fittingness “need not involve desert.” He draws two analogies: (1) one might appropriately feel pained that one failed to meet one’s standards for chess playing when one understands that one’s substandard performance is due to factors beyond one’s control, while this pain is not deserved; (2) it would be appropriate . . . to feel the pain of grief upon the death of a loved one, while this pain is not deserved.33 Pereboom theorizes that the reason the pain in all three cases is not basically deserved is that it is not “appropriately imposed,” and he suggests, following Hilary Bok, that such appropriate imposition is linked to the infliction of punishment (in these cases, on the self).34 I have denied that basically deserved treatment must be thought of as retributive in this way.35 So how can I  distinguish between fitting responses that are, and are not, basically deserved? Consider the two analogies that Pereboom discusses. Basic desert is a matter of fittingness between a(n) (im) moral action that reveals in some way one’s deep self (e.g., a will one has reflectively endorsed, or one’s non-conscious but identity-grounding attitudes) and a particular response. Since in neither the chess case nor the grief case is there any (im)moral behavior, there cannot be any basically deserved response. In the regret case, in contrast, one has (by hypothesis) behaved immorally, and so one may in fact basically deserve certain responses to one’s behavior. But I  agree with Pereboom that pain is not something that one basically deserves in response to immoral behavior if one’s behavior is not

Can an Unfree Person Have a “Deep Self?” 83 free. While protest may cause pain, it is not the pain that is deserved, but the protest itself. Protest is not necessarily painful to experience, and protest does not aim at causing pain in its target, but at communicating a certain message, making plain to its target a certain truth: the kind of person she has revealed herself to be, in acting as she has.36 Notes 1 Peter van Inwagen, An Essay on Free Will (New York: Oxford University Press). 2 See a summary of various versions of this argument in Robert Kane, A Contemporary Introduction to Free Will (chap. 4) (Oxford: Oxford University Press, 2005). 3 Gregg Caruso, Free Will and Consciousness: A Determinist Account of the Illusion of Free Will (chap. 12) (Lanham: Lexington Books, 2012). 4 Leigh Vicens, “Theological Determinism.” Internet Encyclopedia of Philosophy (2014). 5 Leigh Vicens, “Human Freedom and the Inevitability of Sin” in Peter Furlong & Leigh Vicens (eds.), Theological Determinism: New Perspectives (Cambridge: Cambridge University Press, 2022), 150–64. 6 Peter Strawson, “Freedom and Resentment.” Proceedings of the British Academy 48 (1962): 1–25. 7 This view is inspired by Joseph Butler, Fifteen Sermons Preached at Rolls Chapel, in The Works of the Right Reverend Father in God, Joseph Butler, D.C.L., Late Bishop of Durham (Samuel Halifax (ed.) (Alabama: Carter, 1846). 8 Derk Pereboom, “Forgiveness as Renunciation of Moral Protest” in Brandon Warmke, Dana Kay Nelkin  & Michael McKenna (eds.), Forgiveness and Its Moral Dimensions (Oxford: Oxford University Press, 2021), 90. 9 Ibid., 94. 10 “Free Will Skeptics Can Have Their Basic Desert and Eat it Too.” Journal of the American Philosophical Association 1 (2022): 1–12. 11 Gregg Caruso & Stephen Morris, for instance, characterize basic desert simply as a “backward-looking and non-consequentialist” concept in “Compatibilism and Retributivist Desert Moral Responsibility: On What is of Central Philosophical and Practical Importance.” Erkenntnis 82 (2017): 838. 12 cf. Thomas Scanlon, “Giving Desert its Due.” Philosophical Explorations 16(2) (2013): 101, and Derk Pereboom, Free Will, Agency, and Meaning in Life (Oxford: Oxford University Press, 2014), 2. 13 “Free Will Skeptics Can Have Their Basic Desert and Eat it Too” Journal of the American Philosophical Association 1 (2022): 1–12. 14 Leigh Vicens, “Sin and the Faces of Responsibility” in John Allan Knight & Ian Markham (eds.), The Craft of Innovative Theology (Hoboken: Wiley Blackwell, 2022), 99–113. 15 Derk Pereboom, “Determinism al Dente.” Nous 29(1) (1995): 22, italics added. 16 Derk Pereboom, Free Will, Agency, and Meaning in Life (Oxford: Oxford University Press, 2014), 2. 17 Thanks to Simon Kittle (personal communication) for raising this issue. 18 Justin Capes, “Hard Theological Determinism and Divine Forgiveness Are Incompatible” in Peter Furlong  & Leigh Vicens (eds.), Theological Determinism: New Perspectives (Cambridge: Cambridge University Press, 2022), 165–83. 19 In fact, I think there are several other ways around the problem Capes poses that I will not discuss in the body of this chapter, so as not to distract from my main issue of concern. First, Pereboom might deny that renouncing protest amounts to condoning the protested wrongdoing. For on Pereboom’s view, “protest is essentially confrontational” (2021, 90, italics added). He seems to allow that you

84  Leigh Vicens can still affirm the wrongness of a past action without protesting it if you are not taking a confrontational stance against the wrongdoer, writing, “Some overt specifications of wrongdoing don’t count as protest. In cataloguing instances in which a forgiver has been wronged in the past, she might cite a wrongdoing she has forgiven in a way that doesn’t count as moral protest against the wrongdoer” (2021, 93). Another way around Capes’ objection – for me, at least – is to part company with Pereboom on the nature of resentment and maintain that it is compatible with free will skepticism so that forgiveness might involve renouncing resentment without renouncing protest. Pereboom rejects the “standard view” of forgiveness as essentially involving the renunciation of resentment because he thinks resentment presupposes that the wrongdoer (basically) deserves to be blamed for the wrongdoing, and so is incompatible with free will skepticism. But since I maintain the possibility of basic desert in the absence of free will, it is open to me to maintain the appropriateness of resentment as well as the standard view of forgiveness as involving the renunciation of resentment. 20 Harry Frankfurt, “Freedom of the Will and the Concept of a Person” (1971), reprinted in Gary Watson (ed.), Free Will, 2nd edition (Oxford: Oxford University Press, 2003), 322–36. 21 Cf. John Martin Fischer  & Mark Ravizza, Responsibility and Control: A  Theory of Moral Responsibility (chap. 7) (Cambridge: Cambridge University Press, 1998). 22 Frankfurt (1971), 327. 23 Harry Frankfurt, “Three Concepts of Free Action, II.” Proceedings of the Aristotelian Society, Supplementary 49: 113–25. 24 Michael Brownstein, “Attributionism and Moral Responsibility for Implicit Bias.” Review of Philosophy and Psychology 7: 767. 25 Angela Smith, “Responsibility for Attitudes: Activity and Passivity in Mental Life.” Ethics 115(2) (2005): 237. 26 Brownstein (2016), 773, quoting David Shoemaker. 27 Smith (2005), 237. 28 The first half of this paragraph is largely taken from Vicens (2022). 29 Cf. Neil Levy, who argues that “consciousness of some of the facts that give our actions their moral significance is a necessary condition for moral responsibility” in Consciousness and Moral Responsibility (Oxford: Oxford University Press, 2014), 1. 30 Brownstein (2016), 767. Brownstein similarly argues that a person’s behavior motivated by implicit bias may be “attributable” to her. 31 Interestingly, Pereboom seems to hold to a kind of consciousness condition for the appropriateness of protest. He writes, “that the agent knowingly acted wrongly is part of what makes the protest appropriate. In accord with a broad consensus, it’s the agent’s responsiveness to reasons that’s engaged in central cases of blaming, since blaming confronts its target with moral reasons. Thus, these properties would include the agent’s knowingly having acted wrongly, her being reasons responsive, and her being disposed to moral protest’s realizing the forward-looking aims” (2021, 90, italics added). But this requirement would seem to make Pereboom’s project of establishing the appropriateness of blame-as-protest hostage to empirical questions regarding the extent to which conscious conditions are in fact met. I think Pereboom should reject consciousness conditions for the appropriateness of protest, for protest sometimes aims at making an agent aware of some part of himself that needs moral reform. In personal communication, he has suggested instead that protesting someone who unknowingly does wrong would not count as blame, but instead counsel. 32 Jesse Couenhoven, Predestination: A Guide to the Perplexed (London: Bloomsbury, 2018), 93.

Can an Unfree Person Have a “Deep Self?” 85 33 Derk Pereboom, “Responsibility, Regret, and Protest” in David Shoemaker (ed.), Oxford Studies in Agency and Responsibility 4 (Oxford University Press 2017), 131. 34 Ibid. 35 Vicens (2022). 36 Can regret over one’s immoral but unfree conduct be basically deserved? I am not sure. The pain of the regret is not basically deserved. And regret, unlike protest, would not seem to have a communicative element, since it is self-directed. Regret does involve a judgment that one has done wrong. Can a judgment itself be basically deserved, or only an interpersonally communicated judgment? I leave this question for another day.

Bibliography Brownstein, Michael. “Attributionism and Moral Responsibility for Implicit Bias.” Review of Philosophy and Psychology 7 (2016): 765–86. Butler, Joseph. Fifteen Sermons Preached at Rolls Chapel, in The Works of the Right Reverend Father in God, Joseph Butler, D.C.L., Late Bishop of Durham (Samuel Halifax (ed.)). Alabama: Carter, 1846. Capes, Justin. “Hard Theological Determinism and Divine Forgiveness Are Incompatible.” in Peter Furlong & Leigh Vicens (eds.), Theological Determinism: New Perspectives. Cambridge: Cambridge University Press, pp. 165–83, 2022. Caruso, Gregg. Free Will and Consciousness: A Determinist Account of the Illusion of Free Will. Lanham: Lexington Books, 2012. Caruso, Gregg  & Stephen Morris. “Compatibilism and Retributivist Desert Moral Responsibility: On What is of Central Philosophical and Practical Importance.” Erkenntnis 82 (2017): 837–55. Couenhoven, Jesse. Predestination: A Guide for the Perplexed. London: Bloomsbury, 2018. Fischer, John Martin  & Mark Ravizza. Responsibility and Control: A  Theory of Moral Responsibility. Cambridge: Cambridge University Press, 1998. Frankfurt, Harry. “Freedom of the Will and the Concept of a Person” in Reprinted in Gary Watson (ed.), Free Will (2nd edition). Oxford: Oxford University Press, pp. 322–36, 1971/2003. Kane, Robert. A Contemporary Introduction to Free Will. Oxford: Oxford University Press, 2005. Levy, Neil. Consciousness and Moral Responsibility. Oxford: Oxford University Press, 2014. Pereboom, Derk. “Determinism al Dente.” Nous 29(1) (1995): 21–45. Pereboom, Derk. Free Will, Agency, and Meaning in Life. Oxford: Oxford University Press, 2014. Pereboom, Derk. “Forgiveness as Renunciation of Moral Protest.” in Brandon Warmke, Dana Kay Nelkin & Michael McKenna (eds.), Forgiveness and Its Moral Dimensions. Oxford: Oxford University Press, pp. 83–100, 2021. Pereboom, Derk. “Responsibility, Regret, and Protest.” in David Shoemaker (ed.), Oxford Studies in Agency and Responsibility (volume 4). Oxford: Oxford University Press, pp. 121–40, 2017. Scanlon, Thomas. “Giving Desert its Due.” Philosophical Explorations 16(2) (2013): 101–16. Smith, Angela. “Responsibility for Attitudes: Activity and Passivity in Mental Life.” Ethics 115(2) (2005): 236–71.

86  Leigh Vicens Strawson, Peter. “Freedom and Resentment” Proceedings of the British Academy 48 (1962): 1–25. Van Inwagen. An Essay on Free Will. New York: Oxford University Press, 1983. Vicens, Leigh. “Free Will Skeptics Can Have Their Basic Desert and Eat It Too.” Journal of the American Philosophical Association, 1 (2022): 1–12. Vicens, Leigh. “Human Freedom and the Inevitability of Sin.” in Peter Furlong  & Leigh Vicens (eds.), Theological Determinism: New Perspectives. Cambridge University Press, pp. 150–64, 2022. Vicens, Leigh. “Sin and the Faces of Responsibility.” in John Allan Knight  & Ian Markham (eds.), The Craft of Innovative Theology: Argument and Process. Hoboken: Wiley Blackwell, pp. 99–113, 2022. Vicens, Leigh. “Theological Determinism.” Internet Encyclopedia of Philosophy, 1014. https://iep.utm.edu/theo-det/.

5 The Semantic Case Against Open Theism and Experimental Philosophy Ferhat Yöney

Freewill Theism,1 the view that God granted human beings libertarian (indeterminist) free will, is not a uniform view. The views classified under Freewill Theism conflict with each other on which sense(s) the future is open (e.g., causal, epistemic, and alethic). An important aspect of the debate between different types of Freewill Theism is how ordinary human beings talk about the future. The issue is to clarify what ordinary users of language mean when they assert future contingent propositions. The contention in this task is to uncover whether ordinary language has any bearing on understanding in which sense(s) the future is open, and based on this, to argue that ordinary language is supportive of or undermining any type of Freewill Theism. In this chapter, first, I will summarize three types of Freewill Theism in terms of the sense(s) of the future’s openness they endorse. Second, I  will present semantic arguments proposed against Open Theism which are against the alethic and epistemic openness of the contingent future. Third, I will present some of the work done on experimental epistemology, and draw my own conclusions from these works. Fourth, based on this conclusion, I will suggest a line of argument for Open Theism against the arguments in the second section. Openness of the Future and Three Types of Freewill Theism Freewill theists diverge on what kind(s) of openness of the future they endorse. Alan Rhoda’s distinction between the different senses of the future’s openness is useful to understand the difference between different types of Freewill Theism. Three senses of the future’s openness are as follows2: Causal openness: The future is causally open at time t with respect to state of affairs X and future time t* if and only if, given all that exists as of time t, it is really possible both that X obtains at t* and that X does not obtain at t*. (In other words, whether X obtains at t* or not is, as of t, a future contingent.) Epistemic openness: The future is epistemically open for person S at time t with respect to some conceivable future state of affairs X if and only if for some future time t* neither “X will obtain at t*” nor “X will not obtain at t*” is known by S at t.3 DOI: 10.4324/9781003306191-6

88  Ferhat Yöney Alethic openness: The future is alethically open at time t with respect to some conceivable state of affairs X, and future time t* if and only if neither “X will obtain at t*” nor “X will not obtain at t*” is true at t.4 In terms of which sense(s) of the future’s openness is endorsed, three types of Freewill Theism can be classified as follows: First, a twofold classification: Open Theism and Non-Open Theist Freewill Theism. Open Theism is the view that supports epistemic openness for God to the extent that the future is causally open. Causally open parts of the future are epistemically open for God and they become epistemically settled for (known by) God when the events in question occur (or when their occurrence causally entailed by the current conditions of the universe at any time). Non-Open Theist Freewill Theism (hereafter NOFT) is the view that endorses only causal openness of the future to some extent. According to the proponents of this view, which causally possible scenario will be actualized is perfectly foreknown by God. Their affirmation of epistemic settledness entails affirmation of alethic settledness because knowledge entails truth. There are two formulations of Open Theism supposing in what sense the future is open. The first type of Open Theism is known as Limited Foreknowledge Open Theism (Hereafter LFOT). The supporters of this view endorse only the epistemic openness of the future for God but not the alethic openness for the causally open parts of the future. Accordingly, one of the propositions about the causally open parts in the form of what will and will not happen is true and the other is false. But in contrast to NOFT, these causally open but alethically settled parts of the future are not known by God. This means that there are truths (e.g., future contingents) that are not known by God. Richard Swinburne and William Hasker are supporters of LFOT.5 Another formulation of Open Theism is known as Open Future Open Theism (Hereafter OFOT).6 Supporters of this view maintain all three types of openness of the future mentioned earlier. The future is alethically open to the extent that it is causally open and it is epistemically open for God to the same extent because it is alethically open. Whole of the future is not truly describable in the form of what will happen according to this view. None of the propositions in this form for the causally open parts of the future are true and they are not known by God. The Semantic Case Against Open Theism There are two main strands of the philosophical debate between different types of Freewill Theism. One is metaphysical, which is simply about the relationship between future contingent propositions and what actually exists. It is the controversy about intelligibility of the view that some of the future contingents are true although they are not grounded by something that actually exists. The supporters of OFOT argue against ungrounded truth of future contingents. The proponents of LFOT and NOFT buttress the truth of future contingents, and this means that some of the future contingents are

The Semantic Case Against Open Theism 89 true without being metaphysically grounded by some metaphysical entity which actually exists. Another line of contention in the debate between different types of Freewill Theism is semantic. It is about how the contingent future is talked about and what is meant when the contingent future is alluded to in ordinary language. The point of disagreement in the semantic debate is whether the usage of future tense sentences buttresses or undermines any type of Freewill Theism. In this section, I will present arguments against Open Theism which rely on ordinary language. The first line of argument has implications against OFOT but not LFOT. These arguments provide examples against the view that “[i]t is impossible that the future be alethically settled in any respect in which it is causally open,”7 and bolster the compatibility of alethic settledness and causal openness. Alfred Freddoso proposed this sort of arguments. One such argument, which he calls the retrospective argument, is as follows: Suppose I predict that on the next toss the coin in your hand will come up heads. And suppose for the sake of argument that the coin’s coming up one way or the other is wholly indeterminate, so that prior to the toss the world is not tending, even nondeterministically, either toward the coin’s coming up heads or toward its coming up tails.8 So in the hypothetical case, the probability of coin’s coming neither heads nor tails is higher than 0.5. Freddoso continues as follows: Suppose, finally, that when you toss coin, it in fact comes up heads. In that case it is perfectly reasonable for me to claim that my prediction was true, that is, that I  spoke the truth in asserting beforehand the proposition The coin will come up heads. So it is reasonable for me to maintain that this proposition was true before you tossed the coin.9 This argument concludes that in ordinary language we often speak in a way in which we retrospectively attribute truth to some sentences concerning the contingent future. This provides support for the view that the prior truth of a proposition about the future does not eliminate the causal indeterminacy of the event in question. Similar conclusion can be repeated, Freddoso argues, if the proposition in question was a conditional one If you tossed the coin, it would come up heads.10 Against this argument, Rhoda et al. argue, “[t]he flexibility of colloquial usage nearly always leaves the meaning of statements underdetermined, allowing for more than one plausible interpretation of colloquial linguistic behaviour”11 and Rhoda objects that [I]t does not follow from the fact that English-speakers sometimes retrospectively apply words like “true” to sentences like “X will obtain” (after it has become clear that X has obtained) that the proposition expressed by that sentence was antecedently true.12

90  Ferhat Yöney Instead, they argue that that assertion is not a real prediction about the future. To find out the correct semantic content of the assertion in Freddoso’s example, and justify their objection to his derivation of antecedent truth from the predication of truth, they apply the principle of rational assertibility to the assertion according to which “a person’s claims ought to be interpreted, if the semantic flexibility of his words and the context allow, in a manner that preserves the rational assertibility of these claims.”13 They ask the question about the assertion The coin will come up heads: “Why does the speaker believe this?” and argue, “[t]he most plausible answer . . . is that speaker, for whatever reason, thinks the coin’s landing heads is somewhat more probable than its leading tails.”14 But it is not the case in Freddoso’s hypothetical case. The coin’s landing heads is not probable. From this, Rhoda et al. object that: [T]he statement is not obviously a genuine prediction at all. In saying “The coin will land heads,” the speaker is likely just making an arbitrary choice between heads and tails and has no real conviction either way.15 They argue that correct construal of the assertion “is more like ‘I choose heads,’ an autobiographical claim about the speaker, than it is like ‘I believe the coin will land heads.’ ”16 This objection assumes general applicability of the principle of rational assertibility. If it is applied to Freddoso’s retrospective argument, it is successful at rebutting that argument. I will proceed by assuming the principle’s general applicability and plausibility of it and move on to some other arguments against alethic openness, which rely on ordinary language, with implications against OFOT. Another line of argument involves predictions about contingent future (especially free acts of human beings) where the prediction is probable. Freddoso is among the ones who proposed this kind of argument. The argument which he calls the prospective argument is as follows: If I know that you promised to meet me for dinner at 6:00 PM. and know further that you are a very reliable person, then I am highly justified in asserting that you will in fact meet me at 6:00 P.M. To be sure, my evidence for this assertion consists in my beliefs about your character and in other beliefs about the world’s present tendencies.17 What Freddoso points out is that although the probability on speaker’s side is more than 0.5 (in virtue of his evidence) but below 1.0 (so the agent in question is free to retract from his promise), what is asserted is not a qualified or probabilistic sentence You will probably meet me at 6:00 P.M. But what is asserted is an unqualified or nonmodal sentence You will meet me at 6:00 P.M.18 This means that in ordinary language, we speak about the contingent future as if it was true antecedently at the time of assertion.

The Semantic Case Against Open Theism 91 Swinburne (who is a proponent of LFOT) also deals with the proposal that future contingent propositions are not true until the event in question occurs (or causally necessitated in due course) and finds this unappealing because “[i]n normal use, propositions about a named future time (including claims about any future free actions) are true or false – timelessly.”19 His insistence on using truth in normal use is related to his views on how to use ordinary words (in normal sense or in analogical sense) in theological discourse. If theology uses too many words in analogical senses it will convey virtually nothing by what it says. The claim that he is using a word in an analogical sense must be for the theist a last resort to save his system from a charge of incoherence which would otherwise stick. He must claim that he is using many words in ordinary senses.20 He is reluctant to pursue this point on omniscience and deems it unnecessary and argues that “it would be less misleading to use such an ordinary word as ‘true’ in its normal sense.”21 So, according to Swinburne, if OFOT’s understanding of truth about future contingents is adopted, it will have the effect that the gap between ordinary language and religious language widens. Nevertheless, the supporters of LFOT have less antipathy against the non-truth of all future contingents than the proponents of NOFT. For example, Hasker points out that our ordinary usage of the words does not clearly settle the issue and supposes that “we sometimes say of a prediction that it has ‘come true,’ which is not quite the same as saying that it was true all along.”22 Hasker’s suggestion seems a plausible construal of truth attribution to a prediction within the framework of alethic openness which is not unnatural for ordinary language. But the point is that there is prevalent attribution of truth either retrospectively or prospectively to future contingent propositions, and this fact stands against alethic openness. Similar kind of arguments, which have implications supportive of NOFT (especially Molinism) and against OFOT, involves conditional claims about how human beings would act in certain conditions. These kinds of propositions are called counterfactuals of creaturely freedom which describes what a specific person would freely do if he had been in certain conditions. So once the truth of these kinds of propositions is maintained, it becomes far easier for a proponent of NOFT to vindicate the truth of future contingents. Freddoso proposes a modified version of his prospective argument mentioned previously, this time for the truth of the counterfactuals of creaturely freedom.23 The opponents of Molinism metaphysical concerns that unless the antecedent of the conditional causally entails its consequent (which hinders act’s being free) what is true is If you promised to meet me at 6:00 P.M., you would probably meet me at 6:00 P.M. but not If you promised to meet me at 6:00 P.M., you would meet me at 6:00 P.M. Otherwise the truth of it will be

92  Ferhat Yöney metaphysically ungrounded.24 But this does not work in the semantic debate because, as Freddoso points out, what is spoken as if it is true at the time of assertion is not the former one (the probabilistic proposition) but the latter (non-probabilistic proposition).25 According to these arguments, there is a prima facie case against the view that any part of the future cannot be alethically settled while it is causally open. This is because both arguments (the one concerning the use of future contingents and the other, counterfactuals of creaturely freedom) support the truth of future contingents, which is against OFOT. These arguments constitute the first line of semantic arguments against Open Theism. The second line of argument against Open Theism, which relies on ordinary language, has implications not only against OFOT but against Open Theism in general as well because they are about knowledge. For example, Jonathan Kvanvig notes that human beings surely seem to have the future as the object of their intentional attitudes. We worry over it, hope for it, ponder it, and plan for it. We also seem to know quite a bit about it. To take an example, I surely seem to know that I will go home today at three o’clock, just as I seem to know that I went home yesterday at three o’clock.26 Similarly, William Lane Craig, with regard to the counterfactuals of freedom, tells that “if I  were to offer my wife a plate of liver and onions and a plate of chocolate-chip cookies, I  know which one she would choose as certainly as I know almost anything!” but then he asks “why could not God know us so completely that he knows what we would freely do under any circumstances?”27 Again, it is not so hard to see the move, first, from the fact that humans seem to know counterfactuals of freedom to God’s knowledge of them; then, second, from God’s knowledge of counterfactuals of freedom to God’s knowledge of future contingents. Accordingly, incompatibilists on divine foreknowledge and free will (either they are open theists or not) were careful to distinguish divine foreknowledge from human foreknowledge. For instance, Nelson Pike tells that Smith’s foreknowing an act of Jones, who is an intimate friend of him, does not have similar implications to that of divine foreknowledge of very same act of Jones.28 Hence incompatibilists embraced the asymmetry thesis between divine foreknowledge and human foreknowledge that “God’s foreknowledge undermines human freedom in a way in which human foreknowledge does not.”29 But Craig, objecting this asymmetry, argued that: If Smith really knows that Jones will mow his lawn this Saturday, then he cannot be mistaken; otherwise he would only believe it. So if Jones refrains, then Smith did not foreknow that Jones would mow the lawn; he just mistakenly believed it. But if we suppose that Smith really does know that Jones will mow his lawn on Saturday, then Jones

The Semantic Case Against Open Theism 93 must mow his lawn on Saturday, for if he can refrain, then he has the power to bring about that Smith’s true belief is also false, which is self-contradictory.30 A theological implication of this is the case that there seems to be truths some human beings know but God, who is a cognitively perfect being, doesn’t know. For example, Smith foreknows that act of Jones, who is an intimate friend of him, but God doesn’t know beforehand the very same act of Jones. These kinds of cases create difficulties for Open Theism in general (for both LFOT and OFOT). The supporters of LFOT counter the charge that if there are some truths that God does not know, then God is not omniscient by claiming that there are truths that are not knowable. But this move is not helpful against the charge that stems from the type of the aforementioned cases. In contrast to LFOT’s supporters, these kinds of cases bolster the view that future contingents are knowable by human beings. So there is an argument against Open Theism which relies on ordinary language. These two types of arguments against Open Theism show that how we talk about the future is supportive of NOFT; thus, there exists a prima facie semantic argument against Open Theism. Aforementioned response to Freddoso’s hypothetical case of coin tossing, which involves asserting the flexibility of ordinary language, can be extended to other cases of truth and knowledge attribution in which occurrence of the future events in question is probable. This would mean that the semantic case against Open Theism is not a conclusive argument and I would agree with it. But in my opinion, to be content with claiming that the meanings of the sentences remain undetermined in ordinary usage and just asserting that the usage of them is not precise is not a satisfactory way to respond to the semantic case against Open Theism. This is the case because intuitions, which are our spontaneous responses to hypothetical or real cases, which do not require any cognitive process, are indispensable and the primary inputs for philosophical investigation. These intuitions are to some extent reflected in ordinary language by our use of certain concepts. Therefore, open theists should find a different way to meet these semantic challenges, that is to say, independent grounds to justify the claim that although ordinary language seems against Open Theism, the said cases above which involve truth and knowledge attribution to future contingents need not pose a real challenge against their view. Experimental Philosophy and Epistemic Intuitions In recent years, a new area of research called experimental philosophy has emerged. It is at the intersection of philosophy on the one side and psychology and cognitive science on the other. In the studies of experimental philosophy, how human beings intuit in response to some philosophically relevant cases are tested and measured by methods of psychology and cognitive science.

94  Ferhat Yöney Some of the work conducted in experimental philosophy tries to find out how we intuit in epistemically relevant cases. Jonathan M. Weinberg, Shaun Nichols, and Stephen Stich’s article, which is published in 2001, is at the starting point of epistemologically relevant work in experimental philosophy. In this work, they propose four hypotheses so as to be experimentally tested in order to call into question whether there is a universal core shared by all people about how they intuit in response to epistemically relevant cases. The following are the four hypotheses they presented, and they tested first two of them in that article and last two in subsequent research31: Hypothesis 1: Epistemic intuitions vary from culture to culture. Hypothesis 2: Epistemic intuitions vary from one socioeconomic group to another. Hypothesis 3: Epistemic intuitions vary as a function of how many philosophy courses a person has had. Hypothesis 4: Epistemic intuitions depend, in part, on the order in which cases are presented.32 While the findings presented by Weinberg et  al. support Hypothesis 1 and Hypothesis 2, in recent research, which is carried out with similar methodology but larger samples than the original study of Weinberg et  al., the divergence between different cultural groups and socioeconomic groups is not replicated. There did not appear to be any statistically significant divergence between any of the groups. Instead, an important degree of overlap is observed in the following research.33 With regard to Hypothesis 4, Swain et  al. tested whether our epistemic intuitions about the Truetemp Case, which is proposed as a case against reliabilism, vary depending on which kind of case presented to the participant beforehand.34 One day Charles was knocked out by a falling rock; as a result his brain was “rewired” so that he is always right whenever he estimates the temperature where he is. Charles is unaware that his brain has been altered in this way. A few weeks later, this brain rewiring leads him to believe that it is 71 degrees in his room. Apart from his estimation, he has no other reasons to think that it is 71 degrees. In fact, it is 71 degrees. Please indicate to what extent you agree or disagree with the following claim: “Charles knows that it is 71 degrees in his room.”35 Together with the Truetemp Case, three other thought experiments were presented to the participants. These four thought experiments are ordered in eight different ways and one of them is randomly assigned to each participant out of these eight combinations. One of the three cases presented to participants apart from Truetemp Case is a clear case of non-knowledge, where a

The Semantic Case Against Open Theism 95 man who believes he can predict when a coin will land heads but his predictions are no better than chance. But in that case he gets it right. Another case is a clear case of knowledge where a chemist’s belief relies on her testimony of reading an article in a leading scientific journal and her belief is in fact true. Participants were asked to rate a statement attributing knowledge at the end of each of the four thought experiments (1: strongly disagree, 2: disagree, 3: neutral, 4: agree, 5: strongly agree).36 Swain et al. compared the responses of the participants who receive the Truetemp Case first with (1) the responses of the participants who are first presented with a clear case of knowledge then the Truetemp Case and (2) the responses of the participants who are first presented with a clear case of nonknowledge then the Truetemp Case. While willingness of the participants to attribute knowledge to the Truetemp Case in one of the sequences in which they were presented in the Truetemp Case first is with a mean of 2.64, willingness of participants to attribute knowledge to the Truetemp Case who were presented in Truetemp Case just after a clear case of non-knowledge is 3.31. This, according to Swain et al., shows that participants who are first presented with a clear case of non-knowledge are more willing to attribute knowledge to the Truetemp Case than participants who encounter the Truetemp Case before any other case. Similarly, participants who are first presented a clear case of knowledge just before the Truetemp Case are less willing (with a mean of 2.41) to attribute knowledge to him than the participants who were presented one of the sequences in which the Truetemp Case is presented first (with a mean of 3.00).37 From these observations, authors espouse caution about epistemic intuitions and suggest that intuition-deploying philosophers need to recognize that if their intuitions are sensitive to variables irrelevant to the issues thoughtexperiments are designed to address, then they are ill-suited to do the work philosophers ask of them.38 Jennifer Cole Wright’s research39 is important particularly with regard to Hypothesis 4, and also in relation to the most plausible general conclusion to be drawn about the role and reliability of epistemic intuitions from the research in experimental philosophy. She presented four cases of Swain et al. to the participants in different orders. Differing from Swain et al., she asked the participants for each case whether the specific proposition in the case is known or not (YES or NO), and asked them to rate how confident they were about their answer (0 = not very confident to 5 = very confident).40 With respect to the participants’ willingness to attribute knowledge in the Truetemp Case, the order effect is also observed in Wright’s study. Participants were significantly more willing to attribute knowledge to the case when it immediately followed a clear case of non-knowledge (55%) than when it immediately followed clear case of knowledge (40%). However, susceptibility to the order effect is not observed for clear cases of knowledge and

96  Ferhat Yöney clear cases of non-knowledge. In clear case of knowledge, participants were equally likely to attribute knowledge regardless of which case immediately preceded (79%–84%). Similarly, in the case of clear non-knowledge, participants rarely attribute knowledge (0%–6%) regardless of which case immediately preceded.41 This means that participants’ intuitions are stable in clear cases of knowledge and non-knowledge but unstable in the Truetemp Case where intuitions seem dependent on order effect. Moreover, Wright finds out that participants “themselves tracked this intuitional stability.”42 Participants were significantly more confident in their judgments in clear cases of knowledge (with a mean of 4.5) and in clear cases of non-knowledge (with a mean of 4.4) than they are in the Truetemp Case (with a mean of 3.9). Higher confidence for clear cases of knowledge and clear cases of non-knowledge and lower confidence for the Truetemp Case are also valid regardless of the order in which the cases were presented.43 Wright’s research is important to alleviate Swain et al.’s worry about the usage of intuitions in philosophy to find universal criteria for epistemic concepts: Our judgments about [clear cases] . . . are unlikely to change. Of course, philosophy is often most interesting (and of most value) when it is working “at the margins,” wrestling with unclear and borderline cases.44 But also encouraging and instructing: [P]hilosophers clearly can (and do) have clear/strong intuitions about non-paradigmatic cases. Such cases, while perhaps vulnerable to bias for the general population of reasonable conceptually competent people . . . may nonetheless be stable for most philosophers. After all, philosophers receive extensive training designed specifically to refine and enhance their conceptual mastery.45 Findings of some other research in experimental philosophy that involves Gettier cases in comparison with clear cases of knowledge and non-knowledge are somehow parallel to Wright’s interpretation but they also extend its scope. Christina Starmans and Ori Friedman’s work46 is the first of these that I will mention. Their work includes separate experiments. In one of them, 144 participants are randomly assigned either to a Gettier case, a control case which involves a clear case of knowledge, or a false belief case, which is a clear case of non-knowledge. Participants were asked a knowledge question whether they attribute knowledge in the case presented to them, with choices “really knows” and “only believes,” and a confidence question in their judgment, scaled from 0 to 10. To find out the interval for the participants’ average confidence in each case, the confidence rate is multiplied by +1 if the participant attributed knowledge, and multiplied by −1 if the participant did not attribute knowledge. Participants attributed knowledge with 88%, 72%, and 11% to control

The Semantic Case Against Open Theism 97 case, Gettier case, and false belief case, respectively. Participants who were presented control case, Gettier case, and false belief case made their judgment with an average confidence of 7.27, 4.53, and −7.13, respectively.47 Two other studies on how ordinary people respond to Gettier cases in comparison with clear cases of knowledge and non-knowledge are the ones conducted by Nagel et al. and Machery et al., and their results are similar to some extent to the results of Starmans and Friedman’s work. Participants in these researches are also more willing to attribute knowledge to the clear cases of knowledge than they are willing to attribute knowledge to Gettier cases, and again participants are less willing to attribute knowledge to the clear cases of non-knowledge than they are willing to attribute knowledge to Gettier cases. These comparisons hold for their justification in attributing knowledge to three types of cases. Differing from the results of Starmans and Friedman’s work, participants in the researches by Nagel et al. and Machery et al. attribute knowledge to Gettier cases with a lesser percentage (changing from 35.4% to 11.2%) than participants in the research of Starmans and Friedman.48 These empirical studies on Gettier cases, which involve comparisons with clear cases of knowledge and non-knowledge, confirm and also extend the scope of Wright’s interpretation mentioned earlier. It seems that not only in cases which are non-paradigmatic or non-clear cases of knowledge and nonknowledge, ordinary agents’ epistemic intuitions are not stable and doubtful; they are not reliable in Gettier cases which are clear cases of non-knowledge according to philosophers. This is because, in my point of view, Gettier cases though are not knowledge, they are in a place very close to being knowledge. So, it seems to me that the most plausible interpretation of the output of epistemically relevant experimental philosophy is as follows: (1) There is a universal core for epistemic notions among agents which does not vary drastically from culture to culture or according to socioeconomic status, and stable and reliable for clear cases of knowledge and non-knowledge; (2) epistemic intuitions of ordinary agents and how they use it in ordinary language are not stable and reliable (though not totally misguided) in cases that are not clear cases of knowledge or non-knowledge which are borderline cases of knowledge and non-knowledge; and (3) in borderline cases, intuitions of ordinary agents are not stable and reliable but they are not totally misguided since agents are able to track this instability. The Relevance of Experimental Philosophy to Future Contingents and to the Problem of Divine Foreknowledge and Free Will What is the relevance of these empirical studies and the epistemic status (truth and knowledge) to future contingents and to the problem of divine foreknowledge and human free will? I suggest that ordinary agents’ beliefs that some future contingent events will occur are not paradigmatic or clear cases of knowledge and we have good reasons to suppose so. The first reason

98  Ferhat Yöney is that the truth value of future contingent propositions is philosophically disputed that threads back many centuries. Aristotle was the one who endorsed the non-truth of future contingents. This is important both for philosophers’ perspective and for how ordinary agents who use the word “true” in relation to knowledge. From philosophers’ perspective, if a proposition is not true, this entails failure of knowledge. This is also valid for non-philosophers according to the empirical work mentioned earlier. Non-philosophers also are unwilling to attribute knowledge to cases, which involve false belief. So the philosophical controversy on truth value of future contingents undermines their being clear or paradigmatic cases of knowledge despite their usage as an object of knowledge in ordinary language. Second reason I will propose, loosely related to the first one, stems from the results of how the participants responded to Gettier cases. In almost all the studies in experimental philosophy, the participants were more willing to attribute knowledge to the Gettier cases than they were to the clear cases of non-knowledge, in a statistically significant fashion.49 However, Gettier cases are clear cases of non-knowledge from the philosopher’s perspective. At least some of the ordinary agent’s willingness to attribute knowledge to the Gettier cases undermines their position to be a clear case of non-knowledge although they are clear cases of non-knowledge for philosophers. If ordinary agents’ epistemic intuitions are not reliable in Gettier cases, which are deemed by philosophers as cases of non-knowledge without hesitance, we are more justified to think that future contingents are not clear cases of knowledge given the philosophical controversy on their truth value. This constitutes the second reason not to take ordinary agents’ beliefs that some future contingent events will occur as clear cases of knowledge although in ordinary language human beings talk about them as if they were something they know. Here, I don’t assert the claim that philosophers are the relevant experts in the sense that if they think that the status of knowledge claims in this domain is unclear, then their status is unclear (and similarly, presumably, if they think there is clear knowledge here, then there is) which will make any appeal to intuitions and appeal to ordinary language redundant. This criticism would be valid if my conclusion on relevant work on experimental philosophy entirely relied on the fact that it is uncontroversial for philosophers that Gettier cases do not count as knowledge, but they are deemed knowledge by a significant percentage of the ordinary agents. However, Gettier cases are only a supplement to the results of Wright’s research that though there is intuitional instability in Truetemp cases, which cast doubt to the reliability of these intuitions, this intuitional stability is tracked by the agents and both considerations rely entirely on empirical results. I mean that intuitions and the employment of relevant concepts in ordinary language have indispensable use in any philosophical problem, but if intuitions and the employment of relevant concepts in ordinary language in a specific philosophical problem are shown to be unreliable and unstable, this urges us to approach them with caution. However, this does not mean that philosophers will easily agree on the philosophical problem in question

The Semantic Case Against Open Theism 99 and solve it. I  suggest a similar approach for future contingents, based on my conclusion of the previous section, but this will not mean that the philosophical dispute on the status of future contingents can be easily settled by the philosophers even when ordinary language considerations are ignored. If these considerations are good, open theists can argue against the charge that some of the future contingents are used in ordinary language in a way that they are known by us in the following way: even if some of the future contingents are true and known by any agent, God, or another agent, these propositions are not clear cases of knowledge. If future contingents are not clear cases of knowledge, the way they are spoken of in ordinary language does not pose a challenge against Open Theism. This argument does neither argue for future contingents’ not being true, as OFOT would do, nor suggest that even if they are true they are not knowable, as LFOT would do. Instead, I suggest a way out for open theists with this argument against the charge that considerations that stem from ordinary language contradict with their view. The argument I  proposed so far is effective as a response to the second line of semantic argument against Open Theism, which supports future contingents as objects of knowledge. As mentioned in Section  2, this type of semantic argument is effective against any type of Open Theism, both OFOT and LFOT, because it is about knowledge but not about truth. Similarly, experimental studies mentioned earlier are mostly about agents’ knowledge attribution, along with justification in some of them, but none of the studies are about agents’ truth attribution to the relevant propositions. This makes it harder and more indirect to apply them to the first type of semantic arguments against Open Theism, which concerns the truth of future contingents. However, I think, the defense I suggested earlier for open theists can almost equally be applicable to the first line of semantic argument against Open Theism except the retrospective truth attributions.50 With respect to the future events, which were spoken of in ordinary language that are probable from the speaker’s point of view but still contingent, truth attribution to that proposition by the speaker and knowledge attribution to the speaker are closely related to each other. Concerning these events, which are probable, what makes us attribute truth to them is our beliefs about the conditions and tendencies of the world at the time of assertion which also makes our relevant beliefs justified, whether these beliefs are true or not. Once this truth attribution comes together with belief and justification on the speaker’s side, this results in knowledge attribution, whether the speaker really knows or not. So it is reasonable to claim that once the knowledge of a future contingent event is attributed to an agent in ordinary language, that agent also attributes truth to the relevant future contingent proposition and vice versa. Lastly, I will mention some applications of the way I proposed to respond to the semantic case against Open Theism. The first application is about Swinburne’s concern that not to use the word “truth” in a different way in theological discourse than it is used in ordinary discourse. If the defense I suggested is applied to future contingents on behalf of Open Theism, Swinburne’s

100  Ferhat Yöney concern would be unwarranted. The difference between the meanings of the words “know” and “true” in ordinary language and the usages of them when Open Theism is embraced does not widen the gap between ordinary language and religious language. Instead, when they appear in non-clear cases, fundamental terms of epistemology, like justification, truth, and knowledge, are used in ordinary language loosely and sometimes incorrectly. So for these non-clear cases, there already exists a gap in the usages of fundamental terms of epistemology between the ordinary language and the language of epistemologists. The gap between the ordinary language and the language of Open Theism in the uses of words “know” and “true” is a transfer of the original gap between the ordinary language and the language of epistemologists. (I think that Swinburne’s worry is valid also for knowledge attribution, and my proposal helps for that also.) The second application is about the asymmetry thesis between divine and human foreknowledge and the cases mentioned in the third section. Given that my proposal is utilized, asymmetry thesis need not be endorsed by open theists. Once Open Theism is adopted, this requires tweaking, which is not drastic to my mind, when epistemologically relevant terms are used about future. If the aforementioned future act of Jones is really contingent, then Smith is justified given his evidence in believing that Jones will act in that way, but he does not know whether Jones will act in that way. This means to replace one fundamental term of epistemology with another. If this adjustment is problematic, this is not a problem peculiar to Open Theism. Instead, the problem is more fundamental and prevalent along all the non-clear cases of knowledge and non-knowledge. Conclusion The semantic arguments, which rely on ordinary language, form an aspect of the contention between different types of Freewill Theism in the recent literature on divine foreknowledge and free will. According to this, semantic considerations are against Open Theism and constitute a prima facie case against it. Taking recent work in epistemologically relevant experimental philosophy into consideration, I argued that the most plausible conclusion about epistemic intuitions derived as such is quite moderate that while in clear cases they are reliable, in non-clear cases they are unstable and they should be approached with caution. Given that conclusion, I claimed that we should also approach with caution the way we use future contingents in ordinary language. I think that this constitutes a good line of defense for open theists against semantic considerations.51 Notes 1 I followed David Basinger, The Case for Freewill Theism: A Philosophical Assessment (Illinois: InterVarsity Press, 1996) in using this as an inclusive label for the views which affirm both theism and libertarian free will of human beings. 2 Here, I am assuming the truth of A-theory of time.

