Leibniz on Contingency and : A Friendly Account

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of in the Graduate School of The Ohio State University

By Juan Garcia, MA Graduate Program in Philosophy

The Ohio State University 2018

Dissertation Committee: Tamar Rudavsky, Advisor Julia Jorati, Advisor Lisa Downing William Taschek

© 2018 by Juan Garcia All reserved.

Abstract:

Some philosophical , like metals, reveal their most interesting qualities when stretched. As I see it, the intelligibility and plausibility of the of contingency, as a condition for freedom, is stretched significantly by three of Leibniz’s major philosophical commitments: i) a robust version of the of sufficient reason; ii) strong theological commitments like divine foreknowledge and robust providence; and iii)

Leibniz’s distinctive thesis of complete – the doctrine that every substance has an individual concept which includes predicates denoting everything that ever happen to it. Many interpreters think that these Leibnizian commitments stretch contingency to the breaking point: Leibniz cannot retain a genuine type of contingency. In this dissertation, I argue that these interpreters are mistaken and that Leibniz can accommodate a plausible type of contingency. Two important steps in reaching this conclusion are the following. First, I identify a kind of explanation of rational according to which the explanans does not necessitate the explanandum. Second, I develop an account that makes sense of individual concepts including contingent as contingent. This account both sheds new light on some of Leibniz’s central views on contingency and freedom, and expands our substantive understanding of contingency as a condition for freedom.

The first major goal of my dissertation is to show that Leibniz has remarkable affinities with the

Molinist tradition – something that is commonly overlooked and misunderstood in the secondary literature. I identify two substantive tenets of Molinism and argue that Leibniz endorses a version of each; in , he utilizes them for the same theoretical purposes as Molinists. These two substantive Molinist tenets are: 1) free actions follow contingently from their sufficient conditions; and 2) ’s of what creatures would freely do in different possible circumstances is prevolitional (i.e., prior to God decreeing or willing anything). In Leibniz’s hands, these tenets are significantly molded by his other philosophical commitments, most notably a strong

i version of the principle of sufficient reason. Importantly, Leibniz rejects a libertarian account of freedom and the kind of contingency that it requires, and instead adopts a more traditional Thomistic account of the will, as a rational inclination towards apprehended goodness, whose acts are contingently moved and ultimately explained by the apprehended goodness of the of . From this I conclude that Leibniz is much closer to Molinism than is typically acknowledged. Leibniz is best characterized as a friend – rather than a foe

– of Molinism. I further argue that Leibniz’s views constitute an important development in the of the

Molinist controversy and that they deserve serious consideration in contemporary debates on the plausibility of Molinism.

The second major goal of my dissertation is to defend Leibniz’s views from a traditional challenge that threatens to undermine his ability to retain an intelligible sense of contingency, and thus threatens to undermine the success of the first part of the dissertation. This traditional challenge comes from Leibniz’s notorious thesis that every substance has an individual concept that includes predicates denoting everything that will ever happen to it. The challenge is that it appears that if everything that will ever happen to a substance is included in its individual concept, then it is not possible for the substance to be different from the way that its individual concept has it as . In response to this traditional challenge, I introduce a novel way of reading Leibniz’s account of individual concepts. I argue that an agent’s grounds her modal profile – what is possible, impossible, necessary, or contingent for her – and that this gives rise to a cluster of individual concepts which describe this modal profile. I argue that how an agent would utilize her powers to act settles which individual concept describes everything that will ever happen to her and which individual concepts only describe non- actualized alternative possibilities for her. The kind of contingency accommodated by this model is analogous to the kind of contingency accommodated by contemporary accounts that involve transworld – the thesis that exist in more than one . My account, I argue, makes intelligible Leibniz’s otherwise obscure claim that individual concepts include contingent truths as contingent, and, importantly, it enables us to see how things could have been different for an agent even given Leibniz’s commitment to individual concepts.

ii

These two main goals of the dissertation form a coherent picture. In the first part of the dissertation,

I assume that it makes sense to talk about the very same agent doing something other than what she in fact does; more precisely, I assume it is metaphysically possible for an agent to be otherwise than she is. Leibniz’s version of both Molinist tenets presupposes this assumption. It is only in chapter three that I vindicate, on

Leibniz’s behalf, this important assumption. Furthermore, the developments in the first part of the dissertation play a crucial role in the very articulation of my novel proposal on individual concepts in the second part. Out of the several chapters in this dissertation emerges a unified account of contingency and freedom that has remarkable similarities to traditional Molinism.

I end this dissertation with a chapter in which I address several objections to the advocated proposal on individual concepts. In the process of answering these objections, I illustrate how my proposal fits in the

Leibnizian system more generally. Importantly, I illustrate how my proposal relates to notorious Leibnizian commitments like the principle of sufficient reason and the of explanation, the principle of the identity of indiscernibles, and important aspects of Leibniz’s distinctive , most notably. I argue that the view advanced in this dissertation enables Leibniz to get what he wants – namely an intelligible and plausible sense of contingency as a condition for freedom – without giving up any of his other theoretical commitments.

Leibniz, I conclude, is entitled to an intelligible and plausible conception of contingency as a condition for freedom.

iii

Dedication

To my much beloved late grandfather, a wonderful example of kindness and perseverance, Jacinto Garcia

iv

Acknowledgements

Throughout the years, my work and my life have been enriched by the presence of many outstanding people. Here are some highlights. First, my thanks go to both of my advisors, Tamar Rudavsky and Julia

Jorati, for their dedication to my work; such dedication is highly appreciated. It goes without saying that their many contributions have been indispensable for this dissertation. My thanks also to the other members of the dissertation committee, Lisa Downing and William Taschek, and to the many professors and graduate students at The Ohio State University, including expatriate David Sanson, whose discussions on philosophical matters have enriched my academic career. Particular mention goes to Hope Sample for her and effort reading different drafts of this dissertation and for her insightful comments on them. I am also grateful for the many contributions to my life and career of both Lisa Shabel and William Taschek who have been there from my philosophical infancy, and who have played important roles in my academic development.

Special gratitude goes to my beloved wife, Kate Ellen Garcia, for her love and support. She is a central pillar of our wonderful family, which brings so much joy and to my life. And to our adorable children, David and Dahlia; their innocent smiles and overall zest for life continually rejuvenates my own life and endows it with a clear purpose. Lastly, I would also like to thank my extended family, including my in- laws, for their love and companionship; they have made my life outside of academia a happy one. The years spent toiling at this dissertation have been all the more pleasant because of their presence in it.

Multas gratias

v

Vita

2007 B.A. Philosophy, The Ohio State University

2014 M.A. Philosophy, The Ohio State University

2011 until Present Graduate Teaching Associate, Department of

Philosophy, The Ohio State University

Publications

“, Moral Character, and Alternative Possibilities in ”,

History of Philosophy Quarterly, 35.1: 59-75, 2018.

“Leibniz, a Friend of Molinism”, Res Philosophica, special edition New Frontiers in

Philosophy of , July 2018.

Fields of Study

Major Field: Philosophy

vi

Table of Contents

Abstract i Dedication iv Acknowledgments v Vita vi Table of Contents vii List of Abbreviations x Introduction 1 Philosophical 6 Chapter Summaries 8 Summary of Chapter 1 8 Summary of Chapter 2 10 Summary of Chapter 3 12 Summary of Chapter 4 14

Chapter 1: Leibniz, a Friend of Molinism 17 Introduction 17 Molina and the Doctrine of Middle Knowledge 20 The Basic Tenets of Middle Knowledge 20 Theoretical Motivations for the Molinist Tenets 22 Middle Knowledge at Work 24 A Contrast to Molinism: the Dominican Views 25 Leibniz and Molinism 28 Leibniz on the Doctrine of Middle Knowledge 29 Leibniz on the Substantive Molinist Tenets 30 Leibniz on Condition 31 Leibniz on the Prevolitional Condition 33 Divine Freedom and Providential Control 35 Conclusion 40

vii

Chapter 2: Real Options, Agential Modality, and Inclining Reasons 41 Introduction 41 Metaphysically Possible Options as Real Options 44 The of Choice and Real Options 48 Real Options and Agential Modality 51 Potential Objection and Reply 57 The Reality of Choice and Merely Inclining Reasons 62 Merely Inclining Reasons and Explanations 63 Agential Modality and Moral Necessity 66 Conclusion 68

Chapter 3: Leibniz on Contingency and Individual Concepts 70 Introduction 70 Individual Concepts and the Superessentialism Challenge 72 The Intuitive Challenge 72 Arnauld’s Argument for Superessentialism 73 Individual Concepts and 75 Essence and the Logical Strategy 78 Answering the Demarcation Question for the Logical Strategy 79 A Leibnizian Response to Superessentialism 85 Including Contingent Truths in Individual Concepts 85 Unexercised Powers and Individual Concepts 86 Comparison with Contemporary of Modality 89 Cluster of Concepts and Modal Profile 91 Primitive Force, Multiple Individual Concepts and PSR 94 Privileged Concept and Counterfactual Dependence 101 Including God’s Possible Decrees in Individual Concepts 103 A Leibnizian Response to the Argument(s) for Superessentialism 107 Conclusion 108

Chapter 4: Objections and Replies 109 Introduction 109 The Principle of the Identity of Indiscernibles Violation Objection 111 Basic Reply to PII Violation Objection 112 The Modal Status of PII 113 God’s Power to Violate PII 116

viii

Leibniz on Theodicy 117 The Theodicy Compromised Objection 119 The Initial Response: Superessentialism is Unnecessary 120 Theodicy not Compromised 122 The Buck-Passing Objection 124 Response to Level One: Spectrum of Possible Precisifications 124 Second Level of the Buck-Passing Objection 132 God’s Essential Nature and Subjunctive Conditionals 133 God’s Essential Nature and PSR 134 God’s Actions are Contingent 140 Conclusion 143

Conclusion 145 Bibliography 147

ix

List of Abbreviations

Works by Leibniz and Translations thereof

A Sämtliche Schriften und Briefe. Ed. Deutsche Akademie der Wissenschaften. Darmstadt, Leipzig, Berlin: Akademie Verlag, 1923-. Cited by series, volume, and page.

AG Philosophical Essays. Ed. and transl. Roger Ariew and Daniel Garber. Indianapolis: Hackett, 1989. Cited by page number.

C Couturat, Louis, ed. Opuscles et fragments inedits de Leibniz: Extraits des manuscrits de la Bibliotheque royale de Hanovre. Paris: F. Alcan, 1903.

CD Causa Dei asserta per justitiam ejus, cum caeteris ejus perfectionibus, cunctisque actionibus conciliatam (1710). G 6:439–462. English translation from S 114-145. Cited by section number.

CP Confessio Philosophi: Papers Concerning the Problem of , 1671-1678. Edited and Translated by Robert C. Sleigh, Jr., B. Look and H. Stam. New Haven: Yale University Press, 2005.

DM Discourse on Metaphysics (1686). G 4:427–463; English translation from AG 35-68. Cited by section number.

DPG Dissertation on and Grace. Edited and Translated by Michael J. Murray. New Haven: Yale University Press, 2011. Cited by section number.

FC Refutation Inedite de Spinoza. Foucher de Careil, Paris 1857. Cited by section number.

G Die philosophischen Schriften von Gottfried Wilhelm Leibniz. Edited by C. I. Gerhardt. Berlin: Weidmann, 1875–90. Reprinted Hildesheim: Georg Olms, 1978. Cited by volume and page.

Grua Textes inédits d’après des manuscripts de la Bilbliothèque provinciale d’Hanovre. Edited by Gaston Grua. Paris: Presses Universitaires de France, 1948.

L Philosophical Papers and Letters. Translated and edited by Leroy E. Loemker, Kluwer Academic Publishers, 1989. Cited by work, section and page number.

LA The Leibniz-Arnauld Correspondence, With Selections from the Correspondence with Ernst, Landgrave of Hessen-Rheinfels. Translated by Stephen Voss. Yale University Press. Cited by page number.

x

LC G. W. Leibniz and Samuel Clarke: Correspondence. Edited by Roger Ariew. Hackett. 2000.

LGR Leibniz on God and Religion, Edited and translated by Lloyd Strickland. Bloomsbury Academy, 2016.

M Monadology (1714). Cited by section as in G 6.607-623; English translation from AG 213-225.

MP Philosophical Writtings, Edited by G. H. R. Parkinson, translated by Mary Morris and G. H. R. Parkinson. London: Everyman’s Library, 1973.

NE New Essays concerning Understanding, G 5:41-509; Translated by Alfred Gideon Langley. New York: The Macmillan Company, 1896. Cited by section number.

PT Philosophical Texts. Edited by R. S. Woolhouse and R. Francks. New York: Oxford University Press, 1998. Cited by page number.

S Monadology and other Philosophical Essays. Edited and translated by P. and A. M. Schrecker. Indianapolis: Bobbs-Merrill, 1965. Cited by page number.

SLT The Shorter Leibniz Texts, Edited and translated by Lloyd Strickland. New York: Continuum. Cited by page number.

T Theodicy. (1710). G 6:102–365; English translation E. M. Huggard. Theodicy. La Salle: Open Court, 1985. Cited by section number.

W Leibniz: Selections. Edited by Philip P. Wiener. Scribners, 1951. Cited by page number.

WFI Leibniz’s ‘‘New System’ and Associated Contemporary Texts. Edited by R. S. Woolhouse and R. Francks. Oxford: Clarendon Press, 1997. Cited by page number.

WFII Philosophical Texts, Edited and Translated by R. S. Woolhouse and R. Francks. Oxford: Oxford University Press, 1998. Cited by page number.

Other Authors

AF , On Divine Foreknowledge (Part IV of Concordia), Translated and introduction Alfred Freddoso, Ithaca: Cornell University Press, 2004. Cited by page number.

xi

B Domingo Bañez et Diego Alvarez, Apologetica fratrum prædicatorum in provinciâ Hispaniæ sacræ theologiæ professorum, adversus novas quasdam assertiones cujusdam doctoris Ludovici Molinæ nuncupati, Madrid, 1595. Cited by section number.

Co Luis De Molina, Liberi arbitrii cun gratiae donis, divina praescientia, providentia, praedestinaione et reprobatione concordia. Ed. Johann Rabeneck. Oña and Madrid: Collegium Maximum S. I. Cited by section number.

CT Saint , Corpus Thomisticus: Sancti Thomae de Aquino Opera Omnia. Ed. Enrique Alarcon. www.corpusthomisticum.org/iopera.html

E Benedictus de Spinoza, The in A Spinoza Reader: The Ethics and Other Works. Translated and with introduction by Edwin Curley. Princeton University Press, 1994. Cited by book and section number.

HB and Bramhall on and Necessity, ed. Vere Chappell. Cambridge: Cambridge University Press 1999. Cited by section numbers.

OO Francisco Suarez, Opera omnia. 28 vols. Ed. Carolo Betron. Paris: Vives, 1856-78. Cited by volume and section.

SCG Saint Thomas Aquinas, Summa Contra Gentiles. In CT. Cited by book, chapter and section number.

ST Saint Thomas Aquinas, Summa Theologicae. In CT. Cited by part, question, and article numbers.

xii

Introduction

One of the most taxing topics in Leibniz scholarship is Leibniz’s account of contingency. Leibniz insists that contingency is a condition for freedom. He writes, for example, “I am of the opinion that our will is exempt not only from constraint but also from necessity.”1 This is not an isolated remark. In many writings, when broaching the topic of freedom, Leibniz begins by noting that freedom requires contingency, and then devotes much of his impressive intellectual powers to elucidating the nature of both freedom and contingency.

These are recurring topics throughout Leibniz’s life. Furthermore, Leibniz describes “the question of the Free and the Necessary,” as one of the two “labyrinths” of the human , one that “perplexes almost all the human race.”2 Many are led astray here, Leibniz warns us.3 It is clear that the topic of contingency, as a condition for freedom, captivated and even vexed Leibniz throughout his career.

Part of what makes this topic so taxing in Leibniz scholarship is that Leibniz vacillated much, throughout his career, regarding how best to capture this valuable condition on freedom. He advocated multiple, and apparently independent, strategies to account for contingency without clearly or definitely opting for one of them or without clearly explaining their interrelationship, if any.4 Given this plethora of strategies, at the end of the day it is not definitively clear which strategy, if any, represents Leibniz’s final word.

1 T § 34. Translations are my own unless a translation is cited. 2 T Preface, p. 53. 3 T Preface, p. 53. 4 I address some of these strategies throughout this dissertation, and take it as a strength of the picture advocated here that I can incorporate several of these strategies into a coherent whole. The secondary literature on Leibniz’s attempts to ground contingency is extensive. Some examples include Abraham 1969, Adams 1994, Baxter 2000, Blumenfield 1988, Burms 1979, Carriero 1995, Frankel 1984, Griffin 1999 and 2013, Grimm 1970, Hunter 1981, McNamara 1980, Meijering 1978, O’Leary-Hawthorne and Cover 1990 and 1992, Savage 1998, Vailati 1986, and Wee 2006.

1

Interpretative proposals on this topic, like the one developed in this dissertation, must, therefore, be advanced with these exegetical limitations in mind.

A further complication, for Leibniz scholarship on this topic, is that Leibniz endorsed several theses which seem to preclude contingency – most notably, a strong version of the principle of sufficient reason,

Leibniz’s distinctive doctrine that every substance has an individual concept that includes predicates denoting everything that will ever happen to it, and strong theological commitments like divine foreknowledge and robust providence. Many commentators have that these commitments do away with any intelligible sense of contingency in Leibniz’s system. It is a main task of this dissertation to argue that these commentators are mistaken and that Leibniz is indeed entitled to an intelligible, and even plausible, account of contingency as a condition for freedom even given these seemingly conflicting theoretical commitments.

The account on contingency and freedom advanced in this dissertation, as an interpretation of

Leibniz’s views, enjoys remarkable similarities with Molinism – the view that God has prevolitional knowledge of contingent subjunctive conditionals of freedom, or of the form “if agent S were in circumstances C, then S would freely phi”. These similarities will strike many Leibniz scholars as surprising.

For, one important target in Leibniz’s discussions on contingency and freedom is sixteenth century Jesuit priest

Luis de Molina, from whom Molinism gets its name. Molina’s account of freedom is a robust kind of libertarianism, and Leibniz thinks that this kind of “freedom of equipoise,” as he calls it, is “a chimera , which contravenes the great principle of reason”.5 Leibniz’s objection is that robust libertarianism requires that the conditions for action not explain, or causally determine, free actions, and thus that this view requires violations of his much treasured principle of sufficient reasons.6 It would be contrary to wisdom to demand a kind of freedom that violates this principle, Leibniz insists, providing a criticism reminiscent of Dennett’s famous objection that libertarian freedom is not worth wanting.7

5 Leibniz to Jaquelot 28 April 1704, G 3.471/ WFI p. 180. 6 Leibniz to Jaquelot, G 3.471-3, T §§ 175, 199, 303, 349, Grua 271, 176-7, 280. 7 Dennett 1984.

2

Partly because of Leibniz’s castigation of robust libertarianism, and the kind of contingency his view requires, it is commonly believed in the secondary literature that Leibniz is a traditional foe of Molinism. This classification of Leibniz as a foe of Molinism has further support because Leibniz explicitly distanced himself from the doctrine of middle knowledge – a distinctive tenet of Molinism.8 Leibniz insists: “Thus we can see that, in order to account for the foreknowledge of God, one may dispense with both the middle knowledge of the Molinists and the predetermination which a Bañez or an Alvarez (writers otherwise of great profundity) have taught.”9 These are initially persuasive reasons to think that Leibniz is indeed an opponent of traditional

Molinism.

Yet, I will argue, this received view is not quite correct. Contrary to this common reading of Leibniz, a major goal of this dissertation is to argue that Leibniz’s views on contingency, freedom, divine foreknowledge, providence and grace are remarkably close to traditional Molinism. The general sketch of my argument for this conclusion is the following. Robust libertarianism, which Leibniz clearly rejects, is not constitutive of Molinism as such; robust libertarianism is rather one of the main purported theoretical benefits of Molinism.10 Leibniz,

I will argue, endorses an account of contingency, as a condition for freedom, that is not libertarian, but does utilize core theoretical resources of Molinism to secure this alternative kind of contingency.

Here is a brief summary of Leibniz’s preferred account of contingency, as I see it. Leibniz’s preferred kind of contingency as a condition for freedom, I will argue, is grounded in an agent’s ability to determine herself towards a course of action on the basis of her deliberate judgment of this course of action being the best. This kind of power, I will argue, is grounded in a more traditional Thomistic conception of the agent’s will – as a rational inclination towards the apprehended goodness of the object of choice. Importantly, I will

8 As Molina sees it, the doctrine of middle knowledge is a doctrine about God’s knowledge of subjunctive conditionals of freedom. This knowledge is said to be ‘middle’ because, according to Molina, it lies between the two other kinds of knowledge that God is traditionally said to have. These two other kinds of divine knowledge are traditionally known as “God’s knowledge of simple ” and “God’s knowledge of vision.” God’s knowledge of simple intelligence is the kind of knowledge that God has in of knowing His own essence and what follows from His essence, whereas God’s knowledge of vision is the kind of knowledge that God has in virtue of knowing His will (Co IV.52.9-13). 9 T § 47. 10 In fact, this is exactly how both Molina himself, and contemporary Molinist present their views. See, for example, Molina Co IV.52.9-13, Flint 1998 and 1992, and Craig 1990.

3 argue that a choice counts as contingent when it is the result of rational deliberation that takes into consideration more than one possible course of action the agent has this kind of power to bring about. In sum, an agent is free to the extent that she determines herself to do that which she deliberately judges to be the best from several considered possible options that she could have brought about had she come to the deliberative conclusion that these options were best.

The second part of my argument addresses the doctrine of middle knowledge directly. I argue that

Leibniz’s relationship to the doctrine of middle knowledge is complex. Sometimes, instead of distancing himself from the doctrine of middle knowledge, he attempts to reduce it to something else:

What is commonly called middle knowledge is contained in the knowledge of simple intelligence. [CD § 17]

Knowledge of simple intelligence could be taken so as to include middle knowledge. For knowledge of the possibles involves their connections (and thus it includes what would happen when a thing is posited in act), including not only the necessary connections, but also the contingent ones, that is, those which merely incline, for such a series of causes does not obstruct freedom. [DPG § 16a, p. 75]

Conditional foreknowledge or middle knowledge arises out of the divine knowledge of simple intelligence or of the possibilities, even with respect to those that are not brought into actuality, just as it was explained above. [DPG § 49a, p. 135]

Sometimes it even appears that Leibniz is just endorsing middle knowledge:

If, however one wants a knowledge midway between the knowledge of simple intelligence and the knowledge of vision one could conceive both the knowledge of simple intelligence and middle knowledge differently from the common usage. In this case one could assign to middle knowledge not only the knowledge of conditional future events but, generally, the knowledge of all contingent possibles. Thus knowledge of simple intelligence would be taken in a more restricted sense, namely, as dealing with possible and necessary truths, while the knowledge of vision would deal with contingent actual truths. Middle knowledge and knowledge of simple intelligence would have this in common, that they both deal with possible truths, while middle knowledge and the knowledge of vision would both deal with contingent truths. [CD § 17]

Given passages like these, it is best to conclude that Leibniz’s response to the doctrine of middle knowledge is more complicated than mere rejection. Here is a general sketch of my interpretation defended in this dissertation. Leibniz endorses the substantive tenets of middle knowledge and rejects aspects of this doctrine he sees as immaterial. To better appreciate this, consider the following more careful articulation of the basic tenets constitutive of the doctrine of middle knowledge, as Molina sees it (note: these are my labels):

4

Contingency Condition: subjunctive conditionals of freedom are contingent.11

Prevolitional Condition: God knows subjunctive conditionals of freedom prevolitionally (i.e., prior to God actually willing anything, or prior to any actual ).12

Middle Position Condition: God’s knowledge of subjunctive conditionals of freedom is different from, and in some sense between, God’s knowledge of simple intelligence and vision.13

The doctrine of middle knowledge gets its name from the Middle Position Condition, but, as Leibniz sees it, the philosophical substance lies in the Contingency and Prevolitional conditions. I suggest that the passages in which Leibniz distances himself from the doctrine of middle knowledge are better understood as mere rejections of the Middle Position Condition. Leibniz’s grounds subjunctive conditionals of freedom in God’s knowledge of simple intelligence instead. Furthermore, and crucially, I argue that Leibniz endorses both the

Contingency and Prevolitional conditions as uses them for similar theoretical purposes as Molinists. Given all of this, I conclude, Leibniz is best characterized as a friend – rather than a foe – of Molinism.

Essential to both Leibniz’s versions of the Contingency and Prevolitional conditions is the assumption that Leibniz is not committed to superessentialism – the view that for any object O and P, if O has P it is metaphysically necessary that O has P. If Leibniz is committed to superessentialism, then the unexercised powers that ground the kind of contingency that matters for freedom would be essentially unexercised powers, and subjunctive conditionals of freedom would be essentially true, if true, and essentially false, if false. These consequences are devastating for the main project of this dissertation, for they both undermine a tenable conception of contingency as a condition for freedom. It is the second main goal of this dissertation to argue that Leibniz is not committed to superessentialism. To secure this conclusion, I advocate a novel way of understanding individual concepts which opens up conceptual space, in Leibniz’s system, for something analogous to transworld identity,14 and thus for an intelligible sense of metaphysical contingency. In the first couple of chapters of this dissertation, it is a mere promissory note that Leibniz is entitled to metaphysical

11 Co IV.52.9-10. 12 Co IV.52.9-10, IV.53.1.20. 13 For more details about these different kinds of divine knowledge, please see footnote 8. 14 The contemporary thesis, grounded in analysis of modality in terms of possible world semantics, that individuals exist in more than one possible world.

5 contingency and is thus not committed to superessentialism. It is not until chapter three that I fulfill this promissory note on Leibniz’s behalf. All in all, the picture advocated in this dissertation, I believe, shows that

Leibniz has the theoretical resources to retain an intelligible and even plausible sense of contingency as a condition for freedom.

Philosophical Methodology

A few words about methodology are in order. Much of this dissertation exemplifies what I shall call

‘pure historical exegesis’ – that is, careful reading of a text, understood in its proper historical context, with the goal of understanding what the author of the text intended to convey. As I see it, the main consideration in pure historical exegesis is understanding, and articulating, what is explicit in the text – typically by synthesizing different parts of the text and finding different forms of packaging these views in informative ways. In this dissertation I aim to understand Leibniz’s views regarding contingency and freedom, and I advance what I take to be careful articulations of some of these views by the standards of pure historical exegesis.

My dissertation also goes beyond pure historical exegesis, however. I also undertake a project I shall call ‘rational reconstruction’ at a couple of central junctures in this dissertation. By ‘rational reconstruction’ I mean the development of a model which is not explicitly stated in the text, but which, if accepted, enables the reader to understand what is explicitly stated in the text or to understand what is implicit in the text. Rational reconstructions make more precise ideas that are underdetermined in the text, and so go beyond what is explicitly stated, but not in a way that does damage to what is stated either explicitly or implicitly in the text.

When relying on rational reconstructions to understand a text, attributing the rationally reconstructed model to the author of the text should be done only tentatively and with important qualifications. The most important qualification is that by its very nature rational reconstructions involved more than what is explicitly found in the text.

6

The precise rules for assessing what makes a rational reconstruction an adequate one, if there are any, are difficult to articulate. Here are what I take to be some general guidelines for adequate rational reconstructions, however. As I see it, a rationally reconstructed model, much like pure exegesis, should stay as close to the text as feasible; that is, the model should take seriously the philosophical motivations stated in the text, and should not introduce philosophical ideas which are clearly in conflict with anything explicitly or implicitly stated in the text. In going beyond the explicit content of the text, however, it is acceptable to introduce philosophical ideas which help make sense of what is stated in the text. A example, in this dissertation, is my introduction of something analogous to transworld identity. It is clear that Leibniz did not explicitly state something as precise as this doctrine; though it is not clear to me that he did not implicitly rely on something like it. However, I argue that by explicitly incorporating this philosophical tool into the rationally reconstructed model, we can make sense of what Leibniz did explicitly say about contingency without undermining anything else he said or without undermining a clear implication of something else he said.

Additionally, I am also concerned with philosophical plausibility – in the general sense of a credible contender to the . This concern with philosophical plausibility is secondary to both pure historical exegesis and rational reconstructions: concerns of philosophical plausibility play a role only after both of these deeper interpretative desiderata have been met. At least, those are the parameters of the project of this dissertation as

I see it. My philosophical investigations of Leibniz’s texts assume the intrinsic of historical scholarship, but also aim to uncover tenable philosophical ideas that are worth considering in themselves. I do not, however, do much in terms of defending Leibniz’s views. I merely intend to uncover interesting philosophical ideas which can in some sense be attributed to Leibniz and which also enjoy some degree of philosophical plausibility.

A final point about methodology concerns what I have called the ‘philosophical stretches’ on the tenability of contingency as a condition of freedom.15 As I see it, the theoretical commitments that stretch contingency are not to be understood as mere theoretical obstacles for a tenable account of contingency as a condition for freedom; they are better understood, rather, as theoretical considerations that help assess the

15 I borrow the imagery of ‘stretching’ a philosophical to reveal interesting qualities from Adams (1980: 84).

7 philosophical tenability of contingency itself, at least as Leibniz sees it. This is so because, as I see it, Leibniz is a system builder; that is, as Leibniz understands philosophy as a systemic enterprise: whether a precise and nuanced philosophical point fits well within more general philosophical considerations is of central importance for the tenability of that very philosophical point. For instance, and apropos the goals of this dissertation, whether an account of the nature of contingency as a condition of freedom is tenable depends upon how it relates to more fundamental theoretical commitments like a strong version of the principle of sufficient reason, the complete individual concept doctrine, and strong theological commitments like divine foreknowledge and robust providence.

I chose to discuss the kind of contingency that matters for freedom in Leibniz’s views precisely because this topic requires addressing a cluster of interrelated topics – the nature of modality, the nature of explanation, the demands of a strong version of the principle of sufficient reason, the nature of the will, and the constraints on a plausible account of freedom imposed by strong theological commitments. Again, from Leibniz’s perspective, the plausibility of an account of the kind of contingency required for freedom depends upon how well it fits with these more general philosophical and theological considerations. Thus, writing a dissertation on this narrow topic afforded me the opportunity to engage, at least to some extent, with a wide variety of more general philosophical topics, for the very plausibility of the narrow topic depends upon its interrelation to these more general ones.

Chapter Summaries

Summary of Chapter 1: Leibniz a Friend of Molinism

Molinism is the view that God has prevolitional knowledge of contingent subjunctive conditionals of freedom—that is, of propositions of the form “if agent S were in circumstance C, then S would freely ɸ”.

Molinism was first advanced by sixteenth century Jesuit Luis de Molina, from whom it gets its name, and it

8 gave rise to fierce theological controversies between Jesuits and Dominicans regarding the nature of freedom and its relationship to traditional theological commitments like divine foreknowledge and robust providence.

This genesis of the Molinism controversy received much fruitful attention in the secondary literature. There is a crucial development in the history of this controversy, however, that has been mostly neglected and even misunderstood in the secondary literature. This crucial development was advanced by Leibniz. In chapter one,

I aim to fill some of this gap in the secondary literature by developing Leibniz’s novel contribution to the history of the Molinism controversy. Leibniz challenges some central assumptions driving the controversy, and in so doing he opens conceptual space for a plausible and novel position that deserves a place in the contemporary discussion.

Leibniz’s views on this topic have received some attention in the secondary literature, but mostly to portray him as a traditional opponent of Molinism. This classification is not without reason. Leibniz objects to robust libertarianism, one of the main purported theoretical benefits of Molinism, because it violates his much-championed principle of sufficient reason. Furthermore, Leibniz also seems explicitly to dismiss a distinctive thesis of Molinism, namely the doctrine of middle knowledge. He writes, for example, “Thus we can see that, in order to account for the foreknowledge of God, one may dispense with both the middle knowledge of the Molinists and the predetermination which a Bañez or an Alvarez (writers otherwise of great profundity) have taught.”16 It is, thus, no surprise that Leibniz is commonly classified a foe of Molinism.

In this chapter, I argue that this standard view is not quite correct. I identify the two substantive tenets of Molinism. First, the connection between the conditions for free actions and these free actions is a contingent one: free actions follow contingently from their sufficient conditions. Second, God knows what creatures would freely do in different possible circumstances prevolitionally – that is, prior to God willing anything. I argue that Leibniz himself endorses a version of both substantive Molinist tenets and utilizes them for theoretical purposes similar to those of Molinists. In particular, I illustrate how Leibniz’s version of these Molinists tenets allows him to advance a novel and intriguing account of that is much closer to traditional

16 T § 47.

9

Molinism than it is typically acknowledged in the secondary literature. I conclude that Leibniz is much closer to Molinism than is typically acknowledged. Leibniz is best characterized as a friend – rather than a foe – of

Molinism.

Summary of Chapter 2: Real Options, Agential Modality, and Inclining Reasons

One of Leibniz’s more perplexing statements regarding freedom is that free actions are contingent because they are brought about by reasons which merely incline and do not necessitate these free actions.

Leibniz pens “Now every effect is determined by its causes and their dispositions, such that there is always some reason why it exists rather than not,… the reasons that determine a free cause are never necessitating but only inclining, and to that extent the indifference or contingency in them is preserved.”17 I have labeled this doctrine Leibniz’s “merely inclining reasons” doctrine. Part of the puzzle with Leibniz’s merely inclining reasons doctrine is that these merely inclining reasons are meant to explain why the actions they bring about are the way they are and not otherwise without also thereby necessitating them. That is, part of the puzzle here is that Leibniz insists that merely inclining reasons are sufficient to meet the demands of his strong version of the principle of sufficient reason without necessitating that which they bring about and fully explain.

In chapter two, I make sense of Leibniz’s merely inclining reasons doctrine by advancing a version of what Robert Adams has labeled ‘the reality of choice’ Leibnizian strategy to ground contingency.18 The basic idea, undergirding the reality of choice strategy, is that the contingency of a choice is preserved by the fact that the considered but not-chosen possible options open to the agent are not rendered impossible by this choice.19

I argue that the reality of choice strategy enables us to see how Leibniz’s three conditions for freedom – intelligence, spontaneity and contingency – come together in a unified account. Intelligence is essential to freedom because free actions are deliberate actions. Spontaneity is essential to freedom because free actions

17 On God and Man, G 3.36/ LGR 297. 18 Adams 1994: 20. 19 Adams 1994: 20-3.

10 are the result of the kind of rational self-determination that follows deliberation. Contingency is essential to freedom because free actions are the result of deliberate weighing of the apprehended goodness of several options that the agent has the agential power to bring about. In sum, an agent is free to the extent that she determines herself to do that which she deliberately judges to be the best from several options that she could have brought about had she come to the deliberative conclusion that these options were best.

I argue that the reality of choice strategy makes room for a kind of power to do otherwise that grounds the contingency of free actions. Importantly, the kind of power that matters for freedom, on Leibniz’s views, is the power the agent has to determine herself on the basis of the deliberated judgment of the best. My suggestion is that the agent must have this kind of power with respect to more than the actual course of action; generally speaking, to have ‘real options’ – the kind of possible options that for freedom – is for there to be options that the agent has this kind of power to bring about. In other words, the kind of ‘possibility’ that matters for possible options is grounded in the power the agent has to determine herself on the basis of deliberate judgments about the best.

Essential to this unified account of freedom is a Leibnizian commitment regarding what it is for an option to be open to an agent. I suggest that what it is for an option to be open to an agent, in the sense that matters for freedom, is for this option to be the sort of thing that can be taken into consideration in the process of deliberation, and, importantly, for an option to be the sort of thing that can be taken into consideration is for this option to be the sort of thing that an agent can come to recognize as good and for the agent to be rationally inclined to bring this option about on the basis of this perceived goodness. For an option to be open, in the way that matters for freedom, is, in sum, for that option to be such that the agent can determine herself to bring that option about if the agent comes to the deliberative conclusion that that is the best considered course of action.

Finally, and importantly, in making sense of Leibniz’s merely inclining reasons doctrine, we will thereby also make sense of Leibniz’s version of the Contingency Condition – the Molinist tenet that subjunctive conditionals of freedom are contingent. As noted in chapter one, Leibniz holds two seemingly conflicting

11 claims regarding these subjunctive conditionals of freedom: i) the connection between conditions for action, specified in the antecedents of subjunctive conditionals of freedom, and free actions, specified in the consequents, is a contingent connection; and ii) the conditions for action explain the free actions. In chapter two I argue that the advocated account of the reality of choice strategy makes intelligible how merely inclining reasons can explain free actions, and in so doing it also makes intelligible Leibniz’s version of the Contingency

Condition.

Summary of Chapter 3: Leibniz on Contingency and Individual Concepts

One of the most vexed topics in Leibniz scholarship is Leibniz’s account of contingency. Leibniz insists that freedom requires contingency, yet he also endorses several theses which, at least initially, seem to entail that substances have all of their properties essentially – i.e., superessentialism. One such important thesis is Leibniz’s notorious doctrine that every substance has an individual concept that includes predicates denoting everything that will ever happen to it. This doctrine is referred to as ‘the complete individual concept’ doctrine in the secondary literature. Despite explicit protestations from Leibniz, many readers have thought that

Leibniz’s views do indeed commit him to superessentialism. In chapter three of this dissertation, I argue that this somewhat common reading of Leibniz’s commitments is mistaken. I present a plausible argument from

Leibniz’s complete individual concept doctrine to superessentialism, and I argue that Leibniz has the theoretical tools to retain an intelligible sense of contingency in his system. The key idea I develop is Leibniz’s insistence that he is not committed to superessentialism because contingent truths can be included in individual concepts precisely as contingent. My suggestion is that truths can be included as contingent in individual concepts by concepts including two kinds of predicates: a) predicates denoting what the agent will in fact do; and b) predicates denoting unexercised powers to do otherwise that agents also possess. By including both of these kinds of predicates, it is intelligible to think that contingent truths can be included in individual concepts as contingent.

12

This suggestion by itself however does not yet answer to superessentialist challenge, for it is consistent with this proposal that these unexercised powers to do otherwise be essentially unexercised powers. In order for these powers to ground legitimate contingency – the kind required for freedom – it must be possible for these powers to be exercised. In order to ground these kinds of possibilities, I suggest that the content of individual concepts be counterfactually dependent upon how agents utilize their causal power. More precisely, an agent’s individual concept includes predicates denoting what she will in fact do, but she retains the power to do otherwise such that were she to exercise this power her individual concept would have included a predicate denoting this alternative possibility instead.

The final part of my proposal includes a model which makes sense of this kind of counterfactual dependence without doing violence to the plausible thesis that concepts are individuated by their content and so that predicates included in concepts are essentially included in those concepts. I suggest that we understand

Leibniz’s discussions of ‘possible-Adams’ – which in some sense represent alternative possibilities for Adam – as discussion of a of individual concepts each of which represents a complete way things could go for Adam.

One individual member of this cluster represents how things will actually go for Adam if created – I call this concept the ‘privileged’ member of the cluster – and the other members of the cluster represent unactualized possibilities for Adam. Finally, and crucially, I understand the postulated counterfactual dependence between the content of an agent’s individual concept and how she utilizes her powers in terms of a counterfactual relation regarding which member of the cluster gets to be the privileged one. It is how an agent would utilize her powers that determines which member of the cluster gets to be the privileged one. For example, consider concepts C1 and C2 both of which are members of Adam’s cluster of individual concepts. Suppose C1 includes the predicates ‘ɸ-ing’ and ‘having the power not to ɸ’ and C2 includes the predicates ‘not-ɸ-ing’ and ‘having the power to ɸ’. Further suppose that because of the way that agent S would utilize her powers, if created, C1 is the privileged member of S’s cluster of individual concepts – that is, if S were created it would be C1 that would describe everything that will happen to S and all other members of the cluster would only describe unactualized alternative possibilities. My suggestion is simply that given that S has the power to not ɸ, as specified in C1, were S to exercise this power, it would have been the case that C2 would have been the

13 privileged member of S’s cluster of individual concepts – and thus it would have been C2 that would have described everything that will happen to S if created. This account, I suggest, allows Leibniz to answer the argument for superessentialism and retain an intelligible sense of contingency in his system.

Summary of Chapter 4: Objections and Replies

In chapter four, I illustrate how the account painted in the dissertation up until that point fits with several other central Leibnizian commitments. I do so by addressing three important objections to my proposal.

These objections are the following. First, postulating a cluster of individual concepts associated with a single substance might seem to compromise Leibniz’s principle of the identity of indiscernibles by making it possible for God to instantiate more than one member of this cluster and thus create the same individual more than once and with different properties. This would clearly be a violation of Leibniz’s principle of the identity of indiscernibles for there would be discernible which are nonetheless identical. Second, by postulating a cluster of individual concepts my account seems to do away with a part of Leibniz’s theodicy because things could have been different for a given substance and yet God allowed this substance to suffer, thus seemingly opening legitimate grounds for complaining to God for creating the substance this way instead of otherwise.

Third, it appears that my account postulates a problematic buck-passing element because it postulates that

God’s possible free decrees are included in individual concepts. This objection has two levels. On the first level, the mere fact that the contingency of human actions depends upon the contingency of divine actions is taken to be problematic, for it is hard to see how God’s actions being contingent contributes to human freedom.

The second level of the objection is that the contingency of human actions is lost if this contingency depends upon the contingency of divine actions. This is so, the objection continues, because God’s essential nature and the principle of sufficient reason render God’s actions necessary.

My basic response to the first objection is to point out that my proposal does not introduce any tensions or problems for Leibniz that are not already present without it. Importantly, my proposal does not imply any actual violation of the principle of the identity of indiscernibles; it merely implies that God has the

14 power to violate this principle. I further provide reasons for thinking that this implication is not as unpalatable as might originally seem – one important suggestion is that Leibniz is already committed to the principle of the identity of indiscernibles being metaphysically contingent, and if so there is no new problem for Leibniz here.

My basic response to the second objection is that Leibniz’s theodicy is not compromised because

Leibniz is in a position to provide a theodicy analogous to the one traditional Molinists provide. That is, a given agent has no grounds for complaining to God because reality is the way it is partly on the basis of which subjunctive conditionals of freedom are true and which are false, and, again, all of this is settled prevolitionally, and, thus, outside of God’s control. In the terminology of this dissertation, part of what makes the privileged concept be the privileged member of the cluster is grounded in the substance’s own essence and what the substance would do in different possible circumstances, and thus not the sort of thing for which God can be held accountable.

My basic response to the first level of the third objection is the following. First, I insist that it is still underdetermined whether the proposal in chapter three entails that the contingency of human acts depends on the contingency of divine acts in any problematic way. There is indeed a kind of ontological dependence that is postulated in that proposal, but whether it amounts to a problematic dependence relation is still unspecified: the problem remains a mere prima facie problem. One of the goals of this chapter is to develop this prima facie problem into a clearly problematic picture and, then, to distinguish this picture from the one I am advocating on Leibniz’s behalf. The main philosophical move involves noting that there is a whole spectrum of possible precisifications of the general proposal articulated at the end of chapter three. A good way of articulating this spectrum is in terms of explanation: on the one end, all monadic actions are explained by the nature of the monad and God’s causal activities are mere conditions sine qua non; at the other end of the spectrum God’s causal activities fully explain all monadic actions and the nature of monads is a mere condition sine qua non. It is clear that the precisification at the latter end of the spectrum includes a problematic buck-passing element; whereas there is no such problematic element involved with the precisification at the other end. I argue that

Leibniz’s actual views lie somewhere in between these extremes, and leave it open whether this amounts to a problem for his views.

15

My basic response to the second level of this objection is two-fold. First, I argue that even if God’s actions are indeed metaphysically necessary, this would not undermine the contingency of human freedom.

The contingency of human acts is preserved by the Contingency and Prevolitional conditions, which would not be undermined if God’s actions were metaphysically necessary. The second part of my response is that Leibniz is not committed to the claim that God’s actions are metaphysically necessary. I argue that the account of inclining reasons developed in chapter two can also be applied to God, and can thus enable us to see how

God’s actions can be contingent even given His metaphysically necessary nature.

16

Chapter 1:

Leibniz, a Friend of Molinism

INTRODUCTION

Molinism is the view that God has prevolitional knowledge of contingent subjunctive conditionals of freedom—that is, of propositions of the form “if agent S were in circumstance C, then S would freely ɸ”.20

This view has seen a resurgence in contemporary .21 There is again a live controversy over whether Molinism has the theoretical tools to plausibly reconcile a robust libertarian account of freedom with traditional theological commitments like divine foreknowledge and a robust account of divine providence.22

Like in most other philosophical debates, much light can be shed on the crucial philosophical ideas driving this controversy by looking at the historical context in which those ideas were first articulated and defended.

Molinism was first advanced by sixteenth century Jesuit Luis de Molina, from whom it gets its name, and it gave rise to fierce theological controversies between Jesuits and Dominicans regarding the nature of freedom and its relationship to traditional theological commitments. This genesis of the Molinism controversy has already received fruitful attention in the secondary literature.23 There is a crucial development in the history of this controversy, however, that has been mostly neglected and even misunderstood in the secondary literature.

20 I spell out this view in more detail below. 21 (1974: Ch IX, and 1977) reintroduced Molinism to develop a compelling defense in response to the , and in doing so he placed Molinism back into mainstream philosophical discussion. 22 See for example, Craig 1988, 1990, 1994, Fales, 2010, Fisher 2008, Gaskin 1993, 1994, 1998, Hasker 1986, 1989, 1992, 1999, Perszyk 1998, 2011, 2013, and Flint 1988, 1992, 1998, 1999, Adams 1977, 1991, Perszyk 1998, and O’Connor 1992. 23 See for example, Freddoso 1998, Kaphagawani 1999, Smith 1966, Murray 1995, 1996, 2004, Craig 1988: Ch. 7-8.

17

This crucial development was advanced by Leibniz. In the present chapter, I aim to fill some of this gap in the secondary literature by developing Leibniz’s novel contribution to the history of the Molinism controversy.

Understanding Leibniz’s positions on these theological debates is intrinsically valuable, to be sure. Yet, I will argue, it can also can also shed important light into the contemporary debate. Leibniz challenges some central assumptions driving the controversy, and in so doing he opens conceptual space for a plausible and novel position that deserves a place in the contemporary discussion.

Leibniz’s views on this topic have received some attention in the secondary literature, but mostly to portray him as a traditional opponent of Molinism.24 This common classification is not without reason. Leibniz objects to robust libertarianism, one of the main purported theoretical benefits of Molinism,25 because it violates his much-championed principle of sufficient reason.26 It would be contrary to wisdom to demand a freedom that violates this principle, Leibniz insists, providing a criticism reminiscent of Dennett’s famous objection that libertarian freedom is not worth wanting.27 Furthermore, Leibniz also seems explicitly to dismiss a distinctive thesis of Molinism, namely the doctrine of middle knowledge.28 He writes, for example, “Thus we can see that, in order to account for the foreknowledge of God, one may dispense with both the middle knowledge of the

Molinists and the predetermination which a Bañez or an Alvarez (writers otherwise of great profundity) have taught.”29 This is not an isolated remark by Leibniz.30 Given all of this, it is no surprise that Leibniz is commonly classified as a foe of Molinism.

24 Bruntrup and Schneider 2013: 97, 99, Kaphagawani 1999: Ch 6, 7, Begby 2005: 84, Griffin 1999: 330-1, 2013: Ch. 6, 7, Greenberg 2005, Davidson 1996: 104-5, 2005: 238, 2011, and Kaphagawani 1999. There are couple of exceptions to this trend: Knebel (1996) and Anfray (2002) argue that Leibniz endorses a version of middle knowledge. 25 This already constitutes an important departure from the common way of framing the debate regarding Molinism. As we shall see below, Leibniz utilizes core Molinist tenets to defend a non-libertarian conception of the kind of contingency that is required for freedom, thus carving intelligible conceptual space for these tenets divorced from libertarianism. It is worth noting that some contemporary libertarians find Molinism objectionable precisely because they think the kind of libertarianism that is accommodated by Molinists is not robust enough: Hasker 2011, and Zimmerman 2009, for example. 26 Leibniz to Jaquelot, G 3.471-3/ WFI p. 180, T §§ 175, 199, 303, 349, Grua 271, 176-7, 280. 27 Dennett 1984. 28 I discuss this doctrine in more detail below. I also address the passages where Leibniz seems to distance himself from this doctrine. 29 T § 47. 30 Origin of Contingent Truths, A 6.4.1660-61 n.145, CD §§ 16, 17.

18

Yet, this common appearance is not entirely correct. It is my contention that closer inspection of

Leibniz’s views reveals that he has remarkable affinities with the Molinist tradition. In fact, I shall argue that

Leibniz himself endorses a version of two tenets essential to Molinism. These two substantive tenets of

Molinism are the following. First, the connection between the conditions for free actions and these free actions is a contingent one: free actions follow contingently from their sufficient conditions. Second, God’s knowledge of what creatures would freely do in different possible circumstances is prevolitional (i.e., prior to God actually willing anything). In other words, God’s knowledge of subjunctive conditionals of freedom is prior to or independent of God’s volitions.31 Because Leibniz endorses a version of each of these tenets, I conclude that far from being a foe, Leibniz is best characterized as a friend of Molinism.32

My plan is the following. In section one, I present the basic tenets of traditional Molinism – I use

Molina’s views as representative. I also contrast these views with Dominican Domingo Bañez, as representative of traditional Dominican views. In section two, I argue that Leibniz endorses a version of both Molinist tenets; in Leibniz’s hands, these tenets are significantly molded by his commitment to a strong version of the principle of sufficient reason. I illustrate how Leibniz’s version of these Molinists tenets allows him to provide a plausible account of providential control over creation. In doing so, it becomes clear that Leibniz’s views in this domain depart in several crucial respects from traditional Molinists views, like Molina’s, while still preserving the core of the two Molinist tenets. Thus, Leibniz’s views constitute a novel and plausible development in the history of the Molinism controversy.

31 For illuminating discussions on this second tenet of Molinism see: Murray 1995, 1996, 2004, and 2005; Davidson 1996, and 2005; and Freddoso 1988. It is worth noting that early in his career Leibniz explicitly denied this tenet; he does this in De Libertate, Fato, Gratia Dei et Connexis, Grua 306-322 and A 6.4:1595-1612. 32 I use the admittedly vague category of ‘friend’ to gesture towards the important similarities between Leibniz’s views and those of Molinism. In the main text below, I articulate those similarities more carefully, yet a more precise classification of Leibniz’s views, in this domain, requires careful attention to the different versions of Molinism, especially as they developed in the early part of the seventeenth century. I illustrate some of these connections in section 2.3, but I take the project of more precise classification of Leibniz’s views to be a slightly different project than the one I undertake in this chapter.

19

1. Molina and the Doctrine of Middle Knowledge

1.1 The Basic Tenets of Middle Knowledge

Before addressing Leibniz’s views, it is important to present the basic shape of the philosophical debates he saw himself as entering and in relation to which he defined his own views. Leibniz directly addresses

Molina’s views,33 so this is a good place to begin. As Molina sees it, the doctrine of middle knowledge is a doctrine about God’s knowledge of subjunctive conditionals of freedom. This knowledge is said to be ‘middle’ because, according to Molina, it lies between the two other kinds of knowledge that God is traditionally said to have. These two other kinds of divine knowledge are traditionally known as “God’s knowledge of simple intelligence” and “God’s knowledge of vision.”34 God’s knowledge of simple intelligence is the kind of knowledge that God has in virtue of knowing His own essence and what follows from His essence, whereas

God’s knowledge of vision is the kind of knowledge that God has in virtue of knowing His will.

On the one hand, it is traditionally thought that what follows from God’s essence are all and only necessary truths, and so that the objects of God’s knowledge of simple intelligence are all and only necessary truths.35 Furthermore, because these truths are settled prior to any of God’s actual decisions or volitions, God’s knowledge of them is called ‘prevolitional.’ Molina talks about these propositions being true ‘prior’ to God’s will. It is illuminating to state this priority using temporal language and thus think of these propositions as being true before God wills anything.36 As Molina sees it, however, this temporal language is only metaphorical, for Molina endorses the traditional doctrine that God is atemporal. The more accurate way of capturing the relevant priority is as ontological independence: these truths do not ontologically depend on God’s will. Rather,

God’s will is constrained by these truths which He knows prevolitionally.

33 DPG § 16, T §§ 38-47, Scientia Media, A 6.4.1373-4. 34 God’s knowledge of simple intelligence is also known as ‘God’s natural knowledge’ and God’s knowledge of vision is also known as ‘God’s free knowledge.’ Molina himself uses this terminology (Co IV.52.10) and it is commonly used in contemporary discussions. 35 ST Ia.14.9, SCG 1.66.8. 36 Co IV.52.10.

20

God’s prevolitional knowledge is best understood in contrast to His postvolitional knowledge. God has the latter kind of knowledge in virtue of knowing His own will; that is, God has this kind of knowledge by willing. In general, God has postvolitional knowledge that p by willing that p.37 This kind of knowledge is labeled ‘postvolitional’ because it is posterior to or dependent upon God’s will or volitions. It is again illuminating to use temporal language and say that God has postvolitional knowledge after willing. Though, again, this is merely a metaphorical way of taking about ontological dependence.

In sum, God’s knowledge of simple intelligence is said to be both prevolitional and necessary. God’s knowledge of vision, on the other hand, is the kind of knowledge that God has in virtue of knowing His own will or more precisely by willing. Traditionally, the objects of God’s knowledge of vision are all and only contingent truths.38 God’s knowledge of vision is thus both postvolitional and contingent.

God’s knowledge of subjunctive conditionals of freedom lies between these kinds of knowledge because it fits into neither category and yet it shares characteristics with both: it is prevolitional, like God’s knowledge of simple intelligence, and it is contingent, like God’s knowledge of vision.39 As Molina sees it, then, the basic tenets constitutive of the doctrine of middle knowledge are the following (note: these are my labels):

Contingency Condition: subjunctive conditionals of freedom are contingent.

Prevolitional Condition: God knows subjunctive conditionals of freedom prevolitionally (i.e., prior to God actually willing anything, or prior to any actual volition).40

Middle Position Condition: God’s knowledge of subjunctive conditionals of freedom is different from, and in some sense between, God’s knowledge of simple intelligence and vision.

The doctrine of middle knowledge gets its name from the Middle Position Condition, but, as already alluded to, I will argue that the philosophical substance of Molinism lies in the Contingency and Prevolitional conditions. In the next section, I will present Molina’s theoretical motivations for advancing these tenets.

37 Co IV.52.9, 13. 38 ST Ia.14.9, 12, SCG 1.66.8. 39 Co IV.52.9-10. 40 Co IV.52.9, IV.53.1.20.

21

1.2 Theoretical Motivations for the Three Tenets

As Molina sees it, the Contingency Condition shields freedom from the threat of causal necessitation.

This condition requires that the connection between the conditions for action, specified in the antecedents of subjunctive conditionals of freedom, and the free actions, specified in the consequents, is a contingent one.

Molina understands this kind of contingency in the following way: “a given future state of affairs is called contingent… because it rules out… the fatalistic and extrinsic necessity that results from the arrangement of causes.”41

As Molina sees it, the kind of contingency that matters for freedom requires lack of necessitation from causes and their arrangements – I shall call this kind of contingency ‘causal contingency.’ More precisely, a state of affairs S is causally contingent if and only if it is possible for S to obtain and possible for S not to obtain given all its causal conditions and their arrangements.42

The robustness of this kind of causal contingency depends on what should be included as conditions for free action. It is crucial for Molina’s understanding of causal contingency that the conditions for action include everything but the act of the will itself.43 As Molina sees it, freedom requires the absence of necessitation from causes external to the agent and from causes internal to the agent which are external to the will. Thus, Molina insists, a will is said to be free only if nothing external to it necessitates its acts; that is, freedom requires that given all conditions external to the will it is possible for the will to act and possible for the will not to act.

This kind of causal contingency matters for freedom, Molina insists, because it opens conceptual space for understanding the will as a self-determining faculty. Molina insists that the self-determining nature of the will is protected only if all the conditions for action are insufficient to causally determine which of the

41 Co IV.47.2/ AF pp. 86-7. 42 I will not further pursue the sense of ‘possible’ that is relevant for causal contingency. 43 Co I.2.3.

22 possibilities will be actualized – that is, only if the connection between the conditions for action and the act of the will is causally indeterministic. Rather, freedom requires that it is the will that causally determines which of the alternative possibilities gets actualized. A bit more precisely, the will is a self-determining faculty if and only if it is the will itself that determines which of the possibilities available to it (given all conditions for action excluding only the act of the will) is actualized. As Molina sees it, then, freedom requires a self-determining will, and the self-determining nature of the will requires causal indetermination and causal contingency. This is the core of the theoretical motivation for the Contingency Condition as understood by Molina.

The core of Molina’s theoretical motivation for the Prevolitional Condition, on the other hand, is the following. Molina thinks that it is not enough for freedom that the will is able to act or not to act given all the conditions for action. It must also be the case, Molina insists, that how the will would act in different possible circumstances is not settled by divine decree. In other words, Molina thinks that if subjunctive conditionals of freedom have the that they have due to divine decrees or volitions, then human freedom would be compromised. As Molina sees it, then, the Prevolitional Condition is essential for protecting human freedom because it prevents God’s will from having undue control over human actions. Human freedom is preserved only if subjunctive conditionals of freedom have their truth independently of God’s will.

The final tenet of middle knowledge is The Middle Position Condition. As far as I can tell, the only theoretical motivation Molina presents for this condition is that neither God’s knowledge of simple intelligence nor His knowledge of vision can contain His knowledge of subjunctive conditionals of freedom. And the reasons for this are the ones we have already discussed. Subjunctive conditionals of freedom must be contingent and God’s knowledge of them must be prevolitional, and so this kind of knowledge fits into neither traditional category of God’s knowledge. This fact about Molina’s theoretical motivation for advancing this third tenet clearly supports my contention that this tenet is not philosophically substantive. This will be quite clear in Leibniz’s own views. Before looking at Leibniz’s views, however, it is worth our while to see the doctrine of middle knowledge at work.

23

1.3 Middle Knowledge at Work

The doctrine of middle knowledge enables Molinists to reconcile libertarian freedom with several traditional theological doctrines. For example, it explicates how God can foreknow creatures’ future free actions. God foreknows what will freely do by knowing (via middle knowledge) what humans would freely do if they were placed in different circumstances, and by knowing (via knowledge of vision) in what circumstances He will place them. The integrity of human libertarian freedom is not threatened by God’s foreknowledge because humans are free to act or not to act in the causally indeterministic circumstances in which they find themselves in just the way that Molinist think is required for freedom.44 As we have already seen, this is guaranteed by the Contingency Condition.

There is an important complication to this basic picture, however. It depends on more than merely

God’s will whether an agent S will find herself in some circumstance C. Importantly, C includes free of other agents S1…Sn, and ex hypothesi God has no direct control over these free choices. This is, of course, no problem for the Molinist; it merely makes the Molinist explanation of divine foreknowledge more complex.

For God to know what S will freely do, God also has to know what agents S1…Sn will freely do, but God knows this as well. His middle knowledge also extends to agents S1…Sn. By knowing both what agents S1…Sn would freely do together with His knowledge of vision, God knows in what circumstances S will find herself; and, as we have seen, by knowing the latter together with His knowledge of vision, God knows what S will freely do.

And, of course, the same is true for God’s knowledge of the circumstances in which agents S1…Sn will find themselves, for these circumstances also include free choices of other agents, and so on. This makes God’s foreknowledge incredibly complex, but this is no problem for an infinite mind, so it is ultimately no problem for the Molinist’s account. In sum, for God to foreknow what a single agent will freely do requires knowing what many other agents will freely do, but the doctrine of middle knowledge provides God with the tools to do so while respecting the libertarian freedom of all agents involved.

44 Co IV.52.10, 29.

24

The doctrine of middle knowledge also enables Molinists to retain a robust account of divine providence. The general doctrine of divine providence is the doctrine that everything that happens in creation happens because it is part of God’s plan and is ultimately within His control.45 According to the Molinist, then,

God has control over everything that happens in creation by having control over the circumstances in which creatures will find themselves and by knowing what they would freely do in those circumstances. That is, God’s prevolitional knowledge of contingent subjunctive conditionals of freedom together with His control over which antecedents of these conditionals will be actualized suffices to provide a robust account of divine providence, Molina insists.46 We will revisit providential control in section 2.3.

1.4 A Contrast to Molinism: the Dominican Views

Before looking at Leibniz’s contribution to the controversies surrounding Molinism, it is helpful to present the basic shape of the views advocated by the main opponents of traditional Molinism, namely traditional Dominicans. I will use Domingo Bañez as representative of this alternative set of views.

Bañez postulates a kind of divine grace that causally necessitates human actions but which is nonetheless compatible with human freedom. According to Bañez, God has the power to provide divine premotions which are intrinsically efficacious – that is, premotions which always bring about their intended result (i.e., efficacious), and which do so only in virtue of what they are in themselves and not in virtue of what they are in relation to something else (i.e., intrinsic).47 These divine premotions are to be included in the conditions for action, and thus will ensure that an intended human action will come about. Bañez argues that

45 This is just the core of the doctrine of Providence. This core is typically developed into a more robust doctrine by different or theologians. Molina, for example, insists that God’s providence also provides the end, or that for the sake of which, everything in creation exists (Co V.1.1.). 46 Co V.4.10. It is important to note that this kind of providential control has limitations imposed on God by the truth values of subjunctive conditionals of freedom. If it is true that S would freely ɸ if in C, then it is not possible for God to create S in C and have S not freely ɸ. Plantinga (1974: Ch. IX, 1977) relies on this kind of limitation on God’s power to advance his free will defense. Flint (1998: Ch. 2) presents, as far as I can tell, the most detail articulation of the way in which middle knowledge limits providential control to make room for freedom. 47 B 1.9.1-3, 1.9.7, 1.12.2, 1.14.3.

25 in his infinite power God can necessitate a free human action even if it escapes human comprehension how this can be.48

According to his picture, then, God’s knowledge of subjunctive conditionals of freedom is included under God’s knowledge of vision. That is, God knows the truths of subjunctive conditionals of freedom by willing that they be so: these conditionals are true because God wills it. God’s knowledge of these conditionals is thus postvolitional: God foreknows what creatures would freely do if placed in different circumstances precisely by knowing whether or not He will provide or fail to provide necessitating premotions in those circumstances.49 Divine providence is thus quite robust: God has direct control over everything in creation, including what humans would freely do. God has the freedom and control over whom to grant divine aids, and

He also has the freedom and control over who would freely accept these graces. Furthermore, in His distribution of graces and execution of His divine plans, God invites no uncertainty and takes no risks precisely because God has the power to causally necessitate human free actions. God has infallible foreknowledge of how creatures will respond to divine graces by having power to provide graces that are impossible to resist.

God’s will to provide these necessitating premotions suffices to ground infallible foreknowledge and providential control over creation.50

This is the basic shape of the contrasting accounts of divine foreknowledge and providence advocated by Dominicans. Because the Dominican views make explicit use of divine graces to account for subjunctive conditionals of freedom, it is worth our while to see how the Molinists accommodated divine graces. Doing so would give us a fuller account of the kinds of debates that informed Leibniz’s views.

According to Molina, the legitimacy of human freedom requires that nothing necessitate free actions, and this, Molina insists, applies to divine graces as well.51 That is, as Molina sees it, if God’s graces necessitated

48 B 1.16.7. Bañez accuses Molina of relying too much on his own understanding regarding the divine. The mere fact that Molina cannot conceive how God can causally necessitate a free human action does not count against God having this power. It is impious to rely on one’s own conceptual powers in this way, Bañez insists. 49 B 1.14.3, 1.16.2. 50 For brief and illuminating discussions on Bañez’s views in the secondary literature see Murray 1995, 1996, 2004, and 2005, Freddoso 1988: “Introduction,” and Davidson 1996. 51 Co IV.52.10.

26 human actions, these actions would not be free. For the traditional Molinist, then, divine assistances or graces are not necessitating: it is possible for these graces to fail to bring about their intended results. Furthermore, and crucially for our purposes, whether this possibility is actualized depends on the way free agents utilize their freedom. Humans retain the freedom to refuse divine graces, Molina insists. God’s graces are to be included in the antecedents of contingent subjunctive conditionals of freedom much like other conditions for action, and these conditionals including divine graces are to be understood as retaining the same kind of causal contingency as those that do not include these graces in their antecedents.52

As Molina sees it, God’s graces are intrinsically sufficient (they lack nothing in themselves for bringing about the desired effect) but only extrinsically efficacious (they bring about the desired effect only together with the extrinsic fact of the creature’s free choice).53 It is possible for intrinsically sufficient graces to fail to bring about their intended result, and the actualization of this possibility is up to the Molinist self-determining will.

However, this modal fact does not undermine the efficacy of intrinsically sufficient graces, nor does it invite any risk or uncertainty in God’s plans or distribution of these graces. God knows with that things will turn out in accordance with His plans by knowing with certainty His own causal contributions to creatures and by knowing with certainty contingent subjunctive conditionals of freedom. When God distributes merely extrinsically efficacious graces, He knows with certainty that these graces will not fail because He knows with certainty via His middle knowledge that the possibility of these graces failing will not be actualized by the creature’s self-determining will. We can thus see how in the Molinist picture God can dispense intrinsically sufficient graces which respect the robustness of libertarian freedom while also being extrinsically efficacious and compatible with infallible divine foreknowledge and a robust account of divine providence.

This concludes my sketch of the debate between Jesuits and Dominicans that Leibniz entered and which informed his views on this topic. I turn next to Leibniz’s novel contributions.

52 Co IV.53.30. 53 Co III.40.4-5, IV.53.30.

27

2. Leibniz and Molinism

Leibniz addressed the controversy between Jesuits and Dominicans on multiple occasions. He even wrote an entire treatise, Dissertation on Predestination and Grace (DPG), with the goal of reconciling analogous versions of these views in the Protestant tradition: Calvinists (whose views are roughly analogous to Dominican views) and Arminians (whose views are roughly analogous to Jesuit views). Leibniz presents a very interesting position. On the one hand, Leibniz agrees with the Dominicans that God has the power to provide intrinsically efficacious graces which are nonetheless compatible with human freedom.54 This is an important concession, yet, Leibniz insists, against the Dominicans, that these intrinsically efficacious graces are not necessitating, otherwise they would preclude human freedom.55 Additionally, Leibniz also insists that God’s wisdom requires that He not rely on this kind of intrinsically efficacious graces unless it is necessary.56 As a result, and despite this important concession, Leibniz thinks that in the vast majority of cases, God dispenses only extrinsically efficacious graces. Furthermore, and crucially for our purposes, despite accepting that God has the power to dispense intrinsically efficacious grace, Leibniz does not rely on these kinds of graces to ground the truth of subjunctive conditionals of freedom in God’s will or volitions; that is, Leibniz refuses to construe God’s knowledge of subjunctive conditionals of freedom as postvolitional. Instead, Leibniz sides with Molinists in insisting that God’s knowledge of subjunctive conditionals of freedom is prevolitional. Because of this, Leibniz ends up siding with the Molinists on most of the disputes regarding divine foreknowledge and divine providence.57

54 DPG § 4. 55 DPG § 4. 56 DPG § 4. 57 As Michael Murray notes in DPG Introduction, p. xxxvii.

28

2.1 Leibniz on the Doctrine of Middle Knowledge

A main thesis of this chapter is that Leibniz is best characterized as a friend of Molinism. This thesis might first appear to be a surprising one, for Leibniz sometimes seems to distance himself from the doctrine of middle knowledge. Leibniz writes, for instance:

[God’s] knowledge of actual things, that is, of the world produced into and all past, present, and future states of the world, is called knowledge of vision. It differs from the knowledge of simple intelligence of this same world considered as merely possible only in that it contains, added to the latter, the reflexive knowledge whereby God knows his decree to produce it into actual existence. Nothing more is needed as a foundation for the divine foreknowledge.58

There are more passages in which Leibniz articulates the same point.59 Yet, it would be a mistake to read these passages as rejections of all tenets that constitute the doctrine of middle knowledge. As I see it, the best way of reading these passages is as merely rejecting Middle Position Condition – that God’s knowledge of subjunctive conditionals of freedom lies between His knowledge of simple intelligence and His knowledge of vision.

As we have seen, the Contingency and Prevolitional conditions constitute the philosophical substance of Molinism, and Molina himself seems to have endorsed The Middle Position Condition only because he thought it followed from the Contingency and Prevolitional conditions. Following tradition, Molina thought that God’s knowledge of simple intelligence included all and only necessary truths.60 Leibniz, however, breaks from this tradition and allows that God’s knowledge of simple intelligence includes contingent truths.61 Leibniz writes: “Knowledge of simple intelligence could be taken so as to include… not only the necessary connections, but also the contingent ones, that is, those which merely incline.”62

58 CD § 16/ S p. 116. Note: I have modified the translation a bit to better fit the terminology of this chapter. I translated ‘scientia’ as ‘knowledge’ instead of Schrecker’s ‘science.’ 59 CD § 17, A 6.4.1789, Grua 349, Origin of Contingent Truths A 6.4.1660-61 n.145. 60 Co IV.52.9-10. 61 Carriero (1995) makes the same point. He writes “Leibniz’s ‘compromise’, however, involves a quiet but important break with the tradition over God’s knowledge of simple intelligence. Leibniz includes contingent truths (e.g., ‘Judas will betray Christ’) under God’s knowledge of simple intelligence, whereas both the Molinist and their adversaries restricted such knowledge to necessary truths (e.g., ‘It is possible that Judas will betray Christ’).” (p. 6). 62 DPG § 16a.

29

Because Leibniz departs from this tradition he no longer has much philosophical motivation for endorsing the Middle Position Condition. As a result, he sometimes tries to ground God’s knowledge of subjunctive conditionals of freedom in God’s knowledge of simple intelligence. He says: “The knowledge commonly called middle knowledge is contained in the knowledge of simple intelligence.”63 There are many similar passages.64

At other , Leibniz is willing to accept something like the Middle Position Condition, but treats this issue as a merely terminological one. He suggests, for example:

If, however, one wants a knowledge midway between the knowledge of simple intelligence and the knowledge of vision…. one could assign to middle knowledge not only the knowledge of conditional future events but, generally, the knowledge of all contingent possibles. Thus knowledge of simple intelligence would be taken in a more restricted sense, namely, as dealing with possible and necessary truths, while the knowledge of vision would deal with contingent actual truths. Middle knowledge and knowledge of simple intelligence would have this in common, that they both deal with possible truths, while middle knowledge and the knowledge of vision would both deal with contingent truths.65 Here Leibniz is willing to accept that God’s knowledge of subjunctive conditionals is in some sense between God’s knowledge of simple intelligence and knowledge of vision. Importantly for our purposes,

Leibniz is treating endorsing the Middle Position Condition as something of mere terminological convenience.

Thus, it seems that for Leibniz this tenet is not a substantive one. Hence, it is most reasonable to treat the passages in which Leibniz seems to reject the doctrine of middle knowledge as merely rejecting that there is anything of philosophical substance in the Middle Position Condition. This rejection, of course, should not be treated as also rejecting either of the two substantive Molinist tenets.

2.2 Leibniz on the Substantive Molinist Tenets

In the previous subsection, I argued that the passages in which Leibniz seems to distance himself from the doctrine of middle knowledge are best understood as denying any philosophical substance to the Middle

63 CD § 17. 64 DPG §§ 16a, 49a, T § 42, A 6.4.1789. 65 CD § 17/ S pp. 116-7. Note: I have modified the translation a bit to better fit the terminology in this chapter. I translated ‘scientia’ as ‘knowledge’ and ‘media’ as ‘middle’ instead of Schrecker’s ‘science’ and ‘intermediate.’

30

Condition Position. In the present subsection, I will argue that Leibniz endorses a version of the two substantive tenets of Molinism, namely the Contingency and the Prevolitional conditions.

2.2.1 Leibniz on The Contingency Condition

The first substantive tenet of the doctrine of middle knowledge, The Contingency Condition, states that the connection between the conditions for action specified in the antecedent and the free action specified in the consequent is a contingent one. Leibniz explicitly endorses this tenet:

When we propose a choice to ourselves, for example, whether to leave or not, it is a question whether, with all the circumstances, internal or external, motives, , dispositions, impressions, passions, inclinations taken together, I am still in a state of contingency, or whether I am necessitated to take the choice to leave, for example, i.e., whether in fact this truth and determined – in all these circumstances taken together, I will choose to leave – is contingent or necessary. I reply that it is contingent… And assuming that by freedom of indifference we understand a freedom opposed to necessity (as I have just explained it), I agree with that freedom.66

Here Leibniz describes a conditional whose antecedent includes all conditions for a given free action and whose consequent describes this free action, and clearly states that he takes the connection between these to be a contingent one. Leibniz’s account of this kind of contingent connection, however, is significantly different from the accounts of traditional Molinists like Molina. As we have seen, it is essential to Molina’s understanding of freedom that it requires the connection between conditions for action and free actions to be causally indeterminate. Leibniz, on the other hand, thinks that such a radical contingency is problematic because it violates his strong version of the principle of sufficient reason. For the strictures of this principle to be satisfied it must be the case that the free actions specified in the consequents be explained by the conditions for action specified in the antecedents: these conditions for action must explain why the free action is the way it is and not otherwise. This, however, is compatible with free actions being contingent, Leibniz insists, because these conditions for action merely incline but do not necessitate these actions. Leibniz writes:

Freedom of the will consists in that wherein we act not only spontaneously but also in a deliberated way; nor are we necessitated to what we decide, but only inclined to it…. Now every effect is determined by its causes and their dispositions, such that there is always some reason why it exists rather than not,… the

66 Letter to Coste, On Human Freedom, G 3.403/ AG p. 194.

31

reasons that determine a free cause are never necessitating but only inclining, and to that extent the indifference or contingency in them is preserved.67 Leibniz’s basic philosophical here is that determination and necessitation come apart. That is, conditions for action fully explain, and thus determine, those free actions, but this determination need not amount to necessitation. What this doctrine of inclination without necessitation amounts to is itself a huge topic in Leibniz scholarship, and the details need not detain us here. It will suffice for our present purposes to adumbrate the core idea underlying what I take to be a plausible interpretation of Leibniz’s account of contingency as inclination without necessitation.68

The basic idea of my interpretation is based on Leibniz’s account of the will, which I think is best understood as returning to a more traditionally Thomistic account. As Leibniz sees it, the will of a rational agent is a rational inclination whose strength is proportionate to the apprehended goodness of the object of choice. Importantly, rational inclinations in the agent just are merely inclining reasons, and these rational inclinations are constitutive parts of what it is for an agent to be rational in a value-responsive sense. If the apprehended goodness is maximal, the choice of a rational agent is necessitated by the goodness of the object of choice, and if the apprehended goodness is not maximal the choice of a rational agent is only contingently inclined (i.e., not necessitated) by the goodness of the object of choice.

Crucially for our purposes, rational inclinations together with the apprehended goodness of the object of choice suffice to explain actions qua rational actions in a way that meets the strictures of the principle of sufficient reason. And this kind of explanation is one that does not require that the explanans necessitate the explanandum, for value-responsive does not require that rational action be necessitated by the apprehended goodness of the object of choice. The rational inclination of a rational agent together with the apprehended goodness of the object of choice determines rational action, qua rational action, but it does not necessitate it.

67 On God and Man, G 3.33/ LGR p. 297. 68 I dedicate chapter two of my dissertation to defending and elaborating this view.

32

The details of this account need not detain us here, what matters for our purposes is that Leibniz endorses a version of the Contingency Condition. Yet, Leibniz’s views depart from traditional Molinism in important ways, and this is manifested in Leibniz’s account of the contingent connection between conditions for action and free actions which meet the strictures of the principle of sufficient reason. Despite these differences in details, the crucial fact for our purposes is that Leibniz endorses a version of the Contingency

Condition.

2.2.2 Leibniz on the Prevolitional Condition

The second substantive Molinist tenet is the Prevolitional Condition. This tenet states that God knows subjunctive conditionals of freedom prior to His will being involved or prior to any volition. In this section I will argue that Leibniz endorsed a version of this tenet.