The Semantic Case Against Open Theism 101 3 Alan R. Rhoda, “Generic open theism.” Religious Studies 44(2) (2008): 227. 4 Rhoda, Generic open theism, 229. 5 Richard Swinburne, The Coherence of Theism (Oxford: Oxford University Press, 1977) and William Hasker, God, Time, and Knowledge (Ithaca: Cornell University Press, 1989). 6 I borrowed these two labels and abbreviations (LFOT and OFOT) from Benjamin H. Arbour, “Future Freedom and the Fixity of Truth: Closing the Road to Limited Knowledge Open Theism.” International Journal for Philosophy of Religion 73(3) (2013): 189–207. 7 Alan Rhoda, “The Philosophical Case for Open Theism.” Philosophia 35(3–4) (2007): 303. 8 Alfred J. Freddoso, “Introduction”, in Luis De Molina (ed.), On Divine Foreknowledge: Part IV of the Concordia (Trans. A. J. Freddoso) (Ithaca: Cornell University Press, 1988), 71. 9 Freddoso, “Introduction,” 71. 10 Freddoso, “Introduction,” 73. 11 Alan R. Rhoda, Gregory A. Boyd & Thomas G. Belt, “Open Theism, Omniscience, and the Nature of the Future.” Faith and Philosophy 23(4) (2006): 444. 12 Rhoda, “The Philosophical Case,” 308. 13 Rhoda et al., “Open Theism, Omniscience,” 442. 14 Rhoda et al., “Open Theism, Omniscience,” 445. 15 Rhoda et al., “Open Theism, Omniscience,” 445. 16 Rhoda et al., “Open Theism, Omniscience,” 445. 17 Freddoso, “Introduction,” 71–72. 18 Freddoso, “Introduction,” 72. 19 Swinburne, The Coherence, 175. 20 Swinburne, The Coherence, 70. 21 Swinburne, The Coherence, 175. 22 William Hasker, “Foreknowledge and Necessity.” Faith and Philosophy 2(2) (1985): 127. See also Richard Swinburne, The Coherence of Theism, 2nd edition. (Oxford: Oxford University Press, 2016), 191. Recently, Hasker abandoned LFOT and embraced OFOT. See William Hasker, “Future truth and freedom.” International Journal for Philosophy of Religion 90(2) (2021): 109–19. 23 Freddoso, “Introduction,” 73. 24 Robert Merrihew Adams, “Middle Knowledge and the Problem of Evil.” American Philosophical Quarterly 14(2) (1977): 110–11. 25 Freddoso, “Introduction,” 73. See also Thomas P. Flint, Divine Providence: The Molinist Account (Ithaca: Cornell University Press, 1998), 78 and Jonathan L. Kvanvig, The Possibility of an All-Knowing God (New York: Palgrave McMillan, 1986), 138. 26 Kvanvig, The Possibility, 2. 27 William Lane Craig, The Only Wise God: The Compatibility of Divine Foreknowledge and Human Freedom (Eugene OR: Wipf and Stock Publishers, 2000/1987), 139, 145. 28 Nelson Pike, “Divine Omniscience and Voluntary Action.” The Philosophical Review 74(1) (1965): 40–45. 29 John Martin Fischer, “Freedom and Foreknowledge.” The Philosophical Review 92(1) (1983): 76. See also Hasker, “Foreknowledge and Necessity,” 124–5. 30 Craig, The Only Wise God, 53–54. 31 Jonathan M. Weinberg, Shaun Nichols & Stephen Stich, “Normativity and Epistemic Intuitions.” Philosophical Topics 29(1–2) (2001): 429–60. 32 Weinberg et al., “Normativity,” 437–38. 33 Jennifer Nagel, Valeria San Juan & Raymond A. Mar, “Lay denial of knowledge for justified true beliefs,” Cognition 129(3) (2013): 658; Hamid Seyedsayamdost, “On Normativity and Epistemic Intuitions: Failure of Replication.” Episteme

102  Ferhat Yöney 12(1) (2015): 95–116; Minsun Kim & Yuan, “No Cross-Cultural Differences in the Gettier Car Case Intuition: A  Replication Study of Weinberg et  al. 2001.” Episteme 12(3) (2015): 355–61; Edouard Machery, Stephen Stich, David Rose, Amita Chatterjee, Kaori Karasawa, Noel Struchiner, Smita Sirker, Naoki Usui & Takaaki Hashimoto, “Gettier Across Cultures.” Nous 51(3) (2017): 645–64. 34 Stacey Swain, Joshua Alexander  & Jonathan M. Weinberg, “The Instability of Philosophical Intuitions: Running Hot and Cold on Truetemp.” Philosophy and Phenomenological Research 76(1) (2008): 138–55. 35 Swain et al., “The Instability of Philosophical,” 154. 36 Swain et al., “The Instability of Philosophical,” 141–43, 154–55. 37 Swain et al., “The Instability of Philosophical,” 144–45. 38 Swain et al., “The Instability of Philosophical,” 153. 39 Jennifer Cole Wright, “On intuitional stability: The clear, the strong, and the paradigmatic.” Cognition 115(3) (2010): 491–503. 40 Wright, “On intuitional stability,” 492. 41 Wright, “On intuitional stability,” 492–4. 42 Wright, “On intuitional stability,” 493. 43 Wright, “On intuitional stability,” 493–94. 44 Wright, “On intuitional stability,” 500. 45 Wright, “On intuitional stability,” 500. 46 Christina Starmans and Ori Friedman, “The folk conception of knowledge.” Cognition 124(3) (2012): 272–83. 47 Starmans and Friedman, “The folk conception,” 274–76. In one of the separate experiments similar results were replicated by Starmans and Friedman 276–77. 48 Nagel et  al., “Lay denial of knowledge”, and Machery et  al., “Gettier Across Cultures.” 49 Similarly, they are less willing to attribute knowledge to Gettier cases than they are to the clear cases of knowledge. 50 In retrospective truth attributions, rational assertibility could be utilized. 51 This chapter was written during the postdoctoral research period at the University of Leeds in 2019 under the supervision of Mark Wynn which was supported by the Scientific and Technological Research Council of Turkey (Türkiye Bilimsel ve Teknolojik Araştırma Kurumu – TÜBITAK) [Grant Number 1059B191700516]. I would like to thank Mark Wynn and Patrick Todd for their comments on the earlier drafts of the chapter.

Bibliography Adams, Robert Merrihew. “Middle Knowledge and the Problem of Evil.” American Philosophical Quarterly 14(2) (1977): 109–17. Arbour, Benjamin H. “Future freedom and the fixity of truth: Closing the road to limited foreknowledge open theism.” international Journal for Philosophy of Religion 73(3) (2013): 189–207. Basinger, David. The Case for Freewill Theism: A Philosophical Assessment. Chicago: InterVarsity Press, 1996. Craig, William Lane. The Only Wise God: The Compatibility of Divine Foreknowledge and Human Freedom. Eugene: Wipf and Stock Publishers, 2000/1987. Fischer, John Martin. “Freedom and Foreknowledge.” The Philosophical Review 92(1) (1983): 67–79. Flint, Thomas P. Divine Providence: The Molinist Account. Ithaca: Cornell University Press, 1998.

The Semantic Case Against Open Theism 103 Freddoso, Alfred J. “Introduction” in Luis De Molina, On Divine Foreknowledge: Part IV of the Concordia (trans. Alfred J. Freddoso). Ithaca: Cornell University Press, pp. 1–81, 1988. Hasker, William. “Foreknowledge and Necessity.” Faith and Philosophy 2(2) (1985): 121–57. ———. God, Time, and Knowledge. Ithaca: Cornell University Press, 1989. ———. “Future truth and freedom.” international Journal for Philosophy of Religion 90(2) (2021): 109–19. Kim, Minsun & Yuan Yuan. “No Cross-Cultural Differences in the Gettier Car Case Intuition: A Replication Study of Weinberg et al. 2001.” Episteme 12(3) (2015): 355–61. Kvanvig, Jonathan L. The Possibility of an All-Knowing God. New York: Palgrave Macmillan, 1986. Machery, Edouard, Stephen Stich, David Rose, Amita Chatterjee, Kaori Karasawa, Noel Struchiner, Smita Sirker, Naoki Usui & Takaaki Hashimoto. “Gettier Across Cultures.” Nous 51(3) (2017): 645–64. Nagel, Jennifer, Valeria San Juan & Raymond A. Mar. “Lay Denial of Knowledge for Justified True Beliefs.” Cognition 129(3) (2013): 652–61. Pike, Nelson. “Divine Omniscience and Voluntary Action.” The Philosophical Review 74(1) (1965): 27–46. Rhoda, Alan. “Generic Open Theism and Some Varieties Thereof.” Religious Studies 44(2) (2008): 225–34. ———. “The Philosophical Case for Open Theism.” Philosophia 35(3–4) (2007): 301–11. Rhoda, Alan R., Gregory A. Boyd & Thomas G. Belt. “Open Theism, Omniscience, and the Nature of the Future.” Faith and Philosophy 23(4) (2006): 432–59. Seyedsayamdost, Hamid. “On Normativity and Epistemic Intuitions: Failure of Replication.” Episteme 12(1) (2015): 95–116. Starmans, Christina & Ori Friedman. “The Folk Conception of Knowledge.” Cognition 124(3) (2012): 272–83. Swain, Stacey, Joshua Alexander & Jonathan M. Weinberg. “The Instability of Philosophical Intuitions: Running Hot and Cold on Truetemp.” Philosophy and Phenomenological Research 76(1) (2008): 138–55. Swinburne, Richard. The Coherence of Theism. Oxford: Oxford University Press, 1977. ———. The Coherence of Theism (2nd edition). Oxford: Oxford University Press, 2016. Weinberg, Jonathan M., Shaun Nichols & Stephen Stich. “Normativity and Epistemic Intuitions.” Philosophical Topics 29(1–2) (2001): 429–60. Wright, Jennifer Cole. “On intuitional Stability: The Clear, the Strong, and the Paradigmatic.” Cognition 115(3) (2010): 491–503.

6 Freedom, Even If God Decrees It Fr. James Dominic Rooney, OP

W. Matthews Grant1 holds that “there is no incompatibility between our acts being caused by God and their being free in the libertarian sense.”2 A typical view of that incompatibility between freedom and “divine causality” comes from Plantinga: If I  am free with respect to an action A, then God does not bring it about or cause it to be the case either that I take or that I refrain from this action. . . . For [if He does] then I am not free to refrain from A, in which case I am not free with respect to A.3 Grant argues, instead, that it is possible to reconcile God’s causal sovereignty with libertarian freedom by denying an assumption that God causes by means of some factor, property, or decree intrinsic to Himself. (I will refer these as “divine decrees.”) Grant calls this assumption the Intrinsic Model (IM) of divine agency.4 IM will (Grant claims) inevitably lead to God’s causality determining an action because IM introduces factors (decrees) that are “logically sufficient” to cause the creature’s act. Grant proposes that a model of divine agency on which God causes the free actions of His creatures in virtue of no intrinsic property or decree, the Extrinsic Model (EM), avoids this problem. On EM, since there are no intrinsic properties or divine decrees involved in the process, God’s causality does not introduce some factor that is both prior to and logically sufficient for bringing about the creature’s action. Further, God’s act and the creature’s act are “simultaneous necessary conditions” for each other.5 Grant holds that this relationship between God’s causing the act and the act’s occurrence preserves created agents’ ability to do otherwise and ultimate responsibility for their acts, given that the creature has “counterfactual control” over what God does (i.e., if the creature did not choose to act, God would not cause the act).6 EM is thus compatible with libertarian freedom, Grant proposes, as it gives creatures full counterfactual control over God’s causal activity. Pace Grant, I take two classical theories of grace associated with Domingo Banez and Luis de Molina and argue that these are exempt from Grant’s attack on intrinsic models of divine causality. The classical theories of grace DOI: 10.4324/9781003306191-7

Freedom, Even If God Decrees It 105 can remain libertarian in just the same sense Grant’s account is. Grant proposes that God’s causal activity and the free act of the creature are mutually necessary and sufficient for each other’s occurrence. Yet a divine decree can be understood to bring about a creaturely act under the same conditions. Further, on both Banezian and Molinist accounts of God’s causality, God is able to give counterfactual control to creatures regarding their response to God’s grace. These accounts illustrate that the problems for libertarian freedom do not reside in the mere fact that God causes an action, but rather in the manner in which God’s causality brings about human free acts. In fact, I  argue that Molinism and Banezianism offer responses to that grounding problem, whereas Grant does not. I conclude by comparing the theoretical virtues of each of these approaches in order to provide a better account of what it is to embrace theological compatibilism. The Extrinsic Model Drawing on Robert Kane,7 Grant argues that a free act has two necessary conditions: a free act is one where “there is no factor both prior to and logically sufficient for the act.”8 [1] “Prior to” is a logical or causal priority; a relation whereby “b asymmetrically depends on a, whether or not a temporally precedes b.”9 [2] “A is logically sufficient for b just in case it is not possible for a to exist (or occur, or obtain) without b’s existing (or occurring, or obtaining).”10 The first condition is intended to make clear that some factor could be logically, even if not temporally, prior to what is caused: for example, my espresso mug is logically, not temporally, prior to it being colored. The second condition is to be distinguished from cases where a factor a is necessary for factor b, but does not entail the existence of factor b necessarily. My espresso mug is a necessary, but not sufficient, condition for the mug’s whiteness; whereas my mug’s being white is a (trivially) sufficient condition that the mug is colored. Grant presupposes a “strong” view of divine causal sovereignty: God causes the being of all entities and does so continuously.11 Citing Aquinas, Grant takes this notion of God’s causality to extend even to free acts.12 This is to say that free creatures and their acts are both immediately or directly caused by and dependent upon God.13 Grant argues that God’s causality, while prior to free actions, is necessary but not logically sufficient for the occurrence of those acts. God’s causality appears to be logically sufficient for the occurrence of what God causes only because of a problematic assumption. This assumption is that God causes free acts (e.g., the free act E) by means of “some real, intrinsic property, feature, or state of God in virtue of which God causes E, and which state would not exist were God not causing E.”14 This is the “Intrinsic Model” of divine agency (IM). IM is problematic because it introduces a factor that is not

106  Fr. James Dominic Rooney, OP only prior to free acts but also logically sufficient to cause their occurrence: the divine decree (i.e., a distinct property, feature, or state of God) which decree cannot exist without the caused entity also obtaining. Grant argues that this latter fact means that IM is incompatible with libertarian freedom because IM introduces both a prior and logically sufficient factor that necessitates a free action (the decree) and so entails determinism. Grant therefore proposes an Extrinsic Model of divine agency (EM) which holds there is “nothing in virtue of which God causes” a free act.15 God causing the free act E only involves: God, E, God’s reason for causing E, and the causal relation between them.16 Grant goes through all of these ingredients and argues that none is a factor that would be both prior and logically sufficient for E’s occurrence, thus arguing that no factor entails determinism. E is (trivially) logically sufficient for itself, but not prior to itself. God’s reason for causing E is prior to, but not logically sufficient to cause E because God is free. Despite having reasons to cause E, God is “free to refrain from bringing E about.”17 The core of Grant’s analysis is that God’s act of causing E is logically sufficient for E, but it is not prior to E. This is because God’s act has E as a “constituent,” and so cannot be prior to E to act as a cause; God, not God’s act, is the cause of E. If we consider that act, or causal relation, without E as a constituent, then (Grant notes), this relation is neither causally prior nor sufficient for the occurrence of E.18 Grant claims that God’s act of causing E and E are then “logically necessary and logically sufficient for each other.”19 This is because God’s act of causing E is a relation, and “relations are not prior to their relata.”20 Further, “God’s act of causing my act depends on my act as on an essential constituent”21 so that God’s act to bring about E is logically simultaneous with or dependent on E. Thus, God’s act and E are “simultaneous (or concurrent) necessary conditions for each other” because “neither God’s act nor the creature’s act can be prior to the other.”22 Grant believes this logical codetermination means that a created agent has “counterfactual power” over what God causes with respect to their free acts.23 Because neither God nor His reasons differ in worlds with or without E, all antecedent factors to the agent’s choice remain the same; “were I to do otherwise, God’s act of causing my act would not occur.”24 Grant argues that I have control over God’s causing my act: “My act is an ontologically necessary condition for God’s causing my act. Without my concurrent co-operation in performing my act, God’s act of causing my act does not occur.”25 What leads to E actually occurring (and God causing E) is thus that creatures decide to E. If this is true, Grant concludes, EM is compatible with libertarian freedom. Throwing a Wrench Into the Works Bracketing concerns about God’s universal causal sovereignty and whether the construal of libertarian freedom is sufficient, there is something else missing in Grant’s account. Michael Almeida argues that what has gone wrong

Freedom, Even If God Decrees It 107 is that Grant’s position involves an equivocation on whether God’s act of causing E is prior to E. The relation, God causes E, is not prior to God and E only in the sense that God causes E is not temporally prior to the relata, God and E. The relation obtains only when the relata exist. But Grant has already concluded that explanatory priority does not require temporal priority, so God’s causing E can be prior to E in the relevant explanatory sense despite the fact that the relation is not prior to E in the irrelevant temporal sense.26 However, I do not think this is quite accurate. Grant’s position is that God’s act of causing the creature’s free act E cannot be prior to E because it has E as a constituent. In light of this distinction, he holds that “my act is an ontologically necessary condition for God’s causing my act. Without my concurrent co-operation in performing my act, God’s act of causing my act does not occur.”27 God’s act of causing E and E are then “logically necessary and logically sufficient for each other.”28 The point about priority then seems to be that only the joint decision of both God and the free created agent together is sufficient to determine their action. Despite Grant’s claim that this means that God’s causing an act E is logically sufficient but not prior to the action, what seems to be at issue is that God’s causal activity is necessary but not individually sufficient to bring about E. I take it that this is a fair way to put the point that, because the creature’s action E is a constituent of God’s causal relation to it, the causal relation does not come into existence without the creature choosing to bring about E. The creature’s decision to E is (consequently) logically or explanatorily prior to God causing E. While the causal relation by which God is causing E is logically sufficient for the occurrence of E, as it cannot be the case that God is causing E and E doesn’t occur, it is the creature that is explanatorily prior for determining whether E will occur and thus whether God causes E. It seems to me that what is doing the critical explanatory work on Grant’s account is then that, when God causes a free human action, God’s causality is ultimately necessary but not sufficient for bringing about that action. This seems to me another way to express Grant’s claim that the creature’s action has explanatory priority over God’s causal relation to that action. Notice, however, that such a response is available to two traditional accounts of God’s causality of free acts which do involve divine decrees. Take two very simple characterizations of such theories as follows: 1 Molinism: “God possesses and makes providential use of comprehensive knowledge of what creatures would freely choose in any possible situation, but God has no control over the facts about what free creatures would choose.”29

108  Fr. James Dominic Rooney, OP 2 Banezianism: There are no truths about what a creature would do independent of God’s causal decisions; “God knows what someone will do in the future because he eternally decrees either to cause or permit the human action.”30 In both theories, God actualizes a world in which humans freely perform their intentional actions. On the Molinist view, God has no direct control to actualize one set of free decisions rather than another – He can only use knowledge of what an agent would do to actualize a world in which the free decisions of that agent figure. On the Banezian view, God does have direct control over whether to actualize some set of free decisions of an agent. Historical defenders of both views adopted a classical view of divine simplicity, where divine decrees, that is, God’s decisions on what possible worlds to actualize, are identical with God.31 Thus, strictly speaking, neither are ‘intrinsic models’ of divine causality, even though they use the notion of a divine decree to model whatever is happening when God freely chooses to actualize one world rather than another.32 Nevertheless, on neither theory is there a requirement that God’s causality and His decrees must be logically sufficient for the occurrence of the creaturely decision that they cause, in the sense of being explanatorily prior to and causally sufficient for their occurrence. In fact, simply put the “divine decree” of either theory in the place of how Grant treats “God causing E” on his own extrinsic model. On either view, God does not choose to actualize a world in which I choose to E unless I also choose to E in that world. Similarly, God’s causing me to E clearly has E as a constituent, where neither God’s causing me to E nor me E-ing will occur without the other. Even for the Banezian, who holds that there are no truths about whether I would E independent of God’s decision to cause me to E, does not require that God’s decision be both prior to and logically sufficient for me E-ing. God’s decision to cause me to E could have been, for example, a joint decision of both me and God, where (while it would not be true that my action is logically prior to God’s causal relation) their relation does not involve being either logically or explanatorily prior to the other. Where Grant’s theory limps is that it provides no account of how God’s causality is insufficient for the occurrence of E. Generally, it seems problematic to explain the way in which (on EM) God’s causing an action is an intentional act on God’s part, since EM seems to imply that God’s action is not informed by something like an intention. In the classical tradition, these are of course the role of “divine decrees.” But, more specifically, Grant claims that God and the creature are concurrent in producing the whole of their effect where both are sufficient and necessary – for example, God is the sufficient cause of my choosing to become a friar, and I am the sufficient cause of my choosing to become a friar.33 Grant seems to think that both of these are not contradictory claims because we are only jointly sufficient, not individually so, for my decision. But the very question is how these two things are jointly and

Freedom, Even If God Decrees It 109 not individually sufficient. This is because, for instance, my action is a constituent of God’s action – if God is sufficient to bring about my E-ing, then it looks like any way in which I am a sufficient cause of my action would be subsumed under God’s causing. Either God is a sufficient cause or not.34 It would be problematic for the classical theist, for example, if the account requires that I am able to exercise causal influence on God, given the way in which my choice to freely E seems to be explanatorily prior in the account. There would otherwise appear to be an overdetermination problem, which is precisely the question at issue in reconciling God’s causality with my free agency. What we need is an answer for how God is causing the action with me. Conclusion: Avoiding Determinism Consider by analogy how God can know truths about what a creature freely does, despite being impassible and eternal. Garrigou-Lagrange famously claimed, in regard to God’s knowledge of free actions, that “God is either determining or determined, there is no other alternative.”35 Eleonore Stump responded to this claim about God’s knowledge by denying that knowledge of some fact about what a creature freely does requires that God be causally acted upon by the creature – God can be eternal and know the free actions of creatures, without either causing them or being causally affected by those actions.36 Kevin Timpe develops the suggestion by pointing out that contemporary account of truth-making, the relation holding between a true proposition and whatever that makes or necessitates the truth of that proposition, is not a causal relation.37 Consequently, the fact that some creature’s action would make a proposition (that God knows) true does not require that God is causally affected by that creature. And this way of understanding God’s knowledge permits God to be providentially responsive to creatures in time insofar as God’s action is responsive not in being metaphysically dependent on creatures, but rather that, whatever God did, He did because of the creature’s free decision; “God is responsive in that had the human agent done otherwise, then God would also have done otherwise.”38 We can characterize what it is to be a ‘theological determinist’ in light of the claims about God’s knowledge: theological determinism is the view that God’s decisions are individually and totally sufficient to account for all the contingent truths about creaturely actions.39 Yet an account of how God’s decision-making is responsive to the free agency of a creature, and how it relates to what God brings about, is precisely what Grant rules out on his extrinsic model. By contrast, the classical theories can each offer a vision of how God’s intention to be responsive to creatures is what ensures or accounts for how God’s causality is not (individually) sufficient for the occurrence of my free action, and then, consequently, in virtue of which I might have counterfactual control over my free actions despite God causing them. These theories propose models of God’s decrees or decision-making process in order to highlight facts about God’s intentions that are supposed to explain

110  Fr. James Dominic Rooney, OP how God’s actualizing or causing some set of my free decisions (directly or indirectly) nevertheless is not sufficient to determine my decision. For the Molinist, God intends to respect a creature’s free decisions in His providential plan, and so the model of divine decision incorporates God’s making decisions on the basis of knowledge about creaturely free decisions. Even when God is necessary to cause my decisions, that is, actualizing the worlds in which I make them, He does not determine the truths in light of which He made the decision to actualize that world. For the Banezian, too, God’s decision-making process in bringing about my free action is supposed to explain how that act is free. God’s intention, for the Banezian, is to help creatures produce a free act, perfecting their freedom – it is this fact about how God has caused my action that explains why it remains free.40 Further, God’s readiness to assist me in bringing about decisions provides me with alternative possibilities for action; thus, Aquinas notes: “What we can do with the Divine assistance is not altogether impossible to us. ‘What we can do through our friends, we can do, in some sense, by ourselves.’ ”41 Notes 1 Grant’s position was laid out in two related papers: W. Matthews Grant, “Divine Universal Causality and Libertarian Freedom.” in Kevin Timpe and Daniel Speak (eds.), Free Will and Theism (Oxford: Oxford University Press, 2016), 214–33; “Can a Libertarian Hold That Our Free Acts Are Caused By God?” Faith and Philosophy 27(1) (2010): 22–44. He has since collected the papers into a book, but I will freely refer to the papers as the views are the same; Matthews Grant W., Free Will and God’s Universal Causality: The Dual Sources Account (London: Bloomsbury Academic, 2019). 2 Grant, “Divine Universal Causality,” 215. See Free Will and God’s Universal Causality, 1–10. 3 Alvin Plantinga, The Nature of Necessity (Oxford: Oxford University Press, 1974), 171. Cited in Grant, “Divine Universal Causality,” 215. 4 Grant refers to this in “Divine Universal Causality,” as the “Popular Model” or PM, but in later writing it is called the “Intrinsic Model.” The latter terminology is clearer for our purposes. 5 Grant, “Divine Universal Causality,” 222. 6 Grant, “Divine Universal Causality,” 229. 7 C.f., Robert Kane, A Contemporary Introduction to Free Will (Oxford: Oxford University Press, 2005), 5–6. 8 Grant, “Divine Universal Causality,” 218. 9 Grant, “Divine Universal Causality,” 217. 10 Grant, “Divine Universal Causality,” 217. 11 Grant, “Divine Universal Causality,” 214; “necessarily God directly causes all entities distinct from himself for as long as such entities exist.” God’s causality involves two aspects: ontological dependence (Grant, “Divine Universal Causality,” 215: “[all entities] are conserved by God”) and ‘causal sovereignty’ (Grant, “Divine Universal Causality,” 214: “[God] is immediately present to all as an agent.”). 12 Thomas Aquinas, Summa contra gentiles, trans. James F. Anderson (Notre Dame: Notre Dame University Press, 1975), Bk. II, Ch. 15 (6): “Everything other than God . . . must be referred to Him as the cause of its being.” 13 Grant, “Divine Universal Causality,” 214. Grant makes no distinction, for example, between an act “having being” and “existing as” an act.

Freedom, Even If God Decrees It 111 4 Grant, “Divine Universal Causality,” 219. 1 15 Grant, “Divine Universal Causality,” 220. 16 Grant, “Divine Universal Causality,” 220. 17 Grant, “Divine Universal Causality,” 220. 18 Grant, “Can a Libertarian,” 34. 19 Grant, “Divine Universal Causality,” 222. 20 Grant, “Divine Universal Causality,” 221. 21 Grant, “Divine Universal Causality,” 228. 22 Grant, “Divine Universal Causality,” 222. 23 Grant, “Divine Universal Causality,” 229. 24 Grant, “Divine Universal Causality,” 229. 25 Grant, “Can a Libertarian,” 43. 26 Michael Almeida, “Review of: W. Matthews Grant, Free Will and God’s Universal Causality: The Dual Sources Account.” European Journal for Philosophy of Religion 12(4) (2020): 242. (Referencing Grant, Free Will and God’s Universal Causality, 6). 27 Grant, “Can a Libertarian,” 43. 28 Grant, “Divine Universal Causality,” 222. 29 Kenneth Pearce, “Are We Free to Break the Laws of Providence?” Faith and Philosophy 37(2) (2020): 159. 30 Thomas Osborne, “Thomist Premotion and Contemporary Philosophy of Religion.” Nova et Vetera (English Edition) 4(3) (2006): 607–8. 31 Classical proponents held that divine decrees would be identical with God’s essence. Mark Spencer lists some prominent defenders of the identity of God’s essence and decrees in his “Divine Causality and Created Freedom.” Nova et Vetera (English Edition) 14(3) (2016): 932; [I reproduce his list exactly] John Capreolus, In I  Sent, d. 45, q. 1, sC. 2, ad 1 and 4 Aureoli (vol. 2, 587–89); Bañez, In I ST, q. 19, a. 2–3, 10, 604, 607–08, and 644–46; John of St. Thomas, CP, Logica, q. 23, a. 1 (vol. 1, 635–36); In I ST, q. 19, d. 5, aa. 4–5 (117–29); Salmanticenses, In I ST, De voluntate Dei, d. 7, dub. 7, ss. 4–5 (vol. 2, 119–23). 32 And, even assuming, contrary to classical theism, that the divine decree is a separate entity from God, Grant claims that the causal relation of God causing me to E is also distinct from God. See Grant, Free Will and God’s Universal Causality, 58–59. 33 Grant, Free Will and God’s Universal Causality, 38–39. 34 See a similar objection by Simon Kittle, “W. Matthews Grant, Free Will and God’s Universal Causality: The Dual Sources Account.” Faith and Philosophy 37(3) (2020): 377. 35 Reginald Garrigou-Lagrange, God, His Existence and His Nature: A  Thomistic Solution of Certain Agnostic Antinomies (volume 2, trans. Dom Bebe Rose) (London: B. Herder Book Co., 1936), 546. 36 Eleonore Stump, Aquinas (New York: Routledge, 2003), 120–2. 37 Kevin Timpe, “Truth-making and Divine Eternity.” Religious Studies 43 (2007): 305–7. 38 Timpe, “Truth-making and Divine Eternity 310. 39 Pearce, “Are We Free to Break,” 158. 40 See Reginald Garrigou-Lagrange, Grace (trans. Dominican Nuns of Corpus Christi Monastery) (St. Louis: B. Herder Books, 1952), 248–9; David Oderberg, “Divine Premotion.” International Journal for Philosophy of Religion 79 (2016): 218–20. 41 Both from Summa Theologiae (translation English Dominican Fathers, Benzinger Bros., 1920), I–II, q. 109, a. 4, ad. 2. The quotation from Aristotle is from the Nicomachean Ethics III, 3.

Part II

Community

7 Free Together On Christian Freedom and Group Ontology D. T. Everhart

Introduction Many philosophical and theological discussions on free will focus on agents being the source of the decisions they make, focusing almost entirely on the free will of individuals. Recent work in analytic philosophy on the metaphysics of groups can offer a different approach to theological questions of free will. In this chapter, I look at the function of the Christian’s freedom in Pauline epistles, giving special attention to our belonging to others in the body of Christ. The purpose of human freedom for the Apostle Paul is made clear in his ecclesiological ethics. Paul uses freedom not as an isolated individual, but as something one person either takes up or lays down for the sake of the other. This is rooted in Paul’s relational view of human persons: that human persons are essentially constituted by their relations to others. This view of personhood implicates the ontology of groups as the social context of interpersonal relationships. By integrating analytic philosophical categories of social ontology and theological categories of relational personhood, I  can offer an account of corporate freedom. Because many of the commonly offered accounts of free will are only concerned with the selfdetermination of the autonomous agent, social ontology can offer an interesting way of reframing these concepts to better account for Paul’s vision for human freedom. Paul’s vision for freedom in Christ and his relational view of persons reframe self-determination in a helpful way. Freedom on this account takes on a more social character than purely individualistic conceptions. I argue that Christian freedom is concerned with the freedom of the group and the ways in which individual agents will either use or give up their freedoms for the sake of the whole. By mapping group agency and the role of individual agents onto the Pauline conception of freedom in Christ, I demonstrate the unique character of Christian freedom as primarily freedom for the other. I  then show this conception of freedom to offer a unique answer to key concerns of free will debates, namely, the concern for self-determination operative in sourcehood accounts of libertarian free will. This conception doesn’t undermine the self-determination of human agents, but rather socially situates that DOI: 10.4324/9781003306191-9

116  D. T. Everhart self-determination in relational dependence on other persons. Thus, I argue that (1) human persons are the most free they can be when they are in community with other human persons and God; (2) in community, other persons non-causally influence the self-determination of individual agents; and (3) this does not undermine self-determination, but only absolutely autonomous senses of self-determination. The flourishing of free individual human agents occurs in belonging to one another in community; we are free when we are free together. Free Will and Agency Briefly Outlined Before delving into the Apostle Paul’s thoughts on the nature and role of freedom in the Christian’s life, we ought to lay out some considerations about the debates to which his vision for freedom has something to say. This will help us to identify where certain libertarian accounts of free will can be helpfully reframed in light of Pauline categories of personhood and freedom in Christ. Free will, in contemporary philosophical and theological literature, is generally considered in light of two components: alternate possibilities and self-determination. The former is fairly self-explanatory. Traditionally, libertarian free will has been described in terms of a particular agent making a choice when they could have chosen otherwise. This has been called the Principle of Alternative Possibilities, which states that “a person is morally responsible for what he has done only if he could have done otherwise.”1 The free-will corollary of this, according to Kevin Timpe, is that “a person only has free will if he could have done otherwise.”2 The possibility of choosing otherwise implies that the decision was not determined, and thus made freely. More recently, some advocates of libertarian free will have abandoned this component. One reason often offered for this is something called F-cases. Harry Frankfurt used F-cases to “drive a wedge between responsibility and alternative possibilities.”3 Frankfurt set out to untangle these concepts because responsibility is thought to be only possible where a choice is freely made and the presence of alternative possibilities has historically been thought necessary for libertarian free will.4 The idea was to describe cases in which an agent could freely make a choice, but for which they had no alternative possibilities. Let us take the following example case from Linda Zagzebski: Black, an evil neurosurgeon, wishes to see White dead but is unwilling to do the deed himself. Knowing that Mary Jones also despises White and will have a single good opportunity to kill him, Black inserts a mechanism into Jones’s brain that enables Black to monitor and to control Jones’s neurological activity. If the activity in Jones’s brain suggests that she is on the verge of deciding not to kill White when the opportunity arises, Black’s mechanism will intervene and cause Jones to decide to commit the murder. On the other hand, if Jones decides to murder White on her own, the mechanism will not intervene. It will merely

Free Together 117 monitor but will not affect her neurological function. Now suppose that when the occasion arises, Jones decides to kill White without any “help” from Black’s mechanism. In the judgment of Frankfurt and most others, Jones is morally responsible for her act. Nonetheless, it appears that she is unable to do otherwise since if she had attempted to do so, she would have been thwarted by Black’s device.5 In the case given, Jones seems to be responsible for killing White, even though Jones had no other possibilities. Zagzebski argues that this is because Jones’ decision to kill White and the non-intervention of the machine indicate that Jones would have killed White even with alternative possibilities.6 We can thus make a qualitative distinction between Jones killing White because the machine intervened and Jones killing White because of her own decision sans machine intervention. What makes one responsible for a particular action, according to Zagzebski, is not alternative possibilities, but rather once we have identified who or what makes an act happen we have identified the potential bearer of responsibility for it. In Frankfurt cases in a non-deterministic world the agent makes her act happen in as ultimate a sense as you like.7 This rejects the notion that alternate possibilities are necessary for free will, leaving us with the other generally agreed-upon component of free will: the power of self-determination. Sourcehood accounts only require an agent’s choice to be self-determined in order for their will to be considered free. For such accounts, the ultimate source or cause of the particular action or choice is the agent who made it. As Timpe puts it, “according to [this] approach, free will is primarily a function of an agent being the source of her actions in a particular way . . . nothing outside of me is the ultimate explanation of my action or choice.”8 Another way of putting this has been to say that the particular action or choice is ultimately self-determined by the agent, as opposed to being ultimately determined by external forces and factors. This is key to understanding sourcehood accounts of libertarian freedom, as many philosophers and theologians presume that “people cannot be morally assessed for what is not their fault, or for what is due to factors outside of their control.”9 While this principle is not assumed in all sourcehood accounts of libertarian free will, and Timpe’s definition could include a variety of “particular ways” in which an agent is the source of their action inclusive of external factors, it is nevertheless a prevalent assumption to which Pauline categories of personhood and freedom offer challenge. As I will demonstrate, Pauline freedom includes factors which appear to undermine self-determination. This is because these factors are external to the self and yet are necessary for a given agent to act freely.