An important complication for Leibniz’s views in this domain, and another departure from traditional

Molinism, is that his strong version of the principle of sufficient reason requires that all effects can in principle be known and made intelligible on the basis of their causes. This applies to future contingents as well. Leibniz insists: “[It] is both most true and implied in our view… that even future contingents are known determinately from their causes.”69 This understanding of the principle of sufficient reason has important consequences for understanding God’s knowledge. As Leibniz sees it, it is not enough that God knows all truths, God must be in a position to understand why all truths are the way they are and not otherwise on the basis of their causes.

More directly relevant for our purposes, God’s knowledge of subjunctive conditionals of freedom requires that

God know the free actions specified in the consequents on the basis of the conditions for action specified in the antecedents.70 Leibniz writes, for example:

69 DPG § 16a. 70 This marks a crucial between Leibniz’s views and those of Francisco Suarez, another great defender of middle knowledge. Suarez relies on the principle of conditional excluded middle as sufficient grounds for God’s knowledge of subjunctive conditionals of freedom. This principle is the principle that for any pair of conditional proposition of the form “If it were the case that P, it would be the case that Q” and “If it were the case that P, it would be the case that not Q” one of these conditional propositions is true. See Suarez, Tractatus de gratia Dei, II, cap. VII, § 21, OO vol. 7, 94.

33

Yet there remains the difficulty urged by some against divine , namely how can God know what another mind will choose according to the pleasure of its own free will? ... [We can] solve the problem without difficulty, for since God foresees contingent things from his own free decrees, he will also know from those what the state of a free mind deliberating about some choice will be at any given time, i.e., how the arguments for each side will appear to it. Therefore he knows on which side of those presented the greater good or evil will be found, and hence what a mind will freely but certainly choose. From this it is also straightforwardly obvious how God knows what any free mind would choose if it were to find itself in any situation which nevertheless will not actually occur.71 Here Leibniz argues that God knows what a free agent would do on the basis of knowing how things would appear to this agent. How things would appear to an agent is part of the agent’s deliberation process, and thus part of the conditions for free choice.

It is crucial to note that Leibniz is not merely talking about what will in fact happen to a substance, but also about “any situation which nevertheless will not actually occur.” That is, Leibniz is not just talking about future contingents; he is also talking about subjunctive conditionals some of whose antecedents will not be realized. God knows what agents would freely do even in situations in which those agents are never going to be. As Leibniz sees it, God has knowledge of free actions on the basis of their conditions for action; or, in other words, God has knowledge of the truth of subjunctive conditionals of freedom by knowing how the free actions specified in the consequents are explained by the conditions for actions specified in the antecedents.

This understanding of the implications of Leibniz’s version of The Prevolitional Condition, on the basis of the strictures of the strong version of the principle of sufficient reason, also marks a significant departure from traditional Molinists views like Molina’s.

Leibniz utilizes his version of The Prevolitional Condition in much the same way as traditional Molinist do—namely, to account for fundamental theological commitments. For example, Leibniz uses this view to illustrate how God distributes the graces required for salvation. He notes:

Undoubtedly, it must be conceded that God foresees conditionally how someone would use his free choice, were certain aids are afforded; and relying on knowledge of that, along with knowledge of all others, He renders his decisions concerning the division of humanity with respect to salvation.72

71 Rationale of the Catholic Faith § 7, A 4.4.2319-20/ LGR pp. 74-5. 72 DPG § 9d.

34

This passage sounds as if it were written by a traditional Molinist. God’s decision to elect some people for salvation depends upon His knowledge of what people would freely do if afforded different kinds of divine graces or aids. In traditional Molinist fashion, in this passage it seems clear that God’s knowledge of subjunctive conditionals is prevolitional and so it limits God’s actual decisions. And God’s foreknowledge about salvation itself is postvolitional – the result of deciding what to do given the constraints provided by the prevolitional knowledge about what possible creatures would do. Hence, it is clear that Leibniz endorses a version of the

Prevolitional Condition and utilizes it to account for fundamental theological doctrines in much the same spirit as traditional Molinists.

2.3 Divine Freedom and Providential Control

So far, I have shown that Leibniz endorses a version of both substantive Molinist tenets, and so that he is best characterized as a friend, rather than a foe, of Molinism. What still needs to be established is that

Leibniz’s views in this domain constitute a novel and plausible advancement in the history of the Molinism controversy, rather than some minor variation. A good way of taking a firm step in this direction is by presenting how Leibniz’s views accommodate divine providential control over creation. In the process of articulating Leibniz’s account of providential control, crucial details of Leibniz’s version of the two substantive

Molinist tenets will come to the fore and their novelty and unique strengths will become more apparent.

At first sight, it may seem that Leibniz is committed to a very attenuated version of providential control.

This is so because Leibniz insists that God’s knowledge of a single complete individual concept provides God with enough to know everything that is ever going to happen to the corresponding substance, if created.73 Leibniz writes:

[S]trictly speaking it is neither the foreknowledge of God, nor his decisions, that determines the sequence of things, but the mere comprehension of possibles in the divine understanding; or the idea of this world, seen as a possibility prior to the decision to choose and create it. It is therefore the nature of things themselves which

73 DM § 9.

35

produces their sequence, prior to all decisions [of God]; God chooses only to actualize that sequence, the possibility of which he finds ready-made.74 Here Leibniz insists that God’s decisions do not determine the sequence of things, but rather, that God merely decides to actualize a “ready-made” possibility found in His intellect. This seems to deprive God of active control over anything that happens to any given possible substance beyond creating or not creating it as found in His intellect. This indeed seems like a very thin conception of providential control.

However, Leibniz also insists that God’s plans, resolutions and deliberations are essential to the very building of possible worlds. Leibniz insists:

So no human could fail to happen as it actually has happened, once the choice of [creating] Adam is made; but not so much because of the individual notion of Adam, although that notion involves it, but because of God’s plans, which also enter into that individual notion of Adam, and which determine that of the whole of this universe, and consequently both that of Adam and that of the other individual substances of this universe. For each individual substance expresses the whole universe of which it is a part according to a certain relation, thought the interconnectedness which exists between all things because of the links between God’s decisions and plans.75 In passages like this one, it seems quite clear that Leibniz wishes to endow God with a robust amount of active control over creation and the very building of the content of possible worlds. In fact, Leibniz seems to give

God’s plan some important explanatory priority. He insists that Adam’s complete individual concept is what it is, or contains what it contains, “because of God’s plans” which are part of its content.

Leibniz seems to be in an uncomfortable position here. On the one hand, Leibniz insists that complete individual concepts are in some sense already made in God’s intellect prior to God’s actual free decisions. On the other hand, Leibniz also insists that God actively plans and decides what goes into the actual world. In the quoted passage, for example, Leibniz says that it is “God’s decisions and plans” which are partly constitutive of Adam’s complete individual concept itself, and by implication the entire possible world which Adam mirrors.

Thus, Leibniz wants to say that God’s knowledge of the content of entire complete individual concepts is in some sense prevolitional, for it is simply ‘found’ in His intellect, yet there is also an important sense in which

74 Leibniz letter to Jaquelot, September 1704, G 6.558/ WFI p. 188. 75 Leibniz to Arnauld 14 July 1686, A 2.2.73-4/ WFII pp. 107-8.

36

God’s plans, decisions, and resolutions are constitutive of the very content of these complete individual concepts and have explanatory priority.

I believe that Leibniz can have it both ways. The key idea that enables him to do so is that included in possible worlds, and in complete individual concepts themselves, are God’s free decisions considered as possible, or possible free decrees.76 Put differently, the basic idea here is that in the very prevolitional construction of a possible world in God’s intellect, God’s will is not entirely absent. The construction of a possible world does take place prior to any of God’s actual decisions or decrees, but, crucially, this process includes God’s possible free involvement with the created world. Leibniz writes:

God’s free decisions, considered as possible, enter into the notion of the possible Adam, and the actualizing of these same decision is the cause of the actual Adam. I agree… that possibles are possible before any of God’s actual decisions, but not without sometimes presupposing those same decisions considered as possible. For the possibilities of individuals or of contingent truths involve in their notion the possibility of their causes, namely God’s free decisions.77 As Leibniz sees it, Adam’s complete individual concept already includes God’s free decisions considered as possible, or God’s possible free decrees. Furthermore, these possible free decrees are included in the conditions for human action, and thus are included in the antecedents of subjunctive conditionals of creaturely freedom.

God’s knowledge of these conditionals is thus prevolitional in the sense that they have their determinate truth value prior to God’s actual will being involved or prior to any of God’s actual volitions. However, God’s will in not entirely absent from the determination of these conditional truths, for God’s possible free involvement with these possible creatures is already included in the content of His prevolitional knowledge. That is, Leibniz only endorses the Prevolitional Condition with respect to God’s actual will, but not with respect to God’s possible will.

76 Anfray (2002) points out interesting similarities between Leibniz’s views on divine possible decrees and the views of early seventeenth century Jesuit Gabriel Vazquez. 77 Leibniz to Arnauld 14 July 1686, A 2.2.73/ WFII p. 107.

37

This mixed view provides Leibniz with some theoretical tools not available to Molina. Crucially for our purposes, on Leibniz’s view God can have prevolitional knowledge of subjunctive conditionals of freedom about Himself (i.e., “if God were in C, God would freely ɸ”). Leibniz notes:

God knows future absolute things because he knows what he has decreed, and futura conditionata [subjunctive conditionals] because he knows what he would have decreed. And he knows what he would have decreed because he knows what would have been best in that particular case, and he would have decreed the best.78 Leibniz does not utilize our terminology here, but I think that the most natural reading of this passage is that

God has postvolitional knowledge of future contingents, and prevolitional knowledge of subjunctive conditionals about Himself.79 If so, Leibniz’s view allows that the antecedents of subjunctive conditionals of human freedom include what God would do, and not merely what God could do.80 In other words, Leibniz has the theoretical tools to allow for two kinds of possible divine decrees to be included in complete individual concepts: a) those representing what God could freely do in circumstance C; and b) those representing what

God would freely do in circumstance C. Allowing possible divine decrees of kind (b) as ingredients in the very building of complete individual concepts, and thus also possible worlds, enables Leibniz’s God to have a significant amount of control in the very prevolitional building of possible worlds. This kind of control in the process of building possible worlds is analogous to the kind of active control that the God of traditional

Molinists has in arranging creation.

78 Scientia Media, A 6.4.1374/ CP 133. 79 Griffin (2013: Ch. 7) provides a different reading of this passage. Griffin thinks that Leibniz is saying that God knows what He would do by willing it, and so that God’s knowledge of subjunctive conditionals about Himself is postvolitional. 80 Molina can accommodate the latter (Co IV.52.31.), but not the former. Molina insists that God’s knowledge regarding His own will is postvolitional (Co IV.52.13.). Molina worries that God’s freedom would be compromised if He knew what He would do in a particular set of circumstances prior to actually willing anything. Such infallible prevolitional knowledge about Himself would render His decisions necessary and thus undermine His freedom, Molina insists (Co IV.53.19.). This claim does not introduce a double standard for human and divine freedom, however. The worry is a general one. Any agent who knows prevolitionally, and with certainty, what she would do in a particular set of circumstances is thereby deprived of the contingency of her actions and thus of her freedom (Co IV.52.12, IV.53.19.). Humans cannot, in principle, have this kind of knowledge about themselves (Co IV.52.11.), however, so middle knowledge places no legitimate threat to human freedom. It is important to note that Suarez, another great defender of middle knowledge, disagrees with Molina here. As Suarez sees it, the principle of conditional excluded middle (see footnote 51) suffices for God to have prevolitional knowledge about Himself as well – otherwise it would constitute a limitation of His omniscience, Suarez insists. See Suarez, Tractatus de gratia Dei, II, cap. VII, §§ 20-21, cap. VIII, § 5, OO vol. 7, 93-4, 98. Suarez’s views thus mark a position midway those of Molina and Leibniz.

38

To better appreciate this, consider the following comparison. As the traditional Molinist sees it, prior to creation God had prevolitional knowledge of a set of essences of possible creatures and a set of conditional truths about them (subjunctive conditionals of creaturely freedom). God has a significant amount of control, for He is free to create any subset of essences, and place them in any circumstances He wishes, depending on

His plans for them. Furthermore, a single essence is compossible with an infinite number of ways in which it can come together with other essences in different circumstances. In actively deciding which essences to actualize, and in what circumstances to place them, God is actively building the entire actual world in accordance with His plans.81

At first glance, Leibniz’s God seems to face a significantly different creation situation. However, an analogous process of building a possible world takes place prevolitionally in God’s intellect in Leibniz’s picture.

By including possible divine decrees of type (b) in the very prevolitional process of building a complete possible world, Leibniz’s God has a kind of prevolitional control over the very content of complete individual concepts themselves, for this process includes what God would freely do. That is, the very content of complete individual concepts themselves depends partly on God’s possible will (including, importantly, what God would do). God decides, or wills, which complete possible world to create, but importantly, the very content of possible worlds depends on God’s possible free decrees of both type (a) and (b).

The best of all possible worlds, as found in God’s intellect, is best partly because it includes prevolitionally what God would freely do, or possible divine decrees of type (b). God wills to create this world because it is best, but possible divine decrees of type (b) are partly constitutive of the best possible world being the best. Because the very content of the best possible world depends on God’s possible will, God has a kind of control in arranging creation. The kind of control that Leibniz’s God has in arranging creation, via His possible free decrees of type (b), is analogous to the kind of control that the traditional Molinists’ God has, via

His actual will. Divine freedom as control over the content of creation seems preserved to similar extents on both views, though in importantly different respects.

81 Flint (1998: Ch. 2) presents a precise and detailed account of this basic picture.

39

Conclusion

There is again a live controversy surrounding the plausibility of Molinism. In contemporary discussions, Leibniz is commonly classified as a foe of Molinism. Leibniz’s explicit rejection of the main purported theoretical benefit of Molinism, namely a robust libertarian account of freedom, together with some remarks in which Leibniz seems to distance himself from the doctrine of middle knowledge, seem to provide ample support to this common interpretation of Leibniz. I have argued, however, that closer inspection

Leibniz’s texts reveal that his views are much closer to Molinism than is commonly acknowledged in the secondary literature, for Leibniz endorses a version of both substantive Molinist tenets – namely, the

Contingency and Prevolitional conditions. Leibniz’s version of these tenets is significantly molded by his commitment to a strong version of the principle of sufficient reason, and, as such, is not to be simply assimilated to traditional Molinism. Instead, Leibniz’s views constitute a novel and plausible development which deserve a place in contemporary discussions about the plausibility of Molinism.

40

Chapter 2: Real Options, Agential Contingency, and Inclining Reasons

INTRODUCTION

Jorge Luis Borges memorably described time as a garden of forking paths.82 This is simple and intuitive: the series of events of the past up until the present are represented by a single line that ‘forks’ into multiple lines representing multiple possible futures. This imagery has been used by several philosophers to motivate an insight central to freedom.83 The insight is that freedom requires possible options, or alternatives, from which the agent can choose. Furthermore, these possible options must be open to the agent in such a way that it is up to the agent which of them is actualized. Returning to Borges’ imagery, the insight regarding freedom is that which of the forking paths, representing possible futures, becomes present is within the control of free agents – that is central to what freedom is, it seems.

Like many other philosophers, Leibniz wishes to incorporate this insight into his account of freedom.

Regarding God’s freedom, Leibniz writes: “For God chooses among the possibles, and for that very reason he chooses freely, and is not compelled; there would be neither choice nor freedom if there were but one course possible.”84 God is not in time, so His free choices do not fit as neatly into the picture suggested by Borges’s imagery, but the underlying insight regarding freedom is indeed advanced by Leibniz: God’s freedom requires options from which to choose. This is also true of human freedom. Leibniz insists: “When there are several

82 Borges 1944. 83 Kane 2007: 5-7, Fischer 1994: v, 3-6. 84 T § 235.

41 paths, one has the freedom to choose…. But if one found oneself in a narrow street, between two high walls, there would only be one possible path, and this represents necessity. By this we see that… freedom… [requires] the faculty of choosing among several possibles.”85 A condition for freedom, both human and divine, is, then, that there be possible options or alternatives from which the agent can choose.

The crucial question for Leibniz, and for any wishing to incorporate this insight on freedom, is about the correct sense of ‘possible’ according to which these options, or paths, are possible for the agent in a way that matters for freedom. Different philosophers try to answer this question by introducing more precise discussions of a power to do otherwise, or a power to initiate or not initiate causal-chains leading to actions, or a two-way power with respect to the agent’s volitions, or agent-causation, or contingency or lack of necessitation as a condition for freedom, to name a few.

In this chapter I will begin by labeling the correct sense of ‘possible options’ that is required for an adequate account of freedom. I wish to distinguish between a) the options that are possible for an agent in the correct sense of ‘possible option’ that is required from freedom, and b) the options which are possible for an agent in some other sense of ‘possible option.’ I will use the expression ‘real options’ to refer to options of type (a), and not of type (b). As I use the term, to say that a course of action is an option for an agent is to be neutral about whether the agent will take this course of action or whether this course of action will remain an unactualized possibility. That is, real options include courses of action that will be taken by the agent and courses of action that are merely possible for the agent to take, in the sense that maters for freedom, but that will remain untaken. With this terminology we can now ask: what is it for a possible option to be a real option?

Or what is the right criterion to distinguish mere possible options from real options? In this chapter I will identify a conception of alternative possibilities in Leibniz’s texts that is a plausible candidate for being the kind of possible options that matter for freedom – i.e., real options. The main question I will be pursuing in this chapter is what kind of possible options are real options, according to Leibniz.

85 Leibniz to Gerhard Wolter Molanus, A 1.17.611.

42

I will argue that a plausible account of real options can be provided on the basis of what Robert Adams has labeled ‘the reality of choice’ strategy to ground contingency.86 The basic idea is that the contingency of a choice is preserved by the fact that not-chosen possible options open to the agent are not rendered impossible by this choice.87 I will develop this basic idea into what I take to be credible candidates for real options in

Leibniz’s system. An important further theoretical benefit of the account I will be developing is that it would make sense of Leibniz’s otherwise enigmatic doctrine that free actions are contingent because they are brought about by merely inclining and not necessitating reasons – which I shall label Leibniz’s ‘merely inclining reasons’ doctrine.88 That is, I will argue that the reality of choice strategy makes intelligible Leibniz’s merely inclining reasons doctrine. I will conclude that these strategies together make intelligible the kind of contingency that

Leibniz holds between conditions for actions and free actions; in other words, I shall argue that these strategies make intelligible Leibniz’s version of the Contingency Condition – a substantive Molinist tenet that we encountered in the previous chapter.

The plan is the following. In section 1, I will argue that merely metaphysically possible options are insufficient to count as real options, for Leibniz. In section 2, I will present an interpretation of the reality of choice strategy that makes conceptual space for a plausible sense of real options in Leibniz’s system. Finally, in section 3, I will argue that the advocated account of the reality of choice makes intelligible Leibniz’s merely inclining reasons doctrine. The basic philosophical move, in this final section, will be to identify a kind of adequate explanation of rational action that requires real options and that also meets the constraints of Leibniz’s strong version of the principle of sufficient reason.

86 Adams 1994: 20. 87 Adams 1994: 20-3. 88 As far as I can tell, both of these strategies to ground contingency in Leibniz’s system have received little attention in the secondary literature. Torralba (2005) utilizes contemporary discussions on philosophy of action to elaborate the merely inclining reasons doctrine. Jorati (2017a: 123-32) also discusses this strategy. Murray (2004) makes explicit use of this doctrine, but he sees it as merely a precursor to Leibniz’s account of moral necessity, which is the topic Murray himself is investigating. Furthermore, Adams (1994) and Jorati (2017a: 123-32) briefly discuss the reality of choice strategy, and this doctrine receives fuller treatment in Lagerlund and Myrdal 2006.

43

Section 1: Metaphysically Possible Options as Real Options

A common interpretation of Leibniz has it that the only kind of contingency that is required for freedom is metaphysical contingency89 – something is metaphysically contingent if it is not metaphysically necessary, and something is metaphysically necessary if its opposite implies a .90 In our terminology, then, it is a common interpretation of Leibniz that all it takes for possible options to be real options is for these options not to be metaphysically necessary.

This interpretation is common for a reason. Leibniz explicitly says, and on several occasions, that freedom requires lack of logical or . Leibniz writes, for example:

I have shown that freedom, according to the definition required in the schools of theology, consists in intelligence, which involves a clear knowledge of the object of deliberation, in spontaneity, whereby we determine, and in contingency, that is, in the exclusion of logical or metaphysical necessity. (T § 288)

This is an illuminating passage. In it, Leibniz cites the three conditions of freedom – intelligence, spontaneity, and contingency – as he is fond of doing,91 and he also appears to give a gloss of the basic meaning of each of these conditions. Importantly for our purposes, the gloss on contingency only makes reference to a lack of logical or metaphysical necessity. What the passage explicitly says is only that metaphysical necessity rules out freedom, but it does not seem unreasonable to conclude also that, as far as Leibniz is concerned, the only kind of contingency that matters for freedom is metaphysical contingency. Thus, it is not unreasonable to conclude, from passages like this one, that, in Leibniz’s system, metaphysically possible options count as real options.

I take it that few philosophers who endorse the insight on freedom with which we began – that freedom requires options from which to choose – would find this Leibnizian response attractive.92 It seems quite intuitive that the mere fact that an option is metaphysically possible is not enough to make this option a credible

89 See for example, Blumenfeld 1988, Burms 1978, Frankel 1984 and McNamara 1990. 90 I argue in chapter 3 that Leibniz is entitled to this kind of contingency. 91 On God and Man, G 3.36/ LGR 297, A 6.4.1407/ SLT p. 92, Letter to Coste, On Human Freedom, G 3.393, T 34, 45, 65-67. 92 Molina, for example, motivates his own account of the kind of contingency that matters for freedom by contrasting it with mere metaphysical contingency (Co 4.47.2). Molina assumes that the reader will immediately recognize metaphysical contingency as insufficient for freedom, and relies on this presumed response by the reader to motivate his own preferred account of the kind of contingency that matters for freedom.

44 candidate for a real option – i.e., the kind of possible option that matters from freedom. Fortunately, I think that we do not need to read Leibniz as claiming that metaphysically possible options are real options. There is a more promising reading of passages like the quoted one, I think.

As I see it, part of the temptation for reading these sorts of passages as advancing metaphysical contingency as a condition for freedom includes understanding these three conditions as independent conditions that happen to come together in the case of freedom. That is, part of the temptation is to think that these conditions come apart and can occur independently of each other, and freedom is just the special case in which they do come together. Leibniz, however, has a more unified account of how these three conditions come together in freedom. Two sentences after the quoted passage, Leibniz continues: “The free substance is self-determining and that according to the motive of the good perceived by the understanding, which inclines it without compelling it: and all the conditions of freedom are comprised in these few words.”93 In this passage,

Leibniz does not appear to present three independent conditions, but rather a single or unified account that has three aspects or elements. Importantly for our current purposes, then, we should understand the kind of contingency that matters for freedom not merely as metaphysical contingency as such, but rather as metaphysical contingency further constrained by the other conditions on freedom. This combination of the three conditions gives rise to a more restricted and intelligible kind of contingency that matters for freedom.

I will provide more details of this unified account of the three conditions in the following section. A general description will do for now. One of the conditions of freedom is that of spontaneity, or self- determination. Jorati (2017a) has persuasively argued that the kind of spontaneity that matters for freedom, on

Leibniz’s view, is self-determination on the basis of the apprehended goodness of the object of choice – which she labels ‘rational spontaneity.’94 More precisely, rational spontaneity is the kind of spontaneity whereby an agent determines herself to a course of action on the basis of her deliberate judgment of the best. This

93 T § 288. 94 Jorati (2017a: 52). I am running together two categories that Jorati separates. She defines ‘rational spontaneity’ as the kind of self-determination which is exempt from non-rational influences. The other category she labels ‘rational .’ She defines ‘rational teleology’ as the kind of end-directedness, or teleology, that is based on the agent’s explicit judgment of the best (p. 75). As she sees it, then, both rational spontaneity and rational teleology come together, and “correspond” (p. 75), in voluntary or free actions.

45 characterization is important for our purposes, for it clearly illustrates how the two conditions of intelligence and spontaneity come together in a single characterization. The account is not merely that freedom is the special case in which spontaneity and intelligence happen to come together; but rather the stronger claim that what it is for the kind of spontaneity to be the kind of spontaneity that matters for freedom essentially involves reference to intelligence.95 My aim in this chapter is to take this insight and extend it to the case of contingency as a condition for freedom. More precisely, my aim in this chapter is to elaborate on this general account by making it explicit how contingency is also interconnected with the other two conditions on freedom. In brief, the kind of contingency that matters for freedom is the kind of contingency that is required for rational spontaneity – that is, the kind of contingency that matters for freedom is the kind of contingency that is required for an agent to determine herself on the basis of her deliberate judgment of the best.

My general proposal is the following. The basic thought is simply that the agent must have the power to bring about more than one course of action – this is at the heart of the insight with which we began this chapter. Importantly, the kind of power that matters for freedom, on Leibniz’s views, is the power the agent has to determine herself on the basis of her deliberated judgment of the best. That is, the power that matters for freedom is the power associated with rational spontaneity. My suggestion is that the agent must have this kind of power with respect to more than the actual course of action; generally speaking, to have real options is for there to be options that the agent has this kind of power to bring about. In other words, the kind of

‘possibility’ that matters for possible options is grounded in the power the agent has to determine herself on the basis of her deliberate judgments about the best. As we will see in the following section, this power is ultimately grounded in the agent’s will – as a rational inclination towards apprehended goodness.

This general sketch is, of course, significantly more detailed than what can be directly read from the quoted passages from T § 288. My aim, at this stage, is not to argue that all of this can be gathered from these sorts of passages: it certainly cannot. My aim, for now, is simply the weaker one of insisting that passages like

95 Jorati herself relies on this and insists rational spontaneity counts as a different kind of spontaneity, in Leibniz’s system, precisely because of this essential reference to deliberate judgments of the best. She identifies two other kinds of spontaneity which she labels ‘metaphysical’ and ‘agent’ spontaneity (2017a: Ch. 2). The details of what make these kinds of spontaneity different kinds of spontaneity need not detain us here.

46 those of T § 288 need not demand that we understand Leibniz as thinking that merely metaphysically possible options count as real options. If these passages are understood as articulating a unified account of the three conditions on freedom, the kind of contingency that emerges from them is more restricted than mere metaphysical contingency as such.

Furthermore, and importantly, the sketched account does enjoy a significant amount of textual support, as we will see in more detail in the next section. For now, here are a couple of passages that fit the general contours of my basic proposal quite well – and thus which conflict with understanding merely metaphysically possible options as real options. In these passages, Leibniz explicitly advances a kind of power to do otherwise grounded in the agent’s will as a condition for freedom. In the NE, for example, Leibniz writes:

Ph. We find in ourselves the power to begin or not to begin, to continue or to end many actions of our and many of our body, and this simply by a thought or choice of our mind, which determines and commands, so to speak, that such a particular action be done or not done. This power we call Will. The actual use of this power is called Volition; the cessation or production of the action which follows such a command of the soul, is called voluntary, and all actions done without such direction of the soul is called involuntary.

Th. I find all that very good and just. However, to speak more fairly, and to go perhaps a little farther, I will say that volition is the effort to tendency () toward what is considered good and against that considered bad. (NE § 5) In this passage Philalethes, the character representing Locke’s views, clearly states that the will is a two-way power over one’s actions – “the power to begin or not to begin, to continue or to end” many actions.

Theophilus, the character that represents Leibniz’s own views, goes on to straightforwardly accept this characterization as “very good and just” before adding that the will by its very nature is oriented towards apprehended goodness. This passage clearly presents, at least in general terms, all the ingredients of the account

I have advanced: the will is a rational inclination towards apprehended goodness, and the will is or at least grounds the agent’s powers over several options from which the agent can choose.

There are other similar passages. In DM, for example, Leibniz writes:

For speaking absolutely, our will is in a state of indifference, in so far as indifference is opposed to necessity, and it has the power to do otherwise, or to suspend its action altogether, both alternatives being and remaining possible. (DM § 30, AG 61)

47

This passage also clearly advances a power to do otherwise, grounded in the agent’s will, as a condition for freedom. In our terminology, it seems from these sorts of passages that, for Leibniz, real options are those options that an agent has the power to bring about by determining herself on the basis of her deliberate judgment of the best. This is a more plausible sense of real options than merely metaphysically possible options as such. It is the task of the following section to elaborate on this general account.

Section 2: The Reality of Choice and Real Options

Essential to my proposal about what kinds of possible options count as real options, in Leibniz’s system, is the strategy that Adams calls ‘the reality of choice’ strategy.96 In this section I will present Adams’s version of this strategy and will argue that, as stated, it does not advance a more plausible sense of real options than that presented by mere metaphysical contingency. I will conclude this section by modifying Adams’s proposal in a way that I think makes the modified proposal a more plausible candidate for real options in

Leibniz’s system.

As Adams sees it, the reality of choice strategy involves two crucial ingredients. First, choice, both human and divine, is itself explanatorily relevant. Given that choice is an act of the will, and that the will is a rational inclination towards the apprehended goodness of the object of choice, the fact that choice is explanatorily relevant enables Leibniz to include teleological explanations as legitimate explanations regarding reality. More specifically, the divine choice to create the best is a crucial part of the explanation of why created reality is the way it is, and so goodness and teleology are crucial parts of the explanation of why reality is the way it is and not otherwise. This first ingredient in the reality of choice strategy enables Leibniz to distance his views from the ‘blind’ necessitarianism of Spinoza – a necessitarianism that is not guided by any goodness in creation nor by the wisdom of the Creator responding to this goodness. This insistence on the explanatory role of divine choice in creation does indeed enable Leibniz to distance his views from those of Spinoza, but

96 Adams 1994: 20-4.

48 only by rejecting the blindness aspect of Spinoza’s blind necessitarianism. The metaphysical necessitarianism part of Spinoza’s views remains untouched by this ingredient in the reality of choice strategy. As Adams himself points out, if this is all Leibniz has to offer in defense of contingency in his system, it seems reasonable to conclude that Leibniz’s views are just as necessitarian as Spinoza’s.97

As I see it, it is the second ingredient of the reality of choice strategy that can help Leibniz retain a more plausible sense of real options in his system. According to this second ingredient, a choice counts as contingent because the options not selected are not rendered impossible by this choice. Leibniz writes, for example, “The decree to create is free: God is prompted to all good; the good, and even the best, inclines him to act; but does not compel him, for his choice creates no impossibility in that which is distinct from the best; it cause no implication of contradiction in that which God refrains from doing.”98 More precisely, then, as

Leibniz sees it, choice is the result of practical deliberation, which essentially involves weighing the goodness of multiple options that are in some sense open to the agent, and, importantly, the options not taken remain possible even though unchosen. As stated, this second ingredient of the reality of choice strategy seems initially promising. However, the same crucial question remains, what is the sense of ‘possible’ according to which these not-taken options are possible, and remain possible after the choice? We need an answer to this question to assess whether this strategy enables Leibniz to have a more plausible sense of real options.

Robert Adams, and other scholars, think that what Leibniz has in mind here is his notorious doctrine of per se possibility, or possibility in itself.99 According to this reading, then, the alternatives not chosen are possible in themselves prior to the choice to bring about some other alternative, and remain so after this choice.

More precisely, the choice C1 to bring about option O1 does not render a distinct option O2 impossible in itself, neither before nor after C1 is made; this is so, intuitively, precisely because the kind of possibility in itself that

O2 enjoys is due to its intrinsic qualities alone, and not extrinsic properties, or , like being or not being chosen. Leibniz himself characterizes possibility in itself in the following way: “we have defined as in its nature

97 Adams 1994: 21. 98 T § 230. 99 Adams 1994: Ch. 1 § 1, Langerlund and Myrdal 2006.

49 possible anything that, in itself, implies no contradiction”.100 In Leibniz’s mind, then, per se modality is closely connected with metaphysical modality. Presumably what it is for something to imply a contradiction ‘in itself’ is for this something to imply a contradiction in virtue of its intrinsic qualities alone – that is, blocking or ignoring extrinsic properties or relations. Adams thinks, for example, that at least some extrinsic properties, like being worthy of divine choice, must be blocked if one is to adequately assess whether something implies a contradiction in itself, and thus ultimately whether this something is possible in itself.101

Regardless of the precise details required for assessing whether something implies a contradiction in itself, and thus whether this something counts as possible in itself, the close connection between metaphysical modality and per se modality seems to render per se possible options no more plausible candidates for real options than merely metaphysically possible options.102 If so, the second ingredient of the reality of choice strategy is unsuccessful for our purposes. More precisely, if possibility in itself is the best way of understanding the kind of possibility that is had and retained by not-chosen alternatives, then the second ingredient of the reality of choice strategy fails to advance a more plausible candidate for real options in Leibniz’s system. I think, however, that this is not the case. That is, I think that there is a more promising way of filling in the sense of ‘possible option’ that is required by the second ingredient of the reality of choice strategy.

I will argue that part of what makes a possible option a plausible candidate for being a real option is the modality governing whether the agent can choose the option. More precisely, part of the plausibility of real options depends on whether an option is possible not only in itself, but, crucially, whether the option is possible relative to the agent – i.e., whether the option can be chosen by the agent. I will argue that there is conceptual space in Leibniz’s system for precisely this kind of modality governing what an agent can choose. The first main step in my proposal is to note that there is nothing essential to the agent or essential to possible options that makes it impossible for the agent not to choose these options. What rules out the agent choosing a given

100 On Freedom and Possibility, Grua 277/ AG 21. 101 Adams 1994: 15. 102 In fact, depending on the details of what it is to imply a contradiction in itself, an option being possible in itself might be even less plausible candidate for real options than merely metaphysically possible options. This is would be the case, for example, if it turns out that an option that is possible in itself is also metaphysically impossible all things considered.

50 option O1 is something nonessential to both the agent and O1 – namely the existence of another option On that the agent deems better than O1 in a given case of rational deliberation.

I think that this fact, about what rules out an agent choosing an option, matters because it opens up conceptual space for an intelligible sense of the modality governing what an agent can choose that, I will argue, can be utilized to develop a plausible account of real options in Leibniz’s system. The basic idea behind the modality I have in mind is the following: because there is nothing essential to option O1 or to agent S that rules out S choosing O1, there is a deliberation situation D1 in which S comes to the conclusion that O1 is the best option under consideration, and thus D1 results in S choosing O1. And so, there is a plausible sense of it being possible for S to choose O1. If the deliberation D1 is sound, it would require that there are no alternatives better than O1 under consideration, but this is of course no obstacle to the proposal precisely because there is nothing essential to O1 or to S that requires that S take into consideration another option On that is better than O1 in the process of deliberation. I will develop this basic thought in the next subsection.

2.1 Real Options and Agential Modality

I shall label the kind of modality governing what an agent can choose ‘agential modality.’ In this section

I will develop a basic sense of agential modality that I think will make conceptual space for plausible candidates for real options in Leibniz’s system. As already stated, the basic idea behind agential modality is that whether it is possible (or impossible, or necessary) for an agent S to choose an option O1 depends upon the essential qualities of both S and O1. As I see it, the crucial essential properties are the following: a) the (apprehended) goodness of a given option; and b) the rationality of the agent. I will briefly sketch what these essential qualities amount to, for Leibniz, and on the basis of these I will articulate an intelligible sense of what it is for an option to be open to an agent. Finally, I will rely on this definition of what it is for an option to be open to agent to articulate agential modality and real options.

51

Leibniz’s conception of what it is for an agent to be rational is, at least in general terms, a traditional

Thomistic one. Leibniz accepts a faculty psychology according to which the rationality of an agent consists in the agent possessing the faculty of the intellect, as a capacity for apprehending truths and deliberating about what to do, and the faculty of the will, as a rational inclination towards the (apprehended) goodness of the object of choice.103 In the case, the intellect of an ideally rational agent recognizes, or apprehends, the correct degree of goodness of a given option, and the strength of the rational inclination is proportionate to the (apprehended) degree of goodness in the possible object of choice, or possible option. Unfortunately, not all agents are ideally rational agents; rather, agents can fail to be ideally rational in several respects: their intellect can fail to apprehend correctly the degree of goodness in a given option; or the strength of the will can fail to be proportionate to the degree of apprehended goodness in the possible object of choice; or the passions of the agent can get in the way of the rational elements culminating in action. It is important to note these failures in rationality to adequately describe human agents, but, as Leibniz sees it, an adequate description of perfectly free actions requires a description of a fully rational agent.104 The more a given agent falls short of this ideal description the less free this agent is, and conversely the closer to the ideal not only the more rational but the freer the agent is. Leibniz puts this point about degrees of freedom thus: “Therefore true freedom of the mind consists in recognizing and choosing the best; and as we established above, that one is freer the more knowledge one has…; and the more one is accustomed to following reason, the more determined one will be to carry out what one judges the most reasonable.”105

With this basic sketch of the nature of rationality, here is my basic proposal for what it is for an option to be open to an agent. The basic idea is that what it is for an option to be open to an agent is for this option to be the sort of thing that can be taken into consideration by the agent in the process of deliberating about what to do. As Leibniz sees it, an essential aspect of the process of deliberation is the intellectual weighing of the apprehended goodness of each of the alternatives open to the agent. He writes, for example: “A measure

103 T §§ 22, 33, 311, 327. 104 DPG § 35b, A 6.4.1406-9/ SLT 92-3, 105 A 6.4.1409/ SLT 93.

52 of freedom is necessary for punishments and rewards, and this is why there is an intellect that compares and weighs against each other and also a faculty of inclining and willing in accord with one’s deliberations”.106 This intellectual aspect of the process of deliberation culminates in a final judgment by the intellect regarding which considered course of action is best. The process of deliberation, however, involves more than a merely intellectual aspect of recognizing and weighing the apprehended goodness of the considered options; as Leibniz sees it, the process of deliberation also essentially involves an appetitive aspect that he describes as a struggle of rational inclinations, or appetitions. In this context, Leibniz makes a distinction between antecedent and consequent wills. An antecedent will is just a rational inclination towards the apprehended goodness of an option under consideration in the process of deliberation. As Leibniz sees it, the appetitive aspect of the process of deliberation essentially involves the coming together of several antecedent wills in a kind of struggle that results in the consequent will – the one that settles the action for the agent.

Leibniz describes this appetitive aspect of the process of deliberation and the resulting consequent will thus:

“Now this consequent will, final and decisive, results from the conflict of all the antecedent wills, of those which tend towards good, even as of those which repel evil; and from the concurrence of all these particular wills comes the total will.”107 The process of deliberation, then, includes both an intellectual and an appetitive aspect, for Leibniz. In the cases of rational action, at the end of rational deliberation the agent’s intellect issues a final judgment regarding which course of action is best, and this judgment is followed by a consequent will to bring about the course of action judged best.108

I have suggested that what it is for an option to be open for an agent is for this option to be the sort of thing that can be taken into consideration in the process of deliberation, so for an option to be open is for this option to be the suitable object of both the intellectual and appetitive aspects of deliberation. A bit more precisely, then, I suggest that what it is for an option to be the sort of thing that can be taken into consideration in the process of deliberation is for this option to be the sort of thing that an agent can come to recognize as

106 DPG § 42c. 107 T § 22. 108 As already mentioned, agents can fall short of this ideal in several respects. The further from this ideal the less rational and free is the action.

53 good and for the agent to be rationally inclined to bring this option about on the basis of this perceived goodness. This, in general terms, is what it is for an option to be open for an agent.

More precisely, in the ideal case, I shall say that an option O1 is perfectly open to an agent S if and only if

S apprehends the correct degree of goodness in O1, and S possesses a rational inclination whose strength is proportionate to the degree of apprehended goodness in O1. Most agents to some extent or other fall short of this ideal, lamentably. Given this unfortunate fact about most agents, it is important to define a different sense in which an option is open to an agent. In the non-ideal case, I shall say that an option O1 is imperfectly open to an agent S if and only if S apprehends some goodness in O1 but S is mistaken about the degree of goodness in

O1, or S has a rational inclination towards O1 but the strength of S’s rational inclination fails to be proportionate to the apprehended degree of goodness in O1. Given that most agents fail to be ideally rational, in one way or another, I think that both senses of being open to an agent are important for agential modality. I combine then in the following simple way. I shall say that an option O1 is open to an agent S if and only if O1 is either perfectly open or imperfectly open to S.

With this basic sketch of what it is for an option to be open to an agent, we are now in a position to articulate agential modality – the kind of modality that governs whether an agent can choose an option in the sense that matters for freedom. I shall say that it is agentially possible for S to choose option O1 if and only if O1 is open to S, and there is a deliberation situation Dn in which S comes to the conclusion that O1 is the best considered option. I shall say that it is agentially impossible for S to choose option O1 if and only if O1 is not open to S, or there is no deliberation situation Dn in which S comes to the conclusion that O1 is the best considered option. I shall say that it is agentially necessary for S to choose option O1 if and only if O1 is open to S, and there is no option On, distinct from O1, that is open to S. Finally, I shall say that agent S chooses option O1 in an agentially contingent way if and only if S chooses option O1 in actual deliberation situation D1, and there is at least another option On, distinct from O1, that is both open to S and taken into consideration in D1.

My commitment to deliberation situations, in the definition of agential modality, is meant to be quite thin and liberal. For there to be a deliberation situation, relevant for the purposes of agential modality, what

54 matters is merely that such a deliberation be consistent with i) the essential properties of an agent S’s rationality, namely S’s intellect and will, and ii) the (apprehended) intrinsic goodness in O1. In other words, for there to be a deliberation situation D1, relevant for the purposes of agential modality, it is enough that D1 be metaphysically possible – that is, it is enough that D1 does not imply a contradiction.109 Crucially, I take it that there is a large number of options, and combinations thereof, which are consistent with (i) and (ii), or, in other words, I take it that there is a large number of options which are metaphysically possible given (i) and (ii). More precisely, then, for any finite set of options O consistent with O1 and with S’s intellect and will, there is a deliberation situation Dn where S considers O1 along with every Oi member of O and comes to a conclusion about which considered options is best.

A few clarificatory comments are in order. First, the definition of agential necessity makes no reference to deliberation situations. This is an important omission. A major thesis of this chapter is that deliberation by its very nature requires agential contingency, or, put differently, that deliberation by its very nature requires the weighing of the apprehended goodness of at least two agentially possible options under consideration. If this thesis is correct, agential contingency is a condition for deliberation in general, and thus agential necessity rules out legitimate cases of deliberation.110 Second, there are two different senses of agential impossibility that are included in the presented definition. The first one has to do with an option not being open to an agent. The basic idea here is that rational action requires deliberation and deliberation is only about options open to the agent, so if an option is not open to an agent, there is an important de facto sense according to which the agent cannot choose this option. The second sense of agential impossibility has to do with an option that possesses so little apprehended goodness in it (perhaps zero) that it is never fit to be the object of deliberate choice

109 As promised in section 1, here it becomes explicit how I can incorporate the passages in which Leibniz only makes reference to lack of metaphysical necessity as a condition for freedom without thinking that metaphysical contingency as such counts as a credible candidate for real options in Leibniz’s system. 110 From this clarification it follows that the term ‘agential necessity’ is a bit of a misnomer. itself requires agential contingency, because agency requires deliberation, so if an action is agentially necessary it is not really an action and there is really no agency involved in bringing it about. The expression ‘agential necessity’ seems best suited to capture something like the following. It is agentially necessary for an agent S to choose option O1 if and only if O1 is open to S, and for every deliberation situation Dn, S comes to the conclusion that O1 is the best considered alternative. The alternatives of ‘loving God’ or ‘being happy’ might arguably be credible cases of agentially necessary courses of action for any agent, in this difference sense of agential necessity. I will not pursue this issue farther.

55 precisely because there is no deliberation situation in which the agent comes to the conclusion that this option is the best considered option. Lastly, it is typically enough to define a given kind of contingency, in a modality

M, by defining it as not necessary according to M [contingentlym p ≡ ¬□mp] or as it being possible for it not be according to M [contingentlym p ≡ ◊m¬p]. My definition of agential modality does not lend itself to this simple inter-definition of agential contingency with agential non-necessity or with the agential possibility of its negation or opposite, but, I believe, it remains quite intuitive nonetheless.

With this sketch of agential modality, we are now in a position to shed some light on what the power to do otherwise, that is required for freedom, can be for Leibniz. I do not intend to advance a full account of what the power to do otherwise amounts for to for Leibniz, but I hope to illustrate at least one important aspect of it. According to the proposal that I am advancing here, the power to do otherwise is grounded in the agent’s will. More precisely, an agent S has the power to bring about an option O1 if and only if i) it is agentially possible for S to choose O1, ii) S takes O1 into consideration in actual deliberation, and iii) S has a rational inclination towards O1.111 Because this kind of power is explained partly in terms of agential modality, I shall call it ‘agential power’. As defined, an agent S can choose an option O1 only if S has the agential power to bring about O1. Now, for an agent S to have the power to do otherwise, in the sense that matters for freedom, I submit, is for S to have the agential power to bring about at least one of the not-chosen but considered options in a given deliberation situation.

With this understanding of agential powers, we can now articulate what real options amount to for

Leibniz. For a possible option O1 to count as a real option just is for an agent S to have the agential power to bring about O1. We can now also make more precise Leibniz’s insistence that freedom requires that the agent have options from which to choose; the more precise way of putting this point is to say that freedom requires deliberation which itself requires the weighing of at least two real options; or, alternatively, freedom requires that the agent deliberates about at least two options which the agent has the agential power to bring about. As

111 Condition (iii) is redundant. As I have defined agential possibility, it requires that the option be open to the agent, and the option being open to the agent requires that the agent have a rational inclination towards it. I made condition (iii) explicit merely to illustrate the importance of the will for agential powers.

56 the reality of choice strategy has it, then, the choice C1 to bring about option O1 thus counts as agentially contingent partly because it is the result of a kind of deliberation that involves the weighing of several options that the agent has the agential power to bring about, and C1 does not render any of the not-chosen options to be agentially impossible or in any other way undermine the agent’s agential power over these not-chosen but considered options.

We are now in a position to see how the three conditions of freedom come together in a single account, on Leibniz’s picture. Intelligence is essential to freedom because free actions are deliberate actions. Spontaneity is essential to freedom because free actions are the result of the kind of rational self-determination that follows deliberation. Contingency is essential to freedom because free actions are the result of deliberate weighing of the apprehended goodness of several options that the agent has the agential power to bring about. In sum, an agent is free to the extent that she determines herself to do that which she deliberately judges to be the best from several options that she could have brought about had she come to the deliberative conclusion that these options were best.

2.2 Potential Objection and Reply

There is at least one important reason for thinking that the interpretation of the reality of choice strategy I advanced in the previous subsection should not be attributed to Leibniz. This reason is the following.

Leibniz sometimes explicitly says that the not-chosen alternatives are incompatible with the agent’s will.

Regarding God’s decision to create the world, Leibniz writes, that a not-chosen possible world “remains possible in its own nature, even if it is not possible with respect to the divine will.”112 Passages like this one can indeed be reasonably read as saying that there is no room in Leibniz’s system for something like agential modality, as I describe it, and thus no room for my interpretation of the reality of choice strategy, and ultimately no room for my account of real options. In this section, I will argue that this is not the correct way of reading

112 On Freedom and Possibility, Grua 277/ AG 21.

57 passages like the quoted one. I will argue that the correct way of reading these passages is perfectly compatible with my interpretation of agential modality and the reality of choice strategy.

It is quite clear how passages in which Leibniz states that the not-chosen alternatives are incompatible with the agent’s will fit with, and even support, Adams’s interpretation of the reality of choice strategy – namely, the interpretation that it suffices for possible options to count as real options that these options be possible in themselves. If this is the correct way of reading Leibniz’s reality of choice strategy, we should expect Leibniz to insist that not-chosen alternatives are possible only in themselves, and thus not also possible relative to the agent’s will. In some sense the agent has the power to bring those non-chosen alternatives about, but it is the will of the agent that rules them out: they are compatible with the agent’s power, but not the agent’s will.

Essential to my proposal in the previous subsection, however, is that part of the plausibility of real options is precisely that they are compatible with the agent’s will, and thus that there is a sense of ‘possible’ according to which it is possible for the agent to choose the not-chosen alternatives. My main interpretative move, here, is to say that Leibniz is indeed committed to saying that the not-chosen alternatives are incompatible with one sense of the ‘will’ of the agent, but importantly, that that sense is not the one that is required for my account of agential powers and ultimately my interpretation of the reality of choice strategy.

My interpretation of the reality of choice strategy is compatible with the correct interpretation of these passages by Leibniz.

The crucial question is, then, what does Leibniz mean to deny and how does this relate to agential modality and my interpretation of the reality of choice strategy? Here is my proposal. Leibniz meant to deny that not-chosen alternatives, qua not-chosen alternatives, are compatible with the consequent will, qua consequent will. That is, Leibniz meant to deny that it is possible for the agent to choose option O1 after concluding, in the process of deliberation, that a different option On is best. To better appreciate this, consider a deliberation situation D1 in which agent S comes to the conclusion that option O1 is the best considered option. As we have seen, an important aspect of the process of deliberation, as Leibniz sees it, is deliberation as a struggle between antecedent wills striving to bring about the open alternative to which they are conceptually

58 connected and resulting in a triumphant antecedent will which becomes the consequent will for the agent.113

From this appetitive aspect of the nature of deliberation, it follows that part of the result of deliberation D1 includes not only the judgment that O1 is best, but also the triumphant and now consequent will W1 that is a rational inclination towards O1. What Leibniz denies is, then, that at the end of D1 the consequent will W1 is compatible with considered but not-chosen options Oi…On, other than O1. That is, at the end of D1 it is consequent will W1, which is conceptually connected to option O1 as the option it is striving to bring about, that is incompatible with the not-chosen but considered options Oi…On. As Leibniz sees it, then, for agent S to choose an option Oj, other than O1, at the end of D1 is tantamount to S acting on the basis of antecedent will Wj, distinct from the consequent will W1, despite the fact that it was will W1 that was triumphant in the process of deliberation, but this is to undermine the very rational nature of the will as a rational inclination towards apprehended goodness.

Another way of understanding my response to this objection is the following. Leibniz is rejecting a conception of the will as a self-determining faculty that has the power to choose any of the options under consideration even after the intellect has judged one of those options to be the best one. This is the ‘voluntarist’ conception of the nature of the will.114 The voluntarist conception of the will is contrasted with the

‘intellectualist’ conception. According to the latter, once the intellect reaches a conclusion at the end of the process of deliberation, or issues its ‘last practical judgment,’115 the will inevitably follows this judgment. For things to be otherwise, the intellectualist insists, is for the will to be irrational, but the will by its very nature is a rational inclination towards apprehended goodness – and thus not the sort of thing that can determine itself towards anything other than the goodness as apprehended by the intellect. Once the intellect issues its last

113 This characterization of the appetitive aspect of the process of deliberation is a bit simplistic, but it will do for our purposes. Part of what makes it simplistic is that it presupposes that the individually strongest rational inclination is inevitably the one that will become the consequent will. This, however, need not be the case. Several rational inclinations can come together to give rise to a new complex inclination (T § 22), or non-rational inclinations can get in the way of rational inclinations (T § 305), or consideration of general can play a role (T § 337), etc. These further complications are important for understanding what Leibniz takes the nature of deliberation to be, but are only indirectly relevant for the purposes of this dissertation. 114 For an insightful history of this doctrine and its main opposition ‘’ see Kent 1995, especially chapter 3, and Murray 1996, 2004 and 2005. 115 This is a term of art during the medieval period.

59 practical judgment, the will, by its very nature, inevitably follows this judgment and wills the option judged best, the intellectualist insists. Leibniz is well aware of this controversy, and falls squarely on the side of the intellectualist.116

A main reason for Leibniz’s theoretical decision here is that the voluntarist conception of the will violates his much-treasured principle of sufficient reason. The conditions for the voluntarist self-determining will to act are insufficient to explain this act; as the voluntarist sees it, this is necessary for protecting the freedom of the will. That is, if these conditions for action were sufficient to explain the will’s act, then these conditions would determine the will’s act, and this would deprive the will of its self-determining capacity.117 It must be the will, and not the conditions for action, that ultimately determines which possible course of action is taken, if the will is to be free, the voluntarists insist. This lack of explanation between the act of the will and the conditions for action violates Leibniz’s insistence that everything must have an explanation for why it is the way it is rather than otherwise, and thus violates Leibniz’s version of the principle of sufficient reason.

I propose, then, that we read Leibniz’s claim that the non-chosen alternatives are incompatible with the agent’s will as denying this voluntarist conception of the will as a self-determining faculty. If this is indeed what Leibniz is denying, then Leibniz is not to be understood as denying the proposed account of agential modality or the interpretation of the reality of choice strategy. My proposal on agential modality and interpretation of the reality of choice strategy are not committed to the voluntarist conception of the will, and so do not entail or require that which Leibniz denies when he denies that not-chosen alternatives are compatible with the agent’s will. The proposed account of agential powers is compatible with Leibniz’s intellectualism. To better appreciate this, consider the following. In general, a given antecedent will W1, that is a rational inclination

116 T §§ 22, 311, 325. There are passages which seem to paint a different story. Sometimes Leibniz writes: “we do not always follow the last judgement of practical understanding when we resolve to will” (T § 51). This seems to be a paradigmatic statement of the voluntarists position, but that reading would be too facile. Rather, Leibniz sometimes uses the word ‘will’ to mean not merely rational inclinations but the final triumphant inclination that results in action and that includes both rational and non-rational inclinations – i.e., the passions. Immediately after the quoted passage, Leibniz continues: “but we always follow, in our willing, the result of all the inclinations that come from the direction both of reasons and passions”. For our purposes, the lesson from these seemingly dissenting passages, I think, is merely that Leibniz uses the word ‘will’ in two different senses and that it is only the one I have isolated in the main text that matters for agential contingency, and thus for freedom. 117 As we encountered in chapter one, Molina’s views on freedom are a version of this voluntarist account.

60 towards option O1, is compatible with an agent S choosing a different option O2 if and only if there is a deliberation situation Dn in which S concludes that O2 is the best considered alternative. That is, an antecedent will W1 (towards option O1) is compatible with an option O2 if and only if it is agentially possible for S to choose O2. Crucially, this is perfectly consistent with it also being true that a consequent will W1 (towards option O1) is incompatible with not-chosen option On, distinct from O1, at the end of deliberation D1 in which the agent concludes that option O1 is the best considered option. That is, denying the voluntarist conception of the will, and the kinds of powers it requires, is compatible with agential powers as I described them. Thus,

Leibniz’s claim that the not-chosen alternatives are in some sense incompatible with the agent’s will should not be read as a rejection of my interpretation of the reality of choice strategy or account of real options grounded in what I have called ‘agential powers.’

With this I conclude what I take to be the heart of this chapter, namely a plausible account of real options grounded in agential powers. The rest of the chapter is dedicated to illustrating what I take to be a very important theoretical benefit of this account. This theoretical benefit is that the proposed account makes intelligible Leibniz’s otherwise obscure claim that free actions are contingent because they are brought about by merely inclining and not necessitating reasons. In making sense of Leibniz’s merely inclining reasons doctrine, we will thereby also make sense of Leibniz’s version of the Contingency Condition – the Molinist tenet that subjunctive conditionals of freedom are contingent. As we noted in the previous chapter, Leibniz holds two seemingly conflicting claims regarding these subjunctive conditionals of freedom: i) the connection between conditions for action, specified in the antecedents of subjunctive conditionals of freedom, and free actions, specified in the consequents, is a contingent connection; and ii) the conditions for action explain the free actions. I will argue in the next section that the proposed account of the reality of choice strategy makes intelligible how merely inclining reasons can explain free actions, and in so doing it also makes intelligible

Leibniz’s version of the Contingency Condition.

61

Section 3: The Reality of Choice and Merely Inclining Reasons

A general challenge to Leibniz’s ability to retain any intelligible sense of contingency in his system is his endorsement of a strong version of the principle of sufficient reason. As we have already seen, Leibniz himself rejected the kind of contingency required by a voluntarist self-determining will precisely because that kind of contingency is ruled out by this fundamental principle. The worry, here, is a general one. The worry is that embracing a robust version of the principle of sufficient reason precludes any intelligible sense of contingency as a condition for freedom; and so that it precludes agential contingency as advanced in this chapter. If Leibniz’s strong version of the principle of sufficient reasons does have necessitarian implications, then agential powers are, at best, necessarily unexercised powers, and as such they are inadequate grounds for plausible candidates for real options in Leibniz’s system. This worry is, thus, an important worry.

This general worry is not new. Both Spinoza and Hobbes, for example, welcomed these purported necessitarian implications of the principle of sufficient reason.118 Leibniz addressed this general worry on multiple occasions and insisted that there were no such necessitarian implications. Everything requires a sufficient reason why it is rather than not, but this does not entail necessitarianism because “these inclining reasons are not necessitating, and destroy neither contingency nor freedom.”119 That is, the sufficient reasons that explain why free actions are the way they are and not otherwise do not necessitate these free actions because they are merely inclining reasons, and merely inclining reasons do not necessitate what they explain. I have called this general response Leibniz’s ‘merely inclining reasons’ doctrine. Some commentators have not been impressed by Leibniz’s merely inclining reasons doctrine. Lovejoy, for example, famously argued that Leibniz, despite protestations to the contrary, was equally committed to the necessitarian conclusions of the principle of sufficient reason, and that Leibniz’s merely inclining reasons doctrine, that meant to block this implication, is “manifestly without logical substance.”120

118 For Hobbes see HB § 31, for Spinoza see E I.p8, p11, p15, p16, p23, and p29. 119 On Freedom and Spontaneity, G 7.110/ SLT p. 94. 120 Lovejoy 1960: 172.

62

As already noted, one of the theoretical benefits of my proposed account of the reality of choice strategy is precisely that it makes intelligible Leibniz’s merely inclining reasons doctrine. Furthermore, by doing so my account of agential contingency makes intelligible how the principle of sufficient reason does not have necessitarian implications in Leibniz’s views, at least with regards to adequate explanations of free actions.

3.1 Merely Inclining Reasons and Explanations

Before presenting how my account of the reality of choice strategy makes merely inclining reasons intelligible, a few words about explanations and reasons are in order. I propose that for Leibniz the word

‘reason’ is a theoretical one. More precisely, I propose that for Leibniz what it is to be a reason is to play a particular theoretical role, namely that of providing an explanation that meets the demands of Leibniz’s principle of sufficient reason.121 In other words, I propose that for Leibniz it is the notion of explanation that is more theoretically fundamental, and the sorts of things that count as reasons for Leibniz are all the sorts of things that provide an adequate explanation – i.e., one that meets the demands of the principle of sufficient reason. This proposal makes sense of Leibniz’s varied use of the word ‘reason’.122

If so, to identify what sorts of things count as reasons in explanations of free actions we must first identify what counts as an adequate explanation of these free actions. In this chapter, and dissertation more generally, I will not be attempting to tackle the deep theoretical question regarding the nature of adequate explanations for Leibniz. The goal for this section is going to be quite more modest than this. I will simply identify the kind of explanation of free actions that Leibniz takes to be an adequate explanation. I will argue that this kind of adequate explanation enables us to see how in the case of free actions merely inclining reasons can meet the strictures of the principle of sufficient reason.

121 My suggestion here is similar to Pereyra’s (2013) suggestion for the principle of contradiction. 122 Sometimes Leibniz uses the word ‘reason’ to refer to requisites, or conditions for the existence of something else (A 6.3.587) or to refer to causes (G 3.36, LGR 297), or to refer to logical proofs (Grua 302, AG), or to refer to perfections or contrastive about perfections (Grua 287, AG 19).

63

The kind of explanation that I have in mind is based on ’s insight regarding the nature of explanation articulated near the end of the Phaedo. Leibniz himself explicitly and enthusiastically endorses this passage in Plato. After arguing against “overly materialistic philosophers” that final causes can be useful in physics, Leibniz continues: “This reminds me of a beautiful passage by in Plato’s Phaedo. This passage agrees marvelously with my opinions on this point and seems to be directed expressly against our overly materialistic philosophers.”123 Plato’s insight is that the adequate explanation for the event of ‘Socrates sitting in prison’ qua rational action is one that must cite the action as being the best amongst its possible alternatives together with Socrates’ appreciation of and appropriate response to this contrastive fact. The other requisites for this event are not required for this kind of explanation. For ease of reference, I will refer to this general kind of explanation in accordance with Plato’s insight simply as “Platonic explanations” of rational action.

Platonic explanations require two different kinds of reasons (in the Leibnizian sense of reasons): one is the apprehended goodness of the object of choice, and the second is the agent’s appreciation and appropriate response to this apprehended goodness. As Leibniz puts it: “We shall study, therefore, both the will and its object, namely, good and evil, which provide the reasons for willing and rejecting.”124

Platonic explanations of rational action require citing the apprehended goodness of the chosen course of action as well as citing the intellect and the will of the agent appreciating and responding appropriately to this goodness in the course of action. As we have seen, Leibniz presents this appreciation and appropriate response in terms of intellectual and appetitive aspects of practical deliberation. More precisely, for Leibniz, the Platonic explanation of S choosing option O1, qua rational action, requires citing the apprehended goodness of O1 and citing S’s last practical judgment of the intellect that O1 is the best considered alternative, and S’s consequent will W1, as rational inclination towards O1. Citing these reasons suffices to adequately explain rational action, for Leibniz.

123 DM § 20, AG 53. 124 CD § 19.

64

Given this Leibnizian version of Platonic explanations of rational action, we are now in a position to understand why the reasons that explain free actions, according to these kinds of explanations, are best understood as ‘merely inclining’ reasons. At bottom, this is so, I suggest, because Leibniz’s version of Platonic explanations requires something along the lines of what I have called ‘agential contingency.’ These kinds of explanations require deliberation on the part of the agent and, I have argued, deliberation itself requires that the agent weigh the apprehended goodness of at least two options which the agent has the agential power to bring about – i.e., deliberation requires agential contingency. In sum, the reasons that explain rational action – namely the apprehended goodness of the chosen option together with the agent’s last practical judgment and consequent will – are best understood as merely inclining reasons because these reasons explain why the agent does what she does without making it agentially necessary that she does so. These merely inclining reasons do not undermine the agent’s agential power to do otherwise, and so do not undermine the fact that it is agentially possible for the agent to do otherwise.

Once this connection between the reality of choice strategy and merely inclining reason is made explicit, it is uncontroversial, I take it, that agential contingency is not undermined or threatened by merely inclining reasons – at least the merely inclining reasons that are involved in proving Platonic explanations of rational action. What is less straightforward, however, is whether these merely inclining reasons meet the demands of

Leibniz’s strong version of the principle of sufficient reason, and thus whether these merely inclining reasons do suffice to explain rational action. This is the more controversial thesis in this context, I take it.

As already stated, I believe that these merely inclining reasons are indeed explanatory in this way. To better appreciate the adequacy of this account, consider the following example. Susan is deliberating about what career path to take. She loves both music and economics, and recognizes the value in pursuing either path: a career in the music world afford her more opportunities for artistic development, appreciation and expression; a career in economics, on the other hand, would afford her a more monetarily prosperous future and the that are associated with it. She adequately assesses her prospects of success in either field, given her talents and aptitudes, and recognizes that she can do well in either career. Upon careful deliberation, then, she reaches the conclusion that a career in music would be better, that is more valuable, than a career in

65 economics. Being a rational agent, Susan is rationally inclined to each of these career paths in a way which is proportionate to their apprehended goodness: she is more inclined towards a music career than a career in economics precisely because she recognizes that the former is better, or more valuable, than the latter. Because of this, Susan chooses a career in music.

How do we explain Susan’s action? Leibniz’s account, I submit, states that pointing to the greater apprehended value of the music career path together with Susan’s last practical judgment and consequent will suffices to explain Susan’s action qua rational action. Citing these merely inclining reasons suffices to explain why

Susan acted as she did and not otherwise – thus meeting the demands of the principle of sufficient reason.

Furthermore, and crucially for our purposes, Susan’s action was agentially contingent: it is still agentially possible for her not to choose a music career path, for she has a rational inclination towards an alternative possibility, a career path in economics, and there is a deliberation situation in which she chooses to pursue a career in economics by coming to the conclusion that this is the better path. Thus, the proposed account of adequate explanations of rational actions does seem to intuitively satisfy the demands of the principle of sufficient reason by explaining why the agent did what she did and not otherwise, while not making the action necessary in the sense that matters for freedom – namely, without making the action agentially necessary. Thus, my interpretation of the reality of choice strategy makes Leibniz’s otherwise obscure doctrine of merely inclining reasons both intelligible and, I may add, plausible.

3.4 Agential Modality and Moral Necessity

Before concluding this chapter, I will like to briefly address a doctrine advanced by Leibniz that is intimately related to merely inclining reasons and agential modality. This is Leibniz’s doctrine of moral necessity. During his mature period, Leibniz insists that there is a kind of necessity that is compatible with freedom. This is the kind of necessity which binds the wise to do that which is best. Or as Leibniz puts it: “it

66 must be said that God, that the , is carried to the best by a moral necessity.”125 This “happy necessity,” as

Leibniz calls it elsewhere,126 is nonetheless compatible with freedom. In sum, for Leibniz moral necessity is a happy necessity which obliges the wise to act in accordance with wisdom but which nonetheless does not rule out the kind of contingency that is required for freedom. This general characterization of moral necessity is clear enough, but the details of Leibniz’s doctrine are quite murky.127 It will not be the aim of this section to try to settle these details. The more modest aim will be simply to elucidate the relationship between Leibniz’s doctrine of moral necessity and agential modality.

The language that Leibniz uses to articulate moral necessity is reminiscent of the language that I used to articulate agential modality. Both rely on the kind of rationality grounded in the appropriate response to the apprehended goodness of the object of choice. Agential modality, however, is insufficient to capture the kind of normativity or determination that is grounded in a course of action being the best one together with its corresponding rational inclination on the part of the agent; agential modality merely assumes that there is such a determination. Moral necessity captures this kind of determination between the best and the wise who choose the best.

Importantly for our purposes, moral necessity does not threaten or undermine agential contingency precisely because moral necessity does not add any new kind of determination that was not already assumed in agential contingency. Moral necessity, or the kind of determination holding between the best and the wise, is compatible with the kind of contingency that matters for freedom, Leibniz insists. He writes, for example:

Otherwise the necessity of good would be geometrical (so to speak) or metaphysical, and altogether absolute; the contingency of things would be destroyed, and there would be no choice. But necessity of this kind, which does not destroy the possibility of the contrary, has the name by analogy only: it becomes effective not through the mere essence of things, but through that which is outside them and above them, that is, through the will of God. This necessity is called moral, because for the wise what is necessary and what is owing are equivalent things; and when it is always followed by its effect, as it indeed is in the perfectly wise, that is, in God, one can say that it is a happy necessity. (T, Summary of the Controversy; Answer to Problem VIII.)

125 T § 132. 126 T, Summary of the Controversy; Answer to Problem VIII. 127 Michal Murray has written a series of papers (1995, 1996, 2004, and 2005) in which he lays down a plausible account of moral necessity as a kind of modality that lies between causal necessitation and the kind of contingency that is championed by libertarians. Sleigh (2009) and Adams (2005) provide radically different models of what Leibniz might mean by moral necessity.

67

Here Leibniz insists that the happy necessity that obliges the wise to do the best “does not destroy the possibility of the contrary” and so preserves the contingency, and ultimately the freedom, of choice. That is, the kind of determination encoded in moral necessity is compatible with the kind of contingency that is required for freedom, Leibniz insists.

Given the advocated account of agential contingency, we are now in a position to state explicitly how moral necessity is compatible with, and otherwise connected to, the kind of contingency that matters for freedom. Here are the basic connections. An option O1 is agentially possible for an agent S only if there is a metaphysically possible deliberation situation D1 in which it is morally necessary that S chooses O1. Also, an agent S chooses option O1 in an agentially contingent way only if it is morally necessary that S chooses option

O1 in actual deliberation situation D1, and there is at least another option On, distinct from O1, that is both open to S and taken into consideration in D1. As these characterizations illustrate, moral necessity is embedded in the more precise characterization of the kind of contingency that matters for freedom. Moral necessity encodes the way in which a rational agent determines herself to do what she judges best from several options that are agentially possible for her.

Conclusion

In this chapter I have advanced a novel way of understanding the reality of choice strategy. I have argued that this strategy admits for a kind of modality governing what agents can choose – I have labeled this kind of modality ‘agential modality.’ I have argued that agential modality enables Leibniz to advance a plausible candidate for real options – the kind of options that are possible for an agent to choose in the sense that matters for freedom. The basic idea is that an agent is free to the extent that she determines herself to do that which she deliberately judges to be the best from several options that she could have brought about had she come to the deliberative conclusion that these options were best. I have argued that agential modality can also help make sense of Leibniz puzzling doctrine of merely inclining but not necessitating reason which nonetheless meet the strictures of Leibniz’s strong version of the principle of sufficient reason. In making sense of Leibniz’s

68 merely inclining reason doctrine we also made sense of Leibniz’s version of the Contingency Condition – the tenet that subjunctive conditionals of freedom are contingent even though the conditions for action, specified in the antecedents, explain the free actions, specified in the consequents of these conditionals. Thus, contrary to what several theoretical commitments might appear to indicate, Leibniz has an intelligible, and even plausible, sense of contingency as a condition for freedom in his system.

69

Chapter 3:

Leibniz on Contingency and Individual Concepts

INTRODUCTION

One of the most taxing topics in Leibniz scholarship is Leibniz’s account of contingency. Many are led astray here, Leibniz warns us.128 Throughout his career, Leibniz vacillated much regarding how best to capture this valuable condition on freedom. He presented multiple, and apparently independent, strategies to account for contingency without clearly and definitely opting for one of them.129 It is clear that the topic of contingency vexed Leibniz throughout his life. Discussions on this topic, in Leibniz’s system, are further complicated by Leibnizian commitments which seem to rule out contingency. A notable example is Leibniz’s complete individual concept doctrine (CICD).130 Beginning with Arnauld, many commentators have thought that if all truths about a substance are included in its individual concept, then it seems that the substance cannot be otherwise than the way its individual concept has it as being. That is, it seems that Leibniz’s CICD entails superessentialism.131

128 T Preface, p. 53. 129 The secondary literature on Leibniz’s attempts to ground contingency is quite extensive. Some examples include Abraham 1969, Adams 1994, Baxter 2000, Blumenfield 1988, Burms 1979, Carriero 1995, Frankel 1984, Griffin 1999 and 2013, Grimm 1970, Hunter 1981, McNamara 1980, Meijering 1978, O’Leary-Hawthorne and Cover 1990 and 1992, Savage 1998, Vailati 1986, and Wee 2006. 130 The doctrine that for any substance S there is an individual concept C that includes predicates denoting every property that S will ever have. See for example Primary Truths, C 519, DM 13-14, G 2.39-41. 131 I understand superessentialism as the thesis that for any object O and for any property P if O has P, then it is metaphysically necessary that O has P. The name “superessentialism” is used in the secondary literature to refer to slightly different views. Mondadori, a famous advocate of attributing superessentialism to Leibniz, defines it thus: for any object O and Property P, if O has P, then it is not possible for O not to have P while still existing/being O, (1973: 83). Look,

70

In chapter two I have advanced an account of the kind of contingency that I think matters for freedom, on Leibniz’s views. I have labeled it ‘agential contingency’. Superessentialism threatens the intelligibility of agential contingency, for if superessentialism is true unexercised agential powers would be essentially unexercised agential powers: it would be metaphysically necessary that these agential powers will not be exercised. This would clearly undermine these powers as legitimate grounds for the kind of contingency that matters for freedom.

The goal of this chapter is to argue that Arnauld is mistaken, and that Leibniz’s CICD does not entail superessentialism. In other words, the goal of this chapter is to deliver on a promissory note made in the introduction of this dissertation: that Leibniz is entitled to metaphysical contingency (for easy of reference I will just say ‘contingency’ from now onward, but it is metaphysical contingency that I am after in this chapter).

This promissory note is central to the goal of this dissertation, for it is presupposed in Leibniz’s version of both the Prevolitional Condition and the Contingency Condition, and it is further explicitly appealed to in the formulation of agential modality itself. Thus, the plausibility of both chapter one and two depend upon adequate fulfillment of this promissory note. In fulfilling this promissory note, I will not be further developing or defending the account of agential contingency at the metaphysical level. The more modest goal of this chapter is merely protecting the plausibility of agential contingency at the metaphysical level from the threat of superessentialism that arises from the logical level, especially from Leibniz’s CICD.

The plan for the chapter is the following. In section one, I begin by clearly articulating an argument for superessentialism from CICD. In section two, I argue that Leibniz is not committed to his argument. The key Leibnizian idea I will develop is that contingent truths can be included in complete individual concepts as contingent. I argue that this can be done by including predicates denoting unexercised powers to do otherwise in complete individual concepts. Furthermore, for these powers to ground legitimate contingency, it must be possible for them to be exercised. To secure this kind of possibility, I suggest that the content of individual concepts is counterfactually dependent on how agents utilize their powers. Furthermore, in order for this

(2013: § 3), also attributes this view to Leibniz, and defines it as: for any individual substance, x, and for any property P of x, necessarily, if x exists, then x has P.

71 counterfactual dependence to respect the plausible claim that concepts are individuated by their contents, I introduce a novel way of reading Leibniz’s account of individual concepts. I argue that an agent’s essence grounds her modal profile – what is possible, impossible, necessary or contingent for her – and that this gives rise to a cluster of individual concepts which describe this modal profile. I argue that how an agent would utilize her powers to act settles which individual concept describes everything that will ever happen to her and which individual concepts only describe non-actualized alternative possibilities for her. This model, I argue, enables Leibniz to retain an intelligible sense of contingency in his system grounded in something analogous to transworld identity.132

1. Individual Concepts and The Superessentialism Challenge

1.1 The Intuitive Challenge

Before presenting the more detailed argument for superessentialism on the basis of Leibniz’s logical commitments, I wish to present an intuitive version of the challenge. This intuitive challenge only relies explicitly on Leibniz’s doctrine of complete individual concepts. Arnauld is one of the first readers of Leibniz who noted that this doctrine seems to preclude contingency, or as he puts it, this doctrine entails “more than fatal necessity.”133 The basic idea here is quite intuitive. If a substance’s complete individual concept includes everything that will ever happen to the substance, then everything that will ever happen to a substance in some important sense must happen to it – that is, in some important sense no substance can be otherwise than how its complete concept has it as being. If so, the line of reasoning continues, Leibniz is not entitled to contingency.

There is something intuitive about Arnauld’s challenge indeed. To be sure, in general, the mere fact that something will be is not sufficient grounds for thinking that in some sense it also must be. Leibniz’s complete

132 Transworld identity is the thesis that an individual exists in more than one possible world. In contemporary discussions on modality, some theorist use this thesis to account for contingency, see Plantinga (1974: Ch. 6). As stated, transworld identity should not be attributed to Leibniz because this thesis presupposes an analysis of modality in terms of possible worlds that Leibniz clearly did not intend. 133 Arnauld to von Hesssen-Rheinfels, 13 March 1686, G 2.15-16, A 2.2.8-9/ WFII p. 98.

72 individual concept doctrine, however, does seem to provide the required intermediary premises to deduce some kind of necessity from mere future truth. It is the first main task of this chapter to articulate these premises and thus to develop Arnauld’s intuitive challenge. Arnauld himself, after being challenged by Leibniz to defend his claims, provides a detailed and sophisticated elaboration of this intuitive challenge which rests on more premises. This elaboration by Arnauld is the topic of our next subsection.