118  D. T. Everhart Freedom and Personhood in Paul Paul’s vision for human freedom is rooted both in his vision for freedom in Christ and his understanding of personhood. To understand the former, we must first understand the latter. Where Paul talks about freedom, he is concerned with the flourishing of persons in the body of Christ. Freedom, for Paul, is understood in terms of not only what human persons are but also what they are created to be. Hence, any Pauline conception of freedom and free will must respect the telos of that freedom in the flourishing of humanity: union with Christ. In union with Christ, human persons and whatever capacities we might think that they have are oriented toward their proper purpose and flourishing. In order to understand the freedom of persons, we ought to understand how freedom contributes to this human telos. Given that Paul’s vision for freedom is focused on this purpose, it should really be asked whether what Paul means by “free” or “freedom” is synonymous with what is meant by philosophers such as Timpe and Zagzebski. Ultimately, the answer is no. What Paul has in mind when he speaks of freedom is broader in scope than what philosophers mean by free will. However, we can say that philosophical free will is included in Paul’s vision for human freedom for reasons explored below. Given that this is included in Paul’s vision, we can explore the implications of that vision for contemporary discussions of free will. Pauline Personhood and Belonging to the Other

To understand Paul’s vision for the freedom of human agents, we should first lay out Paul’s understanding of personhood and the significance of the “other” in that understanding. Fortunately, much of this exegetical work has already been done. Many Pauline scholars recognize the significant threads of relationality in Paul’s conception of persons, emphasizing the formative significance of interpersonal relationship which an “I” has with a “Thou.” This is a key theme of Susan Eastman’s work, Paul and the Person, in which she traces through Paul’s letters the “pattern of talking about persons in which the self is never on its own but always socially and cosmically constructed in relationship.”10 Eastman’s exegesis of Paul’s anthropology implicates both an ontological dependence on relation to the other and a repudiation of cosmic monergism, in which any distinction between the “I” and “Thou” is erased. Rather, the self is a “self-in-relation-to-others,” constituted not in some autonomous sense in which I am what I am regardless of who I am to others (or who they are to me), but in the sense of an embodied individual ever related to the other in social systems which Eastman calls “we-centric space.”11 Within such social spaces, such as the body of Christ, the individual self is understood only in relation to another and is contextualized by a broader “We”. The social context of I–Thou relation is a framework for understanding how the “I” remains distinct from the “Thou” even in ontological dependence on the “Thou” to be an “I.”

Free Together 119 Thus, for Paul, the “other” is essential to the constitution of a self. But what, we must ask, are the others, which constitute a self? C. Stephen Evans captures this view of personhood with an emphasis on the role of the other stating, “ ‘being a self’ [is] a status that is granted by being attributed by others . . . to be a self is simply to be regarded as a self by others.”12 Many accounts hold that this “other” for Paul is God so that a self is a self only when regarded as such by God in close personal (and incarnately embodied) relationship. This new self is indicative of the telos of humanity, a becoming what one is meant to be in union with God through Christ. It is certainly true for Paul that in union with Christ, members of the body are transformed into the image of God (2 Cor 3:18; Rom 8:29–30). Their true selves are formed in relation to God through the incarnate one.13 Yet this is not the whole picture for Paul. This relation we have to Christ takes place in a “we-space”: the body of Christ. For Paul, union with Christ by the Spirit also entails union with one another by the Spirit.14 Eastman argues, “that web of relationships in the body of Christ, not isolated or inward-turned individuals, is the arena in which change happens.”15 The “others” who constitute the self and contribute to its formation, therefore, are persons both divine and human. Because of this understanding of the constitution of selves in relation to others, human persons fundamentally need others to be what they are meant to be. This appears to challenge self-determination in the sense usually meant by philosophers because the self never stands in isolation as a self. Rather, in any instance in which we might argue that the “source” of a particular decision is the self, that self owes what it is in part to others both divine and human. To be a self with agency in the first place, we require relation to others. This seems intuitively to undermine sourcehood accounts of free will. Paul’s understanding of this fundamental need goes even further than this initial requirement. Not only do we need others to become a self, but also we continuously belong to the others whom we are in relation with. This is apparent in Paul’s use of membership language, which he applies both to Christians being members of Christ (1 Cor 6:15; 12:27; Eph 5:30; Col 1:18) and to Christians being members of one another (Rom 12:25; Eph 4:25; Col 2:19). It is noteworthy that passages on membership in one another often follow membership in Christ. Willie Jennings argues that our belonging to Christ is entangled with our belonging to one another, saying: [W]e belong to each other, we belong together. Belonging must become the hermeneutic starting point from which we think the social, the political, the individual, the ecclesial, and . . . the educational. . . . The cultivation of belonging should be the goal of [the body of Christ].16 This is demonstrated in Paul invoking membership in Christ to argue for belonging to and bearing responsibility for one another.17 James Dunn thus argues: “it is a sense of belonging to Christ indeed, but of belonging together with others, with the obvious implication that one without the other would

120  D. T. Everhart make the whole unbalanced and unhealthy.”18 This sense of belonging to Christ together with one another is apparent throughout the familial, adoptive, and temple metaphors which Paul uses in Galatians and Romans. Erin Heim argues that Paul cultivates in these metaphors a strong sense of belonging to God and one another in the unity brought about by the Spirit.19 This belonging, for Paul, implicates a sense of mutual responsibility for others in the community such that Paul can hold entire communities responsible for the sin of an individual and call individuals to take responsibility for others in the community.20 Belonging to one another in belonging to Christ is thus foundational for Paul’s vision of what human persons are created to be in union with Christ. Hence, persons in the body of Christ continue to belong to one another, even in exercising individual agential capacities. Human persons are created with this telos: to flourish in communion with God and one another. For Paul, this means that persons are meant to belong to one another and are bound together in this communion. As Jennings puts it, Jesus created the condition for the crowd, reflecting God’s desire for the gathering. The crowd was not his disciples, but it was the condition for discipleship. . . . In Jesus and the crowd, we see the creator-creature relationship in its most naked, most powerful form.21 This vision of human personhood and what it looks like for human persons to flourish includes human freedom in a way that we might call free will, as shown in the next section. But whatever we say about human free will and flourishing must, for Paul, be contextualized by the belonging of persons to one another in community. Paul’s Vision for Freedom in Christ

Paul’s conception of persons, of individuals who are necessarily socially situated and ever related to the other, grounds his vision for human freedom. Paul talks about freedom in many varied respects, the scope of which is too great for this chapter. Instead, I highlight some key themes of his vision for freedom in Christ and its implications for the telos of human persons to belong to God and one another in community. This provides a basis for assessing and re-framing self-determination. Paul’s assertion that the believer is free in Christ is often, but not exclusively, in reference to freedom from the oppression of sin. Yet sin is not a merely spiritual reality for Paul; it embeds itself in the social contexts of everyday life, the very same social realities that contextualize and contribute to personhood. “Our modern delineation,” argues Esau McCaulley, between spiritual and political evil, when read back into Paul’s thought, is an anachronism. The “present evil age” can be understood to mean

Free Together 121 the demonic evil of slavery in Rome and also rulers’ economic exploitation of the populace. Both were driven by the policies of corrupt Roman leadership, and both were ultimately dictated by spiritual forces.22 The oppression of sin from which Christ frees the believer is an oppression that often occurs at the systemic or group level. This is shown in Paul’s appeal to freedom in Christ throughout Romans and Galatians to argue that Gentiles who are in Christ are free from the constraints of the Torah.23 This is also why Paul’s admonishments for sin and encouragements to resist in the Spirit are often addressed to entire ecclesial communities. Sin, for Paul, is a corporate reality, which must be responded to at a corporate level.24 It follows, then, that freedom from sin is given at a corporate level to the body of Christ. This is, in fact, a large part of Paul’s argument concerning personal or individual freedom to the Corinthian church. Bruce Longenecker writes, “moral chaos is precisely what Paul found among some of the practices of Corinthian Christians. Over and over, they interpreted their freedom in Christ along individualistic lines, without regard to the health of the Christian community.”25 The reduction of freedom in Christ to individual liberties is what set the rampant Corinthian sinfulness apart from Paul’s own vision for what freedom should look like: it is in being bound to Christ and one another in the body of Christ that individuals come to enjoy freedom in Christ corporately. Thus, Longenecker argues that freedom in Christ is not about the free will of the individual, but rather about a freedom from the oppression of sin “resulting in the freedom that equates to enslavement to others in the love that the Spirit inspires within those who are enslaved to God.”26 Because Paul understands freedom more broadly in light of his vision for freedom in Christ, his general conception of freedom is fundamentally corporate as well. It is for this reason that Longenecker argues that “Christian freedom is not individualistically configured. This is what Paul presents in his analogy of the church as ‘the body of Christ,’ with each part of the community playing its part (whether large or small) to enhance the community.”27 Longenecker, and other apocalyptic interpreters of Paul, reads Paul as an anti-individualist due to his communal orientation.28 However, it does not follow that simply because Paul sees the end of freedom to be in the community that he does not recognize or view as significant individual liberty. This competitive dichotomy between individualism and communalism has been thoroughly and convincingly critiqued by Simeon Zahl, who argues that Paul can be (and in fact is) both an individualist and a communalist.29 Freedom in Christ, as Paul construes it, concerns individuals and their actions in ways consonant with the realities of free will and self-determination debated by philosophers and theologians. To erase Paul’s individualism in favor of his communalism is to neglect the significance of Paul’s exhortations concerning individual liberties. Paul writes extensively and significantly about individual liberties and the free decisions of individuals in ways which intuitively indicate some level

122  D. T. Everhart of self-determination. For instance, in the aforementioned example of the Corinthian church, Paul recognizes their free choices as individuals with regard to eating meat, saying that one should determine whether or not to eat meat depending on one’s own conscience. He goes on in this example to offer a list of his own individual freedoms, such as his freedom to take his due from the Corinthian church or his freedom to not be bound to the Torah. In his letters to the Galatian and Roman churches, he also cites his own individual liberties as a basis for Gentile freedom from the Torah. It should be clear, at least with regard to freedom, that Paul is not anti-individualist simply because his vision for freedom in Christ is fundamentally communal. The communal serves to contextualize the individual precisely because of Paul’s understanding of personhood as having a telos in belonging to others. In many of these examples, a particular freedom or individual right is either taken up or given up by an individual for the sake of another. That individual’s decision is constrained by their freedom in Christ. Eastman criticizes the all-too-common readings of Paul which attempt to maintain individual human dignity and responsibility through the language of individual rights, which inevitably come into conflict with others’ rights. Paul’s theology points in a different direction, through an enduring and christologically grounded affirmation of every human body based on the human incarnation of the divine.30 Because the “I” always belongs to “Thou,” such conflicts of individual freedoms are constructive for the true end of freedom in Christ, opening up “endless possibilities of boundary-crossing freedom and life,” as opposed to being a problem to be overcome at the individual level.31 As such, we should read these examples of individual self-determination as being properly understood in the context of community and belonging to others. Let us take a closer look at Paul’s recommendation to the Corinthian church concerning the eating of food offered to idols in 1 Corinthians. Paul makes it clear that some believers, as a matter of individual conscience, are free to partake of meat previously sacrificed to idols. Yet he also states that one should lay down that freedom for the sake of another whose conscience is convicted. Paul readily recognizes something like free will, that is to say a choice made in virtue of the self, had by the individual in and of oneself, yet broadens the scope of freedom to also include the freedom and convictions of another. Because the one with a stronger conscience, the one who is actually free to choose, belongs to the one with a weaker conscience, the one with a stronger conscience is bound in will to their sibling in Christ. Insofar as we might think Paul espouses something like a sourcehood account of free will with regard to these particular personal liberties, they have their telos or purpose in love of the other. That purpose, at least intuitively, seems to bind the will to the other. This telos, according to Jennings, “opens up endless possibilities of boundary-crossing freedom and life . . . however, [it] also

Free Together 123 opens up endless possibilities of boundary crossing slavery and death.”32 Jennings understands this as a tension in our freedom in Christ so that belonging to others can both make us free and hinder our freedom. This tension is precisely what Paul is after. This is shown in his concluding remarks on the idol-meat example. In the following chapter, Paul gives an example of how to treat individual liberties from his own actions toward the Corinthian church. He claims that he was free to take his due from them, and yet he gave up that freedom in order for the Corinthian church itself to be more free to accomplish the work that Christ gave them to do. Within this passage, he rattles off a list of other freedoms, which he has either taken up or laid down for the sake of building up the body of Christ in freedom. The purpose of individual freedom for Paul is to be bound for the sake of the group’s freedom. Personal freedom and self-determination are not ends in and of themselves, but rather serve the telos of human persons flourishing in communion with God and others in Christ. Herein lies one implication of Paul’s corporate conception of freedom in Christ for the individual. For a human person to flourish, that is to say for them to be the most free they can be, they must belong to others in community. The flourishing of the individual, for Paul, comes through the freedom of the whole body in Christ. In many cases, this would seem to compromise individual autonomy because many decisions, which individuals would otherwise make freely, appear to be constrained or partially determined by others to whom the individual belongs in community. For the body to be free in the way that Paul means, the individual has to belong to others in the body, giving up many of their freedoms for the sake of others. We can even conceive of cases in which a particular freedom would be taken up for the sake of another which one does not wish to take up in and of themselves. We are left with two interpretive options for making sense of this. First, we could say that the freedom of the group, in taking priority for Paul over individual freedom, completely undermines the freedom of the individual. This doesn’t seem to be the case, as Paul is still readily concerned with individuals’ choices throughout his letters and the implications of the freedom of the body for individual believers. Second, we can understand this as maintaining a kind of self-determination necessary for free will which has a particular telos or purpose. On this interpretation, the individual is most free when their choices are made for the sake of others to build up the community’s freedom. While this interpretation seems more helpful for reframing sourcehood accounts of libertarian free will, it leaves us with the problem of how belonging to others in community can constrain the freedom of the individual without undermining self-determination. To make sense of self-determination in Paul’s vision, we need to show how the constraint of a telos in the group’s freedom does not undermine the individual’s self-determination and we need to show that belonging to others in the person-constituting way that Paul envisions does not undermine the individual’s self-determination.

124  D. T. Everhart Group and Individual Freedom To make sense both of Paul’s conception of personhood and his vision for human freedom in Christ, I turn to the resources of analytic philosophy and group psychology. Group ontology provides a way to talk about group phenomena in ways that neither reduce the agency of individuals to the actions and intentions of the group nor reduce group phenomena to the sum of the actions and intentions of the group’s individual members. Psychological studies of group behavior give clarity to these philosophical claims, offering scientific explanations of these phenomena supported by empirical evidence. The interplay of group and individual agency in these fields will help explain Paul’s vision for the freedom of persons precisely because it makes sense of the corporate conception of freedom Paul is after without eradicating the personal freedom which Paul also gives reference to. This allows for a selfdetermination of the individual arising from belonging to others which can be non-causally contributed to by the telos of being in communion with others and God. Group Psychology and Individual Agency

One group metaphysic, which will help our purposes here, is that of Stephanie Collins. Collins argues that groups can coordinate the intentions and actions of individual members so that the group itself can come to have a kind of agency, which is not reducible to the sum total of individual agents, which constitute the group.33 In order for a group to be able to have agency in this way, they require some kind of a centralized decision-making process by which the group can ascertain collective goals and coordinate the actions and intentions of its members to make decisions in order to meet those goals.34 In arguing for this, she outlines a particular relationship between group agency and the agency of individual members. She distinguishes between group and individual agency in the following way: [T]he group’s decision is not merely the conjunction of members’ decisions. The members’ decisions were to assent to the collective’s doing such-and-such. By contrast, the collective’s decision was to do suchand-such. The collective’s decision was determined by the members’ decisions, but it is not to be identified with the mere conjunction of them for two reasons. First, it has a different content: the collective’s decision is ‘the collective will do this’. Second, the collective’s decision arose out of two things: the conjunction of member’s decisions plus the fact that they are all committed to the unanimity rule.35 Thus, for Collins, the group’s agency supervenes upon the agency of its individual members. What is meant by this, according to Collins, is that “there can be no change in group-level facts without some change in individual-level

Free Together 125 facts.”36 In other words, the agential group relies on the coordination of its members and their intentions and decisions within that coordination in order to be what it is. As various members contribute to changes in the group, the group has influence on individuals within the group. As the group’s goals and intentions to meet those goals change, so too will the intentions to contribute to those group-level actions change in individual members in virtue of their roles. Collins offers the example of an individual member deliberating on what restaurant to propose that the group go to for dinner. She argues that this is distinct from deliberating over the restaurant that one intends to go to by oneself, but is influenced by the group: that is, your deliberation (over which restaurant to suggest) has been we-framed by you: the constraints and possibilities that you consider are not your own constraints and possibilities, but rather the constraints and possibilities determined by features of group you belong to, where these largely supervene upon (but might be multiply realizable by and depend upon relations between) the features of members of the group.37 This is an excellent example for trying to make sense of Paul’s vision for the group freedom of the body of Christ and its implications for individual Christians. There is still some level of individual deliberation, ultimately leading to self-determination of what restaurant to suggest. However, the decision one comes to is not entirely self-determined. It is also partially determined by the considerations of the group, such as the decided budget for the group’s outing or relative distance from the group’s agreed-upon meeting location. This particular metaphysic of groups gets us to some degree of self-determination, which can be partially constrained by group-level considerations. In addition to this, we ought to further consider the work of Philip Pettit in order to get at the constraints, which Paul sees, on individual freedom in virtue of belonging to others. Collins has gotten us the group-level constraints, so now we turn to Pettit’s work on holism to get at the constraints and possibilities of belonging to one another in the context of community. Pettit’s work on holism is framed within a larger system of social ontology. Pettit splits his system into two key issues, which he takes to be independent of one another. He writes, individualists deny and collectivists maintain that the status ascribed to individual agents in our intentional psychology is compromised by aggregate social regularities. Atomists deny and holists maintain that individual agents non-causally depend on their social relations with one another for some of their distinctive capacities.38 Having taken on Collins’ model of collectives and individuals in the way described earlier, we ought to commit ourselves to Pettit’s view on what he

126  D. T. Everhart calls the vertical issue between individualism and collectivism. Namely, we should be individualists in order to maintain the relationship between group and individual agency as outlined by Collins. Moreover, this claim seems an important one to ground Paul’s exhortations concerning individual liberties. As has been previously argued, Paul is not an anti-individualist, nor does he need to be in order to emphasize the telos of individual freedom in the freedom of the community. As such, I take Pettit at his word on the vertical issue and focus on the horizontal issue: do agents non-causally depend on one another for some of their distinctive capacities? In approaching this issue as an individualist, Pettit has in mind capacities which lend themselves to the self-determination of autarchical agents, which involve “the intentional attempt to shape one’s intentional attitudes, one’s beliefs and desires, with a view to having them satisfy various conditions of rationality.”39 Such capacities, Pettit argues, are essentially interactive enterprise[s]. .  .  . The [capacity having] subject must be in a position to interact with other bearers of the relevant inclination or disposition: herself at later times, or other persons. Without such interaction there are no normal or related conditions on the realization of which the inclination can depend for successfully guiding the subject.40 Pettit goes on to illustrate this with the particular capacity for thoughtful rule following. A subject who follows certain rules has a perfect atom-for-atom copy of themselves that is isolated socially from all contact with others at any past, present, or future time and isolated in time itself from relevant past or future counterparts. Such a creature of the imagination would lack the conceptual resources for thinking or, more generally, for rule following.41 In other words, capacities such as thought require certain reference points in order to function in the ways that concepts like “agency” require. These reference points cannot be acquired solely by the self, but are acquired through interaction with others. The capacities which we generally ascribe to agents and by which agents are able to deliberate on certain choices require social interaction to some extent. Pettit is adamant that this influence that others have on the agent through social relations need not be causal, but rather advocates that “the individual’s capacity to think superveniently depends on her enjoying either interpersonal – and, inevitably social – relations with other individuals.”42 The kind of influence which persons hold over one another partially or subveniently determines their agential capacities.43 What Pettit means by subvenient determination is not sufficient for the realization of a particular decision by an agent. It can only, therefore, partially determine that decision, and still

Free Together 127 requires some self-determination of the agent in question. Hence, the influence of others on a given agent is not causal. Thus Pettit concludes, in order to pursue the shaping of our intentional states, we have to be able to identify propositions and propositional elements as norms by which to guide ourselves. These norms give us our bearings as we seek out the rational things to believe, the rational things to desire, and the rational things to do . . . We have to look to what emerges in interaction with others, not to what remains locked in the intimate enclosures of the self.44 Pettit’s proposal is a modest yet important one. Any sense of self-determination an agent might have will necessarily rely on the formative but non-causal influence of others. Pettit’s holism finds support in recent work on group psychology. Barbora Sipova and Melinda Carpenter demonstrate that many studies show that the cognitive and psychological capacities in virtue of which we act as individual agents are formed in relating to others.45 Much of what we think contributes to our agency is derived from our attending to things together in groups, thus forming a basis of shared knowledge, agency, and attention from which individual agency arises. This is demonstrated in a particularly clear way in studies conducted on infants and small children. In these studies, children tend to perform better at given tasks when conducting the task in an interactive way.46 What these psychologists have concluded is that cognitive capacities such as thinking, deliberation, identification, and intention are formed not independently, but in social interaction with parents and other children. To be agents who freely choose in the self-determined way required by sourcehood accounts of libertarian free will, we require these capacities to be formed and, at least partially, determined by others through social interaction. This approach to the formation of individuals’ agential capacities is grounded in the actuality of individual persons influencing, persuading, leading, and modeling for one another in groups. Said another way, it provides substance to the notion of a socially structured field within the individual. It thereby explains how large numbers of people can act in coherent and meaningful ways, by reference to shared group norms, values and understandings rather than idiosyncratic beliefs.47 Reicher et  al. thus argue that groups provide the appropriate context for Pettit’s holism. They argue that our identification with and participation in groups constitute a bridge between the individual and the social and how it allows one to explain how socio-cultural realities can regulate the behaviours of individuals . . . social identity provides a psychological apparatus that allows humans uniquely to be irreducibly cultural beings.48

128  D. T. Everhart Far from undermining individual self-determination, however, this view of groups and individuals in psychology underpins individual agency and its formation via holism with participation in groups. “(Inter)personal behaviour is not simply underpinned but also made possible by a salient personal identity, just as (inter)group behaviour is both underpinned and made possible by a salient social identity.”49 For human agents to self-determine in ways that sourcehood accounts of libertarian free will require, human agents must be formed and reliant on their relationships to others constrained and made possible by the contextualization of these relationships in social groups. Both of the aforementioned concepts, the individual’s relationship to groups and holism, offer a way of talking about the influence of others on the individual’s self-determination in the context of a group in such a way that does not seem to override or undermine the individuals’ self-determination. Rather, we have a kind of feed-back loop between groups and group members, in which members coordinate to constitute an agential group, causing it to be what it is and continuously contributing to the constitution and ongoing transformation of that group. The group, in turn, sets certain constraints and possibilities on decisions of group members, thus contributing to their self-determination by which they contribute to the agency of the group, and so on. Within this group context, individual members have a similar kind of influence on other members which enables their free self-determination while also setting certain constraints and possibilities on that self-determination.50 Such a feed-back loop depends on individual agents having self-determination to a certain degree. Self-determination, in this sense, relies on belonging to others in group membership. However, it also requires that individual decisions are not entirely self-determined, but adequately rely on these group relations to come about. Maintaining Sourcehood in Paul’s Vision for Freedom in Christ

In light of these claims, we ought to reexamine what counts as sourcehood for the purposes of libertarian free will. To briefly summarize Paul’s vision for freedom in Christ, laid out in holist, individualist, and group terms, individual persons do in fact have some degree of self-determination. However, they non-causally rely on others for those powers of self-determination. Moreover, in belonging to others in groups, individuals are partially constrained in their powers of self-determination, being partially and subveniently determined by their belonging to others. Finally, individuals are the most free when their decisions are self-determined in a way that is coordinated to contribute to the free agency of the group. Certainly some sourcehood accounts will already be consonant with what Paul has in mind. Others, however, will be impossible to the mind of Paul because sourcehood on these accounts is so absolute that any external contribution to the self-determination of the individual will compromise the enterprise entirely. The implications of Paul’s vision for freedom in Christ for sourcehood accounts go beyond simply eliminating some

Free Together 129 accounts, which are too strong to maintain. The telos of human free will in the freedom of the community, and our belonging to others in that community, is, as will be shown, a stronger claim than those sourcehood accounts which allow for external factors in self-determination. One feature common to many sourcehood accounts which Paul’s vision for freedom in Christ poses a challenge to is the concern for external factors which impose upon the self-determination of individual agents. This follows intuitions of philosophers such as Galen Strawson and Thomas Nagel, who think that source conditions outside of the agents control undermine or negate the freedom-level control required for free will.51 One interesting objection in which these intuitions play out is called the luck objection. Neil Levy offers his own version of this objection with what he calls the Luck Pincer.52 To oversimplify a very complicated and important argument, Levy argues that the capacities by which persons make decisions, such as epistemic, emotional, and behavioral dispositions (which Levy calls endowment), are not formed by the person who has them, and so agents lack control over who they are.53 While Levy recognizes the possibility that persons can come to take responsibility for their endowment by modifying it,54 he argues that even this process relies on some relevant form of luck, either in the original endowment received, the circumstances which allow for the modification of the endowment, or some combination of the two.55 This touches on Paul’s vision for relational personhood and freedom in Christ in key ways. First, this sense of endowment that Levy implicates seems to undermine free will entirely, because it undermines responsibility for anything which the agent is not completely in control of, such as the relational constitution of persons. Second, the context of community and belonging to others seems to fit the bill for the circumstances which Levy thinks undermines the taking of responsibility over time by modifying one’s endowment. It would seem, on this understanding of responsibility and free will, that only those agents who are absolutely autonomous and free of external constraint on their decisions are capable of free will. This is positively what Paul’s vision for freedom in Christ rejects. Fortunately, sourcehood accounts need not commit themselves to such absolute construals of self-determination. Robert Hartman’s excellent work defending moral luck offers a critique of this undermining of free will and responsibility via the luck objection. His first tact is to critique the assumption that absolute and complete control of all factors relevant to a decision is required for responsibility of that decision. He writes, There is a surprising dearth of argument for the claim that an agent lacks direct control over an event that is settled by her endowment, her modified endowment, or neither. The consideration that Levy (2011, pp.  90; 90–94) cites is that factors beyond an agent’s control decisively influence her choice. But this is insufficient to demonstrate that the agent lacks direct control over the event, because if the factors

130  D. T. Everhart beyond the agent’s control that decisively influence her choice negate direct control over her action, it follows that an agent has direct control only if she controls all the features internal and external to her agency that contribute to influencing decisively the action that she performs. In other words, the agent must have control over her own constitutive properties and external circumstantial features that contribute to decisively influencing her action to have direct control over her action. But this hyperbolic kind of control is alien to Levy’s (2011, p. 19) own concept of direct control as well as any other plausible concept of it. Thus, appealing to the source condition alone is not a promising argument for the claim that agents lack direct control over actions that issue from their endowment, their modified endowment, or neither.56 Hartman’s critique here undermines the principle that external influence compromises an agent’s responsibility for their actions and decisions and, by extension, their being the source of those actions and decisions. Hartman goes on to show that a commitment to such absolute self-determination would entail that almost no decisions, if any, made by agents could actually be directly controlled by the agent.57 Thus, agents would never truly be morally responsible for any actions or choices and could never be the source of an action or choice in a meaningful way. Finally, Hartman undermines what Levy calls the Responsibility Negation Premise, which states that “constitutive and present luck each negate moral responsibility.”58 Hartman argues that the premise is false through an assessment of each control-mitigating property of both constitutive and present luck. While there is much to unpack within this assessment, one conclusion which Hartman draws is of significant impact for our considerations here: he concludes that “we may have responsibility-level control with regard to some events over which we lack direct control.”59 This means that the level of control we require over any given action or choice in order for it to be made freely does not need to be absolute as the earlier accounts would seem to imply. While Hartman concedes that “each condition of present and constitutive luck individually mitigates responsibility-level control,” he avers that “not even their joint satisfaction negates responsibility-level control.”60 This would seem to indicate that decisions can be made freely by subjects who are not only partially self-determined but also partially determined by external factors. This seems consonant with Paul’s vision for freedom in Christ. In particular, Paul envisions individual human freedom in a way that relies noncausally on others. Thus, the influence of other persons to whom we belong partially determines our decisions, but does not entirely determine those decisions. Some degree of self-determination is maintained. Insofar as we might think such a degreed account of self-determination is sufficient for free will, we can maintain a sourcehood account of libertarian free will in which the

Free Together 131 source of a particular action non-causally relies on others for their powers of self-determination. A Pauline Reframing of Sourcehood

With this having been granted, Paul’s vision for freedom in Christ still has more reframing to do. While accounts such as Hartman permit external influence in ways which do not undermine self-determination, Paul would seem to demand it. Indeed, self-determination for Paul would seem to be impossible apart from our belonging to others in community. Thus, the partial determination of our self-determining powers is essential to a Pauline account of freedom or free will. This is a stronger claim than the permissive claim concerning external influence drawn from Hartman. External influence is a necessary condition for Paul’s vision for freedom. However, this only applies to some kinds of external influence. It is specifically other persons in interpersonal relation, which are so necessary for our self-determining powers. We do not, in Paul’s vision, belong to general circumstances in the same way or to the same degree that we belong to other persons with whom we are in community. It is persons, both divine and human, who constrain our freedom necessarily in this way. So what is it that the persons to which we belong and the groups in which we belong to them contribute to our self-determination? Said another way, given that the constraints, which these external influences place on our selfdetermination, are non-causal, in what ways do they constrain us? Returning once again to Paul’s exhortations concerning the eating of meat previously offered to idols, we see that Paul recognizes certain individual liberties, which are given up or taken up in virtue of belonging to others and with a vision for how this contributes to the freedom of the group as a whole. Again, there are two aspects to this vision, which Paul has for human freedom: the belonging to others and the telos of individual freedom in the freedom of the community. Hence, what determines whether a particular individual liberty ought to be taken up or laid down for the sake of others is how that contributes to the freedom of the group’s agency. In this way, we might think of others as providing or removing necessary (but not sufficient) conditions for a particular free action or choice. Because the conditions provided by others are not sufficient, the provision of these conditions does not necessarily entail that the action or choice is taken, and thus ensures that relation to others does not causally determine the actions of the given agent. It only partially determines those actions. It is still for the agent in question to deliberate on the available conditions and determine, albeit partially, what choice or action to make. As this all takes place in its proper context in community, the group determines which conditions individuals can provide or remove from others within the community. The telos of individual freedom in group freedom is important here. For, we might conceive of a member of the body of Christ belonging to a weaker

132  D. T. Everhart member and still determining to eat meat sacrificed to idols despite the relational constraint. How, one might ask, is this really a constraint and not simply a reason to use our free will in a certain way? Because the telos of that free choice is its contribution to the freedom of the whole community, this member’s selfish decision to act as if they are not constrained by their belonging to the weaker member has the effect of diminishing the community’s freedom as a group. Acting as if atomism is true, as opposed to holism, is to act contrary to our communal nature. Despite the fact that this selfish act was made freely by the individual in question, it results in diminishing the communal freedom for which individual freedom ultimately exists. Remembering that, on Collins’ model, group agency arises from the coordination of individual agents, the act of an individual in contradiction to the constraints of coordination will result in a breakdown in the agency of the group because a group that is not coordinated cannot have agency. A group without coordination cannot make choices as a group and thus is not free. Because the individual belongs to this community, their individual freedom is diminished by the diminishing of the group’s freedom. To illustrate this, let us imagine a ruler of a kingdom, Lerand. For all intents and purposes, let us suppose that Lerand is the source of his own choices, and therefore has free will. However, Lerand has many responsibilities that come with being a ruler, such as a responsibility to protect his kingdom from invaders. These responsibilities, on the earlier model, constrain what he is to do with some of his freely made choices. Now, we can conceive of Lerand ignoring these responsibilities and doing as he pleases with his powers of self-determination. By acting in contradiction to the constraint on his freedom to protect his kingdom, Lerand leaves his kingdom vulnerable to invasion. This inevitably happens, and the kingdom is conquered by another kingdom which takes its conquered victims as slaves. Now many of Lerand’s decisions are determined for him, such as what he does with his day, what he eats, and where he sleeps. His decisions become more constrained in virtue of his choosing to act as if atomism is true and his responsibilities to his kingdom do not constrain his choices. His kingdom is now entirely enslaved and unable to act freely as a sovereign nation, and as this is the context for his own individual freedom, his freedom as an individual is more constrained than it would have been had he acted in accordance with the constraints of his kingship. Lerand is the most free when he acts in accordance with his belonging to his group. This is not to say that every constraint of belonging to a group ought to be followed, as illustrated in the tension of belonging between freedom and slavery that Jennings raises. He writes concerning the building of community, building . . . is a gift given to us by God . . . that speaks our destiny, and it is inescapable. We build either toward life or toward death. . . . It takes a discernment that can see when institutional operations are

Free Together 133 moving in the right direction, spiraling up toward life and away from death . . . toward either captivity or freedom, and of encouraging the motions that aim to form a place of life together.61 Constraint can in fact be more freeing than mere individual autonomy when coordinated toward the freedom of the group rather than its captivity. Free Together The freedom for which human persons were made and in which human persons flourish is not simply the freedom of absolute and merely autonomous self-determination. We are created for community with God and fellow humanity in the body of Christ. As J. B. Torrance puts it, “the triune God is in the business of creating community in such a way that we are never more truly human, never more truly persons, than when we find our true beingin-communion.”62 I take what Torrance means here by being “truly human persons” to be a kind of flourishing. As such, it includes free will. To restate this aspect of Torrance’s point: the triune God is in the business of creating community in such a way that we are never more truly free, never more truly self-determined, than when we find our true freedom in communion. This true communion for which we are created is the body of Christ, a communion that encompasses our belonging to God and to one another in authentically freeing ways. Let us then take stock of how Paul’s vision for freedom in Christ helpfully reframes libertarian sourcehood. First, human persons belong to one another and to God in such a way that we are not free, self-determining agents apart from our non-causal dependence on others. Our sourcehood in virtue of which we have free will relies on our belonging to others in community. Second, in virtue of this non-causal dependence, we cannot say that we are entirely selfdetermined, nor that we are entirely determined by others. Rather, we can say that we are partially and subveniently determined by others and partially self-determined. Insofar as we think this is sufficient for sourcehood, we can maintain that agents belonging to others in community have free will. Finally, our self-determination ought to be thought of as having its telos in the freedom of the group through the constraints placed upon our self-determination by belonging to others. The self-determination of individuals is not an end in and of itself, but rather a means to the end of the freedom of the group. While we cannot say for certain that the telos of belonging to every group is freedom, we can say this of the body of Christ. In the body of Christ, we are constrained by our belonging to others so that the whole body might be free in Christ. This certainly includes degreed accounts of sourcehood, but contextualizes those accounts with their proper end of our being free together. It is for the sake of being free together that Christ has set us free and drawn us into a community of belonging to him and to one another.

134  D. T. Everhart Notes 1 Harry Frankfurt, “Alternative Possibilities and Moral Responsibility.” The Journal of Philosophy 66 (1969): 829. 2 Kevin Timpe, Free Will, 2nd edition (London: Bloomsbury, 2013), 70. 3 Linda Zagzebski, “Does Libertarian Freedom Require Alternate Possibilities?” Philosophical Perspectives 14 (2000): 234. 4 Zagzebski, “Does Libertarian Freedom Require Alternate Possibilities?,” 233–4. While Frankfurt aimed therefore to undermine libertarian free will, Zagzebski goes on to demonstrate that libertarian accounts also do not require alternative possibilities. 5 Zagzebski, “Does Libertarian Freedom Require Alternate Possibilities?,” 234. This is Zagzebski’s adaptation of Frankfurt’s original example of a neurological device which prevents alternative possibilities. 6 Zagzebski, “Does Libertarian Freedom Require Alternate Possibilities?,” 245. 7 Zagzebski, “Does Libertarian Freedom Require Alternate Possibilities?,” 245. 8 Timpe, Free Will, 10. 9 Thomas Nagel, “Moral Luck.” in Mortal Questions (Cambridge: Cambridge University Press, 1979), 25. 10 Susan G. Eastman, Paul and the Person: Reframing Paul’s Anthropology (Grand Rapids: Eerdmans, 2017), 8. 11 Eastman, Paul and the Person, 69. 12 C. Stephen Evans, “Who is the Other in Sickness Unto Death?: God and Human Relations in the Constitution of the Self.” Kierkegaard Studies Yearbook 1997 (1997), 2. 13 Constantine R. Campbell, Paul and Union with Christ: An Exegetical and Theological Study (Grand Rapids: Zondervan, 2012), 406. 14 Campbell, Paul and Union with Christ, 287–8. I argue this at length elsewhere: D.T. Everhart, “Communal Reconciliation: Corporate Responsibility and Opposition to Systemic Sin.” International Journal of Systematic Theology 25(1) (Jan 2023): 144–5. 15 Eastman, Paul and the Person, 182. 16 Willie James Jennings, After Whiteness: An Education in Belonging (Grand Rapids: Eerdmans, 2020), 14–15. 17 Campbell, Paul and Union with Christ, 381. 18 James D.G. Dunn, The Theology of Paul the Apostle (Grand Rapids: Eerdmans, 1998), 406. 19 Erin Heim, Adoption in Romans and Galatians: Contemporary Metaphor Theories and the Pauline Huiothesia Metaphors (Leiden: Brill, 2017), 328. 20 Everhart, “Communal Reconciliation,” 146. 21 Jennings, After Whiteness, 16. 22 Esau McCaulley, Reading While Black: African American Biblical Interpretation as an Exercise in Hope (Downers Grove: IVP Academic, 2020), 60. 23 Bruce W. Longenecker, “Paul’s Assessment of Christian Freedom.” in Freedom (Christian Reflection: A Series in Faith and Ethics) (Waco: Baylor University Press, 2011), 20. 24 Everhart, “Communal Reconciliation,” 22–23. 25 Longenecker, “Paul’s Assessment of Christian Freedom,” 24. 26 Longenecker, “Paul’s Assessment of Christian Freedom,” 27. 27 Longenecker, “Paul’s Assessment of Christian Freedom,” 24. 28 See, Thomas McCall, “Crucified with Christ: The Ego and the Omega.” Journal of Analytic Theology 8 (2020): 2–7. 29 Simeon Zahl, “Beyond the Critique of Soteriological Individualism: Relationality and Social Cognition.” Modern Theology 37(2) (2021): 345.