1.2 Arnauld’s Argument for Superessentialism

Leibniz was not impressed by Arnauld’s intuitive challenge and denies that his complete individual concept doctrine entails that there is no contingency. In response to this intuitive challenge Leibniz appeals to

God’s knowledge of an infinity of ‘possible Adams’ and their similarities as in some sense grounding alternative possibilities for Adam. Leibniz writes:

For by Adam’s individual notion I most certainly mean a perfect representation of a particular Adam who has certain individual conditions which distinguish him from an infinity of possible people, who closely resemble him but who are nevertheless different from him…. Everything that follows from his resolution is necessary only by a hypothetical necessity, and in no way destroys God’s freedom, or that of created . There is a possible Adam with a certain posterity, and an infinity of others whose posterity would be different.134

Leibniz’s response, it must be admitted, is quite opaque. I will later articulate what I take to be the best way of understanding Leibniz’s references to an infinity of possible Adams and their similarities to ‘a particular Adam’ and how this grounds the contingency of some of Adam’s properties. What matters for our current purposes is that Leibniz appeals to an infinity of possible Adams, and their similarities, with the aim of protecting the contingency of some predicates included in Adam’s complete individual concept, and thus denying the plausibility of Arnauld’s intuitive challenge.

Arnauld is not convinced by this response. He thinks that ‘possible Adams’ as described by Leibniz are

‘inconceivable,’135 and relies on them to provide a reductio for Leibniz’s position. Arnauld thinks that

134 Leibniz to von Hessen-Rheinfels, 12 April 1686, A 2.2.24/ WFII p. 100. 135 Arnauld to Leibniz, 13 May 1686, A 2.2.35/ WFII p. 103.

73 corresponding to each possible Adam will be a different individual concept containing different predicates, but finds it unintelligible to think of different individual concepts belonging to the same substance. Only that which individuates a substance should be included in its individual notion. Arnauld writes:

it is as impossible to conceive contradictory predicates in the individual notion of me as to conceive of a ‘me’ different form me. From which we have to conclude, it seems to me, that since it is impossible that I should not have always remained me, whether I had been married or had lived in celibacy, the individual notion of me involves neither the one nor the other of these two states…. This is why, sir, it seems to me, I should regard as involved in the individual notion of myself only that which is such that I would no longer be me, if it were not in me.136 Arnauld’s conclusion is that individual notions should only include “that which is such that I would no longer be me, if it were not in me.” This is, I take it, Arnauld’s way of insisting that all and only predicates denoting essential properties should be included in individual concepts. Arnauld thus thinks that by including predicates that denote all properties a substance will ever have, Leibniz’s view entails superessentialism.

Here is my rendering of Arnauld’s argument. Following Leibniz’s suggestion, suppose there are two possible-Arnaulds each of which depicts a possibility for Arnauld. Consider possible-Arnauld-1 who lives in celibacy his whole life, and possible-Arnauld-2 who gets married. By Leibniz’s complete individual concept doctrine, possible-Arnauld-1’s individual concept, call it ‘C1,’ includes the predicate ‘lives in celibacy for life’, and possible-Arnauld-2’s individual concept, call it ‘C2,’ includes the predicate ‘gets married.’ C1 and C2, therefore, have different predicates. Plausibly, concepts are individuated by their contents, or predicates – call this ‘Individuating Condition 1,’ or ‘IC1.’ By IC1, concepts C1 and C2 are numerically distinct. Plausibly, substances are individuated by their individual concepts – call this ‘Individuating Condition 2,’ or ‘IC2.’ By

IC2, C1 and C2 individuate numerically distinct substances. Therefore, possible-Arnauld-1 and possible-

Arnauld-2 are numerically distinct. This reasoning can be repeated for any possible-Arnauld, which purportedly represents an alternative possibility for Arnauld. Therefore, at most only one such possible-Arnauld is numerically identical to Arnauld. Therefore, these possible-Arnaulds do not represent legitimate alternative

136 Arnauld to Leibniz, 13 May 1686, A 2.2.35-36/ WFII p. 103.

74 possibilities for Arnauld. Therefore, all and only the predicates that are in fact included in Arnauld’s complete individual concept could be had by Arnauld. And this is just superessentialism.

1.3 Individual Concepts and Essences

Leibniz was not moved by Arnauld’s argument. Leibniz insists that his CICD does not entail superessentialism, but, lamentably, his reasoning for this is also a bit opaque. What is clear is that Leibniz insists that superessentialism can be avoided because “notions of individual substances” do “contain contingent truths,”137 and that this fact can be elucidated by appealing to “God’s free decisions (considered as possible)”138 which are also included in substance’s complete individual concepts. That Leibniz’s response to Arnauld’s detailed argument includes both of these elements is uncontestable, but how these elements come together to provide a plausible answer to Arnauld’s argument for superessentialism is not. Providing a response which elaborates these two basic elements in Leibniz’s answer is the main task of this section. The elaboration I wish to present is clearly Leibnizian, but it is also more precise than what Leibniz’s actually wrote. Thus, I present this response as more than pure historical exegesis of Leibniz’s text; this response includes a couple of elements which are best described as ‘rational reconstructions’ as I have characterized them in the dissertation introduction. This response is, thus, better characterized as a precisification of the general, and somewhat underspecified, Leibnizian response to be found in the text.

Before developing this account, I wish to distinguish between two different kinds of projects a faithful

Leibnizian can adopt at this juncture of the discussion: a modest project and an ambitious project. First, a faithful Leibnizian can adopt the modest project of developing Leibniz’s remarks into a plausible rejection of one of the premises of Arnauld’s argument, say IC2. This modest project would indeed suffice to show that

Leibniz is not committed to superessentialism, at least not on the basis of the argument advocated by Arnauld.

137 Leibniz to Arnauld, 14 July 1686/ WFII p. 106. 138 Leibniz to Arnauld, 14 July 1686/ WFII pp. 106-9.

75

An important limitation of this modest project, however, is that it fails to engage with the original intuitive challenge articulated by Arnauld. That is, even if this modest project succeeds in establishing that the more precise argument advocated by Arnauld ultimately fails, the plausibility of the original challenge seems to remain unshaken. Furthermore, and perhaps more importantly, undermining Arnauld’s more precise argument need not generalize to undermining other possible precisifications of the general intuitive challenge.

This limitation of the modest project is a serious one. To illustrate this, I will present a different precisification of the general intuitive challenge that Leibniz’s CICD entails superessentialism. This new argument explicitly relies on a strategy to ground contingency that Leibniz advocated throughout his career.

This strategy aims to ground metaphysical modality in logical notions. The basic idea is the following: a proposition is necessary if its opposite implies a contradiction, and a proposition is contingent if it is true and its opposite does not imply a contradiction. Because this strategy aims to ground metaphysical modality in logical notions (consistency, derivations, , etc.), I shall call it ‘logical strategy.’

Here is the argument for superessentialism on the basis of the logical strategy and Leibniz’s CICD.

Consider a substance S and a property P. Suppose S has P. According to the logical strategy, if S has P contingently, then the proposition “S is not P” does not imply a contradiction, and if S has P necessarily, then the proposition “S is not P” does imply a contradiction. Consider the proposition “S is not P.” Can a contradiction be derived from this proposition? I think so. Given that S has P, from Leibniz’s complete individual concept doctrine it follows that S’s complete individual concept C includes a predicate “being P.”

Furthermore, and crucially, if S’s complete concept C includes the predicate “being P,” then it seems that the proposition “S is P” can be derived from C.139 Now, the proposition “S is P” clearly contradicts the proposition

“S is not P,” with which we began. Thus, we have derived a contradiction beginning with the proposition “S is not P.” Therefore, by Leibniz’s logical strategy, S has P necessarily. Furthermore, this reasoning can be repeated for any substance and any property, so we conclude that for any substance S and any property P, if S has P, then S has P necessarily. This claim just is superessentialism again.

139 This is the crucial move in this argument. I will address its permissibility below.

76

This argument is clearly Leibnizian – it relies only on Leibniz’s CICD and the logical strategy. In contrast to Arnauld’s argument, it does not rely on either IC1 or IC2 or on Arnauld’s more dubious assumption that for any possible-Adam there is a corresponding complete individual concept. Despite the initial plausibility of this new argument, I think that it is also ultimately unsuccessful. My basic response is that, despite initial appearances, this argument makes illicit use of the logical strategy. I will argue, in the next subsection, that the licit application of the logical strategy is intimately connected with a plausible notion of essence as what makes the substance be what it is (i.e., that which answers the Platonic “what is it?” question).

The existence and initial plausibility of this new argument reveals important limitations of the modest project in defending Leibniz’s views from Arnauld’s challenge. A second project a faithful Leibnizian can adopt at this juncture of the discussion is the following. A faithful Leibnizian can develop the considerations brought forth by Leibniz’s remarks into an account that makes it plausible to see how Leibniz can retain an intelligible sense of contingency in his system. At bottom, the superessentialism challenge is that Leibniz’s logical commitments entail that the essence of an individual is ultimately to be identified with its complete individual concept. This second project aims to develop an account in which it is plausible to demarcate a core of the individual concept, that marks the essence of the substance,140 from a superstructure, that marks the properties had contingently by the substance. If this second project is successful, it will develop the tools to answer not only Arnauld’s argument but also other similar arguments for superessentialism based on Leibniz’s CICD. This second project is thus a significantly more ambitious project than the first one. It is this more ambitious project that I wish to adopt in this chapter.

140 Leibniz employs the word “essence” in multiple different ways throughout his career – see for example, CD § 9, DPG § 24a, LC § 9, S p. 110, T §§ 7, 9, 335, On Contingency, Grua 302-6, DM 16, 24, 26, G 2.38. At least on some occasions he does use essence in contrast to “accident” as I do in the main text – see for example, WFII 108, L 606, G 458; Grua 383, LC 37, DM 24. There is also a long tradition of philosophers using the expression “real definition” or “essence” to mean what I mean. See for example, ’s Categories and middle books on Metaphysics, Code 1986, and Fine 1994 and 2005.

77

1.4 Essences and the Logical Strategy

There is an intuitive connection between the notion of an essence and the logical strategy that has important implications for the successful completion of the mentioned ambitious project. This connection is the following. The essential truths about a substance are those and only those whose opposite implies a contradiction and the only kinds of truths that are allowed in derivations for the licit use of the logical strategy are ones which are constitutive of, or in some important sense follow from, the essence of the substance.

This interconnection is illustrated by examples of licit applications of the logical strategy. Aquinas, one of the great proponents of the logical strategy, for instance, uses the following examples to illustrate the licit applicability of this strategy. Aquinas says that the proposition ‘Socrates is sitting’ does not imply a contradiction, whereas the proposition ‘Socrates is a donkey’ does.141 This is so, Aquinas tells us, because the essence of Socrates is to be a rational and the predicate ‘sitting’ is compatible with this essence whereas the predicate ‘being a donkey’ is not. The latter is the case because a donkey is essentially non-rational which is incompatible with Socrates’s essence. Crucially for our purposes, then, the proposition ‘Socrates is a donkey’ implies a contradiction precisely because the predicate ‘being a donkey’ essentially involves a property incompatible with Socrates’s essence.142 Given this we can infer, by the logical strategy, that it is necessarily the case the Socrates is not a donkey. As Aquinas sees it, then, the licit applicability of the logical strategy is intimately connected to the notion of essence.

I wish to end this section by raising what I shall call the ‘demarcation question.’ This question is a question about the proper way to demarcate an intuitive core of a complete individual concept from the rest of the concept. A properly demarcated complete individual concept will have a clear and plausible distinction between a core, which denotes only the essence of the substance, and a superstructure, which denotes the properties had contingently by the substance. Given the intuitive connection between essence and the logical strategy, in the case of a properly demarcated complete individual concept we will be in a position to identify

141 ST 1a.25.3. 142 This is a traditional example and application of the logical strategy. Molina (Co 4.47.2) repeats this same example with the same purpose of illustrating the licit applicability of the logical strategy.

78 the truths whose opposite implies a contradiction, namely those in the core, from those whose opposite does not imply a contradiction, namely those in the superstructure. Thus, answering the demarcation question will not only enable us to make a plausible distinction between a core and as superstructure, but it will also enable us to better understand the licit applicability of Leibniz’s version of the logical strategy.

1.5 Answering the Demarcation Question

The main Leibnizian answer to the demarcation question I wish to investigate is his infamous finite/infinite analysis strategy for grounding the difference between contingent and necessary truths. This strategy was dominant in Leibniz’s thought during the middle period of his career,143 but seems to have fallen out of favor later in his life. Leibniz sometimes introduces this strategy as an independent strategy and sometimes as interconnected with the logical strategy. Thus, it seems that the modest conclusion to be drawn is that the finite/infinite analysis strategy provides answers to the demarcation question at best only during the middle period of Leibniz’s career.

A fruitful way of understanding Leibniz’s finite/infinite analysis strategy is in relation to his distinctive predicate-in- theory of truth. According to this theory all affirmative truths in subject-predicate form are such that the predicate is in some sense included in the subject. In a letter to Arnauld, Leibniz describes this principle thus: “in every true affirmative proposition, whether necessary or contingent, or particular, the notion of the predicate is in some way included in that of the subject.”144 Leibniz expresses his commitment to this principle in no uncertain terms, for he sometimes follows this characterization with

“otherwise I do not know what truth is.”145

Leibniz sometimes insists that the finite/infinite analysis follows from this predicate-in-subject theory of truth. His basic idea here is that the difference between necessary and contingent propositions can be

143 This is the period roughly around the 1680s. 144 Primary Truths, C 519/ AG 31. 145 Leibniz to Arnauld 14 July, A 2.2.56/ LA 81.

79 articulated in terms of the ways in which it can be shown that the predicate is contained in its subject. Leibniz writes: “And with this secret the distinction between necessary and contingent truths is revealed… in necessary propositions… the analysis… arrives at an equation that is an identity…. But in contingent propositions one continues the analysis to infinity through reasons for reasons, so that one never has a complete demonstration,”146 Necessary propositions can be demonstrated with a finite analysis and contingent propositions need an infinite analysis. To better understand this strategy, we need a better understanding of what Leibniz meant by ‘analyses.’

Leibniz, like other thinkers in the 17th and 18th centuries, makes a distinction between different kinds of proofs: analyses and syntheses. An analysis is a proof which begins with a complex conclusion and, by a series of permitted substitutions of terms and expressions,147 it terminates in self-evident axioms, or

“identities.”148 A synthesis, on the other hand, begins with a set of self-evident axioms, or identities, and by a series of permitted substitutions terminates in a desired complex conclusion.149 This broad description of analyses and syntheses is clear enough.150

In sum, at least at a general level of description it is clear what an analysis is for Leibniz. With this sense of analysis Leibniz presents a strategy for explicating the difference between necessary and contingent truths: a proposition is necessary if and only if it can be demonstrated by an analysis with a finite number of steps, and a proposition is contingent if and only if it cannot be demonstrated by a finite analysis – that is, the series of permitted substitutions of terms or expressions goes on infinitely.

146 On Contingency, Grua 302/ AG p. 28. 147 Steward 2014 argues that for Leibniz the only permitted substitutions are real definitions or parts of real definitions of a term. 148 The term ‘self-evident axioms’ is mine. I use it because I think it illustrates general similarities between the function that identities play in Leibniz’s views and the functions that axioms play in contemporary mathematical axiomatic systems. These similarities are, of course, only general. 149 Adams 1994: p. 27. 150 As far as I can tell, however, nowhere does Leibniz clearly lay out the set of all and only the permitted self-evident axioms, or identities, and the permitted substitution principles for either analyses or syntheses. Leibniz does address these issues and gives us into what identities and substitution principles. For example, this is one of the main purposes in his work Primary Truths, C 518-523/ AG 30-34. Unfortunately, the list of axioms and substitution rules is not a complete one, as far as I can tell. So, we are not in a position to actually perform either syntheses or analyses in a way we can be assured is within Leibniz’s expectations.

80

Robert Adams has persuasively argued that a good way of understanding Leibniz’s finite/infinite analysis strategy is as proof-theoretic demonstrability and indemonstrability.151 The main idea here is that, according to this strategy, to say that a proposition is necessary is to say that it is demonstrable, and to say that a proposition is contingent is to say that it is indemonstrable (when “demonstrability” and “indemonstrability” refer only to the restricted kinds of proofs that are analyses, as previously sketched). Adams applies this proto- intuitionistic understanding of contingency as indemonstrability to defend multiple Leibnizian tenets.152

My suggestion is that Leibniz’s finite/infinitely analysis strategy should be understood not merely as a proposal to explicate contingency and necessity by itself; it can also be understood as answering the demarcation question. In terms of the licit applicability of the logical strategy, we can understand this answer to the demarcation question thus: the opposite of a proposition P implies a contraction if and only if P can be proven by a finite analysis, and the opposite of a proposition P does not imply a contradiction if and only if there is no finite analysis for P (i.e., the analysis goes on infinitely). Alternatively, the finite/infinite analysis strategy can be utilized to demarcate the difference between a core and a superstructure of a complete individual concept.

This answer to the demarcation question is thus: for a substance S, and its complete individual concept C, and a predicate “being P” included in C, “being P” is included in C’s core if and only if the proposition “S is P” can be proven by a finite analysis, and “being P” is included in C’s superstructure if and only if the proposition “S is

P” cannot be proven by a finite analysis (i.e., the analysis for “S is P” goes on infinitely).

This certainly provides us with a clearly demarcated complete individual concept, and thus provides us with the theoretical tools to answer at least the second argument for superessentialism. I do not think this answer to the demarcation question will do, however. The main objection I have is that the resulting kind of contingency is not a plausible one.153 That is, understanding contingency as mere proof-theoretic indemonstrability entirely abandons the philosophical insight regarding the intuitive connection between the

151 Adams 1994. 152 Adams 1994: 22-50. 153 At the end of this chapter I will argue that my own answer to the demarcation question can incorporate Leibniz’s finite/infinite analysis account of contingency. I will argue that my account illuminates how the finite/infinite analysis strategy can indeed be seeing as a plausible answer to the demarcation question.

81 logical strategy and the essence of the subject. This answer to the demarcation question fails to do to the intuitive connection between that which makes a substance be what it is (i.e., that which answers the Platonic

“what is it?” question) and the logical strategy, for there seems to be no intuitive connection between being an essence and being demonstrable via finite analysis, or between being part of the superstructure and infinite analysis.154 This is indeed a significant philosophical cost, for, at least with regards to truths about substances, it is the plausibility of this intuitive connection that grounds the tenability of the logical strategy itself.

More specifically, it seems quite clear that the kind of contingency we get from this answer to the demarcation question does not matter for freedom. In other words, it seems clear that the quasi-intuitionistic proof-theoretic account of indemonstrability as contingency, independently of its intrinsic merits, does not explicate the nature of freedom. For, saying that an agent S freely performs an action ф in circumstances C only if the proposition “S ф’s in C” is proof-theoretically indemonstrable is not illuminating what freedom is.155 The resulting account of logical contingency as proof-theoretic indemonstrability is quite ingenious but still largely artificial, or, to use Adams’s own words, it is not “real contingency.”156

I wish to end this section by sketching what I take to be a plausible answer to the demarcation question in Leibniz’s system. The details of this answer will be provided in the following section. My basic suggestion for answering the demarcation question is the following. It is metaphysical reality, and not the logical realm alone, that should guide our judgments regarding which truths are to be allowed in the process of deriving a

154 In the secondary literature, there are several interesting proposals for understanding infinite analyses in such a way that only intuitively contingent truths have infinite analyses. See for example, Hawthorne and Cover 2000, Rodriguez-Pereyra and Lodge 2011, and Seward 2014. If successful, these accounts would show co-extensionality between intuitively contingent propositions and infinite analyses. In the context of answering the demarcation question, however, showing this kind of co-extensionality is not enough, for mere co-extensionality does not secure the intuitive connection between an essence and the logical strategy. 155 Sotnak (1999: 218) and Baxter (2000: 194-5) present similar criticisms. 156 Adams 1994: 29.

82 contradiction for the applicability of the logical strategy.157 That is, our understanding of logical contingency should mirror, and be guided by, our understanding of contingency at the metaphysical level.

At the metaphysical level Leibniz does have a distinction between an essence of a substance and the rest of the properties of the substance which follow from and are explained by this essence. Throughout his career Leibniz articulated this distinction in a number of different ways.158 A major characterization of this distinction is Leibniz’s division of a substance into its ‘primitive force’, ‘derivative forces’, which are just

‘modifications’ of the primitive force, and ‘perceptual states’.159 It is the primitive force, I will argue, that is best understood as a substance’s essence, at the metaphysical level, on Leibniz’s views. These distinctions are perhaps better understood in the context of Leibniz’s mature metaphysics. During his mature period, Leibniz famously postulates mind-like simple substances that he refers to as “monads.”160 Leibniz writes, “a monad, in itself and at a moment, can be distinguished from another only by its internal qualities and actions, which can be nothing but its perceptions (that is, the representations of the composite, or what is external, to the simple) and its appetitions (that is, its tendencies to go from one to another) which are the principles of change.”161 Here Leibniz make a fundamental distinction between appetitions and perceptions. The former are dispositions or tendencies or powers of a substance which in some sense serve as its inner principle of change; whereas the latter are the perceptual states which come and go in intrasubstantial change.

With this distinction between appetitions and perceptual states, we are in a better position to understand Leibniz’s account of primitive forces. Leibniz describes a substance’s primitive force as “a nature or an internal force that can produce in it, in an orderly way all the appearances or expressions it will have, without the help of any created being”.162 Or, in other words, a substance’s primitive force is “the internal

157 By the metaphysical level in this context I mean what is for Leibniz created reality at its most fundamental level, and by the logical level in this context I mean the realm of God’s ideas and their interconnections. 158 Leibniz expresses this core or essence of the substance in terms of ‘entelechies’ or ‘substantial forms’ and seems to think that at bottom they refer to the same underlying reality. 159 New System, G 4.478/ AG 139. 160 A succinct description of Leibniz’s mature metaphysics is presented in his Monadology (M). 161 Principles of Nature and Grace § 2, G 4.598/ AG 207. 162 New System, G 4.486/ AG 144.

83 principle of a substance that brings about change, or the passages from one perception to another”.163 A substance’s primitive force is, for Leibniz, the inner principle of change which explains why the substance undergoes all the change it undergoes and has all the perceptual states it has. A substance’s appetitions are just different modifications of its primitive force, or derivative forces. Thus, a substance’s primitive force is also a fit candidate for answering the Platonic ‘what is it?’ question in Leibniz’s system and a fit candidate for being a substance’s essence in my sense. Leibniz certainly suggests this much when he writes “And so, we must add a soul or a form analogous to a soul, or a first entelechy, that is, a certain urge [nisus] or primitive force of acting, which itself is an inherent law, impressed by divine decree.”164

Crucially for our purposes, for Leibniz the perceptual states of the substance follow contingently from its primitive force and its modifications, and are thus had contingently by the substance.165 Providing a detailed account of this kind of contingency at the metaphysical level, and of what it is for perceptual states to follow contingently from the substance’s primitive force, in Leibniz’s system is an enormous project all by itself.166

Thankfully the details of this metaphysical account of contingency are not needed for answering the demarcation question. What matters is that at the metaphysical level we have a story about an essence of a substance which explains the rest of the properties of the substance. Importantly, we can rely upon this metaphysical story to plausibly demarcate between a core and a superstructure in a complete individual concept.

With this demarcation, we can further distinguish between the truths which can be utilized in the process of deriving a contradiction for the licit applicability of the logical strategy. Furthermore, and importantly for our purposes, this answer to the demarcation question does indeed retain the intuitive connection between an essence of a substance, which explains the rest of the properties of the substance, and the logical strategy. This response, then, captures the basic philosophical insight undergirding the logical strategy itself while sketching of a plausible response to the argument(s) for superessentialism on the basis of Leibniz’s logical commitments.

163 M 15, AG 215. 164 On Nature Itself, G 4.504-16/ AG 162-3. 165 This is a great simplification of Leibniz’s view, but this simplified view will do for our purposes. 166 I tackle part of this project in chapter two of this dissertation.

84

2. A Leibnizian Response to Superessentialism

In this section I will develop Leibniz’s remarks into an account that retains an intelligible sense of contingency and that allows Leibniz to present a plausible response to Arnauld’s arguments. As already noted,

Leibniz’s explicit response to Arnauld’s argument for superessentialism, lamentably, is a bit obscure. Leibniz insists that he is not committed to the superessentialist conclusion and insists that “notions of individual substances” do “contain contingent truths,”167 and that this fact can be elucidated by appealing to “God’s free decisions (considered as possible)”168 which are also included in a substance’s complete individual concept.

That Leibniz’s response includes both of these elements is uncontestable, but how these elements come together to provide a plausible answer to Arnauld’s argument for superessentialism is far from clear. I will present an account that elaborates both of these elements. The account I will present is clearly Leibnizian, but it is also more precise than what Leibniz actually wrote. Thus, I present this response as more than just faithful exegesis of Leibniz’s text. This response is better characterized as a precisification of the general, and somewhat underspecified, Leibnizian response to be found in the text.

2.1 The Basic Response: Including Contingent Truths in Individual Concepts

One of the basic elements in Leibniz’s explicit response to Arnauld’s argument for superessentialism is that contingent truths can be included in complete individual concepts. Leibniz insists that “notions of individual substances… contain contingent truths.”169 In this subsection, I will advance a plausible way of making sense of this suggestion by Leibniz.

167 Leibniz to Arnauld 14 July 1686, A 2.2.71/ WFII p. 106. 168 Leibniz to Arnauld 14 July 1686, A 2.2.72-4/ WFII pp. 106-9. 169 Leibniz to Arnauld 14 July 1686, A 2.2.71/ WFII p. 106.

85

Leibniz’s insistence on contingency as a condition for freedom is many times tied with the power to do otherwise grounding the relevant kind of contingency.170 Leibniz writes, for example:

For speaking absolutely, our will is in a state of indifference, in so far as indifference is opposed to necessity, and it has the power to do otherwise, or to suspend its action altogether, both alternatives being and remaining possible.171 My proposal is to understand Leibniz’s account of contingency as requiring this power to do otherwise. This power is a power over alternative courses of action that are in some sense open to the agent. According to

Leibniz, an agent’s complete individual concept includes predicates denoting which course of action an agent is going to take, and importantly for our purposes, it also includes predicates denoting the power to do otherwise which remains unexercised. A few lines after the quoted passage, Leibniz continues: “it is true, and it is even assured from all , that a certain soul will not make use of this power in such a situation.”172 My basic suggestion is, then, that part of what it is to include contingent truths in complete individual concepts is to also include predicates denoting the unexercised powers to do otherwise. What a substance will do is included in the complete individual concept, but, crucially, so is the fact that the substance could have done differently. In sum, my basic suggestion is that including both of these facts in the complete individual concept is what it is for contingent truths to be included in complete individual concepts.

2.1.1 Unexercised Powers and Individual Concepts

Including contingent truths in individual concepts, by itself, does not yet answer the superessentialist challenge, to be sure. Adding an unexercised power as a further property of a substance only adds a further essential property, if the argument for superessentialism is not blocked. In this case, this power would just be

170 What the power to do otherwise amounts to in Leibniz’s system is a vexed topic. In chapter 3, I argued for what I take to be an important aspect of this power to do otherwise. In this chapter, I will not further enter this debate. The account advanced here should be compatible with different ways of making sense of this condition on freedom. 171 DM § 30/ AG p. 61. 172 DM § 30/ AG p. 61.

86 an essentially unexercised power and as such an inadequate ground for the kind of contingency that matters for freedom.

If this power to do otherwise is to be a legitimate power, and if it is to ground a credible kind of contingency that matters for freedom, it must be possible for it to be exercised. My initial suggestion for

Leibniz to avoid necessarily unexercised powers is for him to adopt a response similar to that endorsed by

Molinists in similar predicaments. Leibniz should insist that the content of an individual concept is counterfactually dependent on how the corresponding substance would utilize her powers to act. That is, an agent’s power to act is not a power directly over the content of its individual concept, but the content of its individual concept is counterfactually dependent upon how the agent would utilize her powers to act. A bit more precisely, in the cases of a free action, an agent S has the power to do otherwise such that if S were to exercise this power, S would have acted differently and as a consequence S’s complete individual concept C’s content would have always been different than the way it actually is.

An important limitation of this initial suggestion is that it seems quite plausible that concepts are individuated by their contents, and so that it is impossible for concepts to have predicates other than the ones they in fact have. But if it is necessary that concepts have the predicates that they actually have, then it seems implausible that having those predicates counterfactually depends on anything else. In order to avoid unpalatable consequences in this domain, I will develop the mentioned account of counterfactual dependence to respect the claim that concepts are individuated by their contents. I will provide this development in subsection 2.3.

Before developing this account, it is important to note that this initial suggestion of counterfactual dependence is in good company. Mainstream Molinists like Molina, Freddoso, and Plantinga, for example, offer analogous proposals in response to the Theological Argument.173 This is the argument that God’s foreknowledge of the future, divine , the necessity of the past, and the transfer of necessity principle preclude the power to do otherwise. In response to this argument, Molinists argue that God’s beliefs

173 I borrow the term “Theological Incompatibilism Argument” from Hasker 2001.

87 about the future are counterfactually dependent on how free agents would utilize their powers to act. They argue that free agents have powers to do other than what they actually do. These powers are not powers directly over God’s beliefs, but God’s beliefs do counterfactually depend on these powers by counterfactually depending on agents would utilize these powers to act. Powers to do otherwise will in fact not be exercised and

God believes that they will not be exercised, but if these powers were exercised God would have always believed that instead. A bit more precisely, a free agent S has the power P to do otherwise than he in fact does such that if S were to exercise P, instead of God believing that P will not be exercised, as He does, God would have always believed that P will be exercised instead.174

Furthermore, several contemporary compatibilists also adopt similar views in response to The

Consequence Argument.175 This is the argument that , the necessity of the past, and the transfer of necessity principle preclude the power to do otherwise. These compatibilists insist that free agents do have the power to do otherwise (in some sense). These powers are not powers directly over the laws of nature, but the laws of nature do counterfactually depend on these powers by counterfactually depending on how agents would utilize their powers to act. A bit more precisely, a free agent S has the power P to do otherwise than she in fact does such that if S were to exercise P, the laws of nature would have always been different than the way they actually are.176

My initial suggestion on Leibniz’s behalf is thus in good company. Despite the fact that free agents do not have direct power over God’s beliefs, mainstream Molinists think that God’s beliefs about the future do counterfactually depend on how free agents would utilize their powers to act. Also, despite the fact that free agents do not have direct power over the laws of nature, several contemporary compatibilists think the laws of nature do counterfactually depend on how free agents would utilize their powers to act. Similarly, I suggest, despite the fact that free agents do not have direct power over the content of their complete individual concepts, this content counterfactually depends on how free agents would utilize their powers to act.

174 See for example, Freddoso 1982, 2004, Plantinga 1986, Mavrodes 1984, 1997. 175 1975, 1983: Ch. 3. 176 See for example: David Lewis 1981, Graham 2008, and Pendergraft 2011.

88

2.1.2 An Interlude: Comparison with Contemporary Metaphysics of Modality

A good way of illustrating this initial suggestion is by comparing it to a similar response in contemporary metaphysics of modality. I believe that the basic Leibnizian response presented in the previous section has important similarities to Plantinga’s thesis that individuals have world indexed properties.177

Plantinga endorses the transworld identity thesis (the thesis that individuals exist in more than one possible world), and explains what is possible for an individual on the basis of what properties that individual has in the different possible worlds in which it exists.178 Some critics of Plantinga argue that individuals are world-bound

(exist in only one possible word) because individuals who exist in different possible worlds have different and incompatible properties in those worlds, but according to the principle of identity of indiscernibles difference in properties suffices for difference in numerical identity.179 Thus, these critics conclude, the principle of the identity of indiscernibles rules out the transworld identity thesis. Plantinga’s response is to postulate world indexed properties: that an individual S exists and has property P in possible world W1 and also exists and has

¬P in possible world W2 is to be understood as S having the world indexed property P-in-W1 and also having the different world indexed property ¬P-in-W2. Crucially, even though property P and property ¬P are incompatible, properties P-in-W1 and ¬P-in-W2 are not. Because these world indexed properties are not incompatible, a single substance can have them both without compromising the integrity of the principle of the identity of indiscernibles.

An analogy with temporally indexed properties helps illustrate the plausibility of world indexed properties. It is clear that the predicates “being P” and “not being P” are inconsistent because the property being P and the property not being P are incompatible. But a temporally qualified version of these predicates is

177 Plantinga 1974: Ch. 6. 178 It is essential to Plantinga’s view, however, that he does not reduce modality to non-modal truths or facts across possible worlds. His account retains modal facts as fundamental. 179 The principle of the identity of indiscernibles states that an object a and an object b are identical if and only if for any property P that a has b also has P. After Leibniz, philosophers have separated this biconditional into each of its conditionals and argued about their individual tenability. Following Leibniz, I do not intend to separate this biconditional.

89 clearly consistent: the predicates “being P-at-time-t1” is clearly consistent with a predicate “not being P-at-time- t2”. This is the case because the property being P-at-time-t1 and the property not being P-at-time-t2 are compatible: these properties can be instantiated in the same individual. It is impossible for Socrates to sit and not sit, but there is nothing problematic for Socrates to sit-at-t1 and not-sit-at-t2. Likewise, it is impossible for Socrates to sit and not sit, but there is nothing problematic for Socrates to sit-in-W1 and not-sit-in-W2.

The analogy with Plantinga’s view is a bit deeper because Plantinga further cashes out these world indexed properties as counterfactuals about the individual: for individual S to have the property P-in-W1 and the property ¬P-in-W2 is for S to be such that if W1 were actual S would have P and for S to be such that if

W2 were actual S would have ¬P. These two counterfactuals are also clearly compatible, so both can be true of a single individual and thus the principle of identity of indiscernibles and transworld identity are both preserved.

Plantinga’s response to the critics’ challenge is surely correct. My suggestion is that an analogous move on Leibniz’s behalf would enable him to retain an intelligible sense of the power to do otherwise without giving up on the claim that concepts are individuated by their contents. The contents of these complete individual concepts are counterfactually dependent on the powers to do otherwise of the corresponding substances, and this very fact is included in these complete individual concepts. This is analogous to Plantinga’s move to postulate world indexed properties which are ultimately explained in terms of counterfactuals holding of individuals.180

180 Mates 1972 also argues that including world-indexed properties in complete individual concepts enables Leibniz to avoid superessentialism while retaining the principle of the identity of indiscernibles. He argues, however, that Leibniz fails to provide a good “criterion for cross-world identification” (p. 116) which would effectively mark the distinction between essential and accidental (contingent) properties, so he opts for abandoning the project and insisting that Leibniz is committed to superessentialism. I take my proposal to identify a substance’s essence with its primitive force to adequately address this challenge in a Leibnizian fashion. Adams (1994: pp. 71-4) also argues that Leibniz can avoid superessentialism, and retain the principle of identity of indiscernibles, by including world indexed properties in complete individual concepts. However, importantly, he does not think that this is Leibniz’s view, but rather thinks that Leibniz mistakenly thinks that the principle of identity of indiscernibles, together with the predicate in subject principle of truth, entail that substances are world-bound. Adams’ main point is to argue that Leibniz is mistaken here. It is a strength of my view, I take it, that I do not attribute confusion to a gigantic intellect like Leibniz.

90

2.2 Cluster of Concepts and Modal Profile

In this subsection, I will develop the answer to the demarcation question adumbrated at the end of subsection 1.5. to make sense of Leibniz’s talk of multiple ‘possible-Adams’ as depicting alternative possibilities for Adam. In the following subsections, I will develop this account to incorporate Leibniz’s suggestion that

God’s possible decrees and included in individual concepts. The resulting account will enable Leibniz to retain an intelligible sense of contingency in his system and provide the theoretical tools to answer Arnauld’s challenge.

I propose that we understand Leibniz’s talk of ‘possible Adams’ as talk of a cluster of concepts which are united by Adam’s primitive force – which, as already mentioned I further propose that, for Leibniz, answers the Platonic question ‘what is Adam?’, and thus corresponds to the intuitive notion of ‘essence’ as that which makes Adam be what Adam is. Adam’s primitive force unites this cluster by this cluster including all and only cluster-members which share a core set of predicates that denote Adam’s primitive force. At the metaphysical level, Leibniz divides a substance into its primitive force, derivative forces (which are just modifications of this force), and the perceptual states that result from and are explained by this primitive force and its modifications.

Essential to my proposal is that, in addition to a core set of predicates denoting Adam’s primitive force shared by all members of the cluster, this cluster further represents all the possible ways in which Adam’s primitive force could be modified, and the states that follow from Adam’s primitive force and these possible modifications.181

Each member of this cluster represents a complete possible sequence of states of Adam as they follow from a complete possible sequence of modifications of Adam’s primitive force – or, for short, each cluster member represents one way things could be for Adam. By representing all possible ways things could be for Adam, this cluster of concepts represents Adam’s modal profile.

A couple of comparisons can illustrate my general meaning here. First, each member of the cluster of concepts roughly correspond to a single complete individual concept as understood by superessentialists.

181 As will be clear below, the picture I am proposing is significantly more complex than this first characterization makes it seem. Importantly, I will argue that the modal profile of a given substance captures everything that it is metaphysically possible for this substance, and, importantly, metaphysical possibility is not constrained by perceptual states of a substance being explained by the substance’s primitive force and its modifications.

91

Alternatively, each member of the cluster roughly corresponds to a set of propositions representing what is true of Adam in a single possible world as understood by those who endorse transworld identity and an analysis of modality in terms of possible worlds.182

Thus, according to the view I’m advocating here, to say that there are multiple ‘possible-Adams’ is to say that there are multiple concepts which are members of this cluster united by Adams’s primitive force. These possible-Adams just are the concepts which are members of the cluster united by Adam’s primitive force each of which represents one complete way things could have been for Adam. These possible-Adams, then, can plausibly be said to represent alternative possibilities for Adam.

A few words of clarification are in order at this juncture. Essential to the proposed picture is that I intend to use the name ‘Adam’ as a rigid designator denoting Adam’s primitive force, and thus Adam himself.183

Thus, the sentences ‘Adam possibly does not eat the apple’ and ‘Adam possibly eats the apple’ correspond to two different individual concepts in Adams’ cluster, but, crucially, both sentences are about the numerically identical Adam. This is the case with all such modal de re attributions, and thus enables us to see why all individual concepts which are members of a cluster correspond to a single substance.