Free Together 135 0 Eastman, Paul and the Person, 179. 3 31 Jennings, After Whiteness, 12. 32 Jennings, After Whiteness, 12. Jennings is here concerned with translation of God’s revelation between an I and Thou, though with an aim for a broader sense of otherness in community in which the self is both bound and free. 33 Stephanie Collins, Group Duties: Their Existence and Their Implications for Individuals (Oxford: Oxford University Press, 2019), 157–9. 34 Collins, Group Duties, 95. 35 Collins, Group Duties, 169. 36 Collins, Group Duties, 74. 37 Collins, Group Duties, 132. 38 Philip Pettit, The Common Mind: An Essay on Psychology, Society, and Politics (Oxford, UK: Oxford University Press, 1996), 118. 39 Pettit, The Common Mind, 175. 40 Pettit, The Common Mind, 177. 41 Pettit, The Common Mind, 178. 42 Pettit, The Common Mind, 178. Pettit also goes on to strengthen this claim by rendering relation to the self across time as insufficient for thought in actuality but not in principle. 43 Pettit, The Common Mind, 171. 44 Pettit, The Common Mind, 190. 45 Barbora Sipova and Melinda Carpenter, “A  New Look at Joint Attention and Common Knowledge.” Cognition 189 (2019): 260–74. 46 Gideon Salter  & Richard Brehany. “Removing Shared Information Improves 3-and 4-Year-Olds’ Performance on a Change-of-Location Explicit False Belief Task.” Journal of Experimental Child Psychology 187 (2019). 47 Stephen Reicher, Russell Spears  & Alexander Haslam S., “Social Identity Approach in Social Psychology.” in Meaghan. S. Wetherell & Chandra Sekhar Mohanty (eds.), Sage Handbook of Identities (London: Sage Publishers, 2010), 48. 48 Reicher, et al., “Social Identity Approach,” 50. 49 Reicher, et al., “Social Identity Approach,” 52. 50 D. T. Everhart, “Communio Dei and the Mind of Christ: Relational Christological Anthropology in Psychological Perspective.” Theologica 6 (2022), 13. 51 See, Galen Strawson, “The Impossibility of Moral Responsibility.” Philosophical Studies 75 (1994): 16–20; Nagel, “Moral Luck,” 35–38; Thomas Nagel, The View from Nowhere (Oxford: Oxford University Press, 1986), 118. 52 Neil Levy, Hard Luck: How Luck Undermines Free Will and Moral Responsibility (Oxford: Oxford University Press, 2011), 84–97. 53 Levy, Hard Luck, 85–86. 54 Levy, Hard Luck, 88. 55 Levy, Hard Luck, 94–97. 56 Robert J. Hartman, In Defense of Moral Luck: Why Luck Often Affects Blameworthiness and Praiseworthiness (New York: Routledge, 2017), 49. 57 Hartman, In Defense of Moral Luck, 50. 58 Hartman, In Defense of Moral Luck, 50. 59 Hartman, In Defense of Moral Luck, 52. 60 Hartman, In Defense of Moral Luck, 55. 61 Jennings, After Whiteness, 76. 62 James B. Torrance, Worship, Community, and the Triune God of Grace (Downers Grove: IVP Academic, 1996), 73.

8 Free Will, Cognitive Biases, and Theology Simon Kittle

Introduction Human agency is the power to act on, and exercise control over, objects in the world. It is what allows us “to direct [our] own lives.”1 Free will is the kind of human agency which, when exercised, makes us morally responsible for the things we decide and do. The kind of moral responsibility in question is basic “in the sense that the agent, to be morally responsible, would deserve the blame or credit just by virtue of having performed the action.”2 I agree with Greene and Cohen that the ordinary concept of human freedom, and the moral and legal practices built on top of that notion are “ultimately grounded in intuitions that are incompatibilist and, more specifically, libertarian.”3 In addition, I  think that “the considerations which motivate incompatibilism in the natural, causal case also motivate incompatibilism in the theological case.”4 As such, in what follows, I assume that free will is incompatible with both causal determinism and theological determinism. I hold that our ordinary concept of agency and free will takes the locus of control relevant to freedom to be conscious choice: the selection by the conscious subject of one from among two or more alternatives (I defend this claim in Section  3). In what follows, I  consider whether certain empirical findings on the nature of human cognition – namely, those to do with heuristics and biases – threaten this view of free will by undermining some of the conditions required for conscious choice. The question to be considered is relevant to those theological traditions which hold that people can be morally accountable to God in the basic desert sense. I take it that this includes the dominant streams of the Western Christian theological tradition, and many other theological traditions besides. Often, such theologies invoke basic desert moral responsibility because they are committed to some form of retributive justice: the idea that agents who are morally responsible for wrongdoing deserve not just blame but also punishment, and that it can be an intrinsic good for such agents to be punished.5 Proponents of such theologies should be concerned by any potential threat to the existence of human agency and free will since if such threats are shown to be genuine, the basis for human moral responsibility before God disappears. DOI: 10.4324/9781003306191-10

Free Will, Cognitive Biases, and Theology 137 Although the implications of such an outcome will vary a little according to the specific theology in question, the implications for most retributivist theologies are likely to be vast because free will “spreads its tentacles all around systematic theology.”6 What Free Will Requires The ordinary concept of free will centers on the notion of conscious choice. As Alfred Mele states, “we would like to believe that we do form intentions, that we make up our minds – or decide – to do certain things and not to do others.”7 We would like to so believe, and ordinarily we do so believe. And who are the agents Mele refers to with “we?” Who are the beings who form intentions and make up their minds? Conscious human subjects. We take ourselves to be conscious creatures who (for the most part) know what we believe, know what we want, weigh up reasons for acting, deliberate, decide, and act. The selection of one action from among two or more possible courses of action is the central characteristic of this control. As Bernard Williams put it, free will requires that a person has “a real choice between more than one course of action.”8 So we decide and act freely only when we are able to decide and act otherwise; minimally, we decide and act freely only if we could have refrained from so deciding and acting. This is what leads Hannah Arendt to write that “[t]he touchstone of a free act is always our awareness that we could also have left undone what we actually did.”9 Conscious Control Is What We Want

In this section, I will argue that as far as agential control goes, it is conscious choice-based control which we do (and should) care about. Consider the following example: (Alyssa) It is Saturday afternoon and Alyssa was due to spend the afternoon with a friend who has cancelled. Alyssa thus faces a decision about what to do instead. Alyssa’s situation is this: she has a few errands which don’t need doing right away but do need sorting in the next few days; she’s nearing completion of an onerous but soon-to-be-beneficial DIY project which could easily be finished off in a single afternoon; she enjoys mountain-biking and recently identified a new trail she’d like to check out; she also likes to spend her spare time chatting with friends and reading, both of which she finds relaxing. Now suppose that Alyssa cannot make conscious choices. She’s just hung up the phone after speaking with the friend she was due to meet and so knows that her friend can’t make it and that she must now figure out what to do instead. But she also knows that she (qua conscious subject) cannot do this. Alyssa (qua conscious self) cannot decide what to do – she doesn’t have

138  Simon Kittle conscious control over the intentions she forms. Alyssa just finds herself doing various things, and then at some point it becomes apparent to her what she’s doing – that is, it becomes apparent to her conscious self what various non-conscious processes within her have “decided” she will do. Sometimes she becomes consciously aware of a judgment about what would be best to do first, and so she gets a bit of warning about what she’s about to start doing. But this is of only minor benefit to Alyssa (qua conscious self) because Alyssa can’t choose to deliberate further instead, she can’t choose to direct her attention here or there, she can’t choose to imagine what it would be like to do this instead of that – all of those would be mental acts of choice that Alyssa qua conscious self would make, but Alyssa can’t exert any choicebased control. Alyssa qua conscious self is a mere passive observer of those things which flow down her stream of consciousness. Nothing Alyssa qua human being does is up to Alyssa qua conscious subject; everything she does is settled (let’s assume) by cognitive and non-cognitive processes which operate unconsciously. I take it that this would not be a desirable situation to be in. Of course, since conscious decision-making can be hard work, we are sometimes tempted to think it is a bit of a burden. And so we might on occasion suppose that it would be quite nice to be able to sit back and just watch the stream flow, as it were, without having to expend any effort or stress deciding what to do. But the novelty would soon wear off. To be a conscious subject without any conscious control would be a scary prospect. To see this, just note that Alyssa’s conscious self – the one who is going to experience both the good and the bad – is entirely at the mercy of the unconscious cognitive processes driving her behavior. If those processes produce behavior which leads to a flourishing, fulfilling life, then Alyssa (qua conscious self) is fortunate; if they produce behavior which leads to destructive behaviors, Alyssa is unfortunate. Either way, however, from the point of view of Alyssa’s conscious self, it is entirely a matter of luck what she experiences. Her condition would not, I submit, be sought out by anyone. The Alyssa example suggests that what we (qua conscious selves) want is some measure of conscious control over what we (qua biological organisms) do. More specifically, what we want, at least sometimes, is a choice about what we think and do. This conclusion therefore imposes a cost on any view of free will according to which it doesn’t require conscious control. One such view is Stump’s Thomist account of free will. Stump is a so-called source incompatibilist: free will is incompatible with determinism, but free will does not require being able to do otherwise. Stump formulates the following two conditions to capture this notion of freedom: (L1’) [an act] is free only if it is not the outcome of a causal chain that originates in a cause outside the agent. (L’) an act is free if and only if the ultimate cause of that act is the agent’s own will and intellect.10

Free Will, Cognitive Biases, and Theology 139 Condition (L1’) rules out a free act’s being determined by factors external to the agent. Condition (L’) ensures that free decisions are not simply the result of random biological processes internal to the agent but must have their ultimate source “in the proper functioning of the agent’s own intellect and will.”11 On Stump’s view, the intellect is that cognitive subsystem which makes judgments concerning what is good while the will is a rational appetite.12 Neither the operation of the intellect nor that of the will requires any form of conscious control. As Stump makes clear, the intellect might generate a belief about what is good without any conscious awareness on the agent’s part.13 And since the will is an appetite, and thus simply generates desires for whatever the intellect judges to be good, it too may operate outside of conscious awareness. Stump writes: [None of the operations of the intellect and will which together produce a voluntary action] need to be in the forefront of consciousness, done with transparent awareness. Any of these acts might be tacit, acts of which we could be made aware only by careful questioning and introspection after the fact. There is . . . a continuum of accessibility to consciousness, from acts done with self-conscious awareness to acts which can be brought up into consciousness only with probing by professional therapists.14 The last two sentences from this quotation make clear that, on Stump’s account, a person may act freely even if they have no conscious awareness at all of what they are doing at the time at which they are doing it; according to Stump, there may be voluntary and freely performed acts which “can be brought up into consciousness only with probing by professional therapists” – which seems to imply that there may be freely performed acts which the agent is not consciously aware of performing (at the time at which they are performed). Ekstrom’s event-causal libertarian account of free will may also allow for free decisions of which the agent is not conscious. On Ekstrom’s account, a decision occurs when some set of the agent’s “preferences, convictions, desires, values and beliefs” non-deviantly cause in the agent an intention to act, and where that causation is not the result of manipulation or coercion.15 But since an agent may have preferences, desires, values, and beliefs of which she is not conscious, it is unclear what in this picture would require the involvement of the conscious self (for more on this point, see my commentary on Ekstrom’s book16). In any case, whether or not Ekstrom’s account allows particular free decisions to be made without the agent’s conscious awareness of those decisions, the conclusion I am seeking to draw here from the Alyssa scenario is that it is a defect of any account which does allow this. What the Alyssa example shows, in other words, is that we want conscious control over (at least some of) what we think and do and that conscious control is essentially tied up with the agent performing mental acts of selection among different options.

140  Simon Kittle The conclusion that the sort of control we value is conscious, choicebetween-alternatives-based control is also supported by our moral and legal practices. If someone performs an action (even a complex, goal-directed action) and we become convinced that the person did not consciously choose to perform the action, we tend not to hold the person responsible. A vivid example of this is the case of Kenneth Parks.17 In the early hours of May 24, 1987, Parks drove 15 miles to his in-laws’ house. He entered the house and attacked them, killing his mother-in-law. He then drove to a police station whereupon he said that he felt he may have killed someone. Parks was charged. He pleaded not guilty on the grounds that he was suffering from non­insane automatism – very roughly, that he was asleep. Parks was eventually acquitted. Assuming that Parks was indeed suffering from a non-insane automatism, his acquittal seems to me to be exactly the right response. Similar things can be said of more run-of-the-mill cases. If someone (say) scares another member of the family while sleepwalking, we don’t blame the sleepwalker because we recognize that the sleepwalker could not have done any differently. If someone is under a great deal of stress, or has just been bereaved, or is in a great deal of pain, and then does something wrong, we modulate their responsibility in accordance with the mitigating factor because we recognize that the factor in question made it harder to choose to do otherwise. Or at least, my claim here is that we should so modulate the person’s responsibility. And to the degree that one shares these intuitions, one has reason to agree with my basic contention that the control relevant to free will is that associated with the possession of a choice between alternatives. What Does Conscious Choice-based Control Require?

If the possession of a conscious choice is to bestow significant control on an agent, the agent must possess a range of related cognitive abilities. This is because the ordinary concept of human agency and freedom is committed to what John Doris calls reflectivism, which is: [The] doctrine according to which the exercise of human agency consists in judgment and behavior ordered by self-conscious reflection about what to think and do. Typically, this doctrine is associated with a corollary: the exercise of human agency requires accurate reflection. In an exercise of agency, as construed by reflectivism, a person correctly divines the beliefs, desires, and other psychological states relevant to her decision, makes her decision in light of these states (sometimes called her reasons), and acts accordingly.18 Reflectivism states that ordinary human decision-making requires accurate, self-conscious reflection, which in turn requires accurate access to the beliefs, desires, and other attitudes which are relevant to the choice at hand. Reflectivism – or something close to it – is plausibly thought to be required

Free Will, Cognitive Biases, and Theology 141 by the ordinary concept of free will because we conceive of the choices we make as being, at least to some degree, rational. And rational agency embodies a commitment to what Bhaskar calls the axiological imperative: the idea that “we appraise reasons qua beliefs for consistency, truth, coherence,” and their suitability for being acted on – we don’t just treat reasons as yet another natural cause.19 If, then, possessing a choice is going to bestow control, at a minimum the following two conditions will need to be met: Accessibility: The agent must have conscious access to those of their own beliefs, desires, and other propositional attitudes which are relevant to the decision, and this access must be reasonably comprehensive. Processing: The agent must be able to process – weigh, compare, evaluate, investigate, revise, etc. – the attitudes relevant to the decision, and this processing must exhibit certain integrity. These conditions are specified functionally; they correspond roughly to the first two “very general capacities” required for freedom that O’Connor lists as accepted by many compatibilists and incompatibilists alike and which “seem compelling from a subjective, pretheoretical point of view.”20 Separating these conditions out in this way is not essential to the point being made here; nevertheless, doing so helps us to get clear about what is needed for choice-based conscious control. What I will suggest in the remainder of this section is that, to the degree an agent fails to satisfy these conditions – and both of them can be satisfied to a greater or lesser degree – to that degree is that agent’s conscious, choice-based control compromised. Let us begin with Accessibility. To see how this condition might fail to be satisfied, return to the Alyssa example. Suppose now that Alyssa (qua conscious self) is able to make conscious choices. So Alyssa gets off the phone with her friend and she must decide what to do. Alyssa (qua conscious self) starts to deliberate about what to do – that is, she chooses to direct her conscious attention to the question of what to do, with the intention of making up her mind. Suppose that Alyssa entirely fails to bring to mind any of the salient options regarding her Saturday afternoon. She sits there patiently, but nothing at all comes to mind. Clearly, if Alyssa suffered such a failure of Accessibility with respect to her salient options, the power to make a conscious choice about what to do would be much less useful. Alyssa may still pick something to do, perhaps working from first principles or very general beliefs about what people tend to do on Saturdays, but in such a scenario, it’s likely her decision would not be integrated with her current projects, her hobbies, her life plans, and so on. Alternatively, Alyssa might experience a partial failure of Accessibility. Perhaps Alyssa directs her attention to the question of what to do, and she remembers only that she has some errands to run and that she has a DIY project she needs to finish. After recalling these two options, Alyssa asks herself,

142  Simon Kittle “And was there anything else I could do?” but nothing else comes to mind. Such a partial failure of Accessibility compromises her control to lesser (but not insignificant) degree. Both of these failures concern a lack of conscious access to the relevant attitudes. Another way that Accessibility might fail is if Alyssa’s recall of the attitudes were inaccurate; that is, if Alyssa could recall the attitudes, but doing so corrupted them or associated attitudes in some way. For instance, if Alyssa recalled that she had some errands to do but erroneously believed that the errands didn’t need doing anytime soon, or mistakenly judged that they’d take two full days to complete – such mistakes would compromise the appropriateness of the choice made. Similarly, if Alyssa remembered about wanting to try out a new bike trail but, when considering the pros and cons of going for a bike ride, concluded that she’d hurt her knee on her last ride and needed to give it more time to heal (when in reality her injury had healed months ago), then her power to choose from among alternatives would be less useful. In such cases, the mismatch between the content of her attitudes and reality makes the control she can exercise through her decision-making much less useful. Alyssa would experience a failure of Processing if, for example, she decided not to contact any of her other friends, not because she had a desire to spend a bit of time alone, but because (say) she happened to be feeling cold at the time of making the decision and when thinking about her options the experience of being cold transferred over to the possibility of spending time with friends such that her friends appeared colder toward her. This would contravene the axiological imperative because Alyssa’s currently being cold is (almost always) an a-rational consideration as far as spending time with friends is concerned. The Alyssa example is a toy example, but it is nonetheless useful in making vivid how an agent must possess a number of other cognitive capacities if an agent’s possessing a choice between options is to be useful. Cognitive Biases and the Fragility of Human Cognition In this section, I present empirical findings which challenge the understanding of free will so far sketched. I focus primarily on the phenomenon of priming. Priming refers to the phenomenon whereby exposure to one thing – the prime – influences some other aspect of an agent’s cognition or action. Almost anything can serve as a prime: words, odors, noises, locations, someone else’s behavior, and more. And many aspects of a person’s mental life and overt behavior can be influenced by a prime. Crucially, even when a subject is consciously aware of the prime, but not aware of it as a prime, it may still influence the person’s behavior. And usually the influence that the prime exerts contravenes Bhaskar’s axiological principle. That is, from a rational point of view, primes are typically irrelevant to the decision made or behavior performed. Primes therefore threaten to undermine the rationality of a person’s thought processes, and thus

Free Will, Cognitive Biases, and Theology 143 the control exercised via rational decision-making. It is for this reason that Daniel Kahneman says that “[s]tudies of priming effects have yielded discoveries that threaten our self-image as conscious and autonomous authors of our judgments and our choices.”21 One of the most famous priming studies was performed by John Bargh and colleagues and is known as the “Florida Effect.”22 Bargh et al. gave participants aged between 18 and 24 a set of five words and asked them to make four-word sentences. One group of participants was given words related to old age, such as “gray” and “wrinkle.”23 At (what the participants thought was) the end of the study, the participants were asked to walk down the hall and complete another exercise. The walk down the hall was the real object of study. What Bargh found was that participants who’d been given words related to old age walked down the hall “significantly more slowly” than the control group. The participants denied that the words they’d seen had had any effect on their behavior and confirmed that they had not had any conscious thoughts about old age while performing the word task. This study involved two forms of priming. The words related to old age primed the concept of old age and then the concept of old age primed the subject’s subsequent walking behavior. Since the participants’ walking down the hall was an intentional action, on the face of it freely performed, this study provides an example where one aspect of a person’s intentional action – the speed of walking – is influenced by a prime of which the subject was entirely unaware. One might suggest that, as impressive as Bargh’s Florida study may be, it nevertheless lacks significance. After all, presumably the participants walking down the corridor had no reason to walk one speed rather than another, so the fact that the speed of their walking was affected by a prime was inconsequential. Indeed, one might go further and suggest that if the participants had had a reason to walk quickly down the corridor (say), their conscious recognition of that reason would have overridden any effect produced by the old-age prime. But while the significance of Bargh’s study, especially for moral action, may be questioned on these grounds, other studies establish that priming effects often pertain to highly significant aspects of decision and action. One area in which the consequences of priming are significant is the criminal justice system. Studies done on primes which produce anchoring affects are of particular import here. Anchoring effects occur when a person who must make a judgment about some quantity sees, prior to making their judgment, a number which influences their subsequent judgment. In one study by Englich et al., German judges (with an average of 15 years of experience) read a case description of a woman who had been caught shoplifting.24 They were then instructed to roll a pair of dice and to report whether they would sentence the woman to a number of months in prison greater or fewer than the total shown on the dice. Finally, they were asked to give the actual sentence they would give the woman. Unbeknownst to the judges, the dice were loaded to come up with either a 3 or a 9. On average, the judges who rolled a

144  Simon Kittle 3 sentenced the woman to 5 months, whereas those who rolled a 9 sentenced the woman to 8 months. In cases like this, where there are two different anchors employed across different trials, it is possible to calculate a measure of how significant the anchoring effect is. The measure, known as the anchoring index, is the ratio of the difference between the judgments and the difference between the anchors expressed as a percentage. An anchoring index of 0% would indicate that the anchor had had no effect; an anchoring index of 100% would indicate that people simply adopted the anchor as their judgment. In the aforementioned Englich et al. case, the anchoring index is thus (8–5)/(9–3) = 0.5 = 50%. An anchoring effect of 50% is remarkable given that (a) the judges are supposed to be drawing on their extensive expertise when delivering sentences and (b) the roll of some dice is clearly irrelevant to what makes for an appropriate sentence. How does anchoring work? One theory is that the judgment is made using the availability heuristic, and this heuristic is affected by the anchor. Kahneman identifies two sorts of cognitive systems: quick, automatic, and unconscious processes which are collectively labeled System 1; and slow, deliberate, conscious cognitive processes collectively labeled System 2. Heuristics are where an automatic System 1 process generates an answer to a difficult question by substituting the difficult question for a much easier one. The hypothesized availability heuristic is thought to be a cognitive process which answers difficult questions about the frequency of some phenomenon or the strength of some character trait by replacing them with (more easily answerable) questions about the ease with which things come to mind. For example, if someone is asked to judge how dangerous flying is – the proper answer to which would require knowledge of flight statistics, passenger numbers, injuries, and fatalities due to flying per year, and so on – a System 1 process will produce an answer based on, for example, how easily the person can recall instances of plane crashes. And since plane crashes are widely reported in the media, people will tend, especially in the weeks and months after a crash, to overestimate how dangerous flying is.25 The fourth study with legal professionals reported by Englich et al. sought to test whether an availability heuristic was being used in the sentencing cases, and the answer seemed to be a “yes.”26 According to Englich et al., the anchor on the dice affects the judges’ sentencing by influencing how easy it is for judges to recall the incriminating arguments offered by the prosecution. Rolling a 9 makes it easier for judges to recall the arguments offered by the prosecution, and this pushes the judges’ sentencing upwards; rolling a 3 makes this harder, and so pushes the judges’ sentencing downwards. More support for the idea that priming effects often have significant real-world consequences comes from a study by Northcraft and Neale and concerned real-estate agents who were asked to visit a property, read an information booklet, and then give an expert opinion on the property’s value.27 The booklet included a listing price; half the real-estate professionals saw a

Free Will, Cognitive Biases, and Theology 145 figure substantially lower than the actual listing price, and half saw a figure substantially higher. After being asked for their valuation, the agents were asked what factors had influenced their judgments. Around 75% of estate agents denied that the asking price had influenced them; indeed, some of them took pride in their expert ability to ignore it. Only, they didn’t ignore it. The listing price in the booklet produced an anchoring effect of 40.9%.28 How much of a problem are priming and anchoring effects for our ordinary view of human agency and freedom? It must be recognized to begin with that there is some dispute concerning the cognitive processes which give rise to the observed biases and errors. While thinkers such as Englich et al.,29 Kahneman,30 and Bargh31 propose that automatic cognitive processes operating outside conscious awareness give rise to many priming and anchoring effects, others dispute this. Newell and Shanks,32 for instance, argue that the evidence suggests that primes and anchors only have significant effects when they are consciously attended to (though not, of course, consciously attended to as a prime), and that therefore explanations of priming and anchoring effects can be given purely in terms of deliberative, conscious cognitive processes. And certainly caution in interpreting the results is warranted. Nevertheless, whatever the nature of the cognitive processes which produce the cognitive biases and systematic errors, there is wide agreement that priming and anchoring effects are “ubiquitous”33 and “robust.”34 Moreover, primes and anchors have a significant effect even when they are obviously irrelevant to the judgment at hand, and thus “anchors do not have their effects because people believe they are informative.”35 This means that primes and anchors are, in effect, a-rational factors which nevertheless influence the contents of propositional attitudes and behaviour which is assessed rationally. Additionally, as the studies by English et  al. and Northcraft and Neale show, the real-world consequences of priming can be both financially and morally significant. Indeed, as Englich pointed out in a review article on priming effects in criminal justice systems, the real-world consequences for criminal justice systems pose a challenge to the neutrality of such systems not just because sentencing decisions may be subject to (say) arbitrary anchors, but because the structure of many criminal justice systems means that it is the prosecution which makes a first sentencing demand, with the result that it produces an anchoring effect on both the defense’s subsequent sentencing request and the judge’s final decision.36 Thus, the structure of the criminal justice system bestows undue power on the prosecution. Returning to those additional cognitive capacities which I suggested are required if the ability to make conscious choices is to bestow a useful form of control, we can see there is good reason to think that priming effects undermine Accessibility and Processing. Accessibility is compromised since primes can affect which items come to mind and how quickly they do so. And this then feeds into (and compromises) Processing by affecting the data used to form subsequent judgments. Other much-studied cognitive biases which yield predictable and systematic errors in human cognition similar to those

146  Simon Kittle described earlier include confirmation bias, framing, exposure effects, Halo effects, and base-rate neglect. Many of these biases and errors compromise to some degree the control that an agent can exercise through the making of conscious choices. Indeed, even just the discovery that the contents of many of our propositional attitudes are generated in part by unconscious cognitive processes which generate systematic errors in predictable contexts threatens to undermine Processing. It may be objected that to conclude from the sorts of empirical findings detailed earlier that human free will is undermined or seriously compromised would be to (a) overgeneralize and – more problematically – (b) rely on contested empirical results. And it must be acknowledged that interpreting empirical results is a perilous task, especially for the armchair philosopher or theologian. Moreover, since the notorious replication crisis in psychology (and beyond) came to light in 2012, matters are even more difficult to detangle. Certainly, the replication crisis should be a warning to anyone wishing to draw philosophical conclusions from empirical results, especially empirical results which appear highly counterintuitive. The replication crisis has revealed that many results in social psychology were nothing more than statistical artifacts, and this may turn out to be true of many more as yet unreplicated studies. On the other hand, it is far from clear that the replication crisis can serve as a global debunking card allowing theorists to dismiss all such results. To see why, consider the large-scale, open science initiative entitled “Many Labs 2” spanning 45 institutions and 36 countries, the aim of which was to test the replicability of 28 previously published results. Many Labs 2 found that only 14 of the 28 studies replicated (assuming a very strict significance criterion (p < 0.0001)).37 Now, to be sure, a replication rate of 50% isn’t great. But since there have been tens of thousands of studies on phenomena such as priming, if 50% of all the studies in the field turn out to be replicable – indeed, even if only 10% of studies in the field turn out to be replicable – that will still leave hundreds of (now replicated) studies which appear to show that various aspects of human cognition and choice can be affected by factors which should not, from a normative point of view, affect the choice. Joshua Shepherd, in an article summarizing purported scientific threats to free will written after the replication crisis came to light, concluded that “research on the role of largely unnoticed situational influences on behavior . . . [is] more difficult to dismiss” than the purported threat to free will posed by the Libet-style studies in the field of neuroscience.38 And summarizing the situation in 2018, Nomy Arpaly observed that “so many studies portray humans as similarly irrational that even if many of them are debunked, it doesn’t take that many of them to justify the belief that there is a lot of unreason in human behavior.”39 Note too that for results such as those outlined earlier to be problematic, one does not need to hold, as Bargh seems to, that they show that the vast majority of human behavior is outside of our conscious control; the sorts of results canvased earlier

Free Will, Cognitive Biases, and Theology 147 threaten human free will to the degree that they show human behavior to be influenced by factors of which we are unaware. I conclude, then, that while the sorts of empirical results discussed earlier should be judiciously and critically assessed, there is some initial support for a picture of human cognition, decision-making, and action according to which, in some contexts, at least some aspects of our decisions and actions are influenced by a-rational factors without our realizing. These results pose a challenge to those incompatibilist accounts of freedom which attempt to do justice to our ordinary concept of free will as focused on conscious, choice-based control. Implications for Theology What implications does the foregoing have for theology? I noted earlier that there are prominent theological traditions which hew very closely to retributivism, the idea that people are responsible in the basic desert sense to God for how they decide and act, and that (at least in some cases) punishment for such wrongdoing is an intrinsic good. Retributivism requires something approaching the ordinary notion of free will. Thus, even if the ordinary notion of agency and free will is an idealization, still, the further we move from it, the less we can legitimately hold people responsible in the basic desert sense. And this means that to the degree empirical findings such as those surveyed earlier demonstrate that people (qua conscious selves) do not make decisions for the reasons they think they do, do not have access to the full range of considerations, or rely on cognitive processes that exhibit systematic errors in given contexts, to that degree is responsibility in the basic desert sense undermined. The correct response for incompatibilists in the face of such evidence is not, I would suggest, to adopt compatibilism. Instead, the correct move is to accept that people are not as responsible for their decisions and actions as we tend to intuitively think. Moreover, the resulting theological picture need not involve a wholesale rejection of those theological anthropologies which endorse retributivism. Retributivism must be rejected (though this need not be seen as any great loss since there are in any case independent reasons for rejecting retributivism; see, for example, Nussbaum40). But according to many such theologies, sin is a much wider category than moral responsibility. And it may be that part of what theologians have attempted to describe with the language of (say) original sin – for example, people being entangled in sin and unable to extricate themselves from their situation – just is the result of our cognitive processes failing in various degrees to meet the criteria outlined earlier. Theological reflection on the ways in which humans are entangled in sin (construed broadly) could be appropriated even if basic desert responsibility and retributivism are found to be untenable; such historical reflections could be construed as descriptively accurate while being normatively mistaken. Human beings are entangled in bad states of affairs in ways which often cannot be escaped; but precisely because such entanglement is unavoidable, human beings are not morally responsible for such entanglement. Such

148  Simon Kittle a revisionary theology would allow us to do justice to the ordinary notion of choice-based free will (to whatever degree it turns out to be possessed) while also being realistic about the human condition.41,42 Notes 1 John M. Doris, Talking to Our Selves: Reflection, Ignorance, and Agency (Oxford: Oxford University Press, 2015), ix. 2 Derek Pereboom, Living Without Free Will (Cambridge: Cambridge University Press, 2001), xx. 3 Joshua Greene and Jonathan Cohen, “For the Law, Neuroscience Changes Nothing and Everything.” Philosophical Transactions of the Royal Society B: Biological Sciences 359(1451) (2004): 1775–85, 1776. 4 Leigh C. Vicens  & Simon Kittle, “God and Human Freedom.” in Cambridge Elements in the Philosophy of Religion (Cambridge: Cambridge University Press, 2019), 32. 5 Alec Walen, “Retributive Justice.” in Edward N (ed.), The Stanford Encyclopedia of Philosophy, 2020th edition (Zalta, 2020), 1. 6 Aku Visala, “Theology, Free Will, and the Skeptical Challenge from the Sciences.” Theology and Science 18(3) (2020): 391–409, 391. 7 Alfred R. Mele, Motivation and Agency (Oxford, New York: Oxford University Press, 2003), 212. 8 Bernard Arthur Owen Williams, Making Sense of Humanity (Cambridge: Cambridge University Press, 1995), 5. 9 Hannah Arendt, The Life of the Mind: 2 – Willing (New York: Harcourt Brace Jovanovich, Inc., 1978), 5. 10 Eleonore Stump, Aquinas (London, New York: Routledge, 2003), 302, 304. 11 Eleonore Stump, Aquinas (London, New York: Routledge, 2003), 304. 12 Eleonore Stump, Aquinas (London, New York: Routledge, 2003), 278. 13 Eleonore Stump, Aquinas (London, New York: Routledge, 2003), 280. 14 Eleonore Stump, Aquinas (London, New York: Routledge, 2003), 291. 15 Laura W. Ekstrom, God, Suffering, and the Value of Free Will (Oxford: Oxford University Press, 2021), 47ff; see also Laura W. Ekstrom, “Toward a Plausible Event-causal Indeterminist Account of Free Will.” Synthese 196(1) (2019): 127–44. 16 Simon Kittle, “Evaluating the Value of Free Will: Comments on Ekstrom’s God, Suffering and the Value of Free Will” (unpublished manuscript, January 28, 2022). 17 Neil Levy, Consciousness and Moral Responsibility, 1st edition (Oxford, New York: Oxford University Press, 2014), 71ff. 18 John M. Doris, Talking to Our Selves: Reflection, Ignorance, and Agency (Oxford: Oxford University Press, 2015), x. 19 Roy Bhaskar, The Possibility of Naturalism: A  Philosophical Critique of the Contemporary Human Sciences, 4th edition, Classical texts in critical realism (London, New York: Routledge; Taylor & Francis Group, 2015), 96. 20 Timothy O’Connor, “Conscious Willing and the Emerging Sciences of Brain and Behaviour.” in Downward Causation and the Neurobiology of Free Will, edited by Nancey Murphy, Ellis George Francis Rayner & Timothy O’Connor, 2009th edition (London: Springer, 2009), 174. 21 Daniel Kahneman, Thinking, Fast and Slow (Allen Lane, 2011), 55. 22 Daniel Kahneman, Thinking, Fast and Slow (Allen Lane, 2011), 53. 23 John A. Bargh, Min-Cheng Chen  & Lincoln  Burrows, “Automaticity of Social Behavior: Direct Effects of Trait Construct and Stereotype-Activation on Action.” Journal of Personality and Social Psychology 71(2) (1996): 230–44.

Free Will, Cognitive Biases, and Theology 149 24 Englich B., “Playing Dice With Criminal Sentences: The Influence of Irrelevant Anchors on Experts’ Judicial Decision Making.” Personality and Social Psychology Bulletin 32(2) (2006): 188–200. 25 Daniel Kahneman, Thinking, Fast and Slow (London: Allen Lane, 2011), 130. 26 B. Englich, “Playing Dice With Criminal Sentences: The Influence of Irrelevant Anchors on Experts’ Judicial Decision Making.” Personality and Social Psychology Bulletin 32(2) (2006): 188–200, 195, 197. 27 Gregory B. Northcraft & Margaret A. Neale, “Experts, Amateurs, and Real Estate: An anchoring-and-adjustment Perspective on Property Pricing Decisions.” Organizational Behavior and Human Decision Processes 39(1) (1987): 84–97. 28 Figure calculated from results in Gregory B. Northcraft  & Margaret A. Neale, “Experts, Amateurs, and Real Estate: An anchoring-and-adjustment Perspective on Property Pricing Decisions.” Organizational Behavior and Human Decision Processes 39(1) (1987): 84–97, 90; cf. Daniel Kahneman, Thinking, Fast and Slow (London: Allen Lane, 2011), 124. 29 Englich B., “Playing Dice With Criminal Sentences: The Influence of Irrelevant Anchors on Experts’ Judicial Decision Making.” Personality and Social Psychology Bulletin 32(2) (2006): 188–200. 30 Daniel Kahneman, Thinking, Fast and Slow (London: Allen Lane, 2011). 31 John A. Bargh, Before You Know It: The Unconscious Reasons We Do What We Do (New York: Touchstone, 2017). 32 Ben R. Newell  & David R. Shanks, “Prime Numbers: Anchoring and its Implications for Theories of Behavior Priming.” Social Cognition 32 (Suppl) (2014): 88–108. 33 Englich B., “Playing Dice With Criminal Sentences: The Influence of Irrelevant Anchors on Experts’ Judicial Decision Making.” Personality and Social Psychology Bulletin 32(2) (2006): 188–200, 498. 34 Englich B., “Playing Dice With Criminal Sentences: The Influence of Irrelevant Anchors on Experts’ Judicial Decision Making.” Personality and Social Psychology Bulletin 32(2) (2006): 188–200, 498; Daniel Kahneman, Thinking, Fast and Slow (London: Allen Lane, 2011), 56; Ben R. Newell and David R. Shanks, “Prime Numbers: Anchoring and its Implications for Theories of Behavior Priming.” Social Cognition 32 (Suppl) (2014): 88–108, 97. 35 Daniel Kahneman, Thinking, Fast and Slow (London: Allen Lane, 2011), 125. 36 Englich B., “Blind or Biased? Justitia’s Susceptibility to Anchoring Effects in the Courtroom Based on Given Numerical Representations.” Law  & Policy 28(4) (2006): 497–514. 37 Richard A. Klein et  al., “Many Labs 2: Investigating Variation in Replicability Across Samples and Settings.” Advances in Methods and Practices in Psychological Science 1(4) (2018): 443–90. 38 Joshua Shepherd, “Scientific Challenges to Free Will and Moral Responsibility.” Philosophy Compass 10(3) (2015): 197–207, 204. 39 Nomy Arpaly, “Comments on Talking to Our Selves by John Doris.” Philosophy and Phenomenological Research 97(3) (2018): 753–57, 754. 40 Martha C. Nussbaum, Anger and Forgiveness: Resentment, Generosity, Justice (New York: Oxford University Press, 2016). 41 I would like to thank Olli-Pekka Vainio, Aku Visala, and Rope Kojonen for organizing and hosting the Helsinki Analytic Theology conference for which this chapter was originally written. 42 This publication was made possible through the support of a grant from the John Templeton Foundation, God and the Book of Nature [Grant ID. 61507]. The opinions expressed in this publication are those of the author(s) and do not necessarily reflect the view of the John Templeton Foundation.