This view is informatively contrasted with a view like Mondadori’s. As Mondadori sees it, Leibniz is a proto-counterpart theorist.184 Mondadori also talks about collections of individual concepts which share different predicates and which, he thinks, help ground modal de re attributions about substances by providing the truthmakers for the sentences expressing these attributions. Importantly, Mondadori argues that for Leibniz the name ‘Adam’ is not a rigid designator, but rather refers to whichever concept-counterpart of ‘our Adam’

(rigid designator expression denoting the Adam in the actual world) makes true the context-dependent de re modal attribution using the name ‘Adam’ (i.e., “Adam possibly does not eat the apple”).185 Mondadori argues

182 Grimm 1970, for example; and to some small extent also O’Leary-Hawthorne and Cover 1992. 183 In contemporary discussions, a ‘rigid designator’ is an expression that denotes the numerically identical individual in different possible worlds – see Kripke 1980. 184 Mondadori 1973, 1975, and 1985. 185 Mondadori 1973: 75-83.

92 that this proto-counterpart theory enables Leibniz to account for an intelligible sense of contingent de re modal attributions about ‘Adam’ which is nonetheless consistent with superessentialism regarding ‘our Adam.’

There are some general structural similarities between Mondadori’s view and the one I am advocating on Leibniz’s behalf. These general structural similarities, however, are predicated upon fundamentally different which ground fundamentally distinct conceptions of contingency as a condition for freedom. Most crucially, perhaps, is the presence of something akin to transworld identity in the view I advocate. Because names are rigid designators, it makes sense to talk about the numerically identical substance in different possible situations – as described by different individual concepts of this substance.

A further difference between Mondadori’s proposal and the one I am here proposing is the semantic function that different individual concepts play. Mondadori’s explicit aim is to rely on ‘possible-Adams’, which are just different individual concepts, to provide an intelligible semantics for modal language by providing truthmakers for modal de re attributions. I do not wish to follow Mondadori’s interpretation here. I think that this kind of semantic story is importantly un-Leibnizian, at least in some respects. Leibniz’s conception of truth is not semantic in our general contemporary sense. Leibniz’s predicate-in-subject principle of truth requires that all true propositions, in subject predicate form, are such that the predicates are in some sense included in the subjects. Arguably this predicate-in-subject principle of truth is, for Leibniz, a logical version of the principle of sufficient reason. This logical principle of sufficient reason requires that the reason for a proposition being true is precisely that the predicate is included in the subject. Somewhat anachronistically, this Leibnizian principle can be compared with Kant’s more restricted thesis that what makes analytical judgments true is precisely that the predicate is included in the subject.

That there is something in reality that corresponds to what true propositions express is something that

Leibniz would be happy to accept. Leibniz, however, would not be impressed by someone arguing that it is this correspondence with reality that explains why a given proposition is true by providing a truth-maker for a sentence in modal language. This semantic story would not impress Leibniz precisely because it fails to provide

93 an a priori reason186 for the truth of the proposition. Leibniz’s principle of sufficient reason at the logical level requires an a priori sufficient reason for its being true – this explanation is ultimately dependent on the nature of the subject of the proposition, Leibniz insists.187

As I see it, this Leibnizian conception of truth is applicable to modal language as well. According to the view I am advocating, the cluster of concepts united by Adam’s primitive force does represent everything that is possible for Adam, but it does not explain why these things are possible for Adam by providing truth- makers for de re modal attributions about Adam. In fact, the cluster does not explain what is possible for Adam at all. Instead, and crucially, it is Adam’s primitive force and the ways in which this primitive force could be modified, and what would follow from this, that explain why the cluster of concepts is the way it is. In other words, it is Adam’s primitive force that explains Adam’s modal profile. This is so precisely because a priori explanation relations at the logical level mirror, and depend upon, a priori explanations relations at the metaphysical level. That is, just like at the metaphysical level Adam’s primitive force (and its modifications) explains the states of the substance, so at the logical level the core predicates denoting Adam’s primitive force explain why the concept has the other predicates it has – namely those that denote the states of the substance which follow from, and are explained by, the substance primitive force and its modifications. Thus, the logical realm just mirrors, and depends up, this explanatory story at the metaphysical level.

2.3 Primitive Force, Multiple Individual Concepts and PSR

In the previous subsection, I introduced the notion of a cluster of individual concepts, representing the modal profile of a single corresponding substance, by saying that these concepts represent the different ways in which the substance’s primitive force could be modified and the perceptual states that would follow

186 For Leibniz, an a priori explanation is more than merely an explanation that can be given independently of ; crucially, for Leibniz an a priori explanation is an explanation of an effect in terms of its causes such that pointing to the causes fully explains why the effect is the way it is and not otherwise. Furthermore, for Leibniz not all causes are efficient causes. Leibniz heavily relies on teleological causes and arguably also on formal causes. 187 G 4.192/ AG p. 270.

94 from these possible modifications. In this section, I wish to fill in some of the details of this proposal. An important issue I wish to address is how the principle of sufficient reason fits with this general proposal.

The concern regarding the fit between a cluster of individual concepts and the principle of sufficient reason is perhaps better articulated as a potential objection. It appears that the principle of sufficient reason allows for only one possible series of perceptual states which follow from a single primitive force and its modifications. Leibniz’s notorious doctrine of ‘traces and marks’ can be recruited to bolster this initial appearance. Roughly put, this doctrine states that at any time a substance in some sense contains ‘traces,’ of everything that has happened to it in the past, as well as ‘marks,’ of everything that will ever happen to it in the future. Leibniz articulates this point in the following way: “Thus when we consider carefully the connection of things, we can say that from all time in Alexander’s soul there are vestiges of everything that has happened to him and marks of everything that will happen to him… even though God alone could recognize them all.”188

Part of Leibniz’s theoretical motivation for this peculiar doctrine is clearly connected to the principle of sufficient reason. Leibniz insists that every effect can be explained in terms of its causes,189 and so that perceptual states of a substance are explained in terms of the substance primitive force and its modifications.

Given this, at least at first glance, it appears that the principle of sufficient reason rules out all but one possible series of perceptual states for a single primitive force, and thus for a single substance.

For the purposes of this dissertation, I do not wish to take a stand on whether the principle of sufficient reason does rule out all but one possible series of perceptual states for a given primitive force. That is, I do not wish for the project of this dissertation to hang on the answer to this question: one can endorse the account of cluster of individual concepts and go either way with regards to this question. I do wish, however, to present a way of making sense of how the same primitive force can be compatible with the principle of sufficient reason and more than one possible series of perceptual states. Doing so would make it clear that the proposal is consistent with either alternative.

188 DM § 8. 189 DPG § 16.

95

Here is my suggestion. First, Leibniz does not think that the series of perceptual states of a substance follow from merely from the substance’s primitive force, and its modifications. Rather, Leibniz insists that these perceptual states also depend upon the ‘initial state’ that God ‘bestows’ upon the substance. Leibniz writes: “Everything occurs in every substance as a consequence of the first state which God bestowed upon it, and extraordinary concourse excepted, his ordinary concourse consists only of preserving the substance itself in conformity with its preceding state and the changes that it bears.”190 Leibniz here notes that the general claim about perceptual states following from the primitive force and its modifications does not apply to the cases of ‘extraordinary concourse’ or supernatural graces. I will bracket the issues regarding supernatural graces for the purposes of this discussion, so the claims I make here should be taken with this important qualification in mind. In any case, the postulation of these ‘initial states’ of a substance opens up conceptual space for a single primitive force to have different possible series of perceptual states precisely by having different initial states. Furthermore, and importantly, the ‘traces and marks’ doctrine is also preserved, for the different possible series of perceptual states would be different from the beginning, and so whatever the story about traces and marks amounts to in the case of one possible series there is no principled reason for not providing a similar story in a different possible series of perceptual states.

It is important to note that Leibniz clearly wanted such possibilities. As we saw in chapter one, Leibniz endorses a version of the Contingency and Prevolitional conditions regarding subjunctive conditionals of human freedom. Part of the theoretical motivation for these tenets was preserving God’s control over creation by allowing God to create the same individual in different possible circumstances.191 Leibniz writes, for example: “God considers what a man would do in such and such circumstances; and it always remains true that

190 G II 91-92/LA 115. 191 This is, of course, only part of the story. The same sort of providential control can be secured by having numerically distinct possible individuals in different circumstances, instead of the numerically identical individual in different circumstances. I do not mean to deny this. Part of the larger theoretical motivation includes safeguarding a more plausible conception of freedom, by providing a more plausible conception of contingency as a condition for freedom, and thus safeguarding human even given divine providential control. Thus, allowing for God to place the same individual in different possible circumstances helps God having providential control while retaining a more plausible sense of human moral responsibility even given this providential control.

96

God could have placed him in other circumstances more favorable, and given him inward or outward succor capable of vanquishing the most abysmal wickedness existing in the world.”192

A plausible way of reading passages like this one is that in considering what a man would do in different circumstances God is considering different possible series of perceptual states which would follow from the man’s primitive force, and its modifications, depending on which initial state God decides to bestow.

Furthermore, that God has the power to ‘place’ the same individual in different circumstances, then, can be understood as God having the power to bestow a different initial state to the substance. By God bestowing a different initial state to a substance, the agent will effectively find itself in different circumstances for action; this is the case not only initially but as the series of perceptual states unfolds. Thus, in bestowing a different initial state to the substance God can adequately be understood to be ‘placing’ the agent in different circumstances for action depending on the series of perceptual states that would follow from this initial state and the primitive force.

An example can help make this proposal more concrete. Consider again the example of Susan we encountered in chapter 2. Susan is considering what career path to take, and her main alternatives are a career in economics and a career in music. She adequately recognizes her aptitudes and talents as well as the goods associated with each of her options. In the actual world, Susan comes to the conclusion that a career in music is best for her because this prospect affords more opportunities for artistic expression and she recognizes that this would contribute more to her well-being than a higher paycheck would – call this deliberation situation D1.

Being a rational agent Susan chooses a career in music on the basis of being the best course of action open to her. In chapter 2, I argued that Susan’s choice counts as contingent because it is the result of a deliberation process in which several options where open to her in such a way that had she come to the conclusion that one of these options were best she would have chosen that one. That is, Susan has the power to determine herself on the basis of her judgment of the best – whether this judgment of the best applies to one option or the other

192 T § 103.

97

– and these kinds of powers regarding different possible courses of action ground the contingency of her choosing the actual course of action. I called these kinds of powers ‘agential powers.’

We can incorporate this story about agential powers grounding contingency to illustrate the present proposal. In the actual world, God bestow upon Susan an initial state S1 which eventually unfolded to include the deliberation situation just described. It was within God’s power, however, to bestow upon Susan a different initial state S2 which would eventually unfold to a slightly different deliberation situation. On this different deliberation situation, let us suppose, it would have been more salient in Susan’s deliberation the economic goods connected with a career in economics. More precisely, instead of concentrating on how the monetary goods of an economics career would contribute to her well-being, she also takes into consideration how these monetary goods can be used to contribute to the well-being of others. In this alternative deliberation situation, call it D2, then, Susan recognizes and emphasizes the kinds of goods that would come about by her having access to more money. She still recognizes that a career in music would contribute more to her well-being than a career in economics, but now she also takes into consideration the good to others that her extra cash would bring in. She recognizes, for example, that the extra cash can alleviate a significant amount of lamentable poverty that plagues many people in the world. Given these considerations, Susan comes to the conclusion that it would be best to pursue a career in economics, and she acts accordingly.

There are several things worth noting about these examples regarding Susan. First, these examples illustrate that the account of agential powers articulated in chapter 2 is a compatibilist account. Several agential powers are present in any deliberation situation – one per open option under consideration, only the power associated with the action judged best will be exercised, and, importantly, for any of the other agential powers to be exercised things would have to have been different. Put differently, the agential powers that are not exercised in a given deliberation situation will only get exercised in a different deliberation situation. Agential powers are not libertarian powers. Second, these examples illustrate how has the power to instantiate different members of a cluster of individual concepts without violating the principle of sufficient reason or the ‘traces and marks’ doctrine. We can think of each of these different possible series of perceptual states, with different initial states, as corresponding to different members of a cluster of individual concepts. Thus, we can see how

98

God’s power to bestow different initial states on a substance is part of the power God has to instantiate different members of a cluster of individual concepts. God’s power to instantiate a given member of a cluster would involve more than merely the power to bestow an initial state, for creation involves more than this, but it will involve this much. For example, if God wishes for Susan to take a career in music, God can instantiate the member of the cluster of Susan’s individual concept that includes the initial state S1 and thus that describes

Susan as taking the career path in music. Alternatively, if God wishes for Susan to take a career in economics,

God can instantiate the member of the cluster of Susan’s individual concepts that includes the initial state S2 and thus that describes Susan as taking the career path in economics. Part of God’s power to instantiate different members of a cluster of individual concepts thus includes God’s powers to bestow different initial states on the corresponding substance. All of this, importantly for our purposes, without compromising the principle of sufficient reason, the ‘traces and marks’ doctrine, or the principle that the perceptual states of a substance follow from the substance’s primitive force and its modifications.

Furthermore, and importantly for the general project of this dissertation, this model enables us to see clearly how the content of the actual world depends upon both what God would do and what creatures would do. The content of the actual world depends upon which initial state God decides to bestow upon the substance; the content of the actual world also depends upon what would follow from the substance’s primitive force, and its modifications, given the initial state God bestows upon it. Crucially, a subset of these facts about what would follow from a substance’s primitive force, given different initial states, are encoded in the subjunctive conditionals of creaturely freedom that God knows prevolitionally. Regarding the examples of

Susan, that the actual world includes Susan taking a career in music depends upon two factors: God’s decision to bestow initial state S1 on Susan, and the subjunctive conditionals of freedom which describe what Susan would do if D1 were actual. Similarly, had God decided to create a different possible world, the one in which

Susan takes the career in economics, the content of that possible world would also depend upon both God’s decision to bestow initial state S2 to Susan and the subjunctive conditionals of freedom that describe how Susan would act if D2 were actual. This, of course, also applies to all deliberation situations that follow from either

S1 or S2, not just D1 and D2, and the subjunctive conditionals that encode what the agent would do in these

99 different deliberation situations. In general, then, the content of “the actual world” (read de dicto) depends upon both what God would do and what creatures would do if placed in different circumstances, much like in traditional Molinist fashion.

This picture about how the same primitive force can be compatible with several possible series of perceptual states and the principle of sufficient reason fits quite well with the rest of the project of this dissertation. I do not wish, however, to make it part of the official project of this dissertation. I wish, rather, to leave open the exact details of how God’s power to place the same individual in different possible circumstances for action fits with the principle of sufficient reason, and more generally whether more than one member of the cluster of individual concepts obeys the principle of sufficient reason. I hope for my general account to be plausibly detachable from these details and thus compatible with several such alternatives.

Furthermore, I also do not wish for the account of cluster of individual concepts to depend upon the account of agential powers. The account of agential powers advanced in chapter 2 is aiming to provide an account of the nature of contingency at the metaphysical level by identifying the kinds of powers that ground the kind of contingency that matters for freedom on Leibniz’s system. The account of cluster of individual concepts in this chapter, however, aims at something significantly different. The goal of this latter kind of account is merely to answer the challenge that commitment to individual concepts does away with an intelligible account of contingency at the metaphysical level. In other words, the cluster of individual concept account is meant to answer a logical challenge to the intelligibility of contingency at the metaphysical level, but answering this challenge does not itself require providing an account of contingency at the metaphysical level, it just require that there be such a story. Thus, it is important that the cluster of individual concepts account does not depend upon a particular conception of powers grounding contingency at the metaphysical level, but remain theoretically independent from, and compatible with, several such stories.

Furthermore, I do intend my account of cluster of individual concepts to capture everything that is metaphysically possible for a substance – its entire modal profile. As I see it, metaphysical possibility is the broadest kind of possibility for Leibniz. Thus, I do intend for my account of clusters of individual concepts to

100 capture cases of clear violations of the principle of sufficient reason, which I take to be metaphysically possible.

Thus, regardless of how many members of the cluster obey the principle of sufficient reason, many members of the cluster will not, for they will encode merely metaphysically possible alternatives for the substance – alternatives which do not imply a contradiction.

2.4 Privileged Concept and Counterfactual Dependence

In subsections 2.1 and 2.1.1, I presented an initial suggestion for understanding Leibniz’s claim that complete individual concepts include contingent truths. The basic idea is that included in individual concepts are predicates denoting unexercised powers to do otherwise and that the content of individual concepts counterfactually depends on how agents would utilize her powers to act. We are now in a position to develop this initial suggestion to accommodate the plausible claim that concepts are individuated by their contents.

The key idea is that there is a privileged member of a cluster of concepts depicting the modal profile of a given substance. I will continue to use Adam as a representative substance. To say that a member of the cluster of concepts of Adam is the privileged member is to say the following: if God were to create Adam, it is the privileged member of the cluster that represents everything that will ever happen to Adam – while the other members of the cluster will only represent non-actualized alternative possibilities for Adam. God, of course, has the power to instantiate members of the cluster other than the privileged one, but subjunctive conditionals of divine freedom select a single member of the cluster as the one that would be chosen if the substance were created. Thus, our first grip on what it is for a member of a cluster to be the privileged member is an extrinsic relation it bears to what God would do were God to create the corresponding substance.

Furthermore, it is privileged members of a given cluster that, in several important respects, will play the role of a substance’s complete individual concept. For example, as already mentioned, if the substance is created it is the privileged member that includes predicates denoting everything that will ever happen to the substance. The privileged member of the cluster also includes predicates denoting unexercised powers to do otherwise. Part of what it is for contingent truths to be included in individual concepts is for both of these

101 kinds of predicates to be included in the privileged concepts. As previously noted, more is needed to rule out necessarily unexercised powers. For this purpose I introduced a counterfactual dependence between the content of individual concepts and how agents would utilize her powers to act. For example, regarding Adam, it must be true that Adam has the power to do otherwise such that if Adam were to exercise this power his individual concept would have always had predicates denoting this alternative possibility instead.

My proposal here is that we should understand the counterfactual dependence of the content of

Adam’s individual concept as counterfactual dependence over which of the members of the cluster is the one that represents what will happen to Adam – i.e., which of the members of the cluster gets to be the privileged one. That is, which member of the cluster gets to be the privileged one counterfactually depends upon how the corresponding agent would utilize her powers to act. This is our second, and independent, grip on what it is for a member of a cluster to be the privileged one. It is both, what God would do and what agents would do, that settles which member of the cluster is the privileged one. This dual dependence need not be puzzling.

In fact, it happens all the time. Whether my wife has tacos for dinner on Tuesday, for example, depends upon both whether I would offer to make tacos for Tuesday, and on whether she would be happy with such a meal, for example. Each of us has the power to do something such that this event will not come about: I can suggest cooking something else, and she can ask me to cook something else, for example. Neither of us have full control over this event and it coming about depends upon both what my wife and I would do.193

To illustrate this part of the proposal, consider two individual concepts C1 and C2 that are members of the cluster of individual concepts representing Adam’s modal profile. Suppose C1 includes the predicate ‘eating the apple’ and C2 includes the predicate ‘not eating the apple.’ Suppose C1 is the privileged concept. Thus, if

Adam were created, C1 would represent everything that will ever happen to Adam, and C2 would represent a mere non-actualized alternative possibility for Adam. Adam will eat the apple, but he also has the power to do

193 The reader might worry that this example is insufficient for the case involving God, and that mere coincidence of subjunctive conditionals of freedom is not informative enough in the divine case. If so, the model presented on the previous subsection involving different initial states God can bestow on the same substance, and what would follow from these initial states can be incorporated here as well. This model would present more details about the relevant kinds of subjunctive conditionals involving God and creatures that come together in identifying the privileged member of the cluster.

102 otherwise. Given that this power is not necessarily unexercised, it is possible for Adam to exercise it. If Adam were to exercise this power to do otherwise, then Adam would not eat the apple, and importantly for our purposes, it would have always been true that concept C2 would have been the privileged concept which would have represented everything that will ever happen to Adam (if Adam were created) and C1 would have only represented a mere non-actualized alternative possibility for Adam. Thus, the counterfactual dependence, grounded in Adam’s powers to act, is ultimately a counterfactual dependence over which of the members of the cluster of concepts gets to be the privileged member and which members of the cluster get to only describe non-actualized alternative possibilities for Adam.

This account of counterfactual dependence respects the plausible claim that concepts are individuated by their contents. For it does not entail or require that individual concepts have, or are capable of having, predicates other than the ones they actually have. What this account requires is merely that it is possible for different members of the cluster to be the privileged member of the cluster. But, crucially, this is an extrinsic property of concepts and not part of their content. Thus, it is compatible with this account that concepts are individuated by their contents. This account, I submit, is a plausible way of making sense of Leibniz’s claim that contingent truths can be included in individual concepts.

2.5 Including God’s Possible Decrees in Individual Concepts

The final element to be included in my proposal is Leibniz’s remark that God’s possible decrees are included in individual concepts, and that this plays a role in securing an intelligible sense of contingency. I will further argue that this is important for determining which member of the cluster is the privileged one. The proposal advocated here is, again, best understood as a precisification of an underspecified text instead of just faithful exegesis.

103

For Leibniz, there is no inter-substantial causation: only God causally interacts with any given substance.194 Furthermore, for Leibniz, there are multiple ways in which God can causally interact with substances: God sustains substances in being, concurs with their activities, and provides different kinds of divine assistances or graces to the substance.195 Importantly, God arranges and plans what to do with a substance precisely by deciding whether to provide or not provide these sorts of possible causal interactions with the substance. God’s possible free decrees thus include all these possible kinds of interactions between

God and a given substance. Leibniz’s main insight, I propose, is that by including God’s possible free decrees in complete individual concepts these concepts would also reflect the different possible ways in which God can causally interact with particular substances and thus the resulting possible ways things could turn out for a substance.

How can God’s possible free decrees be included in individual concepts? To answer this question, the key idea I will like to develop builds on Leibniz’s distinction between predicates being contain and virtually contained in subjects in true propositions.196 Importantly for our purposes, this distinction also applies to complete individual concepts. Some predicates are contained simpliciter and others are merely virtually contained in complete individual concepts. I propose that the crucial difference is that those that are contained simpliciter are predicates which can be deduced from concepts without requiring God’s possible free decrees, and those predicates that are merely virtually contained are those that can only be deduced from concepts with God’s possible free decrees.

Let me try to make this basic picture more precise. I propose to understand the predicates denoting a substance’s primitive force as contained simpliciter in its complete individual concept. Furthermore, I shall say that a substance’s primitive force provides reasons for God to causally interact with the substance in particular ways.197 I further propose to understand the predicates that denote the states of a substance as merely virtually

194 DM § 14. 195 DM §§ 14, 16, 31, DPG § 27. 196 DM §§ 8, 13. 197 My proposal here is similar to and builds upon a key idea developed by Baxter (2000). Baxter also aims to provide an account of conceptual contingency in Leibniz’s system. His account is significantly different from the account of contingency I develop here, however.

104 contained in its complete individual concept. More precisely, I will say that a predicate ‘being P’ is virtually contained in a complete individual concept C of a substance S if and only if a reason R is contained simpliciter in C such that R constitutes a reason for God to bring it about that S has P.198 Thus, what can be deduced simpliciter from C is a conditional with the antecedent expressing reason R and God’s possible free decree to interact with

S, partly on the basis of R, to bring it about that S has P, and the consequent expressing that S has P. For example, to say that a predicate “freely phi-ing” is virtually contained in substance S’s complete individual concept C is to say that the following conditional can be derived from C: “if S were in circumstances M, then

S would freely phi” – where M includes reason R provided by S’s primitive force and also God’s possible causal interactions with S partly on the basis of R with the intent to bring about that S freely phi-s.

Each member of the cluster includes some collection of these kinds of possible free decrees by God.

These possible divine decrees, crucially, come in two varieties: a) possible divine decrees representing what God could do, and b) possible divine decrees representing what God would do. It is possible divine decrees of variety

(b) that help determine which member of the cluster is the privileged one. In other words, the privileged member of the cluster is the one that not only includes merely what God could do (the kinds of causal interactions that God could provide on the basis of R) but it includes what God would do (the kind of causal interactions that

God would provide on the basis of R). Thus, a member of the cluster is the privileged one partly because it includes God’s possible free decrees about how God would causally interact with the reasons provided by the substance’s primitive force.

This distinction between possible divine decrees, of varieties (a) and (b), is available to Leibniz because he endorses a version of a tenet essential to Molinism.199 On Leibniz’s view, God has prevolitional knowledge of subjunctive conditionals of freedom about Himself (i.e., “if God were in C, God would freely phi”). This is an important difference between Leibniz’s views and traditional Molinists.200 Molina himself, for example,

198 Baxter (2000: p. 195) provides a very similar characterization of virtual containment of predicates in concepts. 199 I understand Molinism as the view that God has prevolitional knowledge of contingent subjunctive conditionals of freedom. I defend the thesis that Leibniz endorses a version of both tenets essential to Molinism in chapter one of this dissertation. 200 Another major difference is that traditional Molinism is often championed precisely because it provides a plausible framework in which robust libertarian freedom can be reconciled with orthodox theistic beliefs (robust divine

105 insists that God does not have such prevolitional knowledge about Himself, for such prevolitional knowledge would undermine divine freedom.201 For Leibniz, however, this kind of prevolitional knowledge is compatible with divine freedom. Because of this important difference, traditional Molinists can only include what God could do in the antecedents of subjunctive conditionals of human freedom, but Leibniz can also include what

God would do. That is, the antecedents of Leibnizian subjunctive conditionals of human freedom can include not only what God could free do in some circumstances C, but what God would freely do in C.

It is because of this version of a central tenet of Molinism that Leibniz is able to make a distinction between the mentioned two varieties of possible divine decrees, namely varieties (a) and (b). Crucially, it is because these subjunctive conditionals of divine freedom have determinate truth values prevolitionally that they can be included as part of the content of substance’s complete individual concepts prior to God actually deciding to do anything – crucially, prior to God actually deciding to create and causally interact with any given set of possible substances. This account enables us to understand how each member of the cluster concepts of a possible substance includes God’s possible free decrees, but only the privileged member of the cluster includes the subset of possible divine free decrees that express what God would do (i.e., it includes the possible causal interactions God would provide, if God were to create the possible substance). It is these facts about what

God would to, which are included in individual concepts themselves, together with facts about what agents would do, that make the privileged member of a cluster be the privileged one.

foreknowledge, predestination, grace and motives for election). Crucially, Leibniz rejects this robust account of libertarian freedom. 201 Molina insists that God’s knowledge about what He would do is postvolitional (Co IV.52.13.). Molina worries that God’s freedom would be compromised if He knew what He would do in a particular set of circumstances prior to actually willing anything. Such infallible prevolitional knowledge about Himself would render His decisions necessary and thus undermine His freedom, Molina insists (Co IV.53.19.). This claim does not introduce a double standard for human and divine freedom. The worry is a general one. Any agent who knows prevolitionally, and with certainty, what she would do in a particular set of circumstances is thereby deprived of the contingency of her actions and thus of her freedom (Co IV.52.12, IV.53.19.). Humans cannot, in principle, have this kind of knowledge about themselves (Co IV.52.11.), however, so middle knowledge places no legitimate threat to human freedom.

106

2.6 A Leibnizian Response to the Argument(s) for Superessentialism

We are finally a position to provide a more detailed response to the arguments for superessentialism.

Regarding Arnauld’s version of the argument, Leibniz need not accept Arnauld’s assumption that for every

‘possible-Adam’ there is a corresponding complete individual concept. In fact, according to the picture advanced in this chapter, possible-Adams just are complete individual concepts. It can indeed be granted that these concepts are individuated by their contents and are thus numerically distinct; but this does not imply that

Adam has all the properties that he has essentially. Each of these numerically distinct individual concepts represents an alternative possibility, or way things could be, for the numerically identical Adam.

A second premise of Arnauld’s argument that we are in a position to deny is what I labeled Individuating

Condition 2 – namely that substances are individuated by their complete individual concepts. Essential to the view I am advocating here is that it is a substance’s primitive force that suffices to individuate the substance.

This might initially seem un-Leibnizian, but it is in fact something that Leibniz is independently committed to in discussion about identity through time. That is, Leibniz seems to already be committed to the thesis that it is primitive forces that individuate substances in order to make sense of temporal alteration not changing the numerical identity of substances.202 My proposal here is merely to extend this Leibnizian commitment about individuation from the temporal to the modal case.

We are also in a position to explain why the second argument for superessentialism makes illicit use of the logical strategy. It is truths regarding the nature of a substance’s primitive force, and the reasons it provides for God to causally interact with it, that are allowed in the process of deriving a contradiction for the licit application of the logical strategy. That is, the only truths that are allowed to play a role for the licit applicability of the logical strategy are the ones that are contained simpliciter in the complete individual concept, for it is these predicates that express the essence of the substance at the metaphysical level. The predicates that are merely virtually contained in complete individual concepts, and that denote contingent properties, are not allowed to

202 Several other commentators have noted this. See, for example, Baxter 2000, Adams 1994, and O’Leary-Hawthorne and Cover 1992.

107 play a role in the derivation of a contradiction for the licit application of the logical strategy precisely because these predicates cannot be derived from the complete individual concept without God’s possible free decrees

– individual concepts only contain simpliciter reasons for God to causally interact with this substance to bring about that these predicates hold. This is then a plausible answer to the demarcation question for the licit applicability of the logical strategy.

Conclusion

One of the most vexed topic in Leibniz scholarship is his account of contingency. One important obstacle in this topic is that Leibniz endorses theses which seem to preclude any intelligible sense of contingency. One important such thesis is his distinctive doctrine that every substance has an individual concept that includes predicates denoting everything that will ever happen to the substance. At least initially, it seems that this doctrine entails that substances cannot be otherwise than the way their individual concepts describe them. In this chapter, I have advanced a novel way of understanding Leibniz’s doctrine of individual concepts. I argue that this account enables Leibniz to retain an intelligible sense of contingency grounded in something analogous to transworld identity. If this account is granted, it becomes clear how Leibniz can avoid superessentialism even given his commitment to individual concepts.

108

Chapter 4: Objections and Replies

INTRODUCTION

In the present chapter I will address a few important objections to the general proposal presented in this dissertation. In the process of answering these objections, two important goals will be attained. One, it will become clearer how the proposal fits with other major elements in Leibniz’s thought. Second, it will become clear that there are several possible precisifications of the proposal advocated in chapter three; these different precisifications come with trade-offs between different philosophical and theological Leibnizian commitments. I will also endorse one of these precisifications as my interpretation of Leibniz’s views. The tenability of the content of previous chapters, however, does not depend on the tenability of any one of the relevant precisifications.

The prominent objections I will be engaging with in this chapter are the following. First, postulating a cluster of concepts associated with a single substance might seem to compromise Leibniz’s principle of the identity of indiscernibles – the principle that an object a and an object b are identical if and only if for any property P if a has P, then b has P as well. More precisely, it seems that there being multiple individual concepts of the same substance enables God to instantiate more than one member of this cluster, and thus it seems that this enables God to create the same individual more than once and with different properties. This would clearly be a violation of Leibniz’s principle of the identity of indiscernibles for there would be discernible beings which would be numerically identical nonetheless. I shall label this objection the ‘PII Violation Objection.’

109

The second prominent objection I will be addressing is related to Leibniz’s theodicy, or response to the problem of evil. Important to Leibniz’s theodicy is the claim that creatures have no ground for complaining to God for allowing them to , for if they did not sin, they would not be who they are. Leibniz writes: “But someone else will say, why is it that this man will assuredly commit this sin? The reply is easy: otherwise it would not be this man.”203 Thus, part of Leibniz’s response to the problem of evil is the claim that humans have no grounds for complaining to God for allowing them to sin precisely because sinning is part of their notion and they would not be who they are if they did not sin. A plausible and somewhat common way of reading Leibniz here is as insisting that every property had by an individual is essential to it: the reason an individual would not be who the individual is, if anything were different, is that it is not possible for this individual to be otherwise than the way it is. I shall call this ‘the Superessentialist Interpretation.’ A major task of my dissertation has been to argue against this Superessentialist Interpretation. If my view is correct, this interpretation is false. Unfortunately for my view, then, I can no longer provide the plausible Superessentialist

Interpretation as part of Leibniz’s theodicy. Thus, I must provide an alternative plausible interpretation of

Leibniz’s claims that does not betray the philosophical motivation behind Leibniz’s reply to the problem of evil. I shall label this objection the ‘Theodicy Compromised Objection.’

The final objection I will be addressing in this chapter is the following. Essential to my proposal in chapter three is that God’s possible free decrees are important for individuating different members of the cluster of concepts, and these members of a cluster collectively represent all alternative possibilities for a single substance. Thus, God’s will is involved in distinguishing between alternative possibilities for agents. Because of this it seems that the contingency of human actions depends upon the contingency of God’s free actions.

Thus, it seems that my account is buck-passing in problematic ways.

There are two different levels to this buck-passing objection. On the one level, the mere fact that the contingency of human actions depends upon the contingency of divine actions seems problematic. This is so because contingency is a condition for freedom. Yet, it seems untenable that the fact that God could have

203 DM § 30, see also A 2.2.77.

110 done otherwise contributes to my freedom. That the contingency of God’s actions contributes to God’s actions being free seems intuitive, but that the contingency of God’s actions contributes to my actions being free does not. This seems problematic. The problem appears at a different level as well. Because God’s nature is essentially good and wise, it seems that God’s actions must be in accordance with wisdom and goodness, and thus it seems that God’s actions are in some sense not contingent. But if God’s actions are not contingent, then neither are the human actions whose contingency depends on the contingency of God’s actions. Thus, it appears that contrary to my intended goal, contingency is lost in Leibniz’s system. I shall use the umbrella label

‘The Buck-Passing Objection’ to include both of these seemingly problematic levels.

I will be addressing each of these prominent objections in turn.

1. The PII Violation Objection

At the end of last chapter, I proposed a picture which makes sense of contingency in Leibniz’s system.

Essential to this proposal is the postulation of a cluster of concepts each of which is associated with a given possible substance. Each concept contains a set of predicates that denote the substance’s primitive force; this is the only set of predicates that is shared by all the members of the cluster. The members of the cluster are then differentiated by the other predicates that they have. Each member of the cluster represents a way things could be for the corresponding substance. If any member of the cluster were instantiated, then the very same substance would be created. It is this fact that each member of the cluster is associated with the numerically identical substance that helps retain a more robust and plausible sense of contingency, I argued.

However, it seems that this very fact can also lead to violations of Leibniz’s treasured principle of the identity of indiscernibles. This is so because there is nothing internal to the proposal of this dissertation that entails that God lacks the power to instantiate more than one member of a cluster. If God were to instantiate more than one member of the cluster, then He would be creating the numerically identical substance more than once and with different properties. Thus, there would be numerically identical substances which are nonetheless discernible, contra the principle of the identity of indiscernibles.

111

1.1 Basic Reply to PII Violation Objection

My basic response to this objection is the following. It is indeed true that there is nothing internal to my proposal that entails that God lacks the power to instantiate more than one member of a given cluster.

That is, there is nothing internal to my proposal that prevents God from creating the same individual twice and with different properties, and thus to violate the principle of the identity of indiscernibles. I shall argue, however, that my proposal does not introduce any new problems for Leibniz. In other words, my basic strategy in response to the PII Violation Objection is to point out that the problems that surface in my general proposal, if they are indeed problems, are already present in Leibniz’s system without it.

Before presenting the details of my response, it would be helpful to make a distinction internal to the principle of identity of indiscernibles. The phrase ‘principle of identity of indiscernibles’ tends to be used in

Leibniz scholarship to refer to the following biconditional: an object a and an object b are identical if and only if for any property P if a has P, then b has P as well. In contemporary discussions, however, it has become more common to separate this biconditional into two distinct conditionals: if an object a is identical to an object b, then for any property P if a has P then b has P – that is, identity implies indiscernibility; and, the second conditional states that an object a and an object b are identical only if for any property P if a has P then b has P

– that is, indiscernibility implies identity.204 In the following discussion I will sometimes be addressing these separated conditionals.

Once the biconditional is separated, it is important to ask: which of the separated conditionals is seemingly threatened by my proposal? That is, if God were to instantiate more than one member of a cluster, which of the separated conditionals would be violated? One possibility is that in instantiating two members of a cluster the numerically identical individual would be created twice and with different properties – this is how

204 In contemporary discussions the conditional that states that identity implies indiscernibility is referred to as ‘the principle of the identity of indiscernibles’ and the conditional that states that indiscernibility implies identity is referred to as ‘the indiscernibility of identicals’. Because I already employ the former title to refer to the biconditional, I will not be using this terminology in the main text.

112

I introduced the problem. According to this possibility, the conditional that is would be violated is the one that states that identity implies indiscernibility. Importantly, it is not clear that this possibility also violates the conditional that states that indiscernibility implies identity.

1.2 The Modal Status of PII

If successful, the PII Violation Objection would establish that my proposal is committed to possible violations of Leibniz’s principle of the identity of indiscernibles. That is, my proposal seems to entail that God has the power to violate the principle of the identity of indiscernibles. How bad is it to be committed to this possibility? In this subsection, I will argue that whether my proposal is committed to this possibility depends upon the modal status of the principle of the identity of indiscernibles, and, furthermore, either way it does not constitute an objection to my proposal. On the one hand, if the principle of the identity of indiscernibles is metaphysically contingent, possible violations of this principle are not problematic. On the other hand, if the principle of the identity of indiscernibles is metaphysically necessary, then my proposal does not entail any such possible violation of this principle. In either case, there is no problem for my view here.

The first thing to note is that the modal status of the principle of the identity of indiscernibles is a controversial topic in Leibniz scholarship. In the fifth letter to Samuel Clarke, for example, Leibniz implies that this principle is metaphysically contingent. Leibniz writes: “This supposition of two indiscernibles, such as two pieces of matter perfectly alike, seems indeed to be possible in abstract terms, but it is not consistent with the order of things, nor with the divine wisdom by which nothing is admitted without reason.”205 Here

Leibniz accepts the possibility “in abstract terms” of numerically distinct yet indiscernible entities. What the qualification “in abstract terms” amounts to is not entirely clear from the context. A plausible reading is that this is contrasted with the actual dictates of divine wisdom which demand a sufficient reason for divine action, as Leibniz himself goes on to explain. That is, a plausible reading is the following: God would have no sufficient

205 LC 5.21.

113 reason to create two numerically distinct entities which were indiscernible, yet such is a possibility abstracted from God’s actual wisdom. Leibniz also insists that that which is contrary to divine wisdom does not imply a contradiction and is thus metaphysically contingent.206 If this plausible way of reading the “in abstract terms” qualification in the quoted passage is the correct one, then Leibniz clearly implies that the principle of identity of indiscernibles is metaphysically contingent.

However, this interpretation of the quoted passage is not quite decisive, for in the same correspondence Leibniz also seems to imply that the principle of the identity of indiscernibles is metaphysically necessary. Leibniz also insists that supposing there are indiscernible yet numerically distinct entities is supposing “an impossible fiction.”207 These are strong words indeed. If violations of the principle of the identity of indiscernibles are mere impossible fictions, then the principle is best understood as metaphysically necessary after all. The modal status of the principle of the identity of indiscernibles for Leibniz is, thus, a controversial issue. We do not need to settle this controversy to adequately respond to the PII Violation

Objection, however.

As already mentioned the identity of indiscernibles principle can be separated into two conditionals – one stating that identity implies indiscernibility and the other stating that indiscernibility implies identity. It appears that my proposal on Leibniz’s behalf threatens the biconditional by threatening the conditional that states that identity implies indiscernibility, for this proposal seems to imply that it is possible for God to create a single substance more than once and with different properties by instantiating more than one member of a cluster of individual concepts. However, and importantly, the discussion between Clarke and Leibniz concerns the prospect of two numerically distinct yet indiscernible objects – that is, the discussion directly involves the conditional stating that indiscernibility implies identity. Because these conditionals are separable, it is theoretically consistent to regard one conditional, say that identity implies indiscernibility, to be metaphysically necessary while also holding that the other conditional, say that indiscernibility implies identity, is only metaphysically contingent. Thus, it is possible to read the mentioned discussion between Leibniz and Clarke

206 Grua 289, 326, 479-82, G 3.400, T § 37, LC 5.9. 207 LC 4.6.

114 as merely concerning the modal status of the conditional that states that indiscernibility implies identity, and not also concerning the modal status of the biconditional itself. Thus, even if the proper reading of this text is that Leibniz is committed to the metaphysical contingency of one conditional, this does not imply that he is also committed to the metaphysical contingency of the conditional that seems threaten by my proposal.

Given the uncertainty of the modal status of the principle of the identity of indiscernibles, in Leibniz’s system, my response to the PII Violation Objection will not require that taking stand on this issue. As already mentioned, I think that regardless of the modal status of the principle of the identity of indiscernibles – the biconditional itself – there is a plausible response to the PII Violation Objection. On the one hand, if the principle of the identity of indiscernibles is metaphysically contingent, it seems that possible violations of this principle are unproblematic. On the other hand, if the principle of the identity of indiscernibles is metaphysically necessary, God does not have the power to violate this principle after all. The first point I take to be obvious and thus not in need of defense, but I will return to it in the next subsection. It is the second point, I take it, that seems less straightforward, and thus in need of defense. Here is my argument for it.

There is nothing internal to the proposal in chapter 3 that prevents God from instantiating more than one member of a cluster of individual concepts, so it is consistent with my proposal that God has the power to do so, and even that He in fact does so. These facts, however, do not imply that God has the power to instantiate several members of the cluster and thus create the numerically identical substance several times and with different properties. That is, this kind of power is consistent with the account on chapter 3, but not required or implied by it. Importantly, if it is metaphysically necessary that identity implies indiscernibility, then this fact constrains God’s powers just like any other metaphysical necessity. God has the power to instantiate any member of a cluster of individual concepts He wishes.208 Once He instantiates one of these members, the corresponding substance is created and this places a restriction on God’s powers. Given that it is metaphysically necessary that identity implies indiscernibility (we are assuming), once a substance is created it is at that point

208 This need not be the case, however. Whether God has the power to instantiate any member of a cluster depends upon which subjunctive conditionals of freedom hold of the corresponding agent, and given the Prevolitional Condition, these conditionals are not up to God. Plantinga (1974: Ch. IX) exploits this consequence of the Prevolitional Condition to develop his famous version of the Free Will Defense.

115 no longer possible for God to instantiate a second member of the cluster of individual concepts corresponding to this substance. To do so would be to violate a metaphysically necessary principle, and Leibniz is adamant that God has the power to do only that which is metaphysically possible.209 In short, the metaphysical necessity that identity implies indiscernibility precludes God’s power to instantiate a member of a cluster once He has instantiated a different member. Thus, regardless the modal status of the principle of the identity of indiscernibles, the PII Violation Objection fails.

1.3 God’s Power to Violate PII

As argued in the previous subsection, my proposal in chapter 3 neither entails nor requires that God has the power to violate the principle of the identity of indiscernibles. If this principle is metaphysically contingent, however, my proposal is indeed compatible with this kind of divine power. Perhaps being compatible with the divine power to violate the principle of the identity of indiscernibles is problematic all by itself, even if exercise of this power requires violations of the principle of sufficient reason. Perhaps the mere fact that my view entails that God has this power is problematic all by itself.

My response to this version of the objection is that Leibniz is already committed to a similar kind of power even without my proposal regarding clusters of individual concepts. To better appreciate this consider the following. The mere fact that God has the power to instantiate a complete individual concept once seems to entail that God has the power to do it a second time. If the best way of thinking about complete individual concepts is as structured-blueprint-like ideas in God’s mind, then there is nothing internal to what it is to be a complete individual concept that precludes this concept being instantiated more than once or God having the power to instantiate a complete individual concept more than once.210 If there are restrictions to multiple instantiation of concepts, they come from without those concepts – perhaps it is unwise (violations of PSR) or perhaps it is impossible (if PII is metaphysically necessary).

209 CD § 7, LC 5.73, 5.76. 210 Jorati (2017b: § 2) also emphasizes this point.

116

The proponent of a single individual concept per possible substance is also in the same predicament as my proposal vis-à-vis the PII Violation Objection. Either the principle of identity of indiscernibles is metaphysically contingent, in which case it appears God has the power to instantiate the same complete individual concept more than once. Or the principle of the identity of indiscernibles is metaphysically necessary, in which case God has the power to instantiate any complete individual concept only once. My proposal regarding clusters of individual concepts does not introduce any new problem, vis-à-vis the principle of the identity of indiscernibles, that is not already present in Leibniz’s views.

2. Leibniz on Theodicy

Leibniz’s theodicy is multilayer. Some of the main elements of his theodicy are the following. He makes a distinction between three kinds of , metaphysical, natural, and moral, and provides plausible accounts for why each of these is to be included in creation without God’s goodness being stained.211 Leibniz also argues that the actual world is the best of all possible worlds and that it turns out these kinds of evils are necessary for the best possible world to be the best.212 Because these kinds of evils are necessarily connected with the great good that is the best of all possible worlds, God is justified in allowing them when creating the actual world. The insight here seems quite compelling: surely God cannot be faulted for allowing evils if doing so is necessary for bringing about the greater good that is the best possible world.

In recent discussions on the problem of evil, Eleonore Stump and Marilyn Adams have argued that more than the adumbrated Leibnizian theodicy is needed to vindicate God in allowing evil.213 They argue that the existence of a greater good which is necessarily connected to evil is not enough. The problem of evil also arises at a personal level. It is persons that suffer. God’s actions must not only be justified in an abstract way

211 T §§ 21-34. 212 T §§, 8-11, DM § 3. 213 Stump 1985, and 1990, Marilyn Adams 1999.

117

(by measuring the amount of goodness in the universe), but these actions must also be justified with respect to those that suffer. A way of articulating this point is in terms of persons having legitimate grounds for complaint.

For example, that my is necessarily connected to the greater good of a second person might provide justifying reasons for a third person to allow it to happen. Yet, I might still have legitimate grounds for complaining here. After all, my good was diminished in this transaction. The point here is simply that there being justifying reasons for allowing evil need not also remove all plausible grounds for complaint; the sufferer might still be justified in complaining even in the presence of these justifying reasons for allowing evil. The line of thought that Adams and Stump are pushing, I take it, is that a complete theodicy requires showing that sufferers have no legitimate grounds for complaining to God for allowing them to suffer – I shall label this requirement the “No Grounds for Complaint” requirement.

Stump and Adams argue that a sufferer has no legitimate grounds for complaining to God for allowing their suffering only if their suffering is ultimately connected to their .214 That is, they argue that the No

Ground for Complaint requirements demands showing that God can justifiably allow a person to suffer only if such suffering is ultimately good for that person.215 Both Stump and Adams spend much energy arguing that this more demanding condition on theodicy can indeed be met.216 The details of their arguments need not detain us here, however. What matters for our purposes is that both Stump and Adams think that the way of going about showing that individual persons have no legitimate grounds for complaining to God for allowing them to suffer requires that such suffering be ultimately connected with their good. Sufferers have no grounds for complaining to God only if this is the case.

Leibniz himself does not quite put matters this way, as far as I can tell. In fact, I think that Leibniz would probably deny that God is justified in allowing evils only if it is ultimately good for sufferers. Such suffering being necessarily connected to a greater good suffices to justify allowing it. Regardless of this disagreement, Leibniz does recognize that the problem of evil appears at a personal level. A complete Theodicy

214 Stump 2001: 766. 215 Stump 1985, Marilyn Adams 1999. Robert Adams (1972) also makes a similar point regarding the demands of God’s perfection in creating. 216 Adams 1999, and Stump 2010.

118 requires establishing that individual sufferers have no legitimate grounds for complaining to God for allowing them to suffer. That is, Leibniz seems to grant that a complete Theodicy includes fulfilling the No Grounds for Complaint requirement, for he thinks that God is just and defines justice in terms of lack of grounds for complaint. He writes, for example: “justice is a constant will to act in such a way that no one has a reason to complain to us.”217 This general point advocated by Stump and Adams is one that I think Leibniz would grant.

Fully vindicating God’s holiness includes establishing that there are no legitimate grounds for complaining to

God for allowing evil.

Leibniz addresses the No Grounds for Complaint requirement quite differently from Adams and

Stump, however. Leibniz insists individual persons have no legitimate ground for complaining to God for allowing them to suffer because suffering is included in their notion, or individual concept, and they would not be who they are if they did not suffer in accordance with that which is specified in their individual concept.

2.1 The Theodicy Compromised Objection

It appears that Leibniz’s version of the No Grounds for Complaint requirement for a complete theodicy creates some trouble for my interpretation. To appreciate this, consider the following way of understanding Leibniz’s views on this domain. Leibniz insists that Judas does not have legitimate grounds for complaining to God for allowing him to sin in circumstances C because it is included in Judas’s individual concept that he will sin in C, and thus Judas would not be Judas if this did not happen.218 Why is it that had

Judas not sinned in C he would not have been Judas? An initially plausible answer suggests itself: if Judas had not sinned in C it would not have been Judas because it is an essential property of Judas to sin in C. Thus, not sinning in C is metaphysically impossible for Judas, and this impossibility rules out any legitimate grounds for

Judas to complaint. This is the superessentialist reading of Leibniz’s response of what I’m calling the No

217 Meditations on the Common Concept of Justice § 2/ L p. 569. 218 DM § 30, see also A 2.2.77.

119

Grounds for Complaint requirement for a complete theodicy. This superessentialist reading of Leibniz has several supporters in the literature.219

However, arguing against this superessentialist reading has been a major goal of this dissertation. If my proposal is correct, this superessentialist interpretation of Leibniz is mistaken. It is thus inconsistent with my proposal in this dissertation to provide this interpretation of Leibniz’s response to the No Grounds for

Complaint requirement for a complete theodicy. I must either provide an alternative plausible interpretation of how Leibniz can meet the No Grounds for Complaint requirement or compromise the plausibility of

Leibniz’s theodicy in this aspect. I have labeled this the Theodicy Compromised Objection to my proposal. It is the main task of this section to provide an plausible alternative reading of Leibniz’s version of the No

Grounds for Complaint requirement for a complete theodicy.

2.2 The Initial Response: Superessentialism is Unnecessary

The initial part of my response to the Theodicy Compromised Objection is to weaken the plausibility of superessentialist reading of how Leibniz meets the No Ground for Complaint requirement for a complete theodicy. Importantly, and to the best of my knowledge, Leibniz himself never explicitly articulated the superessentialist reading of this part of his theodicy. That is, Leibniz never explicitly stated, as far as I can tell, that the reason why an agent has no grounds for complaining to God for allowing them to suffer is that it is an essential property of them to suffer.220 In fact, Leibniz did quite the opposite and repeatedly distanced himself from superessentialism.221 Regarding sin, for example, Leibniz insisted that agents have no grounds for complaint precisely because even though it is certain that they will sin, they still have the power not to sin.222

219 Look 2013, Griffin 2013, Mondadori 1973 and 1975. 220 Sleigh (1990: 68) insists on this point. 221 For a list of quotes from Leibniz see Hunter 1981. In response to this criticism, Mondadori (1975) says that Leibniz did not wish to endorse superessentialism, but that his other views committed him to it. A major goal of this dissertation is to argue that Mondadori is mistaken about this. 222 T § 369.

120

More generally, Leibniz repeatedly insisted on contingency as a condition for freedom; he also insisted that free agents have the power to do otherwise, in particular. Regarding the latter, he writes:

For speaking absolutely, our will is in a state of indifference, in so far as indifference is opposed to necessity, and it has the power to do otherwise, or to suspend its action altogether, both alternatives being and remaining possible. (DM § 30/ AG p. 61) From passages like this one, it is clear that Leibniz thinks that it being possible for the agent to do otherwise than they actually do is a condition on freedom. At least at face value, this condition on freedom conflicts with the superessentialist reading of Leibniz’s views. Yet only a few lines after this passage, Leibniz goes on to articulate the same claim that agents would not be who they are if things had gone differently – the very claim that seem to ground the plausibility of the superessentialist reading, that is. Leibniz writes:

But someone else will say, why is it that this man will assuredly commit this sin? The reply is easy: otherwise it would not be this man. For God sees from all time that there will be a certain Judas whose notion or idea (which God has) contains this free and future action. (DM § 30/ AG p. 61) From this passage we get two important claims. In the second sentence, Leibniz presents his theory of complete concepts, which has fueled the superessentialist reading.223 The second important claim is in the first sentence.

Here Leibniz presents a peculiar counterfactual:

C: If Judas were to do other than what he actually does (i.e., sin in C), then Judas would not be Judas.

This is a perplexing counterfactual, so it is best to avoid relying on it.224 I think that the main idea behind this counterfactual can be captured by the following very intelligible counterfactual:

C*: If someone were to do other than what Judas actually does, then that someone would not be Judas.

Superessentialism gives us a way of understanding why C* is true. If someone were to do other than what

Judas actually does, then that someone would not be Judas for the simple reason that everything Judas does he

223 Look 2013, Mondadori 1973 and 1975. This is how I motivated the superessentialist reading in the previous chapter. 224 The most important reason why C is perplexing is that its consequent is impossible and so, according to Lewisian semantic for counterfactuals, this consequent will be false regardless of how the similarity relation between possible worlds is fixed by the context. This is quite puzzling for it entails that C itself will only be true only in a context in which the similarity relation fixes on a possible world in which the antecedent is false. C would be false in all other contexts, for regardless of how the context settles the similarity relation between possible worlds, the closest possible world in which the antecedent is true will always be one in which the consequent is false (for the consequent is false in all possible worlds).

121 does essentially. For example, Judas actually in circumstances C. So, if someone were to not sin in C, we can conclude that that someone is not Judas because it is not possible for Judas not to sin in C.225

However, I think that this is not the correct interpretation. One reason in support of this is that the proximity of these passages makes the superessentialist reading a bit facile. There is not enough time for Leibniz to change his mind about it being possible for the agent to do otherwise being a condition for freedom. The more plausible reading, I think, is that conditional C* does not require superessentialism and that it is compatible with Leibniz’s earlier statements about it being possible for the agent to do otherwise as a condition for freedom. But if this is the case, what should we make of conditional C*? That is, what reading, alternative to superessentialism, is it plausible to attribute to Leibniz here?

Here is my proposal. Contra superessentialism, we do not need to go as far as thinking that it is not possible for Judas not to sin in C, in order for us to make sense of why if someone were to not sin in C, that someone would not be Judas. Instead we only need that the following subjunctive conditional of freedom be true of Judas: “If Judas were in C, Judas would freely sin”. Knowing that if Judas were in C, Judas would freely sin suffices for us to know that if someone were to not sin in C, that someone would not be Judas. This is so for the simple reason that if Judas were in C, Judas would have sinned. Thus, not sinning in C reveals that this agent is not Judas. This is all we need to understand why C* holds. Superessentialism is unnecessary.

2.3 Theodicy Not Compromised

We are also now in a position to see how abandoning superessentialism does not compromise Leibniz’s theodicy. With the proposed understanding of C*, instead, we can also begin to understand Leibniz’s version of the No Grounds for Complaint requirement for a complete Theodicy. Judas has no grounds for complaining to God for allowing him to sin in C because whether Judas sins in C is up to Judas and not up to God. It being true that “if Judas were in C, Judas would freely sin” restricts God’s options. One crucial restriction is the

225 This is exactly how Mondadori, one of the major advocates of interpreting Leibniz as a superessentialist, reads Leibniz (1973: 91-2).

122 following: if God were to create a being who would not sin in C, then God would have to create a being who is not Judas. According to the present proposal this would be the case not because it is essential to Judas that he freely sins in C, but rather just because it is true of Judas that if he were in C, he would freely sin. Thus, creating someone who does not sin in C is in effect creating someone who is not Judas, for Judas would sin if he were in C. Furthermore, and very importantly, this counterfactual of freedom being true of Judas is perfectly compatible with it being possible for Judas not to sin in C, so it is compatible with the denial of superessentialism. In sum, this version of the No Grounds for Complaint requirement argues that the creature has no grounds for complaining to God for allowing them to sin not because it is essential for them to sin but because sinning is what they would freely do if created.

I think this plausible interpretation of conditional C* should be preferred over the superessentialist reading simply because it has a more holistic fit with Leibniz’s texts. Passages in which Leibniz articulates conditional C*, and the like, can indeed be plausibly read as superessentialist, but this reading is no longer a natural one when compared with many other passages in which Leibniz clearly distances himself from superessentialism. As we have noted in this dissertation, Leibniz not only does not explicitly endorse superessentialism, but, in fact, he spends much energy trying to capture contingency as a condition for freedom in his system. More directly relevant for our discussing, after Arnauld expresses reservations about something like conditional C* because it does away with contingency, instead of Leibniz accepting the consequence he denies that it is a consequence. Referring to himself and the future contingent property of taking a trip to Paris,

Leibniz writes “I am uncertain whether I will take the trip, but I am not uncertain that, whether I go or not, I will always be me.”226 Here Leibniz is directly talking about his own convictions instead of directly talking about reality. It is reasonable to think, however, that his point is not merely that of sharing a conviction, but rather that of embracing the content of such conviction. Leibniz, it is clear, I think, does not want the superessentialist reading of his commitment to C*, and, I have argued, he does not need this reading to preserve the No Grounds for Complaint portion of his theodicy.

226 Leibniz’s Remarks on a Letter from Arnauld, June 1686, A 2.2.52/ AG p. 75.

123

3. The Buck-Passing Objection

The final objection I will address in this chapter is what I have labeled the “Buck-Passing Objection.”

According to this objection, by including God’s possible free decrees in the content of complete individual concepts, or more specifically by relying on predicates denoting these divine possible free decrees to help individuate different members of the cluster of concepts, my account entails that the contingency of human free acts depends upon the contingency of God’s free decrees. The first level of this objection is that this dependence relation is itself problematic, for the kind of contingency of human actions that depends upon the contingency of divine actions is obviously not the kind of contingency that matters for human freedom. The second level of the Buck-Passing Objection is that this dependence relation ultimately does away with contingency in Leibniz’s system. This is so because God’s actions are not contingent, the objector contends;

God’s actions are not contingent because God’s actions in some sense follow from God’s goodness and wisdom and these attributes are essential to God. In passing the buck of contingency from the human to the divine realm contingency itself disappears, the objector insists.

3.1 Response to Level One of Objection: Spectrum of Possible Precisifications

The initial response to this objection is that it is still underdetermined whether the proposal in chapter three entails that the contingency of human acts depends on the contingency of divine acts in any problematic way. There is indeed a kind of ontological dependence that is postulated in that proposal, but whether it amounts to a problematic dependence relation is still unspecified. In general, that human acts ontologically depend on divine acts is not something Leibniz and other traditional theists will find problematic; as these thinkers see it, divine conservation and divine concurrence can indeed be construed as unproblematic ontological dependence relations holding between divine and human acts, for example. The ontological dependence relation postulated in the previous chapter involves the contingency of both human and divine

124 acts, and this additional fact does indeed seem to point to a prima facie problematic ontological dependence relation.

Importantly, that there is an ontological dependence relation holding between the contingency of human acts and the contingency of divine acts need not be ultima facie problematic, however. An easy way of seeing this is the following. First, I take it that for the contingency of human acts to depend upon the contingency of divine acts is for i) the actuality of human acts and ii) the fact that it is possible for these human acts to be different to both depend on the contingency of divine acts. There is at least one cheap way of getting this kind of dependence that is unproblematic, however. Both the actuality of human acts and the possibility of them being different depends upon God’s acts of conservation and concurrence. Furthermore, it is orthodox theistic that both God’s acts of conservation and concurrence are freely chosen, and are thus contingent divine acts. Thus, it is possible to articulate a reductive account about how the contingency of human acts depends upon the contingency of divine acts purely on the basis of conservation and concurrence. This kind of reductive ontological dependence relation is of course not problematic for orthodox theists like Leibniz. Furthermore, this kind of reductive account is also consistent with the model presented in chapter three. Thus, whether the contingency of human acts depends upon the contingency of divine acts is ultima facie problematic is still underdetermined for the account articulated at the end of chapter three.

One of the goals of this section is to develop this prima facie problem into a clearly problematic picture and then to distinguish this picture from the one I am advocating on Leibniz’s behalf. The main philosophical move involves noting that there is a whole spectrum of possible precisifications of the general proposal articulated at the end of chapter three. I will argue that this spectrum marks the extent to which there is a problematic buck-passing element in each possible precisification. At one end there is no problematic buck- passing element at all, and at the other end the back-passing element is prominent. I will argue that Leibniz’s actual views lie somewhere between these extremes and thus involve at least a minor buck-passing element. I will not argue for the tenability of this middle ground, but only that this is the best way of reading Leibniz’s actual views in this domain.

125

I think the best way of articulating this spectrum of possible precisifications is in terms of explanation.

More precisely, it is explanation of human actions that serves as the axis around which the spectrum of possible precisifications is arranged. I will be immediately concerned with the explanation of actual human action, but, as will become clear shortly, this kind of explanation will also extend to the explanation of the realization of alternative possibilities and ultimately the explanation of what makes a privileged member of the cluster be the privileged member. The first main point is then that it is explanation of human action that is the axis around which the spectrum of possible precisifications is arranged.

At one end of this spectrum human action is fully explained by the agent’s nature alone. Rutherford’s interesting discussion of what he calls “Leibniz’s principle of intelligibility”227 is a fit example of this kind of reading of Leibniz. The basic idea here is that Leibniz’s principle of sufficient reason demands, at least in some contexts, that there be a ‘natural’ explanation of monadic activity. Natural explanations are the kinds of explanations that only appeal to reasons grounded in, or drawn from, the nature of creatures. At this end of the spectrum, then, God’s contributions to human actions are mere conditions sine qua non which do not play any explanatory role regarding these human actions—much like the presence of oxygen in a building that has been burned to the ground by an Arsonist: the presence of the oxygen is a causal condition without which the burning of the building would not have been possible, but this fact is not explanatorily relevant for the burning down of the building by the Arsonist.

The first important moral to note, for our purposes, is that at this end of the possible spectrum of precisifications there is no problematic ontological dependence between the contingency of divine and the contingency of human actions. Human actions are fully explained by the nature of individual humans and thus which possibility gets actualized – which member of the cluster gets to be the privileged one – is fully explained by human nature. Given the commitments acquired in the previous chapter, we can now say that which member of the cluster gets actualized, if the substance is created, is fully explained by the primitive force of the given substance. Perhaps the best way of making sense of this is by suggesting that the principle of intelligibility

227 Rutherford 1992.

126 only holds in one member of the cluster and this is the way in which a substance’s primitive force explains which member of the cluster is the privileged one. That is, if the explanation of human action is exhausted by its primitive force, it seems quite implausible that this same primitive force would fully explain two different and incompatible possible courses of action – two complete ways things could be for a given substance. Thus, it seems implausible that the very same primitive force, by itself, fully explain more than one member of the cluster of concepts which represent its modal profile.

Another way of putting this point is by noting that a substance’s primitive force provides God with reasons to instantiate the member of the cluster that obeys the principle of intelligibility and not other members other members of the cluster which violate this principle. God in his wisdom and goodness has most reason to instantiate the member of the cluster that obeys the principle of intelligibility and not the other members of the cluster, if God is to create this substance at all.

At this end of the spectrum we find a possible precisification of the general proposal at the end of chapter three. Importantly for our purposes in this context, this possible precisification does not postulate a problematic ontological relation between the contingency of human actions and the contingency of divine actions. In other words, there is no problematic buck-passing element in this possible precisification. As already stated, however, I do not think this is the correct view to attribute to Leibniz. Pace Rutherford, I think

Leibniz’s view does include an element of buck-passing – whether this element is problematic is, again, something I will not try to settle here.

At the other end of this spectrum, it is God’s actions that provide the full explanation of human action.

At this end, the nature of the substance, or more precisely the substance’s primitive force, plays no explanatory role at all; it is a mere condition sine qua non for God’s activity. That which follows from a substance’s primitive force, including human actions, is entirely up to God. Not in itself, but in terms of what which follows from it, a substance’s primitive force is entirely malleable before God’s will and in some loose sense akin to pure potentiality in Aquinas’s views. At this end of the spectrum, then, which member of the cluster is the privileged one is entirely up to God; any alternative possibility is just as compatible with a substance’s primitive force as

127 any other. Rutherford’s principle of intelligibility fails in all of them. At this end, then, the buck-passing element is maximal.

This kind of view is indeed a possible precisification of the general proposal at the end of chapter three.

It is also clearly not what Leibniz wanted. This fact need not be seen as problematic for the general proposal in the previous chapter, for this general proposal does not require this particular possible precisification. It is merely one of its possible precisifications.

Leibniz’s own views in this domain, like in many others, lie somewhere between these two clear extremes.

I think that, for Leibniz, it is both a substance’s primitive force and God’s actions that together explain what follows from this primitive force – which, of course, includes human actions. Leibniz writes:

So all human events could not fail to happen as they have in fact happened, supposing the choice of [creating] Adam to have been made; but not so much because of the individual concept of Adam, although this concept contains them, as because of the plans of God, which also enter this individual concept of Adam, and determine that of this entire universe, and thereupon both that of Adam and those of all other individual substances in this universe – each individual substance expressing the whole universe of which it is a part, according to a certain relation, by the connection that holds among all things because of the linkage among God’s resolutions and plans. (Leibniz to Arnauld 14 July, A 2.2.73-4/ LA 101-103)

In passages like this one, it seems quite clear that Leibniz wishes to endow God with a robust amount of active control over creation and the very building of the content of possible worlds. In fact, Leibniz seems to give

God’s plan some important explanatory priority. He insists that Adam’s complete individual concept is what it is, or contains what it contains, “because of the plans of God” which are part of its content. Leibniz insists that it is “God’s resolutions and plans” are constitutive of the very content of these complete individual concepts and crucially for our purposes that these plans and resolutions have some kind of explanatory priority.

These kinds of passages are a bit vague, it must be admitted, but it is still important to point them out.

They clearly indicate that Leibniz reserves an explanatory role to divine activity in human actions. These kinds of passages, then, count against attributing to Leibniz a possible precisification which has an unrestricted version of the principle of intelligibility in which none of human actions are even partly explained by divine activity. Unfortunately, it seems to me that nothing more precise than this can be derived from these kinds of

128 vague passages. As far as I can tell, the text which most directly addresses the issues that concern us here is

DM § 31. It is worth our time, then, to look at this text in a bit more detail.

The main point of DM § 31 is theological. As is commonly the case with Leibniz, he wants to take a narrow middle path. On the one hand, Leibniz insists that God’s goodness and wisdom require that He acts in accordance with the principle of sufficient reason, so God never does anything without a sufficient reason for it. On the other hand, following orthodox Christian doctrine Leibniz insists that divine graces are unmerited by human agents. That is, the reasons that human natural dispositions towards good works and faith are by themselves, in principle, insufficient reasons for granting or not granting divine graces. Setting the context for the discussion in section DM § 31, Leibniz writes:

God’s graces are totally pure, and creatures have no right to them. However, although it is insufficient in explaining the choices God makes in dispensing these graces to appeal to either absolute or conditional foreknowledge of men’s future actions, at the same time we must not imagine that there are absolute decrees, which have no rational grounds. (DM § 31/ WFII p. 83.)

This passage is deeply embedded in theological discussions essential to Molinism and its opponents. Here

Leibniz articulates the prima facie tension between the nature of divine grace and the demands of the principle of sufficient reason. He also does much more than this. Leibniz also rejects three possible ways to address this prima facie tension.

Leibniz rejects the Calvinist, or Reformed, position of ‘absolute’ or ‘unconditional’ decrees.228 The core of this theological account is that God’s distribution of graces is not to be explained even in part by the intrinsic qualities of the potential recipients of these graces. Instead, God’s sole motivation for choosing to distribute graces is the manifestation of His own glory. Because of this, who gets grace and who does not is irrelevant to

God’s sole motivation for the distribution of graces. Thus, according to this theological position, God simply chooses how to distribute graces independently of any intrinsic qualities of potential recipients. Leibniz tell us explicitly why he rejects this latter alternative. He says that such absolute decrees “have no rational grounds.”

It is thus clear that Leibniz rejects this Calvinist alternative because it violates the principle of sufficient reason.

228 DPG is largely dedicated to providing a plausible alternative in which there are none of these ‘absolute’ or ‘unconditional’ decrees of God to distribute graces.

129

Leibniz also insists that neither absolute nor conditional foreknowledge suffices to explain God’s decisions about how to distribute graces. That is, mere foreknowledge, whether it is of future contingents or of subjunctive conditionals of freedom, is insufficient to explain God’s actions in distributing graces. Leibniz’s main objection to this explanation is that it merely pushes the question elsewhere. To better appreciate this, consider the following example. Suppose agent S1 has natural dispositions favorable towards faith and good works and agent S2 does not. Suppose further that because of these natural dispositions S1 would freely accept divine grace and S2 would not, and that God foreknows all of this. Would God’s decision to grant graces to S1 and not S2 be explained in terms of God’s foreknowledge? Leibniz says ‘no.’ This is so because God himself is in some sense responsible for whatever is good in both S1’s and S2’s natural dispositions. Responding to the foreknowledge explanation of God’s decision, Leibniz writes: “But if that were so, it could be said that those natural dispositions, to the extent that they are good, are also the effect of grace, although in this case ordinary grace”.229 This point is a crucial point for our purposes; God’s decisions to distribute graces cannot be fully explained in terms of responses to the natural dispositions of potential recipients, for whatever is good in these natural dispositions is good because of divine graces, albeit ‘ordinary’ graces. Divine foreknowledge by itself merely pushes the question elsewhere, and is thus not a sufficient explanation of God’s decisions to distribute graces. We still do not know on the basis of what God decided to grant some ordinary graces to S1 that He did not grant to S2.

After distancing himself from these possible solutions to the prima facie tension between the demands of the principle of sufficient reason and the nature of divine graces, Leibniz presents his preferred solution to the prima facie tension. He writes:

I believe then (since we do not know to what extent or in what way God takes account of natural dispositions in the dispensing of grace) that the most precise and the safest thing to say (in accordance with our principles and as I have already remarked) is that among possible beings there must be the person of Peter or of John whose notion or idea contains this whose sequence of ordinary and extraordinary graces, and all the rest of these events and their circumstances, and that from amongst an infinity of other equally possible people it pleased God to choose that person for actual existence. (DM § 31/ WFII p. 83)

229 DM § 31/ WFII p. 83.

130

Here Leibniz appeals to his account of possible worlds and complete individual concepts to answer the mentioned prima facie tension. He argues that in God’s intellect there is a complete individual concept which includes both ordinary and extraordinary graces build into it, and that “it pleased God to choose” some possible people for existence on the basis of the kind of knowledge afforded by these complete individual concepts.

The first thing to note about this response is that it is not at all clear how it amounts to a response to the prima facie tension between the nature of divine grace and the demands of the principle of sufficient reason.

A plausible way of understanding the contribution of Leibniz’s complete individual concept doctrine is that of grounding God’s foreknowledge of both future contingents and, I have argued, subjunctive conditionals of freedom. That is, complete individual concepts and possible worlds are just the theoretical machinery that Leibniz uses to make intelligible how God can have ‘absolute’ and ‘conditional’ foreknowledge.

It is crucial to note, however, that in this context this is not enough. As already noted, it is part of the context of this section that Leibniz thinks that mere foreknowledge is not enough to answer the prima facie tension.

More than mere foreknowledge is required for explaining God’s decisions regarding distribution of graces,

Leibniz insisted a few sentences prior. If Leibniz’s answer is to be successful, besides divine foreknowledge, what else does he think is being added by his possible worlds and complete individual concept doctrines?

Here is my suggestion. I propose to understand this passage as one in which Leibniz is advancing the radical thesis that what is natural to a given substance is partly dependent upon God’s activities. More precisely,

I suggest that Leibniz takes on the bold thesis that part of the very nature of a substance includes some of God’s divine decrees, namely those that correspond to ‘ordinary graces.’ Returning to the terminology of the previous chapter, we can now say that a substance’s primitive force is partly the result of God’s ordinary graces. This is not to invite the miraculous into every human activity, however. I take Leibniz distinction between the ordinary and the extraordinary graces to mark the distinction between that which is natural (i.e., natural dispositions) and that which is supernatural. Extraordinary graces are not constitutive of the nature of an individual – though, these extraordinary graces are also included in the individual notion of this individual. In sum, human actions

131 are not miracles because they are explained by human nature alone, but human nature itself is partly explained by God’s ordinary graces.230

This bold thesis makes sense of Leibniz insistence that his proposal ‘dissolves’ the puzzling question raised by the tension between the nature of grace and the demands of the principle of sufficient reason. The question is dissolved because human nature itself is expanded to include divine decrees built into it. If human nature in some sense includes divine graces, then the sharp distinction between graces and nature seems to disappear. Thus, also does the question about whether it is human nature or pure divine graces that explains divine action. Immediately after the quoted passage, Leibniz adds “After that it seems that there is nothing more to ask, and that all the difficulties disappear.”231

I will not try to assess the adequacy of this answer to the prima facie tension between the nature of divine grace and the demands of the principle of sufficient reason. What matters for our purposes is that it seems reasonable to read Leibniz on DM § 31 as endorsing a precisification of the spectrum that includes some of

God’s divine decrees as explanatorily relevant in the very nature of the substance, and so derivatively, though only derivatively, explanatorily relevant for human action. At a superficial level, this precisification can characterized as fully respecting Rutherford’s principle of intelligibility. This characterization would be a bit deceptive given that, according to this proposal, human nature itself is partly explained by divine activity. It is thus most accurate to characterize this precisification as one that lies in between the presented extremes.

Leibniz, I conclude, does invite a minor element of buck-passing into his action of human agency, in general, and the contingency of human actions, in particular.

3.2 Second Level of the Buck-Passing Objection

The second level of the Buck-Passing Objection is the more ambitious level. If successful, this objection is devastating, for it does away with contingency altogether. The second level of the Buck-Passing

230 This is the definition of miraculous that Leibniz presents in DM § 16. 231 DM § 31/ WFII p. 83.

132

Objection states, once again, that postulating on ontological dependence relation between the contingency of human actions and the contingency of divine actions precludes contingency because God’s actions are not contingent. God’s actions are not contingent, the objection continues, because God’s essential goodness and wisdom make His actions necessary.

In the present section, I wish to articulate more carefully this objection and to provide what I take to be an adequate response on Leibniz’s behalf. My response is two-fold. First, I will argue that even if God’s actions were metaphysically necessary, the contingency of human actions would not be undermined. This is so, I will argue, because it is the Contingency Condition – the first substantive tenet – that secures the contingency of human free actions, and this Condition is not undermined by God’s action begin metaphysically necessary. The second part of my response is that God’s actions are not metaphysically necessary. I will argue that the account of agential contingency developed in chapter two of this dissertation can also be applied to

God, and, thus, the contingency of God’s actions can be preserved in similar respects as the contingency of human actions was preserved in chapter two. I will argue that God’s actions are agentially contingent even given His essentially good and wise nature. Hence, I conclude, the second level of the Buck-Passing Objection fails, and the contingency of human actions is preserved in Leibniz’s system.

3.2.1 God’s Essential Nature and Subjunctive Conditionals of Freedom

The first part of my two-fold response to the second level of the Buck-Passing Objection is that even if God’s actions are metaphysically necessary, this does not imply that human actions are also metaphysically necessary. This is so, I think, because what secures the contingency of human free actions is the truth of the

Contingency Condition – the Molinist tenet that subjunctive conditionals of human freedom are contingent.

The truth, or falsity, of the Contingency Condition is independent from whether God’s actions are metaphysically contingent.

To better appreciate this, consider the following example. Suppose that Adam freely takes the forbidden fruit in circumstances C. Given Leibniz’s endorsement of the Contingency Condition, it follows that

133 the following conditional is both true and contingently so: “if Adam were in circumstances C, Adam would freely take the forbidden fruit.” Let’s call this conditional ‘A.’ Adam has the power to do otherwise, so Adam has the power such that were Adam to exercise this power A would have been false.

Further suppose that God’s actions are metaphysically necessary. If so, whether God brings it about that Adam is in C is also metaphysically necessary – for this is just one of the metaphysically necessary actions of God. Importantly, however, this metaphysical necessity does not make Adams’s own action of taking the forbidden fruit metaphysically necessary. The contingency of A is not undermined by the metaphysical necessity of God’s actions.