150  Simon Kittle Bibliography Arendt, Hannah. The Life of the Mind: 2 – Willing. New York: Harcourt Brace Jovanovich, Inc., 1978. Arpaly, Nomy. “Comments on Talking to Our Selves by John Doris.” Philosophy and Phenomenological Research 97(3) (2018): 753–57. Bargh, John A. Before You Know It: The Unconscious Reasons We Do What We Do. New York: Touchstone, 2017. Bargh, John A., Min-Cheng Chen & Lincoln Burrows. “Automaticity of Social Behavior: Direct Effects of Trait Construct and Stereotype-Activation on Action.” Journal of Personality and Social Psychology 71(2) (1996): 230–44. Bhaskar, Roy. The Possibility of Naturalism: A Philosophical Critique of the Contemporary Human Sciences (4th edition). Classical texts in critical realism. London and New York: Routledge; Taylor & Francis Group, 2015. Doris, John M. Talking to Our Selves: Reflection, Ignorance, and Agency. Oxford: Oxford University Press, 2015. Ekstrom, Laura W. God, Suffering, and the Value of Free Will. Oxford: Oxford University Press, 2021. ———. “Toward a Plausible Event-causal Indeterminist Account of Free Will.” Synthese 196(1) (2019): 127–44. Englich, B. “Blind or Biased? Justitia’s Susceptibility to Anchoring Effects in the Courtroom Based on Given Numerical Representations.” Law  & Policy 28(4) (2006): 497–514. ———. “Playing Dice With Criminal Sentences: The Influence of Irrelevant Anchors on Experts’ Judicial Decision Making.” Personality and Social Psychology Bulletin 32(2) (2006): 188–200. Greene, Joshua & Jonathan Cohen. “For the law, neuroscience changes nothing and everything.” Philosophical Transactions of the Royal Society B: Biological Sciences 359(1451) (2004): 1775–85. Kahneman, Daniel. Thinking, Fast and Slow. London: Allen Lane, 2011. Kittle, Simon. “Evaluating the Value of Free Will: Comments on Ekstrom’s God, Suffering and the Value of Free Will.” Unpublished manuscript, 2022. Klein, Richard A., Michelangelo Vianello, Fred Hasselman, Byron G. Adams, Reginald B. Adams, Sinan Alper & Mark Aveyard et al. “Many Labs 2: Investigating Variation in Replicability Across Samples and Settings.” Advances in Methods and Practices in Psychological Science 1(4) (2018): 443–90. Levy, Neil. Consciousness and Moral Responsibility (1st edition). Oxford and New York: Oxford University Press, 2014. Mele, Alfred R. Motivation and Agency. Oxford and New York: Oxford University Press, 2003. Newell, Ben R., & David R. Shanks. “Prime Numbers: Anchoring and its Implications for Theories of Behavior Priming.” Social Cognition 32 (Suppl) (2014): 88–108. Northcraft, Gregory B., & Margaret A. Neale. “Experts, Amateurs, and Real Estate: An anchoring-and-adjustment Perspective on Property Pricing Decisions.” Organizational Behavior and Human Decision Processes 39(1) (1987): 84–97. Nussbaum, Martha C. Anger and Forgiveness: Resentment, Generosity, Justice. New York: Oxford University Press, 2016. O’Connor, Timothy. “Conscious Willing and the Emerging Sciences of Brain and Behaviour.” in Nancey Murphy, George Francis Rayner Ellis & Timothy O’Connor

Free Will, Cognitive Biases, and Theology 151 (eds.), Downward Causation and the Neurobiology of Free Will (2009th edition). London: Springer, 2009. Pereboom, Derk. Living Without Free Will. Cambridge: Cambridge University Press, 2001. Shepherd, Joshua. “Scientific Challenges to Free Will and Moral Responsibility.” Philosophy Compass 10(3) (2015): 197–207. Stump, Eleonore. Aquinas. London, New York: Routledge, 2003. Vicens, Leigh C.,  & Simon Kittle. “God and Human Freedom.” in Cambridge Elements in the Philosophy of Religion. Cambridge: Cambridge University Press, 2019. Visala, Aku. “Theology, Free Will, and the Skeptical Challenge from the Sciences.” Theology and Science 18(3) (2020): 391–409. Walen, Alec. “Retributive Justice.” in Edward N (ed.), The Stanford Encyclopedia of Philosophy (2020th edition). Zalta, 2020. Williams, Bernard Arthur Owen. Making Sense of Humanity. Cambridge: Cambridge University Press, 1995.

9 Bound Choice and the Mystery of Eternal Punishment – A  Trilemma for Monergistic Accounts of Grace Aku Visala Suppose God is perfect love and therefore wills the salvation of all human persons. Suppose further that God exercises ultimate control over the salvation of all persons. Most Lutherans and Reformed theologians have traditionally held that humans lack significant control over whether they come to salvific faith. It is God who exercises control over the person’s coming to faith, not the person herself. This is the thesis that with respect to salvation humans lack free choice – their choice is bounded or in bondage. Finally, suppose that not all persons are saved. There will be some persons who will suffer eternal separation from God. However, justifying this claim becomes rather difficult, if one accepts the two previous assumptions. If God loves all, wills the salvation of all, and controls the means by which the salvation of all could be achieved, it seems mysterious as to how some persons would end up eternally damned. In an article on the debate between Martin Luther and Erasmus of Rotterdam, Olli-Pekka Vainio and I suggested that this mystery of eternal punishment becomes an especially pertinent problem for those who seek to be faithful to Luther’s account of bound choice.1 Lutherans are explicitly committed to the three assumptions outlined earlier. We estimated that Luther’s attempt to solve the problem in his De Servo Arbitrio (1525) did not succeed. In the chapter at hand, I  want to examine this issue more carefully, outline the problem, and assess whether one particular solution – the appeal to mystery – provides a satisfactorily answer to the problem. First, I will present a step-by-step argument to the conclusion that every created person will be eventually saved. The central premises of the argument are the two assumptions outlined earlier: that since God is perfect love, he wills the salvation of all and that humans lack control over their salvation (bound choice). I will then proceed to examine reasons why Lutherans might want to commit to each of these assumptions. Finally, I  will consider and criticize the appeal to mystery as a solution to the problem.

DOI: 10.4324/9781003306191-11

Bound Choice and the Mystery of Eternal Punishment 153 The Trilemma Here are the three claims: 1 Universal Salvific Will (USW): Because God is essentially self-giving love that always seeks what is best for its beloved, God wills the salvation of every human person. 2 Bound Choice (BC): Humans exercise no control whatsoever over coming to salvific faith. Rather, it is God who exercises ultimate control over whether any person comes to faith. 3 Eternal Punishment (EP): Some humans are damned forever. Let us consider the trilemma more carefully. Philosopher Kevin Timpe is not the only one to notice the potential problem.2 Timpe discusses the problems inherent in what he calls the deterministic accounts of grace (DG), namely, the view according to which God’s grace is the necessary and sufficient condition for the salvation of a created person. Earlier, I called this the thesis of bound choice (BC). It is worthwhile to quote Timpe at length: If God’s volition to extend grace to a human individual is necessary and sufficient for the agent coming to saving faith, then the factor ultimately responsible for whether or not individuals come to faith is God’s will, not the individual’s will. Such a position, however, faces significant dilemma. The proponent of (DG) must either say that God does not will the salvation of all or that He does. The repugnance of the first option should be obvious. If God fails to will the salvation of any, then He (and He alone) is the ultimate source of eternal damnation of these individuals. It is difficult to see how God’s antecedently willing that individuals be eternally damned is compatible with His essential goodness. . . . The second option for (DG), then, leads to universalism. The Church, however, has historically rejected it.3 So, if we keep BC fixed, we face a dilemma. We must give up either USW or EP. Philosopher Jerry Walls has provided a step-by-step argument to the conclusion that endorsing BC will result in rejecting EP. I will briefly examine the argument in order to identify the central assumptions in the reasoning. Walls calls eternal damnation the ultimate test case for our view of moral responsibility and free will. Informally, the argument can be described as follows: First, God’s love is such that he wants to maximize the happiness and flourishing of all persons. God does not hold back on his love in the case of any individual. Second, this flourishing is ultimately secured only with having

154  Aku Visala the right relationship to God. The salvific relationship to God is, for a created person, the highest good. Given God’s love, God wants to give the highest good to all persons. Third, if BC is true and human free will and divine ultimate control over salvation are compatible, God is able to offer this highest good to every person since God can guarantee that every person will freely turn to him and receive his love. Therefore, given God’s love and ability to do so, every person will be saved eventually. It is worth putting the argument more formally4:   1 God truly loves all persons.   2 If God truly loves all persons, then he does all he can properly do to secure their flourishing.   3 Therefore, God does all he can properly do to secure the true flourishing of all persons.   4 The true flourishing of all persons is only secured by in a right relationship with God, in which their nature as free beings is respected and they freely accept his love and are saved.  5 God does all he can properly do to secure the true flourishing of all persons, and the true flourishing of all persons is only secured in a right relationship with him.   6 If God does all he can properly do to secure the true flourishing of all persons and the true flourishing of all persons is only secured in a right relationship with him, then God does all he can properly do to secure a right relationship with all persons.   7 Therefore, God does all he can properly do to secure a right relationship with all persons.   8 Freedom and determinism are compatible.   9 If freedom and determinism are compatible, then God can properly secure a right relationship with all persons by determining all to freely accept his love and be saved. 10 Therefore, God can properly secure a right relationship with all persons by determining all to freely accept his love and be saved. 11 God does everything he can properly do to secure a right relationship with all persons, and God can properly secure a right relationship with all persons by determining them freely to accept his love and be saved. 12 If God does everything he can properly do to secure a right relationship with all persons, and God can properly do to secure the right relationship with all persons by determining all freely to accept his love and be saved, then God will determine all persons freely to accept his love and be saved. 13 Therefore, God will determine all persons freely to accept his love and be saved. 14 If God determines p, then p. 15 Therefore, all persons will freely accept God’s love and be saved. The trilemma leaves us with three obvious options. The first option – let us call it the Calvinist option – is to deny USW. In terms of Walls’ argument,

Bound Choice and the Mystery of Eternal Punishment 155 this would amount to denying either premise 2 or 4. The Calvinist solution would, therefore, attempt to sever the connection between God’s perfect love toward the eternally damned and God’s universal salvific will. God loves all the damned but nevertheless wills their damnation. The burden of this solution is to give an account as to how God could simultaneously will the damnation of some created persons and love them with perfect love.5 The second option – let us call it the Arminian option – would reject BC (premise 8 in Walls’ argument) and explain the truth of EP by invoking human free will. Although human persons cannot save themselves, they exercise control over their coming to faith and holding to faith in a way that is incompatible with God determining the outcome. In other words, human choice is not bound with respect to salvation. Finally, there is the universalist option, which accepts USW and BC and concludes that EP is false – all human persons will be saved at the end. In other words, the universalist would accept Walls’ argument. What I  want to explore in this chapter is a view that Luther and some later Lutherans seem to hold.6 This view does not comfortably fall under any of the three solutions outlined earlier. First, Lutherans have traditionally rejected the Calvinist solution and affirmed USW. This is motivated by the claim that Christ’s atonement has universal scope as well as considerations on the nature of divine love. Luther’s own view on the nature of God’s love suggests that God is essentially self-giving love, who seeks to pour his infinite goodness over the subjects of his love. Second, Lutherans are also forced to reject the Arminian solution since they are committed to BC. The human choice is bound with respect to salvation. Either humans exercise no control over their coming to faith or the kind of control they exercise is compatible with God’s ultimate control.7 In the subsequent sections, I  outline some reasons why Lutherans are tempted to reject both Calvinist and Arminian solutions. But this seems to leave only one option, universal salvation. However, Lutherans have traditionally maintained that the eternal punishment of some created persons is possible, which rules out the universalist solution. But how exactly does the Lutheran solve the problem then? What we see most often is an appeal to mystery or paradox. We are warned against drawing a conclusion from the premises and advised to hold onto the tension between the three claims of the trilemma. What I seek to argue is that appeal to mystery at this point is not appropriate and that such a move will inevitably threaten various religious and theological goods that Lutherans would otherwise want to uphold. Universal Salvific Will and Divine Love I will now provide some reason why Lutherans should find it difficult to reject one or more theses of the trilemma. Let us begin from the first thesis (USW): Because God is essentially self-giving love that always seeks what is best for its beloved, God wills the salvation of every human person. The

156  Aku Visala Lutheran confessions, especially The Formula of Concord, are very clear as to the universal salvific will of God: 7. This Christ calls to Himself all sinners and promises them rest, and He is in earnest [seriously wills] that all men should come to Him and suffer themselves to be helped, to whom He offers Himself in His Word, and wishes them to hear it and not to stop their ears or [neglect and] despise the Word.8 The conclusion is even clearer in the antithesis section, where the following is said: Therefore we reject the following errors: 1 As when it is taught that God is unwilling that all men repent and believe the Gospel. 2 Also, that when God calls us to Himself, He is not in earnest that all men should come to Him. 3 Also, that God is unwilling that everyone should be saved, but that some, without regard to their sins, from the mere counsel, purpose, and will of God, are ordained to condemnation so that they cannot be saved.9 These are “blasphemous and dreadful erroneous doctrines” that must be avoided, apparently, because they compromise on God’s love and universal charity. It is notable that these lines emphasize the honesty of God: God does not simply express His universal salvific will to humans and secretly will something else, but rather He “earnestly wills” or “seriously wills” the salvation of all. The frightened conscience of the believer cannot find solace, if there is even a possibility of God willing the damnation of some human persons. God’s unwavering and universal salvific will is where the believer can find peace. One reason for the commitment to USW is the Lutheran theology of divine love. Very roughly, Luther’s theology of love can be expressed in two points. First, perfect love consists of giving the beloved whatever is best for the beloved. Divine love is charitable giving of what is good for the beloved and what the beloved is missing. Divine love gives existence to the non-existent, life to the dead, salvation to sinners. As Finnish Luther scholar Tuomo Mannermaa has explained, Luther draws a distinction between human love, which seeks what is pleasurable and lovable, and divine love (agape) that rather creates and gives the good to its objects.10 Human love is directed at objects that are considered valuable, lovable, and beautiful. Those good-making features of the objects of love are the grounds for which the love is given. According to Mannermaa, Luther’s radical notion of divine love goes against this. Divine love does not seek the good already inherent in its objects, but rather seeks to make its objects good by charitably donating goodness to them.

Bound Choice and the Mystery of Eternal Punishment 157 Since God is the greatest good, God’s love consists of God’s giving of himself to the beloved. This self-giving is not grounded in any good-making features of its objects, but rather in God’s essential nature. This brings us to the second point. For Luther, “This it means to be God: not to receive good but to give it”11 God himself is the ultimate good. When God loves, he gives himself to the beloved, not because there is something good in the beloved or because the beloved deserves it, but because selfgiving love is God’s nature. According to Luther, God and sinful human persons act in opposite ways. Sinful human persons seek to set up their own “good” and reject God. Sinfulness is essentially selfishness, turning away from God and from fellow human beings and always seeking for one’s own good. A  sinful person is self-centered (lat. incurvatus in se). Divine love is the opposite: it seeks the good of the other and gives this good without considering its own gain or the value of the other. The love is even greater when we realize that sinful humans can do nothing to merit or deserve God’s love. Luther repeats this ad infinitum: God’s self-giving love is a pure gift. In justification, participation in Christ becomes available to the believer, and the believer’s self-centered love becomes transformed by Christ’s self-giving love. Given this account of divine love, it seems impossible to accept the Calvinist solution to the dilemma. The Calvinist solution holds that the connection between God’s love and universal salvific will can be severed. God can love a created person with perfect love but, nevertheless, create her for the purpose of eternal damnation. The problem with this view is clear: from the notions of divine love as self-giving love and God as the source of all perfection and goodness flows the conclusion that God wills what is best for all created persons – the union with God. The greatest flourishing and well-being of the created person are achieved in perfect union with God since God is the greatest good and has created all persons for this goal. Since the giving of the greatest good (God) is essential to God, God cannot fail in this without being God. So, God will always do whatever God can to secure the created person’s union with himself. Given this account of love, it is difficult to see how USW could be avoided. A Free Will Hell? The official doctrine of Lutheran churches has never affirmed universalism. Rather, Lutheran confessions take it as a given that some human persons will be damned forever, or at least it is possible that some human persons will be damned forever. All doctrines of eternal punishment state that (a) there are created persons who will exist eternally, (b) are never saved, and (c) are left to a state of suffering because of the lack of closeness to God.12 Hell is the state where a rational creature is deprived of God’s presence and goodness. Left to her own devices, the damned person only loves herself and exists in a state where sin and evil have full reign.

158  Aku Visala Given USW, we wonder why God would allow for the eternal damnation of some persons. So, the question is: [W]hy would a perfect God, who not only loves His creatures but hates sin, permit (or bring it about) that some of His creatures are eternally marred by both misery and sinfulness, and so fall eternally short of the end for which He made them? Why would a perfect God allow sin to reign forever victorious over His divine purposes in creation, at least in the souls of the damned?13 One obvious answer comes to mind immediately: God risks the possibility of eternal damnation of some persons because God thinks that the world is overall much more valuable when there are created persons with incompatibilist free will in it, rather than created persons without such free will. In other words, God wills a world where there are persons, who can exercise significant moral freedom and freely love God and to whom God can give himself as a gift. This entails that God must risk the possibility of some persons eternally rejecting him. The risk is worth it because the world where there are beings with such free will is significantly more valuable to God than a world where such persons do not exist. This view has many names: the liberal doctrine of hell, free will hell, or the choice model of hell. The aforementioned Jerry Walls argues that hell is an expression of God’s love.14 God prefers a world, where there are persons with incompatibilist free will, because of the goods that moral autonomy of created persons brings to the world. God does all he can to save them, but there is always the possibility that some individuals forever resist God. If God were to determine that all created persons come to faith, there could be no significant moral autonomy and no free human response to God’s offer, for instance. But since God values such a possibility highly, God is willing to allow for the possibility of eternal damnation. As C. S. Lewis has famously put it, “the gates are shut from the inside.” In order for the free will hell to work, a certain kind of human free will must be true, a libertarian free will. Libertarian accounts of free will come in many forms, but they share a common claim: if a person’s action is causally necessitated (or otherwise determined) by factors outside the control of the person, that action is not free. So, if God wants a world in which at least some created persons have libertarian free will, God will then refrain from causally necessitating or otherwise determining at least some human actions. In other words, God refrains from controlling at least some morally and religiously significant human actions such that humans retain the necessary control over their actions to damn themselves. While the free will model of hell provides a promising way of justifying the possibility of eternal damnation, the Lutheran cannot adopt it. The reason is that the Lutheran view of salvation by faith alone entails BC and leads to the rejection of libertarian free will, which is a necessary component of free will

Bound Choice and the Mystery of Eternal Punishment 159 hell. The one who accepts BC must also accept premise 8 in Walls’ argument, namely, that free will and determinism are compatible. In the next section, I will present an argument to this effect. Bound Choice and the Lack of Incompatibilist Control Over Faith In this section, I aim to show that the Lutheran (and broadly Augustinian) doctrine of salvation by faith alone entails the rejection of libertarian free will. In other words, Luther’s insistence of sola gratia and sola fide entails that only God exercises ultimate control over the salvation of all created persons (BC), and it follows from this that it is in God’s power to determine that all created persons are saved. This is because generating salvific faith and upholding salvific faith depend ultimately on God, not on the human being. This means that all created persons suffer from bound choice with respect to salvation, that is, they either lack all control over it or the control they have is compatible with God’s causally necessitating or otherwise determining it. Consider the distinction between what Kevin Timpe calls deterministic and non-deterministic accounts of grace15: DG (deterministic grace): Divine grace is a necessary and sufficient condition for the human response of faith in God. NG (non-deterministic grace): Divine grace is necessary but not a sufficient condition for the human response of faith in God. Timpe writes: [D]oes God’s giving of the grace needed for saving faith to an agent determine that the agent comes to saving faith? If the answer to the question is “yes,” then God determines an agent to have saving faith via giving that individual grace and the giving of the grace will be sufficient for the individual’s coming to saving faith.16 Timpe singles out both Luther and Calvin as representatives of DG. In our article on the debate between Luther and Erasmus, Olli-Pekka Vainio and I have argued that Luther’s response of Erasmus’ challenge entails a deterministic doctrine of grace.17 To support this contention, I draw from philosopher Lynne Rudder Baker’s discussion on Augustinian challenges to libertarian free will.18 First, Baker examines Augustine’s notion of sin. Apparently, Pelagius argued that either sin is involuntary, in which case humans cannot be held responsible for it, or that it is voluntary, in which case it can be voluntarily overcome. The underlying assumption here is that moral responsibility requires control. If a person is unable to exercise control over a sinful action, the person should not be held responsible for it. Conversely, if a person can exercise control over a sinful action, this means that the person can refrain

160  Aku Visala from committing the action. Augustine takes issue with this and challenges the underlying assumption that moral responsibility for sin entails control over sinful actions. For Augustine, God holds humans responsible for both voluntary and involuntary sins. Indeed, humans are born into a sinful state: their moral and spiritual cognition suffers from significant defects. As a consequence, they voluntarily commit sins they cannot refrain from committing. Moreover, they can do nothing in their own power in order to merit any kind of favorable treatment from God. All sinful humans are apt subjects of God’s righteous anger. God would be in his rights to punish all humans infinitely and reject them. However, God has chosen some individuals, by grace, to be saved from this state.19 There is much debate about Augustine’s account of free will.20 Nevertheless, it seems that Augustine’s view is a compatibilist one. An account of free will is compatibilist if it allows for a person’s action to be free even while it is causally necessitated or otherwise determined by factors outside the control of the person. In this sense, it is the opposite of libertarian free will, where no action can be free, if it is causally necessitated or otherwise determined by prior factors outside the control of the agent. According to Baker’s interpretation, Augustine cannot consistently hold anything else than a compatibilist view of free will.21 For Augustine, Adam as the first human did originally possess libertarian free will. Adam exercised ultimate control over whether he committed wrong or right actions. However, Adam chose to exercise his choice by committing wrong actions. As a consequence, subsequent generations of humans were born with bound choice. Humans can make morally relevant choices and be responsible for them, but they lack the ability to commit actions that would merit any favorable treatment from God. Humans inherit a sinful disposition and sinful nature, which inevitably result into sinful actions. Sinful actions are, nevertheless, free and responsible actions in the sense that sinful persons are not forced or coerced to commit their actions.22 The freedom exhibited by these actions, however, is of the compatibilist kind: the sinful action is never free in the libertarian sense, that is, it is always necessitated by prior causes (sinful nature). Further support for the incompatibility of libertarian free will and the Augustinian doctrine of grace comes from Augustine’s account of faith. In order for faith to be a free gift, there can be nothing a person can do to earn it or prepare herself for it. As both Augustine and Luther after him empathetically maintain, faith is a gift from God. It is God who controls to whom faith is given and who holds to it until the end. It follows that “there is no room for free will as libertarians construe it in our salvation, not even in the first step.”23 If humans could exercise libertarian free will with respect to some actions leading to faith or holding to it, it would mean that the sinful will of the human person could block or thwart God’s salvific will. Because of this, Luther famously said that he would not even want free choice over his salvation if it were offered to him. If Luther had libertarian free will with respect to his salvation, it would follow that salvation would be, at least to some extent,

Bound Choice and the Mystery of Eternal Punishment 161 under his own control. This would entail the possibility of failure and selfdamnation. In order to rest assured of one’s salvation with total certainty, one must resign all control and leave it completely under God’s control. One might object and point out that the Augustinian/Lutheran doctrine of grace does require that the justified person herself also acts and wills to commit good actions. After justification, the believer participates in Christ and becomes shaped by him. The believer’s will turns toward actions that are pleasing to God. So, the person is hardly a passive recipient of God’s grace.24 One response to such an objection would be the suggestion that the passivity/ activity –distinction does not track the free will/no free will – distinction. The denier of libertarian free will is not required to say that the person “does not act” or is only a passive patient of God’s action. One can very well say that the justified sinner acts and wills herself and is in this sense active. However, this activity is determined by God’s activity in the person’s will. So, the believer may well act and will what is good but the ultimate control over the will and the actions of the person is elsewhere – in God’s hands. This kind of acting and willing is not free in the libertarian sense. So, the believer might as well act in the process of salvation (whatever that might be) but those actions and the will behind them are determined and controlled by God alone. This is exactly what compatibilist free will is. Baker writes: The hallmark of free will as libertarians construe it is that its exercise is uncaused by anything over which the agent has no control. Therefore, a will whose exercise is caused by God is, by definition, not a free will as libertarians construe it.25 Luther’s view on the Bondage of the Will is subject to some debate.26 My interpretation is that Luther defines free will along incompatibilist lines as the ability to choose between morally robust alternative actions. He then goes on to argue that with respect to salvation, humans can never have access to robust alternative actions.27 He outlines many reasons for this. For our purposes here, I will mention only a few. First, Luther invokes the Augustinian doctrine of sin outlined earlier. Second, Luther presents an interesting reductio argument. If humans could commit some actions that would accrue merit in God’s eyes, it would be possible for humans to save themselves. But because the Gospel tells us that only Christ can save us, we know that humans can do nothing that would merit favorable treatment from God. Finally, Luther also invokes the metaphysical thesis of divine determinism in his support. Divine determinism, according to Luther, is good and salutary because it means that nothing can prevent God’s will from coming to pass. Indeed, it is the only consolation of a sinner that God’s salvific will cannot be hindered or made empty by anything a person does.28 Despite Luther’s charitable spiritual goal, he accidentally creates another problem. The consequence of Luther’s divine determinism and his doctrine of grace is that God’s salvific will cannot be compromised or hindered by

162  Aku Visala anything that humans do. This has the unintended side effect that if someone is to be damned, this is in God’s control as well. If God exercises ultimate control over who is saved, God also exercises ultimate control over who is damned. But if God ultimately controls whether a person is damned or saved and USW is true, then God will save every created person. Shut Up and Stop Asking Questions Given strong reasons to hold onto all the three claims of the trilemma, the Lutheran might want to invoke mystery or declare the whole trilemma as a theological paradox. USW, BC, and EP are explicitly taught by divine revelation, the Lutheran will insist. Perhaps we – limited and sinful human persons – are not in a position to understand the reasons for why some are damned or how God’s love toward the damned are expressed in their punishment. At many points of his De Servo Arbitrio, Luther invokes St. Paul and argues that it is impertinent, prideful, or sinful to seek to resolve the paradox. Whereas the sinner sees the trilemma as a reason to reject and hate God, the believer – enlightened by the Holy Spirit – has faith in God’s goodness and righteousness regardless of the contradiction in God’s self-revelation, or so the argument would go. Luther himself understands mystery in terms of God’s ultimate “otherness”: we can only know of God what God has chosen to reveal about Himself. God, as he is Himself, is unknowable and mysterious to us. Humans are not in the position to demand God to provide some information or ask God for justifications for God’s ultimate decisions. So, the believer will stop her questions and believe what God has revealed. Philosopher Tamler Sommers calls this the “shut up and stop asking questions” – reply.29 Why should we believe this is the correct attitude? In order to assess this response, we should begin by asking when it is appropriate to appeal to mystery in theology. In his informative discussion on the topic, philosopher Tim Pawl suggests that in the face of some fundamental issues in theology appeal to mystery might indeed be appropriate.30 With respect to certain aspects of classical Christology – the hypostatic union, for instance – even the councils of the church appeal to mystery. The point of these appeals, according to Pawl, is the spirit in which one’s theological pursuit is conducted. One is not supposed to be motivated by idle curiosity or a compulsion to rationalize theology. This is a requirement that constrains the motivations of the enquirer. Second, appeals to mystery are appropriate when one is asked to give a positive account of something, and one sees that doing so might be beyond one’s intellectual abilities, but one is aware of no contradiction that result in virtue of not being able to give a positive account.31 It might very well be that certain aspects of theological doctrine are beyond our understanding. This applies especially to situations where we might

Bound Choice and the Mystery of Eternal Punishment 163 want to know how some divine thing, like the hypostatic union, actually works. The fact that we do not have (or are even unable to offer) an account of how it works is not a reason to withhold belief or deny the truth of the claims involved. What I want to suggest is that the trilemma that I have presented earlier is not a case where the appeal to mystery as an enquiry stopper is appropriate. First, I (and many others) have not engaged with this trilemma because of idle curiosity or the compulsion to rationalize but because of genuine concern for our view of God and our own salvation. What is at stake is the perfect goodness of God. Reaching clarity in this crucial issue pertaining to our salvation can hardly be described as impertinent or prideful. I  see no reason why it would necessarily indicate a lack of faith or any other vice, like vain intellectual curiosity or rationalism. Second, the reason why appeal to mystery is an inappropriate response to the trilemma is that we have formed a valid argument with reasonably wellsupported premises to the conclusion that one of the claims we start with (USW, BC, or EP) is false. The trilemma produces a logical contradiction. Appealing to mystery in the face of logical contradiction will lead to damaging and destructive consequences in theology.32 The appropriate attitude to the trilemma is not to shroud it in mystery, but to attempt to alleviate the logical contradiction by trying to show that there is something wrong with the premises of the argument. One need not provide a positive account of, say, how eternal punishment of sinners is compatible with God’s universal salvific will. It is enough to show that these claims are not logically contradictory by undermining the argument presented earlier. So, in order to solve the logical contradiction, something has got to give. What I have tried to demonstrate so far is that the Lutherans are deeply committed to USW, BC, and EP. Walls’ argument derives the falsity of EP from USW and BC. So, there seem to be no premises left with which the Lutheran might disagree, so the appeal to mystery is the only option left. Let us consider the consequences that result from appealing to mystery at this point. I will argue that the consequences would be both theologically and religiously very damaging and undesirable. First, consider our philosophical and everyday view of rationality. We are allowed to rationally believe propositions that are in apparent tension with one another, but not propositions that are contradictory with one another. A theologian might claim that some theological views are “beyond logic” and we should not apply logic to them at all.33 However, the price of this would be very high: all theological argumentation would collapse. Luther and Erasmus, for instance, would have no reason to argue about whether humans have bound will or not. Both could be right: we should believe that the human will is bound and free at the same time. Holding onto logic is not a result of unbelief or pride, but a necessary commitment to a fundamental set of tools without which there could be no theology at all. Consider also what would happen to central theological terms, like “love” and “justice,” if an appeal to mystery were accepted at this point. Here,

164  Aku Visala the question is whether we can derive USW from perfect divine love (Walls’ premises 1–4). The defender of the appeal to mystery would have to reject the analogy between human love and divine love at this point. Luther takes this line of argument so far that he even claims that if we could understand how God’s love and justice were compatible, we would somehow make God less divine: For if God’s righteousness were such that it could be judged to be righteous by human standards, it would clearly not be divine and would in no way differ from human righteousness. But since God is the one true God, and is wholly incomprehensible and inaccessible to human reason, it is proper and indeed necessary that God’s righteousness also should be incomprehensible.34 Taken to the extreme, this line of reasoning leads to the conclusion that we are obligated to believe that God is perfectly righteous and loving without any idea what those terms mean.35 Consider what would happen to the term “love” if we could not draw any kind of analogy between divine love and human love. Walls suggests that in such a case we would no longer understand what love means as a divine attribute. He writes that [T]rue love must seek the true flourishing and ultimate well-being of the beloved as much as it properly can. If theological compatibilists want to insist that God loves unbelievers he may not have elected for salvation, then they must deny this, which means they will be using the concept of love in a deeply idiosyncratic sense.36 Not only would the predicate “love” lose its meaning when applied to God, but it would also follow that at least Mannermaa’s interpretation of Luther’s account of love must be false. As the reader might recall, this account explicitly connects USW with God’s essential nature as self-giving love. If we want to respond to the trilemma by invoking mystery, we must also reject this account of God’s essential nature. This is because USW flows logically from the notion that God is essentially self-giving love. The appeal to mystery would also paint a morally and spiritually problematic picture of the God/human relationship. Let us now assume that God wants us limited and sinful humans to adopt the attitude of “shut up and don’t ask questions.” God would intentionally present himself and his will to us in logically contradictory ways and suggest we have an obligation to have faith in God anyway. This would undermine what we take to be a loving, caring, and constructive relationship between persons. In such a relationship, having access to adequate knowledge about the other person’s good will and love toward the other is paramount. Against this, God would have chosen to relate to sinful humans in another way – by presenting contradictory views of himself and communicated the obligation that all must

Bound Choice and the Mystery of Eternal Punishment 165 believe them. This does not seem like a loving, caring, and constructive relationship we might expect. Such a relationship would also introduce a significant problem with the analogies and metaphors that the New Testament (especially the parables of Jesus) uses about the God/human relationship. Here, God reveals himself as perfectly loving father with an attitude of loving charity and kindness. Of course, God sometimes expresses anger and judgment, but this is to highlight the ultimate goal: the good of all human persons. However, in the case of free will and eternal punishment, the metaphors of the parables break down. We are no longer allowed to expect (and demand) an explanation for an act of God that severely undermines our faith in his charity and goodness. Indeed, demanding an answer from God is considered sinful, prideful, and a sign of unbelief. Instead of God being the loving father, he is now revealed as the narcissist sovereign, who shows very little care as to whether his subjects have a loving relationship to him. God would be the absolute authority, whose commandments and revelation must be believed unquestionably – even to the point that one must have faith in God’s universal love, when everything God does speaks against it. Such a relationship bears little resemblance to the relationship of a loving father and a child, but it is rather close to a hierarchical relationship of subjugation and obedience. The Revealed God and the Unknowable God Another reason for why the appeal to mystery is inappropriate is that it introduces a break between God’s self-revelation and what God actually is. Such a break will seriously undermine some religious and spiritual goods, which Lutheran and other Christians value highly. Distinguishing between God’s revealed and mysterious sides might resonate with the experience of many Christians. Indeed, Luther himself often describes God the Father as a righteous judge and Christ as a sweet savior. It is as if there were two gods: one angry judge and the other the savior – a shield against the angry judge. Following Luther’s interpretation of St. Paul, De Servo Arbitrio draws a distinction between God’s revealed will (universal salvific will) and God’s secret will (salvation for some and damnation for others). Luther further warns the reader against speculating about the latter. In Christ, God’s will toward us is revealed. If there is another will – the secret will of God the Father – it is no business of ours. Luther writes: It is likewise the part of this incarnate God to weep, wail, and groan over the perdition of the ungodly, when the will of the Divine Majesty purposely abandons and reprobates some to perish. And it is not for us to ask why he does so, but to stand in awe of God who both can do and wills to do such things.37 Christ represents God’s universal charity and the will that all shall be saved, and this is what is revealed to humans. However, in His majesty, God might

166  Aku Visala have another, incomprehensible will, which decrees the salvation of some and intentionally leaves others to damnation. It is this will, which humans cannot understand, and they should avoid speculating about it. Here, we have an attempt to solve the dilemma by adjusting USW. First, it is claimed that USW is something that God has revealed about himself. However, because God’s revealed will in Christ might not correspond to the mysterious will of God the Father, USW might not be true after all. The problem is, however, that a contradiction arises between the account of God as a self-giving love and the claim that God might have two wills. There would be people to whom God does ultimately not want to give himself as a gift: these are the ones who God has created and chosen for damnation. Again, if God is revealed to us as essentially self-giving love but there is a possibility that God is ultimately not such love, then USW collapses. In this case, the Lutheran solution would collapse into the Calvinist solution. Distinguishing God’s revealed will and mysterious will from one another will also have other deeply problematic consequences. For one, it allows for the possibility of divine deception, thereby undermining one core aim of Lutherans, namely, that believers can have assurances and can be certain of their salvation. In order for humans to achieve such certainty, they must believe that God’s salvific plan cannot be thwarted by anyone or anything (that God ultimately controls their salvation, BC) and that God has universal salvific will toward all created persons (USW). Christians must believe that God’s promise of salvation is absolute. However, if God’s revealed will can be different from his secret will, then there are no grounds to believe in God’s promise anymore. If I have faith in God’s Word, I take the Word to be certain, because God is trustworthy. That Word is God’s promise to save me despite all my sins. But if there is even a possibility that the Word of God does not express God’s ultimate intentions with respect to my salvation, I cannot trust God, which in turn leads to my losing of my assurance. This is the reason why USW is so important. Theology of the Cross I mentioned earlier that God might intentionally present himself as a vengeful judge in ways that are in contradiction with the model of a loving and charitable relationship. This line of reasoning leads us to one of the most distinctive features of Lutheran theology, the theology of the cross. I do not want to go into the theology of the cross in any detail. Rather, I will simply try to see whether it could be relevant for our dilemma. For Luther, the notion was very important. According to theology of the cross, God intentionally conceals his actions and intentions into their opposites.38 Sometimes, theology of the cross leads to almost mystical conclusions about God: God deliberately hides so that the need for faith becomes greater. Theology of the cross could be used to generate a kind of greater goods defense. According to this interpretation, God intentionally discloses himself

Bound Choice and the Mystery of Eternal Punishment 167 to us in a contradictory way – as both a vengeful judge and a loving father. More importantly, at least the vengeful father part would be false: God is – essentially and necessarily – pure, self-giving love. God would present himself so as to make faith in a loving God more valuable and greater than it would otherwise be. First, the more difficult or demanding faith is, the more valuable and greater it is. Second, if God were to express his will and act in such a way as he actually is, namely, universally loving and charitable, there would be less need for faith. Everyone would happily believe and worship such a God without supernatural help. Therefore, God will express his wrath and eternal punishment so that when people actually have faith that he is a God of love, that faith will be exceptionally great and valuable. It is great and valuable, because it goes so greatly against the apparent facts (that God looks like an unjust judge), as it were. It is unclear whether Luther actually meant anything like this with his theology of the cross. However, it is clear that no greater good defense of this kind will solve our trilemma. Let us suppose (for the sake of the argument) that God would indeed want to make faith more difficult and thereby greater. In other words, let us assume that great faith would be a good that God would want to maximize. This would provide God a reason to reveal himself to us in such a contradictory way. While I find some plausibility in this reasoning, two points can be raised against it. First, given the fact that relating to humans in a contradictory way implies a non-loving, hierarchical relationship of obedience and subjugation, God should not relate to humans in this way if some other way were available to God. If God wanted to maximize the virtue of faith by making it difficult, God surely could have found some other way of doing that (God is omnipotent after all) than communicating to us that he demands us to believe in contradictions. Second, even if this were the only way God could secure the good of virtuous faith, God would not need EP to achieve this good. In order to maximize virtuous faith, it would not be necessary for God to condemn any created person to hell. Exactly the same good could be achieved, if God were to reveal himself in a contradictory way but then save everyone, when the final judgment came. Faith would be made greater, and perhaps the possibility of eternal punishment would function as a deterrent against immoral behavior. It is enough that “such things are preached and published” but no actual punishment would be needed. Indeed, an appeal to the distinction between the revealed God and the mysterious God could lead us into this conclusion. Perhaps God communicates his righteous and punishing side not because he actually is like that, but because of the benefits it might have for our behavior. We just have no way of knowing. However, this would again undermine our trust in God’s intentions and will toward us. If we cannot trust that God’s self-disclosure expresses what God actually wants and intends, there is no way we could ever have assurances (or even do theology, for that matter).