A clean way of seeing this is by nothing that Leibniz also endorses the Prevolitional Condition – the substantive Molinist tenet that subjunctive conditionals of human freedom are known by God prevolitionally, or prior to any of God’s actual volitions. Given the Prevolitional Condition, it follows that subjunctive conditionals of human freedom are true, or false, prior to God’s actual volitions. That God’s volitions are metaphysically necessary, we are assuming, does not render these conditionals metaphysically necessary precisely because their truth value is independent from these volitions. If these conditionals were settled by divine decree – that is, if postvolitionalism were true – then these conditionals would indeed be rendered metaphysically necessary by the fact that God’s actions are metaphysically necessary.

I think that this response suffices to show that the second level of the Buck-Passing Objection ultimately fails. However, it rests on granting that God’s actions are metaphysically necessary – something

Leibniz would vehemently deny. A more ambitious response to this second level of the objection is to insist that God’s actions are also contingent. I undertake this more ambitious project in the next subsection.

3.2.2 God’s Essential Nature and The Principle of Sufficient Reason

The second level of the Buck-Passing Objection depends upon the principle of sufficient reason. This was not explicit in the previous formulation of the objection, but will play a prominent role in this section

134 dedicated to addressing this second level. Throughout history, theists have proposed two radically different ways of understanding the implications of God’s essential nature, in particular His essential nature as good and wise; these alternative ways are necessitarianism, on the one hand, and what I shall call the ‘sheer will’ view,232 on the other. The former view states that everything that is actual is metaphysically necessary. 233 The latter view, roughly, is the view that God’s will is explanatorily rock bottom – in particular, that God’s decision to create what He has created is not to be explained in terms of His essential nature. At best, God’s nature as good and wise provides non-determining reason for action, but it is ultimately God’s sheer will that is the explanation of why created things are the way they are: that He willed to create things the way they are is the explanation for created things being the way they are.

It is commonly thought that the key philosophical commitment that separates necessitarianism from the sheer will view is the principle of sufficient reason.234 This common understanding has it that God’s essential nature together with the principle of sufficient reason entails necessitarianism, and that the only way to avoid necessitarianism, given God’s essential nature, is to abandon the principle of sufficient reason. This is how logical space divides up, this common understanding has it. There is no middle ground between these two radical views, short of giving up on the essential nature of God.

An important thesis of this chapter, and indeed dissertation, is that this common understanding is mistaken; Leibniz has the theoretical tools to carve an intelligible middle ground between these views. Leibniz can retain a system which includes contingency (i.e., the negation of necessitarianism), God’s essential nature as good and wise, and a strong version of the principle of sufficient reason. If this is so, then the second level of the Buck-Passing Objection is ultimately unsuccessful because God’s actions are contingent, and postulating an ontological relation between the contingency of human actions and the contingency of divine actions does not do away with the contingency of human actions.

232 I take this label from Lovejoy 1960: 174. 233 Griffin (2013: 3) also defines necessitarianism in this way. Sleigh (1999: 262) defines necessitarianism as the thesis that all true propositions are necessarily true. For the purposes of this chapter, these definitions can be taken as interchangeable. 234 See for example Lovejoy 1960: Ch. 5.

135

The first main task of this section is to clearly present why necessitarianism seems to follow from

God’s essential nature and the principle of sufficient reason. Because Spinoza is a prominent defender of necessitarianism on the basis of God’s essential nature and the principle of sufficient reason, I shall loosely refer to this kind of reasoning as Spinozistic Rationale. The core of Spinozistic Rationale is the following.

Plausibly God’s essential nature entails that it is metaphysically necessary that God has the nature that He has.

Many theists also believe that God’s nature provides the sufficient reason for God’s existence – here is an application of the principle of sufficient reason. Given that it is metaphysically necessary that God has the nature that he has, and that God’s nature provides the sufficient reason for His existence, it follows that it is metaphysically necessary that God exists. Furthermore, many theists also believe that everything that exists ultimately finds its sufficient reason for existing in God’s nature – here is the second application of the principle of sufficient reason. Given that it is metaphysically necessary that God has the nature that he has, and that everything that exists ultimately finds its sufficient reason for existing in God’s nature, it follows that everything that exists is metaphysically necessary. And this latter claim is just necessitarianism.

As already mentioned, Spinoza is a notorious proponent of this kind of argument.235 Several commentators have thought that Leibniz is committed to this Spinozistic Rationale. Lovejoy, for example, utilizes this kind of reasoning to argue that Leibniz is committed to necessitarianism.236 It is important to note that there is a crucial hidden assumption in this Spinozistic Rationale. I shall call this assumption “Spinoza’s

Assumption.” Here is what it states:

Spinoza’s Assumption: something exists with metaphysical necessity if its sufficient reason for existing exits with metaphysical necessity.

It is this assumption that I wish to challenge on Leibniz’s behalf. Before doing that, however, I wish to articulate a version of Spinozistic Rationale with Spinoza’s Assumption explicitly cited. For our purposes, we can grant the first part of the Spinozistic Rationale and begin with the thesis that God exists with metaphysical

235 E I.p7-8, p11, p16, p29. 236 Lovejoy 1960: Ch. 5.

136 necessity. This first part of the argument is something Leibniz will be happy to concede. Here is the argument that most interests us, which I shall call the “Spinozistic Argument”:

1. It is metaphysically necessary that God exist. 2. Everything that exists has its sufficient reason for existing in the nature of God. 3. Spinoza’s Assumption: something exists with metaphysical necessity if its sufficient reason for existing exists with metaphysical necessity. 4. Therefore, everything exists with metaphysical necessity.

Before explicitly addressing this argument on Leibniz’s behalf, it is important to note that Leibniz himself vacillated much regarding the best way to address this Spinozistic Argument. Leibniz wrestled with these issues for much of his career and explored multiple different responses. In fact, it seems plausible that Leibniz himself at some point endorsed both extremes – namely, necessitarianism and the sheer will view.

Leibniz seems to have endorsed necessitarianism early on in his career. Here is a quote from a letter to

Magnus Wedderkopf in May 1671:

Since God is the most perfect mind, however, it is impossible for him not to be affected by the most perfect harmony, and thus to be necessitated to the best by the very ideality of things…. Hence it follows that whatever has happened, is happening, or will happen is best and therefore necessary, but…with a necessity that takes nothing away from freedom because it takes nothing away from the will and the use of reason. [emphasis added, A 2.2.117] In this passage Leibniz states that from God’s perfection it follows that He will necessarily do the best. From this Leibniz derives the necessitarian conclusion that what God creates is not only best but also necessary. Thus, much like Spinoza, at least at one point seems to have endorsed the Spinozistic Argument.237

Leibniz became dissatisfied with this view, however. In his copy of the letter to Magnus Wedderkopf,

Leibniz wrote the following amendment: “I have since corrected this; for it is one thing for sins to be infallibly going to happen, and another thing for them to be going to happen necessarily.”238 Leibniz, thus, explicitly moved away from necessitarianism.

237 It is also reasonable to read Leibniz’s early work Confessio Philosophi as defending a necessitarian picture. In fact, some philosophers have argued that necessitarianism is Leibniz’s mature and settled view. See, for example, Russell (1945), Mondadori (1973), (1995) and (1985), and more recently Griffin (2013); even Adams (1994) argues that the modal picture of Leibniz’s he is presenting is probably one that Spinoza would not find problematic. The view that I advocate in this dissertation is incompatible with these readings of Leibniz. 238 Quoted in Adams 1994: 11, and in Sleigh et al, 1998: 1262.

137

Leibniz at some later point seems to have also endorsed the sheer will view. Leibniz writes:

For if anyone will ask from me the reason why God decided to create Adam, I say: because he decided to make the most perfect. If you now ask me why he decided to make the most perfect, or why he wills the most perfect (for what is ‘to will’ other than ‘to decide to do’?) I answer that he willed it freely, i.e. because he willed it. (A 6.4.1455/ SLT p. 109.) In this passage, Leibniz does not say explicitly that God’s will is not to be explained in terms of God’s nature, but I take this implication to be quite clear. Thus, I take this passage to clearly reflect Leibniz’s commitment to the sheer will view. We do not need to spend much time understanding the reasons that lead Leibniz to endorse either of these extreme views. The main point for our purposes is simply that Leibniz’s explicit endorsement of both extremes illustrates that these issues are ones that he took quite seriously throughout his career. His vacillation between possible answers illustrates how seriously he took both the Spinozistic Rationale and its assumptions. I think that at least by his mature period Leibniz had abandoned these extremes and had settled on a view midway.

Leibniz articulated what I take to be his settled view on these matters on multiple occasions. A clear presentation of this settled view is the following. Here, Leibniz is explicitly addressing the Spinozistic Rationale for necessitarianism:

He [Spinoza] wrongly holds that the world is an effect of the divine nature…. There is a midpoint between what is necessary and what is by chance, namely, that which is free. The world is a voluntary effect of God, but a voluntary effect due to inclining or prevailing reasons. And even if we imagine the world to be perpetual, it would still not be necessary. For God could either have not created or created the world otherwise; but this was something he did not do. (FC 22-70/ AG 277) Here Leibniz seems to be denying premise 2 [Everything that exists has its sufficient reason for existing in the nature of God] of the Spinozistic Argument. He thinks that Spinoza is mistaken in thinking that everything finds its ultimate sufficient reason for existing (ultima ratio rerum) in God’s nature alone – that is, God’s nature independently of God’s will. God’s will is an important part of the explanation, Leibniz insists. However, merely appealing to God’s will as part of the ultimate explanation of reality does not yet fully answer the

Spinozistic Argument, for it could well be the case that God’s will is itself explained by God’s nature. If so,

Premise 2 of the Spinozistic Argument ultimately stands.

138

Furthermore, I think that Leibniz himself is clearly committed to the thesis that God’s will is ultimately explained by God’s nature. As we saw in chapter two, for Leibniz the will is just the rational inclination towards goodness as apprehended by the intellect. Leibniz also makes the distinction between God’s antecedent will and consequent will. The antecedent will is just an individual rational inclination whose strength is proportionate to the goodness of the object of choice when consider in isolation from other alternatives.239

And the consequent will is just the final rational inclination that results from the coming together of all the antecedent wills taken together in a particular situation. Leibniz describes it thus: “Now this consequent will, final and decisive, results from the conflict of all the antecedent wills, of those which tend towards good, even as of those which repel evil; and from the concurrence of all these particular wills comes the total will.”240 Thus, for Leibniz the will is just the general rational inclination towards apprehended goodness which further divides into antecedent and consequent will. Crucially for our purposes, Leibniz insists that God’s being perfectly good and wise entails that He always acts in accordance with the strongest rational inclination, for that is the good and wise thing to do. That is, contrary to the human case, God always acts in accordance with the strongest rational inclination (which is His will) and this is explained by the fact that God is perfectly good and wise – i.e., by His nature. Thus, for Leibniz God’s will is explained by God’s nature.

Leibniz’s appeal to God’s will as part of the ultimate explanation of created reality is thus best understood as modifying the second premise of the Spinozistic Argument – and thus significantly departing from Spinoza’s . However, it would be a mistake to understand this as effectively blocking the

Spinozistic Argument for necessitarianism. It is the second part that Leibniz’s response that is meant to deliver this more ambitious goal, I think. After modifying the second premise of the argument, Leibniz goes on to articulate why despite God having an essential nature His actions are nonetheless contingent. Leibniz tell us that God’s actions are contingent precisely because God’s voluntary actions are “due to inclining or prevailing reasons.” That is, Leibniz appeals explicitly to his inclining and not necessitating reasons doctrine to safeguard the contingency of God’s actions given God’s essential nature and the fact that everything finds its ultimate

239 T § 22. 240 T § 22.

139 explanation in God (both His nature and will). It is the task of the next subsection to illustrate how this can be so.

3.2.3 God’s Actions are Contingent

The reason why necessitarianism is blocked in Leibniz’s system is precisely because God’s actions are contingent in the sense that is specified by Leibniz’s account of inclining but not necessitating reasons. A clear articulation of this doctrine is the following:

This determination notwithstanding, the agent remains free, not only from coercion, but also from necessity…. God himself, even if he is maximally determined to the good, is nonetheless most free, not only from coercion but also from necessity. Reasons incline the wise man but they do not necessitate him. It would be possible for him to act differently, but it is certain that he will not do so. (DPG 35b) The core idea behind this doctrine is that a reason can be sufficient to explain something, that is it can meet the strictures of the principle of sufficient reason, without also necessitating that something. I have argued, in chapter two, that the best way of making sense of this doctrine is in terms of what I have labeled ‘agential modality.’ The core idea of agential modality is the following.

Agential contingency is the kind of contingency that is grounded in the agent’s powers to determine himself in accordance with his deliberate judgment of which considered course of action is best. An action counts as agentially contingent, I argued in chapter two, in cases in which the choice to bring about this action results from a deliberation situation in which the agent takes at least two possible courses of action into consideration, and the agent has the power to determine himself to any considered option if the agent concludes that this is the best considered option.

I have argued that this account of agential contingency makes sense of Leibniz’s otherwise obscure doctrine of inclining and not necessitating reasons. I here wish to extend this account of merely inclining reasons, grounded in agential contingency, to the case of God, and insist that this account makes intelligible how God can act in an agentially contingent way even given his metaphysically necessary nature. If this account is successful, then the Spinozistic Argument can be blocked on Leibniz’s behalf.

140

To see whether God’s actions are agentially contingent, I need to make explicit a few more details regarding agential modality. First, for an option O1 to be agentially possible for an agent S it must be the case that S recognizes some goodness in O1 and has a rational inclination towards O1. Furthermore, and importantly for our current purposes, it must also be metaphysically possible for S to come to the deliberative conclusion that O1 is the best considered option – that is, the state of affairs of S coming to the conclusion that O1 is the best considered alternative must not imply a contradiction. For an action to be agentially contingent, then, is for this action to be chosen on the basis of the deliberation of at least two agentially possible options.

Part of this account of agential contingency is clearly applicable to God. For any metaphysically possible option O1, God correctly apprehends the degree of goodness in O1, and God also has a rational inclination which is proportionate to this degree of goodness in O1. The potential source of trouble is establishing whether it is metaphysically possible for God to come to the deliberative conclusion that O1 is the best considered option. Given that it is metaphysically necessary that God does not make mistakes, it would be metaphysically possible for God to come to the deliberative conclusion that O1 is the best considered option only if it is metaphysically possible that O1 is the best considered option. Importantly, if it is metaphysically necessary for God to take into consideration all metaphysically possible options in deliberation, then only the best possible world is agentially possible for God. This would be the case because if it is metaphysically necessary for God to take into consideration all metaphysically possible options, then there is only one metaphysically possible deliberation situation for God – namely the one in which God takes into consideration all metaphysically possible options and comes to a conclusion about which is best.

The crucial questions, for our purposes, is whether God’s perfect goodness and wisdom requires that

God take into consideration every metaphysically possible option in deliberation. I hereby suggest that God’s perfect goodness and wisdom only require that God actually take into consideration all metaphysically possible options in deliberation. However, I suggest, God’s perfect goodness and wisdom does not further require that it be metaphysically necessary that God take into consideration all metaphysically possible options in deliberation. If it is metaphysically possible for God to take into consideration only some metaphysically possible options, then there are many metaphysically possible deliberation situations for God. It is, of course,

141 also morally necessary that God actually take into consideration all metaphysically possible options in deliberation, but this kind of moral necessity does not require or entail that it be metaphysically necessary that

God take into consideration all metaphysically possible options and deliberation. And, crucially for our purposes, it is metaphysical possibility that matters for the definition of agential modality. If my suggestion is granted, God’s actions can indeed be understood as agentially contingent in much the same way as, I argued, human free actions should be understood.

This similarity, I think, is best understood as a strength of my proposal. For Leibniz clearly saw God as the exemplar of freedom, and understood human freedom as in some way resembling or imitating God’s own; the closer to this exemplar the more free human agents are. Leibniz puts this point about degrees of freedom thus: “Therefore true freedom of the mind consists in recognizing and choosing the best; and as we established above, that one is freer the more knowledge one has…; and the more one is accustomed to following reason, the more determined one will be to carry out what one judges the most reasonable.”241

Leibniz’s appeal to inclining and not necessitating reason, I suggest, is to be understood in accordance with this model. God always acts on the basis of the strongest rational inclinations because of His essentially good and wise nature makes it morally necessary. God’s essential nature, however, does not make God’s actions agentially necessary, for it is still agentially possible for God to do otherwise given His essentially good and wise nature.

With this basic account of inclining and not necessitating reasons, we are now in a position to understand why Leibniz’s response to the Spinozistic Argument involves restricting Spinoza’s Assumption.

There are a couple of clear exceptions; the sufficient reasons for creation to exist are God’s nature and will, and

God’s nature is metaphysically necessary and the sufficient reason for God’s will, but it is not the case that it is metaphysically necessary that creation exists. More precisely, the sufficient reason for God’s will is God’s nature, and God’s nature is metaphysically necessary, but God’s will is not metaphysically necessary, for it is agentially possible for God to do other than what He will in fact do because of His metaphysically necessary

241 A 6.4.1409/ SLT 93.

142 nature. Another way of saying this is to say that inclining reasons do not impose metaphysical necessity upon that which they bring about and fully explain, and this is, of course, just the canonical way Leibniz articulates his merely inclining reasons doctrine.

Conclusion

In this chapter I have looked at three prominent objections to the general proposal articulated in chapter three of this dissertation. These objections are The PII Violation Objection, The Theodicy

Compromised Objection and The Buck-Passing Objection. The PII Violation Objection states that postulating a cluster of individual concepts allows for potential violations of the principle of the identity of indiscernibles, for it allows for God to instantiate more than one member of the cluster. The Theodicy Compromised

Objection states that postulating a cluster of individual concepts does away with a crucial part of Leibniz’s theodicy, namely the part that states that individuals have no grounds for complaining to God for allowing them to suffer. This part of Leibniz’s theodicy is compromised, the objector continues, because things could have been different for the substance and yet God allowed them to be this way, thus opening legitimate grounds for complaining to God. The Buck-Passing Objection has two levels. On the first level, the mere fact that the contingency of human actions depends upon the contingency of divine actions is taken to be problematic. The second level of the objection is that the contingency of human actions is lost if this contingency depends upon the contingency of divine actions. This contingency is lost because God’s actions themselves are necessary – they are deemed necessary by both His essential nature and the strictures of the principle of sufficient reason.

I have provided what I take to be plausible responses to each of these objections. I have argued that no unnecessary restriction or violation of the principle of the identity of indiscernibles is postulated by my account. I have also argued that the fact that the privileged member of the cluster is the privileged one partly because of how the agent would utilize her causal powers suffices for the creature not to have grounds for complaining to God for allowing them to sin. I have also argued that the kind of dependence holding between the contingency of human actions and the contingency of divine actions is not as problematic as it might

143 originally appear, and importantly, that Leibniz is committed to such a dependence relation. And finally, I have also argued that Leibniz account of inclining and not necessitating reasons for action provides the theoretical machinery to effectively block the Spinozistic Argument for necessitarianism on the basis of God’s essential nature and the principle of sufficient reason.

In the process of answering these objections, important details of the Leibnizian account I am developing have come to the surface. It has also become clearer how the general proposal in this dissertation fits within other crucial aspects of Leibniz’s system – his principle of identity of indiscernibles, crucial aspects of his theodicy, and commitments regarding God’s freedom and the contingency of His acts, most notably. All in all, I believe that this dissertation has made it plausible to think that Leibniz has the theoretical tools to retain an intelligible account of contingency in his system – as he clearly wanted.

144

Conclusion

Philosophical ideas, like metals, reveal some of their most interesting qualities when stretched. The philosophical idea of contingency, as a condition for freedom, is stretched much by three main Leibnizian commitments. These commitments are: i) Leibniz’s strong version of the principle of sufficient reason, ii)

Leibniz’s distinctive thesis that every substance has an individual concept that includes predicates denoting everything that will ever happen to the substance, and iii) Leibniz’s strong theological commitments like divine foreknowledge and robust providence. Many commentators have thought that these Leibnizian commitments stretch the intelligibility of contingency to the breaking point: Leibniz is not entitled to an intelligible conception of contingency as a condition for freedom.

In this dissertation I have argued that these commentators are mistaken. Leibniz does have the theoretical tools to retain an intelligible, and I may add plausible, sense of contingency in his system, despite strenuous philosophical stretches. Essential to the picture I have advocated is emphasizing that Leibniz’s views are remarkably similar to those of traditional Molinists. Importantly, Leibniz endorses the two central Molinist tenets: first, the connection between the conditions for free actions and these free actions is a contingent one: free actions follow contingently from their sufficient conditions (i.e., the Contingency Condition); second, God knows what creatures would freely do in different possible circumstances prevolitionally – that is, prior to God actually decreeing or willing anything (i.e., the Prevolitional Condition). In Leibniz’s hands, these tenets are significantly molded by his other philosophical and theological commitments – most notably a strong version

145 of the principle of sufficient reason. I have argued that Leibniz’s version of these tenets enables him to reconcile a plausible sense of contingency with his strong theological commitments and the principle of sufficient reason.

Essential to my proposal is a novel way of understanding Leibniz’s conception of individual concepts.

I have argued that an agent’s essence grounds her modal profile – what is possible, impossible, necessary and contingent – for her, and that this gives rise to a cluster of individual concepts describing this modal profile.

Importantly, I have argued that it is what an agent would freely do that grounds which of the members of the cluster describes everything that will ever happen to this agent, if she were created, and which members of the cluster only represent unactualized alternative possibilities for this agent. I have argued that this understanding enables Leibniz to get something analogous to transworld identity, and thus to ground legitimate contingency in his system.

The account I have advocated enables us to better understand Leibniz’s views on contingency and freedom, and it also enables us to better understand the remarkable similarities between Leibniz’s views and those of traditional Molinism. Thus, we can better situate Leibniz’s views with respect to his predecessors. All of this I take to be intrinsically valuable historical work. Additionally, the account advocated in this dissertation also expands our substantive understanding of the philosophical tenability of contingency as a condition for freedom, especially as it relates to the strong philosophical commitments that stretch it. We are now in a better position to understand an intelligible and plausible account of contingency as a condition for freedom even given strenuous philosophical stretches.

146

Bibliography

Abraham, E. William. 1969. “Complete Concepts and Leibniz’s Distinction between Necessary and Contingent Propositions”, Studia Leibnitiana, 1.4: 263-279.

Adams, Marilyn McCord. 1999. Horrendous Evils and the Goodness of God. Ithaca, NY: Cornell University Press.

Adams, Robert Merrihew. 1972. “Must God Create the Best?” The Philosophical Review, 81.3: 317-332.

———. 1977. “Middle Knowledge and the Problem of Evil”, American Philosophical Quarterly, 14: 109- 17.

———. 1980. “Pure Love”, The Journal of Religious Ethics, 8.1: 83-99.

———. 1991. “An Anti-Molinist Argument”, Philosophical Perspectives (Volume 5: ), James Tomberlin (ed). 343-354. Atascadero: Ridgeview.

———. 1994. Leibniz: Determinist, Theist, Idealist, Oxford: Oxford University Press.

———. 2005. “Moral Necessity” in Leibniz: Nature and Freedom. Edited by Donald Rutherford and J. A. Cover. 182-193, Oxford: Oxford University Press.

Anfray, Jean-Pascal. 2002. “God’s Decrees and Middle Knowledge: Leibniz and the Jesuits”, American Catholic Philosophical Quarterly, 76.4: 647-670.

Baxter, L. M. Donald, 2000. “Leibniz on Contingent Conceptual Truths and the Arnauld Correspondence”. Studia Leibnitiana, Bd. 32, H. 2 (2000), pp. 191-214.

Begby, Endre. 2005. “Leibniz on Determinism and Divine Foreknowledge”, Studia Leibnitiana, 37.1: 83-98.

Blumenfeld, David. 1988. "Freedom, Contingency, and Things Possible in Themselves." Philosophy and Phenomenological Research, 49.1: 81-101.

———. 1993. “Leibniz and Arnauld: A Commentary on Their Correspondence by R. C. Sleigh,” Review. Philosophy and Phenomenological Reseach. Vol. 53, No. 4 (Dec., 1993), pp. 933-943.

Borges, Jorge Luis. 1944. Ficciones: El Jardin de Senderos que se Bifurcan y Artificios. Buenos Aires, Argentina: Emece Editores.

147

Bruntrup, Godehard and Schneider, Ruben. 2013. “Complete Concept Molinism”, European Journal for Philosophy of Religion, 5.1: 93-108.

Burms, A. and De Dijn, H. 1979, “Freedom and Logical Contingency in Leibniz”, Studia Leibnitiana, Bd. 11. H. 1 (1979), pp. 124-133.

Carriero, John. 1995. “Leibniz on Infinite Resolution and Intra-mundane Contingency. Part Two: Necessity, Contingency, and the Divine Faculties”, Studia Leibnitiana, 27.1: 1-30.

Craig, William Lane. 1994. “Robert Adams's New Anti-Molinist Argument”, Philosophy and Phenomenological Research, 54.4: 857–861.

———. 1990. Divine Foreknowledge and Human Freedom. Brill's Studies in Intellectual History 19, Leiden: E.J. Brill.

———. 1988. The Problem of Divine Foreknowledge and Future Contingents From Aristotle to Suarez. Brill's Studies in Intellectual History 19, Leiden: E.J. Brill.

Code, Alan. 1986, “Aristotle: Essence and Accident,” in Philosophical Grounds of Rationality: , Categories, Ends, R. Grandy and R. Warner (eds.), Oxford: Clarendon Press, pp. 411–439.

Davidson, Jack. 2011. "Leibniz on Free Will", The Continuum Companion to Leibniz, 208-222, Ed. Brandon Look, London: Continuum.

———. 2005. "Video Meliora Proboque, Deteriora Sequor", Leibniz: Nature and Freedom. 234-252, Eds. Donald Rutherford, Jan A. Cover. Oxford: Oxford University Press.

———. 1996. “Untying the Knot: Leibniz on God’s Knowledge of Future Free Contingents”, History of Philosophy Quarterly, 13.1: 89-116.

Dennett, Daniel. 1984. Elbow Room: The Varieties of Free Will Worth Wanting. MIT Press.

Fales, E. 2010. “Is Middle Knowledge Possible? Almost”, Sophia, 50.1: 1–9.

Fine, Kit. 2005, Modality and Tense: Philosophical Papers. Oxford University Press.

———. 1994, “Essence and Modality: The Second Philosophical Perspectives Lecture”, Philosophical Perspectives, Vol. 8, and Language (1994), pp. 1-16.

Fischer, John Martin. 2008. “Molinism”, Oxford Studies in Philosophy of Religion (Volume 1). Jonathan Kvanvig (ed.), 18-43. Oxford: Oxford University Press.

———. Kane, Robert. Pereboom, Derk. And Vargas, Manuel. 2007. Four Views on Free Will. Blackwell Publishing Ltd.

———. 1994. The Metaphysics of Free Will. Oxford: Blackwell.

Flint, Thomas P. 1999. “A New Anti-Anti-Molinist Argument”, Religious Studies, 35.3: 299-305.

———. 1998. Divine Providence: The Molinist Account. Ithaca: Cornell University Press.

148

———. 1992. “, Freedom and Middle Knowledge”, in Our Knowledge of God, Kelly James Clark (ed.), Dordrecht: Kluwer Academic Publishers.

———. 1988. “Two Accounts of Providence”, in Thomas V. Morris (ed.), Divine and Human Action, 147-181. Ithaca: Cornell University Press.

Frankel, Lois. 1984. "Being Able to do Otherwise: Leibniz on Freedom and Contingency." Studia Leibnitiana 16: 45-59.

Freddoso, Alfred J. 1983. “Accidental Necessity and Logical Determinism.” Journal of Philosophy 80: 257-78

———. 1982, “Accidental Necessity and Power Over the Past,” Pacific Philosophical Quarterly, 63: 54– 68.

———. 1988. “Introduction”, In Luis de Molina: On Divine Foreknowledge (Part IV of the Concordia), trans. and ed. Alfred J. Freddoso, 1-81. Ithaca, NY: Cornell University Press.

Gaskin, R., 1994. “Molina on Divine Foreknowledge and the Principle of Bivalence”, Journal of the History of Philosophy, 32.4: 551–571.

———. 1993. “Conditionals of Freedom and Middle Knowledge”, The Philosophical Quarterly, 43.173: 412–430.

———. 1998. “Middle Knowledge, , and Comparative Similarities of Worlds”, Religious Studies, 34.2: 189-203.

Graham, A. Peter. 2008. “A Defense of Local Miracle ”, Philosophical Studies 140: 65-82.

Greenberg, Sean. 2005. "Leibniz Against Molinism: Freedom, Indifference, and the Nature of the Will", Leibniz: Nature and Freedom. Eds. Donald Rutherford, Jan A. Cover, 217-233. Oxford: Oxford University Press.

Griffin, V. Michael. 2013. Leibniz, God and Necessity. Cambridge University Press.

———. 1999. “Leibniz on God’s Knowledge of Counterfactuals”, The Philosophical Review, 108.3: 317- 343.

Grimm, Robert. 1970, “Individual Concepts and Contingent Truths”, Studia Leibnitiana, Bd. 2, H. 3 (1970), pp. 200-223.

Hasker, William. 2011. “The (non-)Existence of Molinist Counterfactuals”, in Molinism: The Contemporary Debate. Ed. Ken Perszyk. Oxford University Press.

———. 2001, “The Foreknowledge Conundrum”, International Journal for Philosophy of Religion, Vol. 50, No. 1/3.

———. 1999. “A New Anti-Molinist Argument”, Religious Studies, 35.3: 291-297.

149

———. 1997. “Explanatory Priority: Transitive and Unequivocal, A Reply to William Craig”, Philosophy and Phenomenological Research, 57.2: 389-393.

———. 1992. “Providence and Evil: Three Theories”, Religious Studies, 28.1: 91-105.

———. 1989. God, Time, and Knowledge. Ithaca: Cornell University Press.

———. 1986. “A Refutation of Middle Knowledge”, , 20.4: 545-557.

Hunter, Graeme. 1981, “Leibniz and the ‘Super-Essentialist’ Misunderstanding”, Studia Leibnitiana, 13.1: 123-132.

Jorati, Julia. 2017a. Leibniz on Causation and Agency. New York: Cambridge University Press.

———. 2017b. “The Contingency of Leibniz’s Principle of the Identity of Indiscernibles”, Ergo 4.31: 899-929.

Kane, Robert. 1996. The Significance of Free Will. New York: Oxford University Press.

———. 2007, “Libertarianism” in Fischer, John Martin, Pereboom, Derk. And Vargas, Manuel. Four Views on Free Will. Blackwell Publishing Ltd.

Kaphagawani, Didier Njirayamanda. 1999. Leibniz on Freedom and Determinism in Relation to Aquinas and Molina. Brookfield: Ashgate.

Kent, Bonnie. 1995. of The Will: The Transformation of Ethics in the Late Thirteenth Century. The Catholic University of America Press.

Knebel, N. Sven. 1996. “Leibniz, Middle Knowledge, and the Intricacies of World Design”, Studia Leibnitiana, 28.2: 199-210.

Lagerlund, Henrik, and Myrdal, Peter. 2006. “Possible Worlds and the Nature of Choice in Leibniz”, Studia Leibnitiana, 38/39.2: 156-176.

Lewis, David. 1981. “Are We Free to Break the Laws?” Theoria 47:113-21.

Look, C. Brandon. 2013, “Leibniz’s Modal Metaphysics” Stanford Encyclopedia of Philosophy entry.

Lovejoy, O. Arthur. 1964. The Great Chain of Being: a Study of the History of an Idea. Cambridge, Mass: Harvard University Press.

Mates, Benson, 1972. “Individuals and Modality in the Philosophy of Leibniz,” Studia Leibnitiana, IV: 81–118.

———. 1986. The Philosophy of Leibniz: Metaphysics and Language. New York, Oxford. Oxford University Press.

Mavrodes, George. 1984. “Is the Past Preventable?” Faith and Philosophy, 1: 131-46.

150

———. 1997. “Omniscience.” In A Companion to Philosophy of Religion. Ed. Philip L. Quinn and Charles Taliaferro, 236-42. Cambridge, Mass: Blackwell.

McNamara, Paul. 1990, “Leibniz on Creation, Contingency and Per-Se Modality”, Studia Leibnitiana, 22.1: 29-47.

Meijering, Theo. 1978, “On Contingency in Leibniz’s Philosophy”, Studia Leibnitiana, Bd. 10, H. 1 (1978), pp. 22-59.

Mondadori, Fabrizio, 1973, “Reference, , and Modality in Leibniz’s Metaphysics”, Studia Leibnitiana, Bd. 5, H. 1 (1973), pp. 74-101.

———. 1975, “Leibniz and the Doctrine of Inter-World Identity”, Studia Leibnitiana, Bd. 7, H. 1 (1975), pp. 21-57

Murray, Michael J. 2005. "Spontaneity and Freedom in Leibniz", Leibniz: Nature and Freedom. Eds. Donald Rutherford, Jan A. Cover, 194-216. Oxford: Oxford University Press.

———. 2004. "Pre-Leibnizian Moral Necessity", The Leibniz Review, 14: 1-28.

———. 1996. "Intellect, Will, and Freedom: Leibniz and His Precursors", The Leibniz Review, 6: 25- 59.

———. 1995. "Leibniz on Divine Foreknowledge of Future Contingents and Human Freedom", Philosophy and Phenomenological Research, 55.1: 75-108.

O’Connor, Timothy. 1992. “The Impossibility of Middle Knowledge”, Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition, 66.2: 139-166.

O’Leary-Hawthorne, John. And Cover, J. A. 1990, “Leibniz on Superessentialism and World-Bound Individuals”, Studia Leibnitiana, 22.2: 175-183.

———. 1992, “Leibnizian Essentialism, Transworld Identity, and Counterparts”, History of Philosophical Quaterly, Vol. 9, No. 4 (Oct., 1992), pp. 425-444.

Perszyk, J. Kenneth. 2013. “Recent Work on Molinism”, Philosophy Compass, 8.8: 755-770.

———. 2011. (Editor), Molinism: The Contemporary Debate. Oxford: Oxford University Press.

———. 1998a. “Free Will Defense with and without Molinism”, International Journal for Philosophy of Religion, 43.1: 29-64.

———. 1998b. “An Anti-Molinist Argument”, Philosophical Studies: An International Journal for Philosophy in the Analytic Tradition, 90.3: 215-235.

Plantinga, Alvin. 1977. God, Freedom and Evil. Grand Rapids: B. Eerdmans Publishing.

———. 1974. The Nature of Necessity. Clarendon University Press.

———. 1986, “On Ockham's Way Out,” Faith and Philosophy, 3 (3): 235–269.

151

Rodriguez-Pereyra, Gonzalo. 2013. “The Principles of Contradiction, Sufficient Reason, and Identity of Indiscernibles”, in Oxford Handbook of Leibniz. Ed. Maria Rosa Antagnazza. Oxford: Oxford University Press.

Rodriguez-Pereyra, Gonzalo, and Paul Logge. 2011. “Infinite Analysis, Lucky Proof, and Guaranteed Proof in Leibniz,” Archiv fur Geschichte der Philosophie, 93: 222-236.

Savage, O. Reginald. 1998, Real Alternatives, Leibniz’s Metaphysics of Choice. Kluwer Academic Publishers.

Smith, Gerard. 1966. Freedom in Molina. Chicago: Loyola University Press.

Sleigh, Robert C. 2009. "Moral Necessity in Leibniz's Account of Human Freedom." Metaphysics and the Good: Themes from the Philosophy of . Eds. Samuel Newlands, Larry M. Jorgensen, 252-271. Oxford: Oxford University Press.

———. 1999. "Leibniz on Freedom and Necessity: Critical Notice of Robert Adams, Leibniz: Determinist, Theist, and Idealist." The Philosophical Review, 108.2: 245-77.

———, 1994. “Leibniz on Divine Foreknowledge”, Faith and Philosophy, 11.4: 547–571.

———. 1990. Leibniz & Arnauld: A Commentary on Their Correspondence. New Haven, London: Yale University Press.

Sleigh, Robert C., Vere Chappell, and Michael Della Rocca. 1998. "Determinism and Human Freedom." The Cambridge History of Seventeenth-Century Philosophy. Eds. Daniel Garber, Michael Ayers. Vol. 2, 1195-1273. Cambridge: Cambridge University Press.

Sotnak, Eric. 1999. "The Range of Leibnizian Compatibilism." New Essays on the Rationalists. Eds. Rocco J. Gennaro, C. Huenemann, 200-223. Oxford: Oxford University Press.

Steward, Stephen. 2014. “Solving the Lucky and Guaranteed Proof Problems,” The Leibniz Review, 24: 29-45.

Stump, Eleonore. 1985. “The Problem of Evil”, Faith and Philosophy, 2.4: 392-423.

———. 1990. “Providence and the Problem of Evil”, in The Oxford Handbook of Aquinas, 401-418. Oxford: Oxford University Press.

———. 2001. “Second-Person Account and the Problem of Evil”, Revista Portuguesa de Filosofia, 57.4: 745-771.

———. 2010. Wandering in Darkness: Narrative and the Problem of Suffering. Oxford: Clarendon Press.

Torralba, Jose Maria. 2005. “La Libertad Possible Acerca de la Nocion Leibniziana de Inclinar sin Necesidad”, Anuario Filosofico, 38.1: 279-312.

Vailati, Ezio, 1986, “Leibniz on Necessary and Contingent Predication”, Studia Leibnitiana, Bd. 18, H. 2 (1986), pp. 195-210.

152

Van Inwagen, Peter. 1983, An Essay on Free Will, Oxford: Oxford University Press.

———. 1975. “The Incompatibility of Free Will and Determinism”, Philosophical Studies 27: 185-99.

Wee, Cecilia. 2006, “Descartes and Leibniz on Human Free-Will and the ability to Do Otherwise”, Canadian Journal of Philosophy, Vol. 36, No. 3 (Sep., 2006), pp. 387-414.

Zimmerman, Dean. 2009. “Yet Another Anti-Molinist Argument”, in Metaphysics and the Good: Themes from the Philosophy of Robert Merrihew Adams. Eds. Samuel Newlands and Larry M. Jorgensen, 34- 92. Oxford: Oxford University Press.

153