168  Aku Visala Concluding Remarks I began the article by outlining of trilemma that consists of three claims: (1) USW: because God is essentially self-giving love that always seeks what is best for its beloved, God wills the salvation of every human person; (2) BC: humans lack control over coming to salvific faith and holding the faith. Rather, it is God who exercises ultimate control over whether any person comes to faith; (3) EP: some humans are damned forever. I went on to outline an argument according to which USW and BC lead into a contradiction with EP. I also canvassed some reasons why Lutherans might be motivated to hold onto all three claims of the trilemma. Finally, I discussed why I think that the preferred response, an appeal to mystery, is not appropriate in this context. While appeals to mystery have their place in theology, they are not appropriate in the context of the trilemma. So, it seems to me that the trilemma still constitutes a significant problem for Lutherans who attempt to navigate it. Notes 1 Aku Visala  & Olli-Pekka Vainio, “Erasmus Versus Luther: A  Contemporary Analysis of the Debate on Free Will.” Neue Zeitschrift fur Systematische Theologie und Religionsphilosophie 62(3) (2020). 2 See, for example, Thomas Talbott, The Inescapable Love of God (New York: Cascade, 2014). 3 Kevin Timpe, Free Will in Philosophical Theology (London: Bloomsbury, 2014), 52–3. 4 Jerry Walls, “One Hell of a Problem for Christian Compatibilists.” in Daniel Speak & Kevin Timpe (eds.), Free Will and Theism (Oxford: Oxford University Press, 2014), 90. 5 For an impressive recent philosophical defense, see Guillaume Bignon, Excusing Sinners and Blaming God: A  Calvinist Assessment of Determinism, Moral Responsibility and Divine Involvement in Evil (Eugene: Pickwick, 2018). 6 For the debates after Luther, see Olli-Pekka Vainio’s chapter in this volume. 7 I am assuming that there are only two options: either God controls the salvation of a created person completely or that person exercises some incompatibilist control over her salvation as well. However, one might also suggest a third option. Perhaps there are some facts that preclude both God and the created person from controlling the process of coming to faith. There is the possibility that there are some metaphysical facts, independent from the will of created persons and the will of God, that constrain God’s creative choice. These facts might have to do with the metaphysics of personal identity, for instance. God’s control over the faith of a created person is constrained by God’s intention to maintain the personal identity of the created person. If God were to instantly give faith to a person, this would introduce a radical break in the identity of the person. This possibility, however, suffers from a significant drawback. Given the fact that God would know these metaphysical truths (whatever they are), God would know in whose case he is able to bring about the desired conclusion (i.e., that the person freely accepts faith and is saved). So, God could create those individuals only. However, it seems that God has indeed created two kinds of individuals – those God is able to save and those God is unable to save. This implies (given USW) that God willingly and knowingly creates persons God knows he is unable to

Bound Choice and the Mystery of Eternal Punishment 169 save. It looks like that God no longer loves those persons in a perfect way since God has decided to create them while knowing that God cannot guarantee their ultimate salvation. Furthermore, it seems that God’s purposes would – in such a scenario – be ultimately defeated by some metaphysical facts independent of God’s creative will. This could be construed as defeat in God’s battle against sin, evil, and death. 8 The Formula of Concord, XI Election, 7. 9 The Formula of Concord, XI Election, 17–19. 10 Tuomo Mannermaa, Two Kinds of Love: Martin Luther’s Religious World (Grand Rapids: Fortress, 2010). 11 Quoted in Jari Jolkkonen, “Eucharist.” in Olli-Pekka Vainio (ed.), Engaging Luther: A (New) Theological Assessment (Eugene: Cascade, 2010), 112. 12 In my discussion of the doctrine of eternal punishment, I follow the classification of John Kronen & Eric Reitan, God’s Final Victory: A Comparative Philosophical Case for Universalism (New York: Bloomsbury, 2013), 12–27. 13 God’s Final Victory, 2013, 14. 14 Jerry Walls, Hell: The Logic of Eternal Damnation (Notre Dame: The University of Notre Dame Press, 1992). 15 Timpe, Free Will in Philosophical Theology, 51–52. 16 Timpe, Free Will in Philosophical Theology, 51. 17 Visala & Vainio, “Erasmus versus Luther.” 18 Lynne Rudder Baker, “Why Christians Should Not Be Libertarians: An Augustinian Challenge.” Faith and Philosophy 20 (2003): 460–73. 19 Baker, “Why Christians Should Not Be Libertarians,” 463–64. 20 See Jesse Couenhoven, Stricken by Sin, Cured by Christ: Agency, Necessity and Culpability in Augustinian Theology (Oxford: Oxford University Press, 2013). 21 For an alternative, incompatibilist interpretation, see Eleonore Stump “Augustine on Free Will.” in Eleonore Stump & Norman Kretzmann (eds.), Cambridge Companion to Augustine (Cambridge: Cambridge University Press, 2001), 124–47. 22 Baker, “Why Christians Should Not Be Libertarians,” 464. 23 Baker, “Why Christians Should Not Be Libertarians,” 465. 24 See, Vainio’s chapter in this volume on the notion of passivity versus activity in Lutheran theology. 25 Baker, “Why Christians Should Not Be Libertarians,” 467. 26 We outline Luther’s argument in Visala & Vainio “Erasmus versus Luther.” See also Miikka Ruokanen, Trinitarian Grace in Martin Luther’s Bondage of the Will (Oxford: Oxford University Press, 2021). 27 For a defense of this interpretation, see Visala  & Vainio, “Erasmus versus Luther,” 323–6. By morally robust alternatives, I mean acting and willing in ways that are pleasing to God. 28 For a contemporary discussion on merits of divine determinism, see Peter Furlong, The Challenges of Divine Determinism (Oxford: Oxford University Press, 2018). 29 Tamler Sommers, “Relative Responsibility and Theism.” in Daniel Speak & Kevin Timpe (eds.), Free Will and Theism (Oxford: Oxford University Press, 2016), 104. 30 Tim Pawl, In Defence of Conciliar Christology: A Philosophical Essay (Oxford: Oxford University Press, 2016). 31 Pawl, In Defence of Conciliar Christology, 89. 32 Pawl, In Defence of Conciliar Christology, 90. 33 See, for example, Jc Beall, The Contradictory Christ (Oxford: Oxford University Press, 2021). 34 AE 252. 35 Vainio argues more extensively for this point in his chapter in the present volume.

170  Aku Visala 6 Walls, “One Hell of a Problem for Christian Compatibilists,” 92. 3 37 Quoted in Visala & Vainio, “Erasmus versus Luther,” 331. 38 Kari Kopperi, “Theology of the Cross.” in Olli-Pekka Vainio (ed.), Engaging Luther (Eugene: Cascade, 2011).

Bibliography Baker, Lynne Rudder. “Why Christians Should Not Be Libertarians: An Augustinian Challenge.” Faith and Philosophy 20 (2003): 460–73. Beall, Jc. The Contradictory Christ. Oxford: Oxford University Press, 2021. Bignon, Guillaume. Excusing Sinners and Blaming God: A Calvinist Assessment of Determinism, Moral Responsibility and Divine Involvement in Evil. Eugene: Pickwick, 2018. Couenhoven, Jesse. Stricken by Sin, Cured by Christ: Agency, Necessity and Culpability in Augustinian Theology. Oxford: Oxford University Press, 2013. Furlong, Peter. The Challenges of Divine Determinism. Oxford: Oxford University Press, 2018. Jolkkonen, Jari. “Eucharist.” in Olli-Pekka Vainio (ed.), Engaging Luther: A (New) Theological Assessment. Eugene: Cascade, 2010. Kopperi, Kari. “Theology of the Cross.” in Olli-Pekka Vainio (ed.), Engaging Luther. Eugene: Cascade, pp. 155–72, 2011. Kronen, John & Eric Reitan. God’s Final Victory: A Comparative Philosophical Case for Universalism. New York: Bloomsbury, 2013. Mannermaa, Tuomo. Two Kinds of Love: Martin Luther’s Religious World. Grand Rapids: Fortress, 2010. Pawl, Tim. In Defence of Conciliar Christology: A  Philosophical Essay. Oxford: Oxford University Press, 2016. Ruokanen, Miikka. Trinitarian Grace in Martin Luther’s Bondage of the Will. Oxford: Oxford University Press, 2021. Sommers, Tamler. “Relative Responsibility and Theism.” in Daniel Speak & Kevin Timpe (eds.), Free Will and Theism. Oxford: Oxford University Press, pp. 99–111, 2016. Stump, Eleonore. “Augustine on Free Will.” in Eleonore Stump  & Norman Kretzmann (eds.), Cambridge Companion to Augustine. Cambridge: Cambridge University Press, pp. 124–47, 2001. Timpe, Kevin. Free Will in Philosophical Theology. New York: Bloomsbury, 2014. Visala, Aku & Olli-Pekka Vainio. “Erasmus Versus Luther: A Contemporary Analysis of the Debate on Free Will.” Neue Zeitschrift fur Systematische Theologie und Religionsphilosophie 62(3) (2020): 311–35. Walls, Jerry. Hell: The Logic of Eternal Damnation. Notre Dame: The University of Notre Dame Press, 1992. ———. “One Hell of a Problem for Christian Compatibilists.” in Daniel Speak & Kevin Timpe (eds.), Free Will and Theism. Oxford: Oxford University Press, 2016.

10 A Theological Three-Body Problem Why Lutherans Are Bound to Struggle With Free Will? Olli-Pekka Vainio

Looking back at his career as a Reformer and theologian, Martin Luther considered De servo arbitrio as one of his most significant works – and to be frank, this work, which stands out as one of the most important and influential works on free will in the history of theology, has been both loved and hated by multitudes. While Luther himself was clearly satisfied with his work, with minor reservations, his colleagues and successors were not as enthusiastic. After Luther’s death, the Lutheran movement had to go through several internal controversies, some of which were directly or indirectly related to the issue of free will. Even if the Book of Concord (1580) was able to settle the controversies, the issue itself has not become any easier over the years.1 In many ways, modern-day Lutherans are still struggling with free will and will probably be condemned to persist in this struggle for some time. Observed from a purely philosophical perspective, Luther ends up between a rock and a hard place in his work since he wishes to maintain both the ideas of individual responsibility and divine determinism. The later Lutheran generations try to remain loyal to Luther’s intentions but remain uneasy about the consequences of his extreme formulations and their metaphysical underpinnings. Effectively, Lutherans subscribe to determinism in theory while being libertarians in practice, thus resolving the paradox. I  think that the doctrine of bound choice is best understood as a tool of spiritual life and a source of mediation of one’s sinfulness. However, this leads Lutherans into a conflicted state from which there seems to be no easy escape. In securing the monergistic doctrine of grace, they are forced to make several hard choices. First, the monergistic doctrine seems to lead to universalism (which they reject). Second, allowing a two-tier approach to theological truths seems to lead to sacrificium intellectus (which they reject).2 Effectively, Lutheran theology needs to avoid not just two, but three extremes. Lutherans typically see themselves as trying to avoid pelagianism and determinism. Searching for a middle space between these extremes, however, creates a new problem, which is theological pessimism resulting from playing fast and loose with core theological concepts, such as divine love and justice. DOI: 10.4324/9781003306191-12

172  Olli-Pekka Vainio Luther’s Original Solution Luther’s De servo arbitrio (1525) has been thoroughly analyzed, criticized, and defended in the past, and here I will only provide a brief outline of his argument.3 Even if Luther castigates those who are perhaps too enthralled by philosophical logic, I will take Luther at his word and try to reiterate his claims as accurately as possible.4 De servo arbitrio should be understood as a part of an ongoing debate.5 To confront his Catholic interlocutors, Luther had written a piece called Assertio omnium articulorum in 1521, in which he had outlined his theory of bound choice, experimenting with ideas that were essentially deterministic.6 This caused Erasmus of Rotterdam, who had previously been fond of Luther ideas, to respond with his De libero arbitrio diatribe sive collatio (1524).7 Luther’s major work is a direct response to Erasmus’s Diatribe. Erasmus later offered a lengthy rebuttal of Luther in his Hyperaspistes (1527), which Luther ignored. Luther’s intention is to offer a pastoral account of Christian life, not a philosophically sound theodicy. Luther recognizes Erasmus’s concerns but decides not to address them as this would signal a false way of entering the genuinely Christian way of thinking. As will become apparent, Luther is not able to divorce himself totally from philosophy, and he likes to make philosophical distinctions to get himself out of quagmires in which he seems to get stuck from time to time. Luther sees the human soul as completely under the control of sin and Satan, which is manifested in us, being slaves to our affects and passions.8 Luther thinks that even asking for help makes us meritorious since the whole of salvation can be attributed to this little act.9 Luther is driven by an extreme Augustinian concern to rob people from all their ways of acquiring merits of any kind. Even if he also used divine foreknowledge as a cause of bound choice (if God knows the future and cannot err, our freedom is non-existent), he later seems to reject this as an argument, apparently realizing that this would effectively make humans mere automatons.10 Luther makes a distinction between voluntas (will) and arbitrium (choice).11 He did not claim that humans do not have will, only that the way they make choices is severely limited. Consider an analogy. We are falling from a plane without a parachute. During the fall, we can flap our arms as much as we want but this does nothing to improve our situation. What eventually kills us is the fall. Likewise, being born into original sin condemns us to a trajectory that will doom us, no matter what we do. So, in the end, we might as well do nothing and our fate would be the same. This distinction allows Luther to say that we are not mindless objects but individuals, which is a fair point to make, but it does nothing to solve the original problem.12 In fact, it might make it even worse as it seems less cruel to burn logs than to burn people, who can understand that they are being burned, unlike logs. Another way Luther dodges the charges of determinism is to acknowledge that we have freedom in matters that are below us, while our power to choose is bound in matters that are above us.13 Thus, I can make a real choice

A Theological Three-Body Problem 173 whether I read Tolstoy or Austen, but I cannot choose not to sin. I can make a choice between walking and riding a bike, but I cannot make myself believe in a merciful God. Luther is also driven by a specific pastoral concern. On the basis of his own experiences as a member of a monastic order, he is looking for a sure ground for the assurance of salvation.14 His solution is to take everything out of our hands and put God in complete control. If God does everything, then nothing is dependent on us. This is a rather extreme move, but it becomes understandable if one realizes that for Luther, the major soteriological question had been whether he had done enough.15 But if there is nothing one can do, then there is nothing to worry about. Of course, this does nothing to resolve the essence of the original problem. If the question before the Reformation was “Have I done enough?”, now the question was “Am I among the elect?” It seemed that little progress had been made. Luther subscribes to a series of theological claims that are hard to fit together. Consider the following: 1 All those who are damned are damned because of their sins (Original Sin). 2 No one can choose to believe in and love God (Bound Choice). 3 Only those believe in God or love him are those on whom God donates his grace (Election). 4 God is all-powerful, all-knowing, and all-loving (Classical Theism). 5 Hell is populated (Anti-Universalism). These propositions create the problem that haunts all monergistic positions: Cur alii et alii non? Why do those believe in God, while those reject God? A stern response would be to admit that God does not will the salvation of all and some Christians choose to believe that.16 Open theists would say that God cannot save everyone because God cannot control the free creation.17 But Luther rejects both solutions and wants to have his Christianity in the form of paradoxes. Christian faith requires that the believers believe that while 1 and 5 are contradictory, this is only an apparent contradiction, which is relieved in the Eschaton. Only in the divine light of Heaven can we understand how all these claims could be true at the same time.18 Trying to solve this paradox in this life will result in some form of heresy. Either we make salvation dependent on human action, or we claim that God is evil. Early Lutheran Reactions Philip Melanchthon

The first systematic exposition of Lutheran theology was Philip Melanchthon’s Loci Communes (1521), which he would rewrite numerous times in the coming decades. Melanchthon was Luther’s right hand, a renowned intellectual and humanist in his own right, who became fascinated by Luther’s ideas, but

174  Olli-Pekka Vainio later in his life distanced himself somewhat from his teacher. In Loci Communes 1521, he presents two arguments for the bound choice that were in line with Luther’s thinking. The first is based on our corrupted affects and the second on God’s foreknowledge. Because our choices are subject to what we desire, and our desires are evil, there can be no free will. Since God knows the future, there can be no truly contingent future either.19 Melanchthon repeats the same monergistic teaching in the subsequent editions of his Loci, and this view is also included in the Augsburg Confession (1530), which was written by Melanchthon.20 However, in the 1530s, he begins to reconsider his position and only after the death of Luther does he begin to speak openly about his more synergistic views.21 While Luther was still alive, Melanchthon did not have much room to maneuver. When Luther heard about Melanchthon’s pontifications, he blurted out that he was espousing Erasmian theology.22 The late Melanchthon went on to modify his earlier teaching concerning the three causes of conversion, which were the Holy Spirit, the Word, and the consenting will of man. Obviously, the last one of these causes required some clarification. In early disputes, Melanchthon said that he means by it the regenerated will that is made able to consent by God’s grace. In his Commentary on Romans (1532), he had said that while the mercy of God is the cause of election, it needs to be accepted and not resisted.23 In Loci 1535, Melanchthon writes: We do not say this to ensnare the consciences, or to deter men from the endeavor to obey and believe, or from making an effort. On the contrary, since we are to begin with the Word, we certainly must not resist the Word of God, but strive to obey it . . . We see that these causes are united: the Word, the Holy Spirit, and the will, which is certainly not idle, but strives against its infirmity. In this manner ecclesiastical writers are accustomed to join these causes. Basil says: “Only will, and God will precede,” God precedes, calls, moves, assists us, but let us beware lest we resist . . . Chrysostom says: “He who draws, draws him who is willing”.24 Finally, in the last edition of his Loci in 1543, Melanchthon argues that “[t]he merciful God does not deal with man as with a block, but draws him in such a way that his will, too, cooperates.”25 The Formula of Concord

The Formula of Concord was the document that brought peace to Lutherans who had started to form parties and argue among themselves after Luther’s death. A series of disputes, known as the Majoristic and Synergistic controversies, specifically concerned the role of will and good works in conversion and salvation. Following the intuitions of late Melanchthon, some Lutherans felt that Luther’s extreme Augustinianism had unsatisfying results. One of

A Theological Three-Body Problem 175 the theologians offering the counterarguments was Georg Major, who argued that good works were necessary for salvation. This can be seen as one way of resolving the paradoxes of Luther. If one could merit damnation by resisting grace, then one should be able to secure salvation through obedience to God. Major’s views were connected to the alleged Synergism espoused by late Melanchthon, and they were both rejected together in the second article of the Formula of Concord.26 Melanchthon’s synergistic line of thought was defended by Johann Pfeffinger in 1555, who argued that human will plays some role in conversion, and we are not like blocks of wood. Pfeffinger’s theses led to a series of exchanges among Lutheran theologians. John Stolz and Nicholas Amsdorf attacked Pfeffinger’s theses, while Georg Major defended him. Another set of disputes followed some years later, when Victorin Strigel opposed Matthias Flacius Illyricus, who was a radical follower of Luther.27 It is not the point of this chapter to analyze these disputes as such. It suffices to note that there were several Lutherans who could not get their heads around Luther’s views and that the issue of human will was painful for the Lutheran reformers. Nonetheless, the formulators of the Formula of Concord managed to get enough people behind its affirmation of Luther’s view and those who had followed Melanchthon effectively lost. The relevant articles in the FC do not rely on divine attributes in arguing for the bound choice, underlining the fallen nature of people instead. The outward freedom in matters below us is affirmed but with regard to the things above us it is said that before man is enlightened, converted, regenerated, renewed, and drawn by the Holy Ghost, he can of himself and of his own natural powers begin, work, or concur in working in spiritual things and in his own conversion or regeneration just as little as a stone or a block or clay. (FC SD II.24) The Formula of Concord also explicitly cites De Servo Arbitrio and endorses it (FC SD II). The Formula of Concord names and rejects four false views: first, the Manichean or Stoic view according to which persons do not have freedom even in earthly things and they are coerced to do evil; second, Pelagianism, which teaches that it is possible without divine help to turn to God; third, The Catholic view that individuals have some kind of freedom to initiate the process of salvation, which is then perfected by the Holy Spirit; and fourth, Synergism, where human will and the Holy Spirit together bring about salvific faith (FC SD, 74–77). In the end, even if the most radical solutions are not adopted, the emerging consensus view was essentially the view Luther had argued for in De Servo Arbitrio. However, the way FC proceeds is not by offering an account of the theory of action but by naming the excesses. What we do not want to teach

176  Olli-Pekka Vainio becomes clear, but the content of our actual teaching remains obscure. This is not the worst way of dealing with the doctrinal disputes, but as I will argue, this can destabilize the doctrinal house of cards. The Inherent Instability of Lutheran Logic So how should we understand the logic of the Lutheran view? There is a particular marching order or hierarchy of theological–philosophical strata that Lutherans wish to follow. The primary level is that of grace. On the basis of Luther’s own experiences, God’s grace cannot depend on anything that is in us; otherwise, our salvation would be insecure. This practical consideration leads to a theological presupposition, which is absolute monergism. God is the sole power behind human salvation, and this power is not shared with human agency in any way (this would be called synergism). Effectively, humans are merely passive in the act of being saved.28 This view has drawn heavy criticism from other Christian traditions. The Counter Reformation objected to this view, claiming that this would lead to harmful consequences since if there was no room for human agency, there could be no consistent theory of Christian life and sanctification.29 This problem has surfaced even in the recent ecumenical dialogues with Lutherans and Orthodox Churches, and Lutherans still need to reassure people that mere passivity does not mean that humans do not have will, only that their will is in bondage. Humans are not like rocks or pieces of wood sailing down the stream, which cannot do anything to change their course or way of being. We are free to make our situation worse, but it is not in our power to improve our fate.30 Granted, the Lutheran view is supposed to be primarily a pastoral view, not an exercise in philosophical logic. Luther was not an analytic theologian in line of Aquinas or Scotus. However, he consistently uses logic and philosophical arguments in developing his own views.31 Luther’s primary concerns lead him to a shaky ground, which he seems to realize, but it is not clear whether he could fathom the depth of the quicksand onto which he had wandered. His solution indeed threatens the whole endeavor of theology as public discipline. This can be illustrated through two crucial divine attributes: love and justice. The Lutheran doctrine of grace is dependent on univocal understanding of divine love so that God’s love is not something totally different from ordinary love. God does not give a stone to a hungry person and call this act love. However, the monergistic doctrine of grace requires equivocal understanding of divine justice. If no one can contribute to his or her salvation as everything depends on God, and there is no ground for salvation in any of us, then it is only God’s random decision who gets saved and who is damned. There is no point asking why Jack was saved while Jill was damned; it is purely God’s inscrutable fiat. In keeping with this, God could in principle save everyone, but again in his inscrutable will, he decides not to do this. Now, from the univocal point of view, this

A Theological Three-Body Problem 177 would be a sign of cruelty and injustice. If I had the means and disposition to save a drowning man, and I chose not to, the earthly courts would rightly condemn me. In fact, the Finnish Law mentions a crime called “the neglect of the act of rescue,” which can take place in such situations, when it is in one’s power to save someone, but one chooses not to act in a salvific way.32 My intention is not to accuse God of neglect, however, as I do not think that this makes any sense theologically unless one wishes to argue that there is an evil god.33 My argument concerns theology as a rational and public endeavor. In his discussion on theological language, Duns Scotus argues that theological concepts need to be understood in a univocal manner if theology is supposed to make any sense in the first place.34 Theology consists of public arguments, which should be assessed by means that are open to all with relevant intellectual skills. For an argument to be rational, all the concepts in the premises and conclusions should be the same so that their meaning does not fluctuate. If they aren’t, the argument makes no sense. From the Scotist perspective, the Lutheran solution requires that one interpret love in a univocal way and justice in an equivocal way, although I think that we cannot understand the meaning of love without also connecting it with justice. If love fundamentally means that I wish good things for the object of my love, it does not make sense to think that it is acceptable if I hand a stone to the object of my love when she asks for bread. The problematic consequence of the Lutheran solution is that it makes theological arguments nonsensical. As noted, I do not argue that God is unjust. Instead, I argue that our trust in theological language holding any truth value drops so low that we might be better off if we just abandoned the whole theological endeavor. This position is called theological pessimism. Many Lutheran theologians argue that Luther’s central aim is to “Let God be God.”35 I  acknowledge that this quite probably is the intention, but the choice itself corrupts the meaning of term “God.” Consequently, we do not know what “letting God be God” means anymore. I understand that Lutherans do not want to be theological pessimists in this sense, even if they might be pessimists in other senses of the word. The way out could be construed by offering a definition of theological language in an instrumentalist fashion. Consequently, theological terms would neither contain their objects nor they could not be taken to be exact descriptions of what they refer to. This solution would be like Bas van Fraassen’s instrumentalist account of scientific theories.36 Even the best scientific theories do not capture or represent reality as it is. It suffices that theory merely helps us to act in the scientific world so that we can achieve our goals. The question about the “fit” of theory with the world is secondary and in fact too difficult to resolve. Moreover, thinking about our theories in absolutist terms could stifle the progress of science since, having arrived at a perfect theory already, we would have no reason to develop a better. Of course, instrumentalism requires some connection to the reality. Otherwise, we could

178  Olli-Pekka Vainio not speak about development in any meaningful sense. But this is not a place for a discussion of the philosophy of science. I am offering instrumentalism simply as a close analogy. In theology, this would mean that our concepts do not really apply to God. Now, there is a major problem with this move. Luther seems to think that theological concepts not only refer to their object but also carry those objects with them. Thus, Luther is a proponent of a kind of a hyperrealist theory of language.37 This is not something that is easy to divorce from Luther’s theology since it grounds much of other interpretations of central doctrines like the real presence of Christ in the Eucharist and the doctrine of justification as deification.38 One could perhaps argue that theological concepts do in fact have this hyperrealist ability, but we just cannot grasp what that means. This seems unsatisfactory as it does not make much sense to make a robust ontological claim while simultaneously arguing that we do not really know what it means. Therefore, I argue that Luther’s position is inherently unstable as he, or those who follow him, may be forced to choose between univocity and theological pessimism. Since opting for univocity does not allow Luther to play the hand he plays in De servo arbitrio, theological pessimism seems a better option. It must be said that this is not an ideal option for Luther either, but it has been a solution that later Lutheran or quasi-Lutheran theologians have adopted. This brings us finally to the title of this essay. In Newtonian physics, the three-body problem refers to our inability to reliably predict the movements of three moving bodies even when we know their trajectories, velocities, and masses. Because of other factors in play, such as gravitational and magnetic fields, the objects will behave unpredictably. Using this physical problem as an analogy, I state that Lutheran theology exists in a similar kind of field of various effects that make it wobbly. A standard simple way of looking at the Lutheran position is to see it as moving on a line with Pelagianism at one end and Determinism at the other, or to use the vocabulary of the Confessions, Manichaeism. Lutherans want to avoid both extremes and stay in the middle, even if they are unable to formalize this position as it is, except by way of naming the extremes and saying that we condemn those positions. Someone could claim that there is no real middle ground between the extremes, but for now I let that slide. For the sake of the argument, let us suppose that there is such a middle position. The third gravitational center is Antirealism, which is born out of paradoxes that are required to bring about the middle position. What follows from this is that the Lutheran position is inherently unstable. It is impossible to pinpoint what it is or how it should be practiced without simultaneously resolving the paradox. And even in where one succeeds in holding on to the paradox, the language used to sustain it is in constant danger of falling into the Kantian black hole, by which I mean the tradition of German Liberal theology and its various strands that are too complex to untangle here. To put it bluntly, German Liberal tradition in the line of Friedrich Schleiermacher

A Theological Three-Body Problem 179 understood theology as an elevated form of anthropology. The focus of theology is the believing subject, not God. If theology speaks about God, it does so only through the human experience.39 This is a reworked idea based on (poorly understood) theology of the cross, according to which God can only be understood through Christ the man.40 Therefore, we should not approach God in his majesty, but only through his incarnate Son. For Luther, this was a major pastoral principle: do not think of God abstractly, but only how he has been revealed in the act of incarnation.41 As noted, Luther was a theological realist and held a traditional view of God, even if he was somewhat suspicious of natural knowledge of God.42 Even if Luther was not an anti-metaphysical thinker, as has falsely been claimed by the Kantian tradition that controlled much of the academic reading of Luther for a long time, his ideas made the slope slippery in ways that he probably did not intend.43 There is only a short step from Luther’s pastoral principle to a deeper anti-metaphysical thesis, which is justified by a commonsensical remark: what good is there having a metaphysical view which cannot be discussed rationally?44 In fact, something like this is affirmed by Friedrich Schleiermacher as he tried to give an account of the role of theological language. Andrew Dole summarizes Schleiermacher’s position as follows: The dogmatic theologian should advance as few claims as possible about the mind- independent reality as the results of her theological work. Such claims must be deducible from the contents of the pious selfconsciousness under the supposition that this is veridical, and should be formulated so as to extend as little as possible into the domain of scientific inquiry. Furthermore, their theological value is restricted to their adequacy as expressions of pious self-consciousness and their mutual coherence; neither their truth value nor their entailments are of independent interest to the theologian.45 At least to me, this seems to be a natural consequence of the Lutheran theology of paradoxes. And, as I have stated, this is a position close to which traditional Lutherans do not like to wander as it undermines much of Luther’s core theological ideas. Conclusion If we observe the history of Lutheran theology and even its contemporary manifestations, we can easily observe how Lutheranism has tried to keep the paradox unresolved while often falling into one of the three black holes. At times, Lutheranism has effectively become a form of Calvinism, despite vocal rejections of such a move. Pietistic forms of Lutheranism have adopted Arminian or Melanchthonian doctrines of will and grace. And it is not a secret that the several generations of Protestant liberal theology arose from within the Lutheran circles.

180  Olli-Pekka Vainio Lutherans have historically been quite busy in avoiding the traditional extremes, in fact so busy that they have not paid enough attention to the third extreme. So, what is the lesson of this story? Engaging in analytic dogmatic theology means that one must make choices, and ideally be aware of the costs of each choice. Furthermore, one cannot abstain from making the choices and there is always a price to be paid. My analysis as such does not claim that Lutherans are in a bad place when they exist, or try to exist, in the middle of three black holes. Maybe this is the best we can do in theology, and the methods of analytic theology tell us exactly that. Abbreviations CA CR FC FC SD LW MW StA  WA

Augsburg Confession (Confessio Augustana) Corpus Reformatorum The Formula of Concord (Formula Concordiae) Solid Declaration of the Formula of Concord (Solida declaratio) Luther’s Works Melanchthons Werke. Studienausgabe D. Martin Luthers Werke. Kritische Gesamtausgabe

Notes 1 Robert Kolb, Timothy J. Wengert & Charles P. Arand, The Book of Concord: The Confessions of the Evangelical Lutheran Church (Minneapolis: Fortress Press, 2000). 2 Theodor Dieter, “Martin Luther’s Understanding of “Reason.” Lutheran Quarterly 25(3) (2011). 3 A good overview of the existing literature is offered by Miikka Ruokanen, Trinitarian Grace in Martin Luther’s Bondage of the Will (Oxford: Oxford University Press, 2021). The German text can be found in Weimarer Ausgabe (WA18) and the English translations are available in Luther’s Works (LW 33). 4 We have provided a more accurate reading of Luther along these lines in Aku Visala and Olli-Pekka Vainio, “Erasmus versus Luther: A Contemporary Analysis of the Debate on Free Will.” Neue Zeitschrift fur Systematische Theologie und Religionsphilosophie 62(3) (2020). See also Andrea Vestrucci, Theology as Freedom: On Martin Luther’s De Servo Arbitrio (Tübingen: Mohr Siebeck, 2019). 5 Robert Kolb, Bound Choice, Election, and Wittenberg Theological Method: From Martin Luther to the Formula of Concord (Grand Rapids: Eerdmans, 2005). 6 WA 7, 308–457. 7 For the English translation of the text, see Erasmus of Rotterdam, “On the Freedom of the Will: A Diatribe or Discourse by Desiderius Erasmus of Rotterdam.” in Gordon E. Rupp (ed.), Luther and Erasmus. Free Will and Salvation (Louisville: Westminster Press, 1969). 8 WA18: 658; LW 33, 98. See also Simeon Zahl, “The Bondage of Affections: Willing, Feeling, and Desiring in Luther’s theology, 1513–25.” in Dale Coulter and Amos Yong (eds.), The Spirit, the Affections and the Christian Tradition (Notre Dame: Notre Dame University Press, 2016). 9 WA 18:733; LW 33, 211. 10 WA 18: 718; LW 33, 189. Ruokanen, Trinitarian Grace, 10–11, 114. 11 Ruokanen, Trinitarian Grace, 35–37.

A Theological Three-Body Problem 181 12 See, for example, Luther’s discussion of various forms of necessity in WA 18, 634; LW 33: 64. 13 WA18: 672; LW 33: 118–119. 14 Susan E. Schreiner, Are You Alone Wise? The Search for Certainty in the Early Modern Era (Oxford: Oxford University Press, 2011), 37–62. 15 Leif Grane, Contra Gabrielem: Luthers Auseinandersetzung mit Gabriel Biel in der disputatio contra scholasticam theologicam 1517 (København: Gyldendal, 1962). 16 This position is held by the TULIP tradition of Calvinism. 17 E.g., Thomas J. Oord, The Uncontrolling Love of God: An Open and Relational Account of Providence (Downers Grove: IVP Academic, 2015). 18 WA18: 766; LW 33: 262. 19 Loci Communes 1521. MW StA II/1: 21–31. 20 See, for example, articles CA II, V, XVIII, and XIX. On the views of early Melanchthon and his general support of Luther against Erasmus, see Timothy J. Wengert, Human Freedom, Christian Righteousness: Philip Melanchthon’s Exegetical Dispute with Erasmus of Rotterdam (New York: Oxford University Press, 1998); Timothy J. Wengert, “Philip Melanchthon’s Contribution to Luther’s Debate with Erasmus over the Bondage of the Will.” in Timothy J. Wengert (ed.), Philip Melanchthon, Speaker of the Reformation. Wittenberg’s Other Reformer (Farnham: Ashgate, 2010). On the historical development of Melanchthon’s views, see Gregory Graybill, Evangelical Free Will: Philip Melanchthon’s Doctrinal Journey on the Origins of Faith (Oxford: Oxford University Press, 2010). 21 For the history of this controversy, see F. Bente, Historical Introductions to the Book of Concord (St. Louis: Concordia Publishing House, 1965). Irene Dingel, “The Culture of Conflict in the Controversies Leading to the Formula of Concord (1548–1580).” in Robert Kolb (ed.), Lutheran Ecclesiastical Culture, 1550–1675 (Leiden: Brill, 2008). 22 “Haec est ipsissima theologia erasmi, neque potest quidquam nostrae doctrinae esse magis adversum.” Quoted in Bente, Historical Introductions, 113. 23 CR 21: 451 (Loci 1535). See also MW StA V, 254, 8–12. (Römerbrief-Kommentar 1532); CR 21: 332 (Loci 1533); MW StA II/2, 597, 3–5 (Loci pracipui theologici 1559). 24 CR 21: 376. “In hoc exemplo videmus coniungi has causas, Verbum, Spiritum sanctum, et voluntatem, non sane otiosam, sed repugnantem infirmitati suae.” Quoted in English in Bente, 125. The same view is repeated in CR 21: 658 (Loci 1543). 25 CR 7: 51. “dass sein Wille auch mitwirket” 26 The strategy of FC was not to name the teachers but only the teachings it rejected. This was meant to diminish internal strife. 27 See Dingel, “The Culture of Conflict”, 46–50. Bente, Historical Introductions, 32–142. Kolb, Bound Choice. 28 On the enigmatic notion of passivity, see Phillip Stoellger, Passivität aus Passion. Zur Problemgeschichte einer categoria non grata (Tübingen: Mohr Siebeck, 2010). 29 Reinhard Hütter, “The Christian Life.” in Kathryn Tanner, John Webster & Ian Torrance (eds.), The Oxford Handbook of Systematic Theology (Oxford: Oxford University Press, 2007). 30 Hannu Kamppuri, Dialogue between Neighbours: The Theological Conversations between the Evangelical-Lutheran Church of Finland and the Russian Orthodox Church, 1970–1986 (Helsinki: Luther-Agricola-Gesellschaft, 1986). 31 As Ruokanen points out (Trinitarian Grace, 43, 72, 117–8, 125), Luther’s solution is based on a series of distinctions, such as things being above or below us, different kinds of necessities, and deus revelatus/absconditus.

182  Olli-Pekka Vainio 2 Finnish Criminal Law 21.15. 3 33 I acknowledge that there is such a debate, but it is not relevant in this context. See, for example, Ben Page & Max Baker-Hytch, “Meeting the Evil God Challenge.” Pacific Philosophical Quarterly 101(3) (2020). 34 Thomas Williams, “The Doctrine of Univocity is True and Salutary.” Modern Theology 21 (2005). 35 For example, Kolb, Bound Choice, 32, 139. WA 18, 685; LW 3, 140. 36 See, for example, Bas van Fraassen, Scientific Representation: Paradoxes of Perspective (New York: Oxford University Press, 2008). 37 Reijo Työrinoja, “Nova Vocabula et Nova Lingua: Luther’s Conception of Doctrinal Formulas.” in Tuomo Mannermaa, Anja Ghiselli  & Simo Peura (eds.), Thesaurus Lutheri: Auf der Suche nach neuen Paradigmen der Luther-Forschung (Helsinki: Luther-Agricola-Gesellshaft, 1987). Eeva Martikainen, Doctrina: Studien zu Luthers Begriff der Lehre (Helsinki: Luther-Agricola-Gesellschaft, 1992). 38 Olli-Pekka Vainio, “Martin Luther on Perception and Theological Knowledge.” Neue Zeitschrift fur Systematische Theologie und Religionsphilosophie 57(1) (2015), https://doi.org/10.1515/nzsth-2015-0005. 39 Friedrich Schleiermacher, The Christian Faith (Edinburgh: T & T Clark, 1999). See also comparison of various theological positions in George A. Lindbeck, The Nature of Doctrine: Religion and Theology in a Postliberal Age. 25th Anniversary Edition (Louisville: Westminster John Knox Press, 2007). 40 Kari Kopperi, “Theology of the Cross.” in Engaging Luther (Eugene: Cascade, 2011). 41 Ruokanen’s Trinitarian Grace is a good defense of this reading. 42 A balanced view of Luther’s natural theology is Ilmari Karimies, “Lutheran Perspective on Natural Theology.” European Journal of Philosophy of Religion (2017). 43 For this kind of trajectory, see for example, Timothy Stanley, Protestant Metaphysics after Karl Barth and Martin Heidegger (Eugene: Wipf and Stock, 2010). 44 For a positive account of this problem, see Vestrucci, Theology as Freedom, 105–6. Vestrucci argues that the hiddenness of God guarantees that our inferred theological concepts do not rise to the level of revelation. 45 Andrew Dole, “Schleiermacher’s Theological Anti-Realism.” in Michael Rea & Oliver Crisp (eds.), Analytic Theology (New York: Oxford University Press, 2009), 150.

Bibliography Bente, F. Historical Introductions to the Book of Concord. St. Louis: Concordia Publishing House, 1965. Dieter, Theodor. “Martin Luther’s Understanding of “Reason.”” Lutheran Quarterly 25(3) (2011): 249–78. Dingel, Irene. “The Culture of Conflict in the Controversies Leading to the Formula of Concord (1548–80).” in Robert Kolb (ed.), Lutheran Ecclesiastical Culture, 1550–1675. Leiden: Brill, pp. 15–64, 2008. Dole, Andrew. “Schleiermacher’s Theological Anti-Realism.” in Michael Rea & Oliver Crisp (eds.), Analytic Theology. New York: Oxford University Press, pp. 136–54, 2009. Grane, Leif. Contra Gabrielem: Luthers Auseinandersetzung Mit Gabriel Biel in Der Disputatio Contra Scholasticam Theologicam 1517. København: Gyldendal, 1962. Graybill, Gregory. Evangelical Free Will: Philip Melanchthon’s Doctrinal Journey on the Origins of Faith. Oxford: Oxford University Press, 2010. Hütter, Reinhard. “The Christian Life.” in Kathryn Tanner, John Webster & Ian Torrance (eds.), The Oxford Handbook of Systematic Theology. Oxford: Oxford University Press, pp. 285–305, 2007.

A Theological Three-Body Problem 183 Kamppuri, Hannu. Dialogue between Neighbours: The Theological Conversations between the Evangelical-Lutheran Church of Finland and the Russian Orthodox Church, 1970–1986. Helsinki: Luther-Agricola-Gesellschaft, 1986. Karimies, Ilmari. “Lutheran Perspective on Natural Theology.” European Journal of Philosophy of Religion 9(2) (2017): 119–38. Kolb, Robert. Bound Choice, Election, and Wittenberg Theological Method: From Martin Luther to the Formula of Concord. Grand Rapids: Eerdmans, 2005. Kolb, Robert, Timothy J. Wengert, and Charles P. Arand. The Book of Concord: The Confessions of the Evangelical Lutheran Church. Minneapolis: Fortress Press, 2000. Kopperi, Kari. “Theology of the Cross.” in Olli-Pekka Vainio (ed.), Engaging Luther. Eugene: Cascade, pp. 155–72, 2011. Lindbeck, George A. The Nature of Doctrine: Religion and Theology in a Postliberal Age. 25th Anniversary Edition. Louisville: Westminster John Knox Press, 2007. Martikainen, Eeva. Doctrina: Studien Zu Luthers Begriff Der Lehre. Helsinki: Luther-Agricola-Gesellschaft, 1992. Oord, Thomas J. The Uncontrolling Love of God: An Open and Relational Account of Providence. Downers Grove: IVP Academic, 2015. Page, Ben  & Max Baker-Hytch. “Meeting the Evil God Challenge.” Pacific Philosophical Quarterly 101(3) (2020): 489–514. Rotterdam, Erasmus of. “On the Freedom of the Will: A Diatribe or Discourse by Desiderius Erasmus of Rotterdam.” in Gordon E. Rupp (ed.), Luther and Erasmus. Free Will and Salvation. Louisville: Westminster Press, 1969. Ruokanen, Miikka. Trinitarian Grace in Martin Luther’s Bondage of the Will. Oxford: Oxford University Press, 2021. Schleiermacher, Friedrich. The Christian Faith. Edinburgh: T & T Clark, 1999. Schreiner, Susan E. Are You Alone Wise? The Search for Certainty in the Early Modern Era. Oxford: Oxford University Press, 2011. Stanley, Timothy. Protestant Metaphysics after Karl Barth and Martin Heidegger. Eugene: Wipf and Stock, 2010. Stoellger, Phillip. Passivität Aus Passion. Zur Problemgeschichte Einer Categoria Non Grata. Tübingen: Mohr Siebeck, 2010. Työrinoja, Reijo. “Nova Vocabula Et Nova Lingua: Luther’s Conception of Doctrinal Formulas.” in Tuomo Mannermaa, Anja Ghiselli and Simo Peura (eds.), Thesaurus Lutheri: Auf Der Suche Nach Neuen Paradigmen Der Luther-Forschung. Helsinki: Luther-Agricola-Gesellschaft, pp. 221–36, 1987. Vainio, Olli-Pekka. “Martin Luther on Perception and Theological Knowledge.” Neue Zeitschrift fur Systematische Theologie und Religionsphilosophie 57(1) (2015). https://doi.org/10.1515/nzsth-2015-0005 van Fraassen, Bas. Scientific Representation: Paradoxes of Perspective. New York: Oxford University Press, 2008. Vestrucci, Andrea. Theology as Freedom: On Martin Luther’s De Servo Arbitrio. Tübingen: Mohr Siebeck, 2019. Visala, Aku & Olli-Pekka Vainio. “Erasmus Versus Luther: A Contemporary Analysis of the Debate on Free Will.” Neue Zeitschrift fur Systematische Theologie und Religionsphilosophie 62(3) (2020): 311–35. Wengert, Timothy J. Human Freedom, Christian Righteousness: Philip Melanchthon’s Exegetical Dispute with Erasmus of Rotterdam. New York: Oxford University Press, 1998.

184  Olli-Pekka Vainio ———. “Philip Melanchthon’s Contribution to Luther’s Debate with Erasmus over the Bondage of the Will.” in Timothy J. Wengert (ed.), (chap. VI) Philip Melanchthon, Speaker of the Reformation. Wittenberg’s Other Reformer. Farnham: Ashgate, 2010. Williams, Thomas. “The Doctrine of Univocity Is True and Salutary.” Modern Theology 21 (2005): 575–85. Zahl, Simeon. “The Bondage of Affections: Willing, Feeling, and Desiring in Luther’s Theology, 1513–25.” in Dale Coulter  & Amos Yong (eds.), The Spirit, the Affections and the Christian Tradition. Notre Dame: Notre Dame University Press, 2016.

Part III

Christology

11 Another Look at the Final Temptation of Christ David Worsley

Could the incarnate Christ sin?1 I argue: Possibly. In what follows, I compare (what I take could have been) Christ’s temptation in Gethsemane with (what I take to be a good candidate for) the temptation faced by both primal and original sinners.2 Assuming the primal sinner to be intellectually flawless, supremely happy, and morally good, I will take them to be as close to impeccable as a peccable creature can be. Indeed, if Christ were to have a peccable human nature, I would assume both Christ and the primal sinner to be in this respect similar.3 With these two assumptions in hand, I suggest a plausible candidate for Christ’s Gethsemane temptation, namely, a structurally similar temptation to that to which the primal sinner succumbed.4 If Christ’s temptation in Gethsemane was sufficiently similar to the temptation faced by the primal sinner, and if we are happy ascribing to the primal sinner the status of “sinner,” we should concede that Christ was genuinely tempted to sin, and not merely, as Richard Swinburne suggests, tempted to do less than the best (but not to sin).5 I begin this chapter with a foray into recent work on the primal sin. I then show one way in which Christ’s temptation in Gethsemane might share structural similarities with the temptation facing the primal sinner. I conclude with some brief reflections on structural dissimilarities between the temptations facing Christ and the primal sinner. Sin: Primal Providing an adequate explanation for Lucifer’s fall is one of the more intractable problems facing the Christian philosophical theologian. Following one line of Church tradition, Lucifer was created supremely happy, intellectually perfect, and morally good.6 And yet, at the very instant after his creation, he – freely and bearing full moral responsibility for his action – sinned.7 This is puzzling for so many reasons. For present purposes, I will limit my discussion to just two problems, helpfully summarized by William Wood in a recent paper on the topic.8

DOI: 10.4324/9781003306191-14

188  David Worsley (1)  The Hard Problem: Understanding “how Satan’s choice to disobey God can be free, motivated, and morally significant.”9 (2)  The Harder Problem: Understanding “how the first sin can be subjectively rational .  .  . from Satan’s own point of view.”10 I will take each in turn. The Hard Problem: Free, Motivated, and Morally Significant Attempts to explain Lucifer’s fall often assume Lucifer had the ability to do otherwise. Nevertheless, an unqualified appeal to Lucifer’s libertarian free will cannot (alone) explain his fall. John Hick notes: The basic and inevitable criticism is that the idea of an unqualifiedly good creature committing sin is self-contradictory and unintelligible. If the angels are finitely perfect, then even though they are in some important sense free to sin they will never in fact do so. If they do sin we can only infer that they were not flawless – in which case their Maker must share the responsibility for their fall and the intended theodicy fails.11 Granting our starting desiderata, that to absolve God from responsibility, Lucifer was created intellectually flawless, morally good, and supremely happy (at least according to the natural order of things), it’s difficult to make sense of Lucifer’s errant free choice. On the one hand, if Lucifer knows what it is good to do, and desires to do that which is good, surely he will only do that which is good. On the other hand, if Lucifer was created with a faulty intellect, or a faulty will, how can he be solely to blame for falling the instant after his creation? Further complicating matters, even if we grant that God is absolved of all responsibility, we will not have shown that Lucifer was morally responsible for his fall; his action could have been inexplicably arbitrary. With these concerns in mind, William Babcock writes: Unless there is some recognizable continuity between agent and act, it will appear either that the evil angels’ will was caused by something other than themselves (some form of compulsion) or that it was completely uncaused (a chance or random outcome). But if there is some recognizable continuity between agent and act, it will appear that the evil angels did not actually share the same initial moral goodness as the good angels and therefore that they were already inclined to evil even before they exercised any moral agency at all. Without continuity, then, the act –. i.e., the first evil will – will not count as the agent’s own; with continuity, the act will not count as the origin of evil.12

Another Look at the Final Temptation of Christ 189 And indeed, in a recent paper outlining the (pre- and) post-Hicksian contours of the discussion, Kevin Timpe concludes “any Christian account of [the] primal sin cannot avoid all arbitrariness.”13 Is Timpe wrong? Nothing I  have come across would suggest otherwise. Nevertheless, I  think a few more details can be fleshed out before throwing in the towel. I will take it that in order for Lucifer’s choice to have been morally significant, Lucifer must have possessed a capacity for what Richard Swinburne describes as “serious free will” such that Lucifer had the ability to freely will some lesser good over a greater good. Concerning such “serious” free will, Swinburne writes: [I]f reasons alone influence action, an agent inevitably does what he believes to be the best, so if desires alone influence action, an agent will inevitably follow his strongest desire. Free choice of action therefore arises only in two situations. One is where there is a choice between two actions which the agent regards as equal best which the agent desires to do equally; which . . . is the situation of very unserious free will. The other is where there is a choice between two actions, one of which the agent desires to do more and the other of which he believes it better to do . . . the more serious the free will and the stronger the contrary temptation, the better it is when the good action is done.14 Minimally, for Lucifer to possess “serious free will,” he must have been capable of choosing between conflicting beliefs and desires. But what form could these conflicting beliefs and desires have taken, given that we’re assuming Lucifer was morally good, intellectually flawless, and supremely happy? Timpe notes that while voluntarist explanations15 and intellectualist explanations16 differ with respect to the origination of sin (occurring respectively in Lucifer’s will or in Lucifer’s intellect17), both broadly agree on the structure of Lucifer’s will: on each, God created Lucifer with what we might describe as an integrated second-order desire for justice or “rectitude” (i.e., the secondorder desire that his first-order desires be properly ordered) and an integrated first-order desire for benefit (i.e., the desire for those things that will lead to his happiness). On this view, rather than Lucifer’s having two conflicting, but equally weighted, higher order desires (which would be a recipe for unserious free will), Lucifer’s higher order and lower order desires may have conflicted, with Lucifer choosing to act on a lower order desire for the sake of his happiness over and against a conflicting higher order desire for justice. Of course, it is not enough to outline how a morally significant conflict might have occurred. In order for such conflict to actually occur, we also need a distinct object for this first evil desire, an object desirable to an intellectually flawless, morally good, and supremely happy creature. This object of this first evil desire is traditionally identified with the beatific vision, namely, that which would have made Lucifer supremely happy according to the supernatural order of things.18

190  David Worsley I’ll use Aquinas as an exemplar within this tradition. On Aquinas’ view, while created supremely happy according to the natural order of things, Lucifer was not created supremely happy according to the supernatural order. Supernatural happiness requires beatific revelation, and Lucifer was created lacking this. Aquinas writes: But the angels did not have from the beginning of their creation that ultimate beatitude which is beyond the power of nature; because such beatitude is no part of their nature, but its end; and consequently they ought not to have it immediately from the beginning.19 And hence: Therefore, the devil’s first sin was that, to attain the supernatural happiness consisting of the complete vision of God, he did not elevate himself to God so as to desire with holy angels his ultimate perfection through God’s grace. Rather, he wanted to attain his ultimate perfection by the power of his own nature without God bestowing grace, although not without God acting on his nature . . . the devil sinned not by desiring something evil, but rather by desiring something good, viz., ultimate beatitude, but not in a fitting manner, that is, not in such a way as to attain it by God’s grace.20 No doubt, desiring the beatific vision is not in itself bad. As Aquinas points out, merely desiring one’s ultimate perfection cannot be bad. So, why would Lucifer’s so desiring be construed as sinful? And if sinful, how, following Wood, could this be seen as subjectively rational from Lucifer’s (apparently) intellectually flawless point of view? The Harder Problem: A Subjectively Rational Object for the First Evil Desire? Consider supernatural happiness. On Aquinas’ view, for anything with an intellect, such final and perfect happiness could only consist in a vision of the divine essence.21 To see why, he writes: [T]wo things must be considered. First, man is not perfectly happy so long as something remains for him to desire and seek. Secondly, the perfection of any power is determined by the nature of its object. The object of the intellect is ‘what a thing is,’ i.e. the essence of a thing. . . . It follows that the intellect attains perfection, insofar as it knows the essence of a thing. If therefore an intellect knows the essence of some effect, whereby it is not possible to know the essence of the cause, i.e. to know of the cause ‘what it is’; that intellect cannot be said to reach that cause simply, although it may be able to gather from the effect the knowledge that the

Another Look at the Final Temptation of Christ 191 cause exists. Consequently, when man knows an effect, and knows that it has a cause, there naturally remains in the man the desire to know about the cause, “what it is.” And this desire is one of wonder, and causes inquiry. . . . For instance, if a man, knowing the eclipse of the sun, considers that it must be due to some cause, and yet not know what that cause is, he wonders about it, and from wondering proceeds to inquire. Nor does this inquiry cease until he arrives at knowledge of the essence of the cause. If therefore the human intellect, knowing the essence of some created effect, knows no more of God than “that He is”; the perfection of his intellect has not yet directly attained the First Cause, and so the natural desire to seek the cause still remains for him. On account of which he is not yet perfectly happy. Consequently, for perfect happiness the intellect needs to attain to the very essence of the First Cause. And thus it will have its perfection through union with God as with that object, in which alone man’s happiness consists.22 Even granting a difference between angelic and human knowledge, the common perfection of their respective intellects remains, I take it, the same. Put more simply, if a person, angelic or human, knows that something exists, but they do not know the essence of that thing, until they come to know its essence, there will remain in them a desire to know. If they remain desirous of something they do not have, that is, if they continue with an unfulfilled desire for knowledge, both their intellect and will cannot come to perfect rest. Furthermore, as each thing has a cause, and as every cause is itself caused by something else, this metastasizing desire for knowledge will bring that person back to the First Cause, that is, to God. And so, without some sort of knowledge of God’s essence (something well beyond mere knowledge that God exists), a person’s intellect and will can never come to perfect rest, but rather, will continue on in a state of perpetual intellectual frustration. At the point of Lucifer’s creation, he may have had as much natural knowledge of God as naturally possible for him to have had, hence his being as supremely happy according to the natural order of things, but he would not have had this knowledge through which his final and perfect happiness would consist. Why did God create Lucifer without such beatific knowledge? While it might be tempting to blame God for creating Lucifer in such a state, had God immediately granted Lucifer beatific revelation, Lucifer would never have had an opportunity to exercise serious free will. He would have been created morally good, with a will inclined to goodness, and given supernatural knowledge of goodness personified. Much like the beatified saints in heaven and the angels who did not fall, beatific revelation would have rendered Lucifer functionally impeccable.23 Lest one be tempted into thinking this lack renders Lucifer intellectually flawed (and so, smuggle in some divine responsibility for Lucifer’s fall), it would have been impossible for Lucifer to attain the same beatific revelation

192  David Worsley shared by members of the Trinity (i.e., comprehension, rather than apprehension, of the divine essence), for God is infinitely knowable, and Lucifer is necessarily neither omniscient nor omnipotent. So, if intellectual flawlessness is a concept that can be sensibly applied to a non-divine person, intellectual flawlessness must be compatible with a limited degree of beatific knowledge (and if not full comprehension, then limited, or no, apprehension, too). Permitting the exercise of serious free will (and so, rendering possible a morally significant relationship between Lucifer and God) might not have been the only reason for creating Lucifer in a state of natural but not supernatural happiness. If God is infinitely knowable (i.e., if beatific knowledge comes in an infinite number of degrees), and if, after receipt of beatific revelation, one cannot grow in the degree of beatific knowledge one has, the time between one’s creation and one’s receipt of beatific revelation is the only opportunity one has to arrive at a greater degree of beatific knowledge. How could a person come to a greater degree of beatific revelation? Let’s return to Aquinas. As he explains matters, in order for a person to know anything, there needs to be a certain union between (1) an intelligible species (by which they know that thing) and (2) a power to know said intelligible species (under which they know that thing). However, it seems pretty clear that a person cannot come to know an infinite God through a created, and therefore finite, intelligible species. Instead, at the beatific vision, God unites himself to a person’s intellect, with his very essence taking the role of the intelligible species. Nevertheless, such union alone is still insufficient for beatific knowledge. God must also gift the beholder a supernatural power to know God’s essence. That is, at the beatific vision God also gives the beholder a created “light of glory,” and he gives them as much of this light as they, through their love for God, desire. On this way of understanding beatific revelation, the greater the beholder’s love for God grows in this life, the greater the power to know they will receive. The great the power to know that they receive, the greater the degree to which they will know God in the beatific vision, and, so, the more intense their beatific enjoyment will be. Thus, one person could come to know God to a greater degree than another in virtue of the fact that pre-beatific revelation, they come to have a stronger desire for God, and so, in the satisfaction of this desire, receive a greater power to know God’s essence. Could a person come to know God to a greater degree once they have beheld the beatific vision? On Aquinas’ account, it seems not. Christina van Dyke sums up Aquinas’s view in the following way: Differences in our will’s dispositions and affections that were formed over the course of our earthly lives thus appear to have a lasting effect. In the beatific vision, our wills rest in eternal and perfect enjoyment of the ultimate end, but the degree of that unchanging enjoyment depends on how we have disposed our wills in this life.24

Another Look at the Final Temptation of Christ 193 Because the beholder receives as much “light of glory” as they, through charity, desire, if, in a contemplative afterlife, there is no opportunity for such charity to increase, then the beholder’s desire for God will not change in strength. If the desire the beholder has for God remains unchanging, their power to know the divine essence will likewise remain unchanged. In Aquinas’ contemplative afterlife, we reach a chicken-and-egg-like impasse when it comes to knowledge of God and desire for God: the beatific beholder cannot increase the degree of their knowledge of God without first increasing their desire for God, but they cannot increase their desire for God without first increasing their degree of knowledge of God. Each person has precisely the degree of beatific knowledge that they desire. Were they to desire a greater degree of knowledge, they would have had it. So, post-beatific revelation, neither desire for nor knowledge of God can grow. With this thought in hand, God has a second reason not to provide Lucifer with immediate beatific knowledge: to give Lucifer an opportunity to increase the degree of beatific knowledge he would receive, and so, increase the intensity of Lucifer’s supernatural happiness. And it is reflection on this point that might indicate where Lucifer has, I think, a genuine choice to make. If there are an infinite number of degrees of beatific knowledge available, with the degree of beatific knowledge one receives unable to increase once received, Lucifer would have had a choice between desiring beatific revelation immediately, or waiting for his desire for God to grow, ensuring a greater degree of beatific revelation in the future. On the face of things, this should be an easy decision. For every moment Lucifer might presently wait, he would everlastingly enjoy a greater degree of supernatural happiness in the future! However, if one has a rational reason to wait such that one’s desire might come to grow, and there are infinite degrees of beatific knowledge available, once one has chosen to wait, there is no rational reason to cease waiting. And, indeed, no guarantee that one will ever receive beatific revelation. Given that God is infinitely knowable, there can be no milestone degree of beatific knowledge to aim for. In effect, the choice to wait amounts to a choice to trust that God will, at some unspecified point in the future, intervene to grant you beatific revelation.25 The alternative to waiting is, of course, to attempt immediate attainment. Now, of course, there is more to beatific knowledge than mere intellectual itch-scratching. Where this is the case, God would be little more than an instrument for a person’s intellectual perfection. Rather, the beatific vision ought to stand as the climax of a person’s loving relationship with God. On the Thomist account of love recently popularized by, among others, Eleonore Stump, love is the function of two desires: the desire for the good of the beloved, and the desire for union with the beloved, where the greater the degree of knowledge of one for another, the greater the degree of union one shares with them. These two desires coalesce at the beatific vision: beatific revelation serves as both a person’s greatest good26 and the vehicle by which that

194  David Worsley person is finally and completely united to God. So, inasmuch as the degree of beatific knowledge determines the degree of union with God one has, the choice to wait and trust God’s timing is a decision to choose love of and friendship with God over one’s immediate (and, possibly, forgoing forever) supernatural happiness. Framed this way, Lucifer’s choice is a little more understandable. The desire for supernatural happiness is a good desire; one all persons should have. If waiting for greater beatific revelation amounts to trusting that God would grant it you at some future point, and if trust in God is impeded (for instance, due to (1) lack of experience of instances where God has been proven trustworthy (granting that the primal sin occurred the instant after Lucifer’s creation), and (2) lack of knowledge of God’s nature, that is, God’s goodness, as it in fact is), the case for the subjective rationality of Lucifer’s decision can be made just a little bit stronger. This, of course, isn’t to say that the primal sin is now explicable. It’s not. It’s not at all clear why some angels would try for beatific revelation immediately and others not. Neither is it clear how the fallen angels thought they could secure beatific revelation (or whether they just thought God couldn’t be trusted to grant them such revelation, or whether they merely wished it were possible). This does hint at an intellectualist solution to the puzzle (insofar as one might point to a lack of beatific knowledge to explain the primal sin27), but nothing more. Nevertheless, I will suggest this way of framing the primal sin helps to shed light both on the original sin28 and on Christ’s temptation in Gethsemane. Christ’s Temptation in the Garden of Gethsemane Consider Christ in the Garden of Gethsemane. Luke 22:39–44 reads: [Jesus] came out and went, as was his custom, to the Mount of Olives; and the disciples followed him. When he reached the place, he said to them, “Pray that you may not come into the time of trial.” Then he withdrew from them about a stone’s throw, knelt down, and prayed, “Father, if you are willing, remove this cup from me; yet, not my will but yours be done.” Then an angel from heaven appeared to him and gave him strength. In his anguish he prayed more earnestly, and his sweat became like great drops of blood falling down on the ground. This scene takes place the night before Christ’s crucifixion.29 One common way of reading Christ’s agony is simply that he didn’t want to undergo the humiliation and physical pain attendant to his forthcoming scourging and crucifixion. Yet this doesn’t seem quite right. Hagiography is rich with Christian martyrs going joyfully to their deaths, even to crucifixion, knowing they would soon be with God. Furthermore, there is no sign that after Gethsemane, Christ experienced this same anguish, at least until his cry of dereliction.

Another Look at the Final Temptation of Christ 195 In her recent book, Atonement, Eleonore Stump makes the following connection between these two events: [I]t seems to me better to connect Christ’s distress in the Garden of Gethsemane with the only part of the crucifixion in which Christ also seems to be in anguish, namely, with the cry of dereliction. One way to understand the cry of dereliction is to suppose that in the experience that gives rise to the cry Christ is opening his human psyche to the psyches of all other human beings, so that their psyches come to be in God through entering the human mind of Christ. In that experience, all the vile, cruel, abhorrent human evil anyone has ever done or wanted to do is available for Christ to experience within himself, in the way that human cognitive capacities for empathy and mind-reading enable, but to a much greater degree, as is possible only for the incarnate Son of God.30 In such a state, Stump adds, Christ might well lose entirely his ability to find the mind of God the Father. For him in that condition then, God would be even more inaccessible than Frodo’s friends were to him when the Black Riders occupied his mind. Furthermore, because in his psychic connection with the evil in every human being, Christ would also have the simulacrum of the stains on the soul accompanying all that evil, and he would feel the moral ugliness of that evil in himself. In that condition, why would he not feel abandoned by God? The ugliness of those stains is a world away from the beauty of God’s goodness, and even the simulacrum of them would be very troubling.31 Christ’s non-culpable loss of his ability to find the mind of God the Father presents an interesting scenario.32 If, hitherto, Christ beheld the beatific vision in his human nature,33 losing this ability may have seen him lose access to this beatific revelation (i.e., in his human nature). Even if this were just a temporary state of affairs (i.e., just during his cry of dereliction), anticipating (or indeed, foretasting) this state would certainly have warranted Gethsemane’s agony.34 With this in hand, we can think about the Gethsemane temptation in a new way: Christ’s temptation mirrors the choice that I suggested may have faced the primal sinner. On the view presented earlier, Lucifer’s choice was between a desire for beatific knowledge (insofar as such was the best thing for Lucifer, as that which would secure his supernatural happiness) and a desire to wait for beatific revelation in God’s timing (insofar as prioritizing a loving relationship with God is the right thing to do). On the view I am now suggesting, in Gethsemane, Christ had to choose between forgoing the cross, and so keeping beatific revelation in his human

196  David Worsley nature (that which would preserve in his human nature a state of supernatural happiness), and submitting in trust to God’s will, to crucifixion and its attendant universal mind-reading (which would see him lose beatific revelation in his human nature for the sake of love of God).35 By Christ’s own admission, he preferred the former over the latter. Indeed, it seems it took a night of prayer, a ministering angel showing him the joy that was to be set before him, and sweat like drops of blood before this choice was confirmed. It seems to me that this more plausibly explains his description in Gethsemane. It goes without saying, of course, that while Christ’s temptation might have been structurally similar to that of Lucifer’s (and, for that matter Adam’s), the outcome was not. Christ prevailed.36 And so, the author of Hebrews could quite literally declare that we do not have a high priest who is unable to sympathize with our weaknesses, but we have one who in every respect has been tested as we are, yet without sin. (Hebrews 4:15) With this, “every respect” extending back to the very temptation led to the fall. However, given this structural similarity between the primal sin and Christ’s final temptation in Gethsemane, were one to treat Christ’s temptation as Swinburne does, as a temptation to do less than the best (but not to sin37), I suggest one ought to treat the primal sin in a similar light: the primal sinner also chose less than the best (but did not sin). But if “the primal less-than-the-best-er” sits uneasily, so similarly should one be uneasy about Swinburne’s claim that this particular temptation of Christ was merely for less than the best. There is a sense, of course, that a desire for beatific knowledge of God for one’s own sake (rather than for God’s sake, and so forth) isn’t bad. It couldn’t be bad and tempt someone with a morally good will. But if we’re willing to chalk up acting on such a desire as sin when the primal sinner so succumbs, I think we should be prepared to have done the same in Christ’s case, too. Gethsemane and the Primal Sin: Some Dissimilarities While these temptations may be structurally similar (albeit mirrored), there are some dissimilarities that temper this comparison. First, in having the beatific vision in his human nature, Christ would have known that God is good, insofar as God in God’s essence is goodness personified. Lucifer wouldn’t have had access to such knowledge, although he may have had very good reason to think such was the case. Christ’s decision to will the will of the Father would not rely on the same degree of trust that Lucifer’s choice may have required. Second, if the beatific vision ensures functional impeccability in the heavenly saints and in the angels who did not fall, surely it would have done so for

Another Look at the Final Temptation of Christ 197 Christ in Gethsemane, too. Perhaps, but neither the heavenly saints nor the good angels are faced with the choice to give up the very thing that ensures their desires remain integrated around the good.38 Furthermore, if the human nature of Christ was functionally impeccable in virtue of his incarnate receipt of beatific revelation, his loss of beatific revelation on the cross may, presumably, have rendered him (briefly) peccable. Third, based on the way I outlined Lucifer’s temptation, one might think Lucifer was faced with the prospect of permanently giving up his opportunity for beatific revelation (particularly, if he did not trust that God would grant this to him, or he thought such revelation wishable, but supernaturally impossible), while Christ was only faced with the prospect of temporarily giving up beatific revelation. True. But thereafter, Christ would now know what it is like not to experience beatific revelation, and that knowledge would, one imagines, remain a permanent scar. Conclusion In conclusion, I have deliberately refrained from commenting on the explicability of the primal sinner’s succumbing to their temptation. I note, however, that if Christ managed to overcome a mirrored version of that temptation, possibly, so could the primal sinner have overcome theirs. And perhaps this state of affairs is sufficient to defuse the general worries associated with the inexplicability of the primal sin. But if this is sufficient to defuse these worries, it does so at the cost of rendering it at least possible that Christ could likewise have sinned.39 Notes 1 Following Timothy Pawl (Timothy Pawl, In Defence of Extended Conciliar Christology (Oxford: Oxford University Press, 2019), I take the question here to be akin to “Did Christ have a human nature that is peccable?”. I note that while I will answer, “possibly,” Pawl answers, “no,” on the grounds that while Christ’s human nature, were it no longer hypostatically united to Christ’s divine nature, would be peccable, Christ’s human nature is assumed by the Word, which is impeccable, and so Christ’s human nature is impeccable. 2 Following Pawl, I’ll understand human temptation as: “An affectively charged cognitive event in which an object or activity that is associated with pleasure or relief of discomfort is in focal attention, yet the object of that desire conflicts with the person’s values and goals.” Pawl, In Defence of Extended Conciliar Christology, 144. I will understand angelic temptation purely in terms of conflict between higher order and lower order beliefs and desires, with only an intellectual and volitional component. 3 Although not, clearly, these temptations are not identical. Minimally, I take it that Christ beheld the beatific vision in his human nature, while the primal sinner did not. 4 I should say something about how I’ll be understanding temptation in this chapter. Following the author of the book of James, I’ll take the kind of temptation I am interested in to be (i) an action-directed desire that (ii) leads to sin. Focusing on this desiderative element for the moment, James (1:14) writes: each person is tempted when he is lured and enticed by his own desire. Now, the first thing to note here is this: if temptation includes an action-directed desire,

198  David Worsley both the action in question and the state of affairs one expects will obtain once the action has been performed must be desirable to the one tempted (see, e.g., ST III, q. 41, a. 4.). Granting this, I take it that the object and activity of temptation must be perceived by the tempted person as attractive or pleasurable; as good, under some description of goodness, and if perceived as such, then sufficiently salient (even if in some respect unconsciously so) such as to render the desired action a live option for that person. I will therefore understand temptation as a version of craving, where craving is a species of the genus of desire. What differentiates temptation from other kinds of craving is that one also believes that this object or activity is in conflict with one’s values and/or goals. Following Tim Pawl, then, I will therefore understand temptation more broadly as something akin to: (i) An affectively charged cognitive event in which an object or activity that is associated with pleasure or relief of discomfort is in focal attention, yet (ii) the object of that desire conflicts with the person’s values and goals. (Timothy Pawl, “Conciliar Christology, Impeccability, and Temptation.” in Johannes Grössl  & Klaus von Stosch (eds.), Impeccability and Temptation: Understanding Christ’s Divine and Human Will (Oxford: Routledge, 2021), 97) Although I think this works reasonably well as a definition for temptation it, too, can be further disambiguated. Continuing his description of temptation, the author of the book of James writes: [this] desire when it has conceived gives birth to sin, and sin when it is fully grown brings forth death. Not all objects of temptation so defined are morally bad, for one can have morally neutral proximate goals (e.g., completing a marathon) or even distal or remote ends (e.g., completing the marathon to please one’s parents), for which failure is perhaps disappointing, but not constitutive of sin. Nevertheless, when the object of temptation conflicts with a person’s relationship with God, we arrive at the particular type of temptation I will have in mind. Of course, our analysis of temptation could continue ad nauseum. In his excellent paper, “Tempting God,” Brian Leftow lists a whole bunch of necessary conditions for temptation not addressed here, but I  think this definition will suffice for present purposes (Brian Leftow, “Tempting God.” Faith and Philosophy 31(1) (2014): 3–23). 5 See Richard Swinburne, Was Jesus God? (Oxford: Oxford University Press, 2006), 46. Not only is this view at odds with Swinburne’s supererogatory approach, taking Christ’s ‘not my will but yours be done’ at face value, I will argue there is a sense in which Christ wanted to do that which he was tempted to do. 6 Gregory the Great, Hom. xxxiv in Evangelia. See also Thomas Aquinas Summa Theologica I, q. 63, a. 7. (Henceforth, ST) 7 Aquinas writes that this probably took place “at once after the first instant of his creation.” ST I, Q. 63, A. 6. 8 Wood writes: The fall of the devil poses the problem of how to explain the very first sin, the initial act by which evil comes to mar God’s wholly good creation. On the traditional Christian account, Satan and the other angels were created completely good, in an environment that was completely good, and with intellects and wills that functioned exactly as designed by God. Nevertheless, Satan rebelled against God, an act so grave that he was justly condemned to Hell, where he suffers the torment of eternal separation from his creator. What can account for such an inexplicable choice? Given Satan’s pre-fallen cognitive and volitional strengths, his sinful choice seems utterly perverse: an act of existential self-harm heightened to an almost infinite degree. Though less frequently addressed, the problem of the first sin poses a challenge to traditional Christian theism that is no less grave than more widely discussed problems like the general problem of evil, the problem of divine hiddenness, the problem of other religions, and so forth. If we are forced to admit that God is responsible for the first sin, then God cannot be perfectly good. Alternatively, if we can only explain the first sin by positing some unintended defect in the created intellect or will, or some unintended source of evil that already

Another Look at the Final Temptation of Christ 199 infects creation, then God no longer seems like a sovereign, all-powerful, creator. It is easy to misunderstand the precise nature of the problem, and so it is easy to embrace various proposals that initially seem like solutions but really are not. One might think that the problem of the first sin is merely another instance of the problem of evil. That is, one might suppose that the question at hand is simply ‘Why would God allow the devil to sin?’ But the problem of the first sin requires us to justify Satan’s choices, not God’s. The relevant question is why Satan would choose to sin, not why God would allow him to sin. William Wood, “Anselm of Canterbury on the Fall of the Devil.” Religious Studies 52 (2016): 223–4. 9 Wood, Anselm of Canterbury on the Fall of the Devil, 224. 10 Wood, Anselm of Canterbury on the Fall of the Devil, 224. 11 John Hick, Evil and the God of Love (New York: Harper and Row Publishers, 1966): 68–9. As Hick notes, this criticism is already found in Schleiermacher. 12 William Babcock, “Augustine on Sin and Moral Agency.” The Journal of Religious Ethics 16 (1988): 46. 13 Kevin Timpe, “The Arbitrariness of the Primal Sin.” Oxford Studies in Philosophy of Religion 5 (2014): 255. Of course this, too, is concerning. Not least because if arbitrary, the specter is raised of a similarly inexplicable fall befalling a perfected saint in the life to come. I have addressed this question at length elsewhere, see David Worsley, “Could there be suffering in paradise?” Journal of Analytic Theology 4 (2016): 87–105. In personal correspondence with Kevin Timpe, Brian Leftow has suggested we need a better understanding of angels before speculating further. He writes, The problem with Satan’s fall . . . is that it’s a whopping big case of akrasia, but we’re trying to make sense of it while denying ourselves the resources we have In the human case for doing so (e.g. passions). This leads directly to its becoming inexplicable. . . . So why not draw the moral that we’ve got the wrong picture of angels, and add some resources needed to make sense of the story? Why think that angels don’t have passions, for instance? Desire almost paralyzes the will to do otherwise (phenomenologically) . . . The stronger the passion, the more likely one goes with it, ceteris paribus. Given a very strong passion, it can become probable that one goes with it. When what is probable happens, then ceteris paribus what made it probable may explain it. Do reasons really give us anything more? (Timpe, The Arbitrariness of the Primal Sin, 253–4) 14 Richard Swinburne Providence and the Problem of Evil (Oxford: Oxford University Press, 1998): 86–7. 15 Most notably, Anselm’s, recently championed by Katherin Rogers in her Anselm on Freedom (Oxford: Oxford University Press, 2008): 98. For further treatments of Anselm’s account, see also Marilyn McCord Adams “St. Anselm on Evil: De casu Diaboli.” Documenti e studi sulla tradizione filosofica medievale 3 (1992): 429–30, Peter King, “Angelic Sin in Augustine and Anselm.” in Tobias Hoffmann (ed.), A Companion to Angels and Medieval Philosophy (Leiden: Brill, 2012), and Thomas Williams, Anselm: Three Philosophical Dialogues (Hackett Publishing Company, 2002). 16 Most notably, Augustine’s, and more recently defended by Scott MacDonald, “Primal Sin,” The Augustinian Tradition (Gareth Matthews (ed.)) (University of California Press, 1998), 110–39. 17 For slightly different variants on these two, see John Peckham on the will corrupting the intellect, Giles of Rome and Peter Auriol on the indirect self-specification of the will, Godfrey of Fontaines on a variant of the intellectualist account, Duns Scotus on Lucifer’s abandonment of Eudaimonism, Aquinas on (alleged) intellectualist determinism, and Henry of Ghent on volitional determinism. Further accounts of each can be found in Tobias Hoffmann, Free Will and the Rebel Angels in Medieval Philosophy (Cambridge: Cambridge University Press, 2021) and Giorgio Pini, “What Lucifer Wanted: Anselm, Aquinas, and Scotus on the Object

200  David Worsley of the First Evil Choice.” in Robert Pasnau (ed.), Oxford Studies in Medieval Philosophy (volume 1) (Oxford: Oxford University Press, 2013). 18 See, for example, Pini, What Lucifer Wanted: Anselm, Aquinas, and Scotus on the Object of the First Evil Choice for a survey of views on this object. 19 ST I, q. 62, a. 1. 20 Aquinas, De Malo, q. 16, a. 3. 21 At this point, of course, one might follow Duns Scotus, and argue such supernatural beatific knowledge was (and is) both naturally and supernaturally impossible. However, even if this was the case, one could nevertheless wish it were possible, with this wish doing the work required in this account. Scotus writes: With regard to the act of willing in the first way [i.e. as an efficacious volition], I say that the [evil] angel could not desire equality with God. In the second way [i.e. by a mere wish], he could, because he could love himself by as much love of friendship as that by which, according to right reason, he ought to have loved God. And nevertheless he could also have desired for himself as much good as he owed to God by love of desire, if one speaks of the act of the will that is called a “wish” (velleitas). Rep. 2.6.1, n. 5. But it seems going down this route downplays the possibility and extent of our own beatific enjoyment. 22 ST I-II, q. 3, a. 8. See also SCG III ch. 50 and ST Suppl., q. 92, a. 3. 23 Given that I will be assuming that Christ had beatific knowledge in his human nature, I  intend to address this point when looking at Christ’s temptation in Gethsemane. 24 Christina Van Dyke, “Aquinas’s Shiny Happy People: Perfect Happiness and the Limits of Human Nature,” Oxford Studies in the Philosophy of Religion Volume 6 (2014): 275. 25 Trusting in God who, without beatific revelation, you do not know to be good – granting divine simplicity and the convertibility of transcendentals, the knowledge of God’s goodness just is the beatific knowledge of God in question! – so Lucifer would have had to trust that God is good, and that God hasn’t just created him with an unscratchable intellectual itch. 26 On Aquinas’ account at least, it is, after all, that in which our final and perfect happiness consists. 27 In a similar respect, it seems the Genesis 3 narrative plays on Eve’s mistaken belief that one ought not even to touch the tree of knowledge of good and evil, a command never given by God. 28 Where we see a structurally similar story repeated, where once again, a certain lacked knowledge was desired over friendship with God. Of course, whatever we are to make of the Genesis 3 narrative, Adam is not portrayed as intellectually flawless (it seems he fails to accurately report God’s initial command to Eve – recall that Eve tells the serpent that they are neither to eat nor touch the tree, rather than merely not to eat it, as God had originally said), or created supremely happy (for God conceded that it was not good for Adam to be alone!), and indeed, without the knowledge of good and evil, perhaps created in a morally odd state. Despite these differences, it is presumably the prospect of the happiness that would accompany such knowledge that caused Eve to eat from the tree of knowledge of good and evil, and, so thought Augustine, the prospect of losing Eve that saw Adam fall (akin, perhaps, to the other fallen angels). Even granting these earlier dissimilarities, though, I take it that the parallels between the primal sin (as discussed earlier), and Adam’s sin, are clear enough. Each sinner was presented with a state of affairs that allowed for their exercise of serious free will, rendering sin a possibility, without the object of either desire being intrinsically morally bad. Plausibly, each sinner might have been created with a higher order desire for justice (an original justice later withdrawn after each fell), but each was also created with a lower order desire for happiness, manifest in a desire for knowledge of God that could be inappropriately acted upon.

Another Look at the Final Temptation of Christ 201 29 I take it that this is the “opportune time” that Lucifer was waiting for after his earlier temptations failed back in Luke 4:13 “When the devil had finished every test, he departed from him until an opportune time.” This is Stump’s view, see Eleonore Stump, Atonement (Oxford: Oxford University Press, 2018), 272. See also Hebrews 12:4: In your struggle against sin you have not yet resisted to the point of shedding your blood. 30 Stump, Atonement, 274. 31 Stump, Atonement, 164–5. 32 Whether one finds Stump’s analysis of the cry of dereliction plausible, it certainly seems as though there was a curse appended to those who hang on trees, a curse that presumably Christ fell under on the cross. See Deuteronomy 21:23 and Galatians 3:13. 33 For a very full defense of this claim, see Simon Gaine, Did the Savior see the Father? Christ, Salvation, and the Vision of God (London: Bloomsbury, 2015). 34 John Henry Newman suggests that Christ was in fact given a taste of the pain he would endure during the cry of dereliction, and it was this taste that prompted his Gethsemane agonies. 35 In Christ’s case, it seems God’s antecedent will was that he maintain beatific revelation, but God’s consequent will was that he subject himself to crucifixion. We can understand Christ as willing God’s antecedent will, but agreeing to God’s consequent will (“not my will, but thy will be done”). It is difficult to see how we might draw a similar antecedent/consequent parallel with the primal sin. 36 In this way, then, Christ echoes Paul’s sentiments in Romans 9:3: For I could wish that I myself were accursed and cut off from Christ for the sake of my own people, my kindred according to the flesh. 37 Swinburne writes: God Incarnate could have chosen at a time to allow himself to make his choice at that time under the influence of temptation to do less than the best. He would then have needed to fight against the temptation not to do that best action; and it would have been possible that he would yield to that temptation and done instead a less good action (and perhaps even a bad action, though certainly not a wrong action). (Swinburne, Was Jesus God, 46) 38 Although perhaps Paul comes close in Romans 9:3: For I could wish that I myself were accursed and cut off from Christ for the sake of my brothers, my kinsmen according to the flesh. 39 After all, Christ stated his will was not to go through with this plan.

Bibliography Adams, Marilyn McCord. “St. Anselm on Evil: De casu Diaboli.” Documenti e studi sulla tradizione filosofica medievale 3 (1992): 429–30. Babcock, William. “Augustine on Sin and Moral Agency.” The Journal of Religious Ethics 16 (1988): 46. Gaine, Simon. Did the Savior see the Father? Christ, Salvation, and the Vision of God. London: Bloomsbury, 2015. Hick, John. Evil and the God of Love. New York: Harper and Row Publishers, 1966. Hoffmann, Tobias. Free Will and the Rebel Angels in Medieval Philosophy. Cambridge: Cambridge University Press, 2021. Hoffmann, Tobias. “Theories of Angelic Sin from Aquinas to Ockham.” A Companion to Angels in Medieval Philosophy. Leiden and Boston: Brill, pp. 283–316, 2012. King, Peter. “Angelic Sin in Augustine and Anselm.” in Tobias Hoffmann (ed.), A Companion to Angels and Medieval Philosophy. Leiden and Boston: Brill, pp. 261–82, 2012.

202  David Worsley King, Peter. “Scotus’s Rejection of Anselm: The Two-Wills Theory.” in Ludger Honnefelder et al (eds.), John Duns Scotus 1308–2008: Investigations into His Philosophy. Munster: Aschendorff, pp. 359–78, 2011. Leftow, Brian. “Tempting God.” Faith and Philosophy 31(1) (2014): 3–23. MacDonald, Scott. “Petit Larceny, the Beginning of all Sin: Augustine’s Theft of the Pears.” Faith and Philosophy 20(4) (2003): 393–414. MacDonald, Scott. “Primal Sin.” in Gareth Matthews (ed.), The Augustinian Tradition. Berkeley: University of California Press, pp. 110–39, 1998. Pawl, Timothy. “Conciliar Christology, Impeccability, and Temptation.” in Johannes Grössl & Klaus von Stosch (eds.), Impeccability and Temptation: Understanding Christ’s Divine and Human Will. Oxford: Routledge, pp. 94–116, 2021. Pawl, Timothy. In Defence of Extended Conciliar Christology. Oxford: Oxford University Press, 2019. Pelser, Adam. “Temptation, Virtue, and the Character of Christ.” Faith and Philosophy 36 (2019): 81–101. Pini, Giorgio. “What Lucifer Wanted: Anselm, Aquinas, and Scotus on the Object of the First Evil Choice.” in Robert Pasnau (ed.), Oxford Studies in Medieval Philosophy (volume 1). Oxford: Oxford University Press, 61–82, 2013. Rogers, Katherin. “Anselm on Eudaemonism and the Hierarchical Structure of Moral Choice.” Religious Studies 41(3) (2005): 249–68. Rogers, Katherin. Anselm on Freedom. Oxford: Oxford University Press, 2008. Stump, Eleonore. Atonement. Oxford: Oxford University Press, 2018. Swinburne, Richard. Providence and the Problem of Evil. Oxford: Oxford University Press, 1998. Swinburne, Richard. Was Jesus God? Oxford: Oxford University Press, 2008. Timpe, Kevin. “The Arbitrariness of the Primal Sin.” Oxford Studies in Philosophy of Religion 5 (2014): 243–58. Van Dyke, Christina. “Aquinas’s Shiny Happy People: Perfect Happiness and the Limits of Human Nature.” Oxford Studies in the Philosophy of Religion 6 (2014): 269–91. Williams, Thomas. Anselm: Three Philosophical Dialogues. Indianapolis: Hackett Publishing Company, 2002. Wood, William. “Anselm of Canterbury on the Fall of the Devil.” Religious Studies 52 (2016): 223–45. Worsley, David. “Could there be Suffering in Paradise?” Journal of Analytic Theology 4 (2016): 87–105.

12 The Impeccable Freedom of Christ Johannes Grössl

Introduction There is an apparent contradiction in one of the core teachings of the Christian tradition. Jesus Christ is said to be truly human and truly divine, although certain attributes of divinity and humanity seem to be incompatible. Some aspects of this contradiction have been intensively discussed within analytic theology, such as timelessness and temporality or omnipotence and creaturely limitation. Notably, little research has been done on the consequences of a libertarian concept of free will on Christology. Most Christians believe – in accordance with the Council of Chalcedon and the Third Council of Constantinople – that Jesus Christ was truly human, yet unable to sin. If, however, as libertarians defend, being human implies the power to choose between morally significant options, it seems that Christ cannot be both truly human and impeccable. Since God is (in most versions of theism) characterized as morally perfect and essentially impeccable, it can thus be concluded that Christ cannot be truly human and truly divine. This explains why insights in Christology offer a strong motivation for traditional Christian theologians to favor compatibilism. Alternatively, they regard the doctrine of the Incarnation as simply mysterious, as Timothy O’Connor observes: Since nearly all Christians accept that Jesus Christ was (and is) incapable of sin, a crucial limit on the scope of his freedom appears to be built into the doctrine. But one ordinarily supposes that he freely chose when and how he graciously dispensed mercy during his earthly ministry. And he is depicted as going to the cross freely, persevering through much anguish in prayer as he anticipates vividly his imminent arrest. Again, the relationship of divine and human natures in the Incarnate Son on orthodox understanding is shrouded in mystery.1 In this chapter, I will present and contrast two viable libertarian solutions to the impeccability paradox: On the one side, the popular supererogation view says that although morally significant libertarian free will (or at least the potentiality for developing it) is essential to human nature, the power to choose DOI: 10.4324/9781003306191-15

204  Johannes Grössl between good and evil is not; in some cases, a human being can only possess the power to choose between performing or omitting a supererogatory free act. Christ’s free decision to take up the cross and not run away in Gethsemane can count as such a decision. On the other side, the dispositional view says that although the potentiality for developing a power to choose between morally significant options (possibly even between good and evil) is essential to human nature, certain external circumstances can prevent this ability to be actualized. Accordingly, Christ only needs to possess the counterfactual power not to take up the cross in order to count as truly human. As a consequence, the accountability of Christ’s actions must be morally traced back to the libertarian free decisions of those who affected his personal character development and even the culture he was brought up in. The Supererogation Solution Given the premise that libertarian free will is essential to human nature, it is challenging to give a coherent account of Chalcedonian Christology. Presuming that libertarian free will can be explicated as the power of an agent to choose between morally significant options, a paradox of Christ’s freedom can be formulated as such2: 1 If x is a person with a truly human nature, x sometimes possesses the power to choose between morally significant options. 2 If x is a person with a truly divine nature, x never possesses the power to choose between morally significant options. 3 There is no person with both truly human and truly divine nature. Chalcedonian Christology, however, teaches that there was or is at least one person with a truly human and a truly divine nature: Jesus Christ. If the Christian wants to uphold libertarianism and avoid affirming a logical paradox, she needs to solve the impeccability problem. Richard Swinburne has proposed a prominent solution to this paradox. According to him, morally significant libertarian free will does not necessarily include a power to choose between good and evil; it may only be actualized as a power to choose between good and even better options. If the ethical theory of supererogation is valid, there are praiseworthy actions whose omission is not blameworthy.3 If one applies this conception of supererogatory libertarian free will to Christ, one can deny his moral perfection without denying his impeccability: (S) Christ possessed free will, but no power to choose evil – only to choose less than the best. Swinburne introduces this concept when he discusses Christ’s temptations: The other temptations were temptations to do less than the best: if Jesus .  .  . had not accepted crucifixion, no one would have been

The Impeccable Freedom of Christ 205 wronged. Hence, . . . God Incarnate could have yielded to a temptation not to do the best; and that is why his doing the best in these situations (if he did) is the work of supererogation which made available an atoning sacrifice for our salvation.4 He thus interprets Christ’s temptation in Gethsemane not as a temptation to do evil, but merely as a temptation to do “other than the best (though not to do wrong).”5 The supererogation view thus rejects premise 2 in the former argument, claiming that a divine person can choose between morally significant options, namely, between omitting and performing a supererogatory good deed. Not only, but especially open theists argue that God’s act of creating the world was a free act in a libertarian sense: God could have not created the world. If God had not created the world, he would only have omitted a supererogatory good deed, but not omitted a moral obligation. Thus, God can be impeccable but still be free in a libertarian sense to create or not create.6 No other kind of freedom is applied to God Incarnate in the supererogation view. Swinburne’s proposal suggests that the power to choose between morally significant options is an essential attribute of both human and divine nature, letting them be compatible at least regarding the issue of free will. If Christ has the power to choose to run away in Gethsemane, or previously form his character in a way that would lead to him running away in Gethsemane, he would not have done anything wrong, but only neglected to do a possible supererogatory good act: sacrificing his life to save humanity. There are two arguments against this solution: First, it seems to be always morally wrong to reject something justifiably perceived to be the will of God. Critics of supererogatory solutions often refer to James 4:17 which says that “anyone who knows the right thing to do and fails to do it, commits sin.”7 Second, if only supererogatory free will, but not the power to do evil, is essential to human nature, the question arises why God did give humans the power to do evil in the first place. It seems that limiting human nature to supererogatory freedom has great, possibly devastating implications for the problem of theodicy. I think the first counterargument can easily be refuted: God might request from us certain deeds which he simply does not command us to do. In these cases, there is no right or wrong option, but only one better and one worse.8 The second counterargument is harder to challenge: Free will theodicies presume that God could not have created free human beings without giving them the power to do evil.9 Thus, such theodicies require a full-fledged libertarian anthropology: that being human in the fullest sense includes the power to choose between good and evil, and not merely between morally significant options. In the following, I  will present an alternative view to reconcile Christ’s freedom and impeccability, which, in another step, can be used to further develop the supererogation view to avoid the second critique by defending (libertarian) free will as a disposition that is sometimes actualized as a power to do evil and sometimes actualized as mere supererogatory freedom.

206  Johannes Grössl Alternative: Free Will as a Disposition Even if freedom is counted as an essential attribute to humanity, what kind of freedom is essential to being human is still open for discussion. Compatibilism allows that a person can be free without possessing the power to choose between opposites; sometimes scholars propose a so-called compatibilist account of freedom. Even libertarians must include certain compatibilist intuitions into their theory if they are aware that not every human being is free at every time of her existence. When discussing the essential attributes of human nature, it is reasonable to distinguish between the potentiality to develop a power, the capacity to exercise a power, and the actual ability to exercise a power.10 As a simple example, we can think of swimming: most human beings, even if they cannot swim, possess the potentiality to learn how to swim (to develop the capacity to swim). Even if a person can (has the capacity to) swim, there might be circumstances (lack of water, being unconscious, a millstone connected to one’s legs, etc.) that prevent her to exercise this capacity.11 If there are no such “masking” or “finkish”12 circumstances, we can speak of an actual ability (in the narrowest sense). Applying this concept to free will, we can think of a 2-year-old child, who can only be regarded as a human being if being human does not require possessing morally significant free will. Most people would, nevertheless, in accordance with a Christian or Kantian anthropology, attribute being human also to children and comatose people, since they all have the potential to be or become moral agents (again). Moreover, most deontologists would attribute being human even to strongly disabled or mentally impaired adults because they might once have had the potential to become moral agents. Accordingly, “dispositional compatibilists” can claim that a person is free when she has the power to do otherwise now or in some possible future or if a counterfactual analysis shows that she could have done otherwise in certain alternative timelines. Here, too, several variants can be discussed and applied to Christology. To actualize a capacity as an ability, a person needs to find herself in a situation in which it can be exercised. If one has the capacity to swim, but never experiences being located in water, one cannot actualize the capacity as an ability. This implies that one is contingently unable to swim. If Christ has morally significant freedom as a power to choose between obligation and inclination, but never finds himself in a situation in which an inclination is strong enough to motivate him to an immoral decision, he could be counted as free but yet contingently unable to do evil.13 Nevertheless, a counterfactual analysis would show that – if he had found himself in different circumstances – he would be in a situation in which making a morally significant decision is not only possible but unavoidable. Thus, proposing that Christ had the capacity to do evil does not require that he ever had the power to bring about a situation in which he has the ability to do evil. (D1) Christ had the capacity to do evil; but there were no possible circumstances available to him in which this capacity becomes an ability.

The Impeccable Freedom of Christ 207 As a consequence of Christ essentially possessing a divine nature (and also against adoptionist Christologies), it is plausible to say that he never in his life had the power to bring himself into a situation in which he has the actual power to do evil.14 But what prevented him to make such a choice? Any intervention from the outside would entail an infringement not only on Christ’s freedom but also on the freedom of all agents surrounding him.15 In order to avoid the question what prevented Christ from actualizing his capacity to sin, one can go further and postulate that he never even developed the capacity to sin in the first place, thus never became a genuine moral agent. He could nevertheless count as being fully human since he could have developed this capacity if grown up in different circumstances. (D2) In Christ was a potential to develop a capacity to do evil; but there were no possible circumstances available to him in which he could actually develop this capacity. Some classical Thomists defend this view by arguing that libertarian freedom is only a “mode of exercising” freedom.16 The situation here is structurally similar (although reflecting the other extreme) to that of a child who has been indoctrinated all of its life and trained to be a child soldier: it is not a moral agent, but – under better circumstance – could have become one. The same holds for disabilities that result from, for example, lack of oxygen at birth. While some humans are, because of unactualized potential, not able to choose the good, a contingently impeccable Christ is contingently not able to choose between good and evil. The idea that Christ is qua human nature in principle able to do evil, but because of the circumstances he lives in never comes into a situation where he can do evil is connected to certain assumptions on the nature of freedom. According to “restrictive libertarianism,” we are only free in some occasions in our lives, and the scope of our freedom is severely limited.17 Usually, we act on our habits and inclinations; only sometimes, we make a rational choice on what option to choose, dependent on our reasons. Usually, we act according to the best reasons available to us; only sometimes, we find ourselves in situations in which we are confronted with incommensurable reasons or possible actions of equal value. Often, we fail to do what we want to do because of lack of willpower; only sometimes, we have the choice to overcome our weakness of will. The more one limits the basic conditions for free decisions, the easier it is to make a case for the existence of a person who never finds himself in a situation to make a free decision between good and evil. The dispositional view on free will in general and applied on Christ’s freedom can be seen as a compromise between compatibilism and libertarianism. Freedom is regarded as a disposition that is on most cases actualized as libertarian free will but in very rare cases as something we can label “compatibilist freedom” with a mere counterfactual power to do otherwise. If we do not want the Logos to put external restrictions on the assumed human nature’s freedom, we need to defend a Christology in which the Logos can

208  Johannes Grössl only assume a human nature whose free will can only be actualized in the compatibilist sense, which again implies that God needs to choose the cultural and individual circumstances for an incarnation very carefully. Or put another way: it is logically impossible for God to assume any human nature or be incarnated as the son or daughter of any arbitrary woman in history. Combining the Supererogation View and the Dispositional View As all solutions to the impeccability paradox, both presented views have their advantages and disadvantages. I must admit that no possible solution to the problem has thoroughly convinced me so far. However, I think the two presented views can be combined to present a new, more convincing solution with more strengths and less weaknesses. (SD) Christ had the capacity to choose between morally significant alternatives, which was contingently actualized as supererogatory freedom; under counterfactual circumstances not available to him, the same capacity would have been actualized as a power to do evil. This theory is based on an anthropology which says that the power to do otherwise is essential to human nature, even if this power develops over time and is not available at all moments of a person’s life: one could call this a dispositional libertarian anthropology. Such an anthropology is compatible with the idea of irreversible character formation, which precludes libertarian free will in an eschatological state. It relies on the hypothesis that there is a disposition in all human beings to bring about a power to do otherwise; whether, when, and what kind of power to do otherwise is actualized, however, depends on the circumstances in which a certain person lives. Now there are three options: either Christ’s freedom remained an unactualized disposition (as the mere dispositional view says), or Christ’s freedom was actualized as supererogatory freedom, or Christ’s freedom was actualized as a power to choose between good and evil. Traditional Christians who affirm the doctrine of impeccability can defend either the first or the second option. The second one might maintain Christ’s freedom to run away in Gethsemane, which (from a libertarian perspective) is a prerequisite for constituting human merit to salvation and thus (at least from an Anselmian perspective) a general prerequisite for salvation of humanity. The third option is, as discussed earlier, incompatible with Christ essentially possessing a divine nature.18 It is important to note that in a dispositional view, Christ’s freedom is formally the same as the freedom of other human beings. The difference lies only in its actualization due to circumstances. This, however, puts a lot of emphasis on the circumstances Christ grew up in. If Christ had had severely morally flawed parents, for example, it would have been very likely for certain vices to be adopted by him. Retrospectively, we can postulate that Christ probably had ideal circumstances that lead him never to come into situations where he had to choose

The Impeccable Freedom of Christ 209 between good and evil, only into situations where he had to choose to perform or not to perform supererogatory good acts. He might have had sufficient knowledge of the good that there never was sufficient reason for him to do evil. He might have had certain genetic dispositions combined with an upbringing and education which instilled into him sufficiently strong good habits so that he had very strong inclinations to do the good, which would have needed very strong opposing reasons to give up, which again were not available to Christ. To make the case even more complicated, we need to evaluate history not only retrospectively but also counterfactually. When God decides to become incarnate in the womb of a specific person, he needs to guarantee that, from this point onward, every possible future history is a history in which the incarnated person would not actualize his freedom as a power to sin. Here, the free will of all persons directly and indirectly interacting with this person must be taken into account and be sufficiently respected. John Sanders notes: In bringing the Messiah into the world, God seeks the cooperation of people of faith, but he does not draw back from situations lacking public respectability. Mary and Joseph are placed in a predicament in which they may either reject God’s desires and keep their public respectability or trust God and subject themselves to possible lifelong disfavor. They place their confidence in God and consent to the risk this entails. God places his trust in them, giving his consent to the risks involved. The incarnation does not come about through sheer overwhelming power but through the vulnerability of being genuinely dependent on some Jewish peasants.19 An omniscient God – even if he cannot know future contingents – has knowledge of all possible future histories; thus, He knows all possible developments of all possible persons God Incarnate could ever interact with; God furthermore knows the extent to which others would influence the Incarnation in certain circumstances. To maintain the hypostatic union, God needs to guarantee that even in the worst possible development, the Incarnation would never actualize her disposition to choose freely in a situation where a choice between good and evil presents itself.20 Still, in order to maintain human freedom and the possibility of merit, it can also be guaranteed that the Incarnation would actualize her disposition to choose freely in a situation where a choice between good and better (or neutral and good) presents itself. Conclusion It is not trivial to defend both a libertarian anthropology and the classical two-nature view in Christology. It seems that in order to reconcile both theories, one needs to reevaluate libertarianism, which is possible without entirely giving up the power to do otherwise as a general prerequisite to freedom. The supererogation view regards the power to choose between morally significant

210  Johannes Grössl options as essential to free will, but not the power to act immorally. In the dispositional view, it is sufficient for attributing genuine freedom that a person could have developed a power to choose between good and evil in certain counterfactual circumstances – even if there is, at a certain point in time, no possible future in which such circumstances may occur. The combined dispositional supererogation view postulates that there is a certain disposition in all humans which, depending on the circumstances one lives in, is not at all actualized, actualized as supererogatory freedom, or actualized as freedom to choose between good and evil. This typology certainly opens up new questions, which should be addressed in further research: (1) In this chapter, impeccability is understood as not being able to act immorally; however, sin and immoral action are not identical. If it were possible for an agent to be impeccable (e.g., understood as being in an unbreakable loving relationship with God) but still possess the power to act immorally, the impeccability paradox could be solved in a different way.21 (2) Catholics may be tempted to attribute supererogatory freedom not to Christ, but to his mother Mary.22 Possibly, only human nature which was freed from original sin contains a disposition to be actualized as pure supererogatory freedom. (3) It is hard to explain how God can – without intervening in history and undermining the freedom of other human persons – prevent Jesus to grow up in circumstances, which would make his freedom to develop into a power to sin. In order to avoid strong interventionism, the dispositional view thus requires a theory of providence or guidance as defended in classical Antiochene Christological views, according to which Christ’s human nature became impeccable not by its unity to the Logos, but through guardiancy and spiritual transformation by the Holy Spirit.23 If Christ is truly human (Council of Chalcedon), possesses a truly human will (Third Council of Constantinople), and was truly tempted in every respect we are (Hebr. 4:15), it is difficult not to attribute some type of freedom to him as well. If freedom is not simply equated with a power to do otherwise (“libertarian free will”) but stands in a certain clearly defined relation to such a power, one can uphold substantial libertarian intuitions while also coherently defending the existence of a human being who possesses “impeccable freedom.” Notes 1 Timothy O’Connor, “Against Theological Determinism,” in Kevin Timpe & Daniel Speak (eds.), Free Will and Theism. Connections, Contingencies, and Concerns (Oxford: Oxford University Press, 2016), 132–41, here 140. Likewise, in this volume, Jesse Couenhoven argues that the impeccability paradox is a strong motivation for a Christian theologian to favor compatibilism.

The Impeccable Freedom of Christ 211 2 Cf. Johannes Grössl, “Introduction,” in Johannes Grössl  & Klaus von Stosch (eds.), Impeccability and Temptation. Understanding Christ’s Divine and Human Will (New York/London: Routledge, 2021), 1–11, here 2. 3 According to Christian tradition, the support given by the merciful Samaritan exceeds the moral obligation of immediate assistance. Later in Church history, many good acts such as martyrdom were considered to be praiseworthy but not obligatory. Cf. David Heyd, Supererogation. Its Status in Ethical Theory (Cambridge: Cambridge University Press, 1982), 15; 20–26. 4 Richard Swinburne, The Resurrection of God Incarnate (New York: Oxford University Press, 2003), 11. 5 Cf. Swinburne, God Incarnate, 99: “We might expect [an incarnate God] to live under conditions of ignorance except of those things he needs to reveal; of unawareness of his extra-human powers except when these are required to fulfill the reasons for his Incarnation; and of the possibility of his yielding to a temptation to do other than the best (though not to do wrong).” Cf. Brian Hebblethwaite, Philosophical Theology and Christian Doctrine (Hoboken, NJ: Wiley, 2005), 69: “[Swinburne] suggests that God incarnate, being who he was, could do no wrong, but that he was genuinely free to act in a variety of good ways, not only in the sense of duties but also in the sense of supererogatory acts (i.e. beyond the call of duty). . . . On Swinburne’s view, Christ’s temptations could only have been temptations to take some other good path than that of the supererogatory call to tread the way of the cross.” 6 Cf. Armin Kreiner, Das wahre Antlitz Gottes. Oder was wir meinen, wenn wir Gott sagen (Freiburg: Herder, 2006), 447. 7 In an e-mail following a discussion on supererogation, John McKinley wrote to me in 2019: “As I remember from our conversation in May at Fuller Seminary, you had said that Jesus’ freedom was in being able to choose such acts that were not commands from God to be obeyed. He could deliberate, choose not to do them, and this would not be sin. What I wonder is if the category of supererogatory acts does not fit with James 4:17, which seems to say that if we know any good that is within our reach to do (along perhaps with the sense that God wants us to do it) and we shrink back from it, ‘to him it is sin.’ I am probably misunderstanding the category of supererogatory acts. If applied to Jesus, then we would have to say that all the goods he did were obligatory and not to do them would be rebellion against the Father’s will (or, the Spirit’s leading).” 8 Cf. Johannes Grössl, “Die Sündlosigkeit Jesu als logische Herausforderung für die christliche Zweinaturenlehre,” in Thomas Schärtl  & Thomas Marschler (eds.), Herausforderungen und Modifikationen des klassischen Theismus 2: Inkarnation (Münster: Aschendorff, 2020), 163–94, here 181–4. 9 Cf. O’Connor, “Against Theological Determinism,” 135–7. 10 I draw the terminological distinction between capacity (which might only be counterfactually exercised) and ability (which can be exercised) from Randolph Clarke; cf. Randolph Clarke, “Dispositions, Abilities to Act, and Free Will. The New Dispositionalism,” Mind 118 (2009): 323–1, 338. Alfred Mele uses different terminology, distinguishing between general and specific practical ability; cf. Alfred Mele, “Agents’ Abilities,” Noûs 37:3 (2003): 447–70, 447f. 11 Cf. Michael McKenna, “Reason-Responsiveness, Agents, and Mechanism,” in David Shoemaker (ed.), Oxford Studies in Agency and Responsibility (volume 1) (New York: Oxford University Press, 2013), 151–84, here 166: “A disposition is masked when its manifestation is concealed in some way. A piece of salt placed in water remains soluble even if, when placed in water, it does not dissolve because it is encased in wax. A disposition is finked when, in just those conditions that would otherwise trigger manifestation, it is altered so as not to have the

212  Johannes Grössl disposition. A glass vase sitting on a shelf undisturbed possesses the disposition of fragility even if, were it knocked over, a wizard would turn it to stone before striking the ground, rendering it not fragile.” 12 Cf. Clarke, “Dispositions,” 327f.: “. . . that abilities to act, like dispositions, can be finkish. A fink can remove an ability in just those circumstances in which it would, if retained, be exercised, or create an ability to act only when such circumstances obtain. For example, a sorcerer might stand ready to remove someone’s ability to speak French, should that person try to do so, or to restore a paralytic’s ability to raise her arm should she try to raise it.” 13 It is disputable whether Jesus was truly tempted if he never felt inclinations strong enough to make him act differently than he does. Swinburne, for example, assumes this to be a prerequisite for temptation. Cf. Hebblethwaite, Philosophical Theology and Christian Doctrine, 69: “[Swinburne] goes on to suggest that felt temptation to do wrong would be illusory temptation. Real temptation must involve desire to take another path.” Hebblethwaite rejects this assumption. 14 Other possible solutions to the paradox, which are not further investigated in this chapter, involve presenting Christ’s divine nature as somehow contingent and logically dependent on Christ’s free decisions, either using Ockhamist or Molinist approaches. Cf. Johannes Grössl, “Christ’s Impeccability,” in James M. Arcadi & James T. Turner, Jr. (eds.), T&T Clark Companion to Analytic Theology (London/New York: Bloomsbury, 2021), 215–9. Here, it is presumed with the Neo-Chalcedonese teaching of an-/enhypostasis that Christ’s identity is essentially connected to his divine nature. 15 Furthermore, it seems plausible that controlling circumstances in order to preclude opportunities for immoral action does not constitute genuine virtue. Cf. Kevin Timpe & Timothy Pawl, “Incompatibilism, Sin, and Free Will in Heaven,” Faith and Philosophy 26:4 (2009): 398–419, 403. 16 Cf. Jean-Miguel Garrigues, La conscience de soi telle qu’elle était exercée par le Fils de Dieu fait homme, 40; cited in Thomas J. White, The Incarnate Lord. A  Thomistic Study in Christology (Washington, DC: CUA Press, 2015), 254: “The human will of Christ itself, while endowed naturally with the same free-will as us, nevertheless does not have an autonomous deliberation (gnome) characteristic of the mode of exercise found in created persons.” 17 For an overview of different versions of restrictivism as defended by Robert Kane, Peter van Inwagen, Richard Swinburne, and others, see László Bernáth, “Why Libet-Style Experiments Cannot Refute All Forms of Libertarianism,” in Bernard Feltz, Marcus Missal, & Andrew C. Sims (eds.), Free Will, Causality, and Neuroscience (Leiden: Brill, 2019), 97–119, here 112–6. 18 There are possible libertarian theories in which Christ is able to sin, but his divinity is contingent upon him freely not sinning. However, these require an Ockhamist view of divine foreknowledge, which is often said to not be compatible with libertarianism since it requires a B-theory of time; cf. Johannes Grössl, “In allem wie wir in Versuchung geführt.” Theologische Modelle zum Verhältnis von göttlichem und menschlichem Willen in Christus (Freiburg: Herder, 2021), 524–41. 19 John Sanders, The God Who Risks. A Theology of Divine Providence. Revised Edition (Grove, IL: Intervarsity Press, 2007), 95. 20 One can see that, in this view, there can be only very rare cultural circumstances in the history of humanity in which an incarnation is possible. But this limitation is indeed an explanatory advantage for the theory: it explains why the incarnation did not happen earlier or at a different point in history. 21 Cf. Grössl, In allem wie wir in Versuchung geführt, 310–35; Jeffrey Siker, Sin and Perfection in Early Christianity (Cambridge: Cambridge University Press, 2015), 20f.: “Sin is . . . also the violation of one’s relationship with God or with people.”

The Impeccable Freedom of Christ 213 22 Cf. Troy Stefano, “Catholica Mater. The Marian Insights of Henri de Lubac,” in John Cavadini & Danielle Peters (eds.), Mary on the Eve of the Second Vatican Council (Notre Dame, IN: University of Notre Dame Press, 2017), 187f.: “Mary as Immaculate Spouse entails personhood and free cooperation with grace. God’s spousal love is not mechanical, nor is it impersonal. God’s love gives birth to persons, and persons are the recipients of God’s love. Only persons can be free, and only persons can be called to a spousal relation with the eternal person of the Word. Mary’s ‘Yes’ to God initiates, from the human side, a history of receiving and living God’s freedom. . . . In the logic of the Incarnation, God did not ‘forcefully’ initiate a history of freedom. De Lubac cites Newman in noting that the early Fathers ‘do not speak of the Blessed Virgin merely as the physical instrument of our Lord’s taking flesh, but as an intelligent, responsible cause of it.’ ” 23 McKinley calls this model “sinless by empowering grace”; cf. John McKinley, Tempted for us? Theological Models and the Practical Relevance of Christ’s Impeccability and Temptation (Milton Keynes: Paternoster, 2009), 140: “Christ’s weak humanity requires the aid of divine grace by the Holy Spirit.” This model was prominently defended in Theodore of Mopsuestia’s Spirit Christology; cf. Cornelia Dockter, “God’s work and human’s contribution,” in Johannes Grössl & Klaus von Stosch (eds.), Impeccability and Temptation. Understanding Christ’s Divine and Human Will (New York/London: Routledge, 2021), 75–93, here 81: “[T]he human being Jesus had to face the temptations of evil and actually did have the ability to sin. Nevertheless, it was not possible for him to succumb to the temptation of sin, because of God’s merciful preservation. In contrast to the normal relationship between God and his creation, in Jesus’ life the gift of grace is a permanent one.”

Index

agency: Christ 34–5, 42, 55–6; divine 9, 28, 31, 44, 104–6; group 17, 124–8, 131–2; individual 2, 4, 14, 18, 27, 37–40, 79, 116–17, 136, 140–1, 176, 188 analytic theology 1, 49, 60, 180, 203 antirealism 178 Aquinas, Thomas 13, 105, 110, 151, 176, 190, 192–3, 198–9 Augustine 37, 159–60, 200 Baker, Lynne Rudder 27–8, 38, 43, 47, 58, 63, 70, 159–61 banezianism 17, 104–5, 108, 110 blame 2–3, 5, 7–8, 11, 14, 28, 30–1, 38, 42, 68, 76–8, 84, 140, 188, 191, 204 causation 13, 48, 57, 139; agent 13, 48, 139; divine 10, 17, 104–5, 107–11 choice 3, 6–7, 11, 13, 18–19, 28–36, 39, 60, 75, 82, 96, 106, 109, 116–17, 122–3, 130–2, 136, 137–8, 140–5, 152–3, 159–60, 168, 171–3, 180, 188–9, 193–5, 198, 201, 207, 209 Christ 10, 15, 19, 20, 33–5, 42, 55, 115, 118–25, 128–30, 133, 155–7, 161, 165–6, 178–9, 187, 194–7, 203–5, 207–9 Christology, conciliar 56, 62–5, 69–70 cognitive bias 96, 136, 142–6 compatibilism 7–8, 11–12, 14, 16, 20, 27, 29, 30, 32–3, 35–6, 37–8, 42, 47–52, 58–9, 63, 66, 77, 83, 105, 147, 203, 207, 210 concurrence 106–8 consciousness 75, 80, 138–9, 179

control 2–3, 5–7, 10–12, 14, 18, 28–31, 44, 48, 52, 57, 75, 80, 82, 104–6, 108, 117, 129–30, 136–45, 152–3, 155, 159–62, 172–3, 179, 212 counterfactuals 13, 35, 91–2, 104–6, 109, 204, 206–11 decision 5, 39, 41, 71, 107–10, 115–17, 121, 124–5, 129–30, 132, 138–43, 145, 147, 162, 176, 193–4, 196, 204, 206–7 deliberation 125, 127, 212 determinism: causal 4–6, 16, 18, 30, 31, 38, 50, 59, 75, 136, 138, 154; divine/ theological 2, 8–10, 12–14, 18–19, 27, 30, 36, 44, 47, 49, 55, 57–8, 68, 70, 81, 106, 109, 136, 161, 171–2, 178, 199 divine: agency (see agency, divine); foreknowledge 13, 34, 88, 92–3, 97, 100, 172, 174, 212; freedom 28–31, 52–3, 67; goodness 9, 28–9, 38–41, 57–9, 155–7, 162–3, 194, 196, 198, 200 dyothelitism 62, 70 Erasmus of Rotterdam 152, 159, 163, 172, 181 evil 3, 9, 11–12, 16, 20, 28, 32, 38–41, 48–9, 57–8, 120–1, 157, 169, 173–4, 177, 188–90, 195, 204–10 foreknowledge see divine, foreknowledge Formula of Concord 156, 171, 174–5, 180 freedom: heavenly freedom 16, 27, 31–4, 53, 191, 197; incarnate (see incarnation); see also

Index  215 compatibilism; libertarianism; incompatibilism free will, divine see divine, freedom free will defense 12, 38–41, 49, 59 free will revisionism 6, 8, 14–15, 61, 148 free will skepticism 7, 11, 14, 16, 18, 48, 75–8, 84

predestination 1, 30, 36, 41 primal sin see sin, primal Providence 10, 13, 49, 76, 107, 110, 210 punishment 2, 7, 10, 12, 16, 18, 21, 41, 76, 80, 82, 136, 147, 152–69 purgatory 32–3, 37, 41, 53–4

grace 1, 9, 16–17, 19, 27, 29–30, 32–3, 36–7, 39, 41, 54, 56–7, 104–5, 153, 159–61, 171, 173–4, 176, 181, 190, 213

rationality 126, 142–3, 163, 194 responsibility, moral 1–3, 6–8, 11–12, 14–15, 30, 33, 38, 39, 41, 48, 58, 68, 70, 76–7, 80–1, 104, 116–17, 119–20, 122, 129–30, 136, 140, 147–8, 153, 159–60, 171, 187–8 retributivism 7–8, 80, 136, 147

hell 28, 38, 41, 57–8, 64–5, 68, 71–2, 157–9, 167, 173, 198 heuristics 144 impeccability 20, 27, 30, 34, 191, 196–7, 203–4, 207–8, 210 incarnation 34–5, 55–6, 61, 122, 179, 203, 208–9, 211–13 incompatibilism 3, 5, 7, 13–14, 16, 18, 20, 28, 30, 42, 43, 47, 50, 52, 66–73, 84, 92, 104, 106, 136, 138, 141, 147, 155, 158–61, 168–9, 203, 208 libertarianism 7–8, 11–12, 16, 21, 27–9, 32–3, 35–42, 47–73, 204, 207, 209; virtue libertarianism 13, 29–39, 42 love 9, 11, 13, 18, 30–1, 39–42, 121–2, 152–8, 162–9, 171, 173, 176, 177, 192–6, 200, 213 Luther, Martin 14–15, 18–19, 82, 152, 155–81 Manicheanism 175, 178 Melanchthon, Philip 10, 173–5, 181 Molinism 13, 34, 91, 105, 107–10 monothelitism 62, 70 Nicene Creed 51, 61

sin, original 9, 11, 36, 147, 172–3, 194, 210 sin, primal 19, 187, 189, 194–202 sourcehood 3–6, 29–31, 48, 115, 117, 119, 122–3, 127–33 Stump, Eleonore 13, 19, 28, 32, 36–7, 109, 138–9, 193, 195, 201 Temptation, of Christ 15, 19, 48, 187, 189, 194–8, 200–1, 204–5, 211–13 theodicy 12, 28, 38–41, 44, 59, 172, 188, 205 theological pessimism 171, 177–8 Timpe, Kevin 3, 12–14, 16, 28–9, 31, 36, 38, 45, 109, 116–18, 153, 159, 189, 199, 212 universalism (universal salvation) 19, 40, 45, 58–9, 64–5, 71, 153, 155, 157, 169, 170, 173

Open theism 13, 17, 87–100, 173, 205

vice 14, 163, 208 virtue 14, 29, 32, 39, 54–5, 60, 70, 105, 167, 212

Pelagianism 32, 57, 63, 159, 171, 175, 178 Plantinga, Alvin 12, 44, 46, 49, 104

Wood, William 60–1, 69, 187, 190, 198