The Catholic University of America Press
  • Self-determination vs. Freedom for God and the Angels:A Problem with Anselm's Theory of Free Will

Anselm is known for offering a distinctive definition of freedom of choice as "the ability of preserving uprightness of will for its own sake." When we turn to Anselm's account of the devil's fall in De Casu Diaboli, however, this idiosyncratic understanding of freedom is not at the forefront. In that text, Anselm seemingly assumes a traditional understanding of free will defined in terms of alternative possibilities for the angels. These alternative possibilities must be present so the angels can engage in 'self-determination.' God, however, does not face alternative possibilities to achieve His self-determination. Anselm thus explicates his notion of free will in terms of three different concepts: his distinctive definition of free choice, self-determination, and the principle of alternative possibilities. Despite attempts (by both scholars and Anselm) to explain how these three concepts are related, I argue that their relationship is problematic. In particular, I argue that Anselm is guilty of conflating and equivocating with regard to these concepts. I further importantly claim that the conflation obscures the fact that his understanding of self-determination calls into question God's excellence over that of the good angels.

Anselm is known for offering a distinctive definition of freedom of choice. At the beginning of De Libertate Arbitrii, he famously denies that "free choice" (liberum arbitrium) consists of "the power to sin and not to sin."1 Instead, it should be defined as "the ability of [End Page 13] preserving uprightness of will for its own sake."2 His primary reason for rejecting the former definition is his assumption that God and the good angels also have freedom of choice. Since they are free but nonetheless cannot sin, the ability to sin cannot be part of the definition of free choice.3 When we turn to Anselm's account of the devil's fall in De Casu Diaboli, however, this idiosyncratic understanding of freedom is not at the forefront. In that text, Anselm seemingly assumes a traditional understanding of free will defined in terms of alternative possibilities. In particular, he claims the angels were faced with the alternative possibilities of sinning and not sinning.4 And the reason these alternative possibilities must be present is so the angels can engage in "self-determination"; self-determination for them requires the alternative possibility of being able to sin. By contrast, Anselm does not argue that self-determination requires the alternative possibility of being able to sin in the case of God.

Anselm thus appeals to three different concepts while explicating his notion of freedom. First, there is Anselm's distinctive understanding of freedom of choice (ADF). Second, there are his repeated appeals to the concept of 'self-determination' (SD). And third, so far as the alternative possibility of being able to sin is concerned, there is an appeal to the principle of alternative possibilities (PAP). Their relationships can be symbolized in the table below.

ADF SD PAP
Unclear relationship to SD Unclear Relationship to ADF Excluded from ADF
Excludes PAP Ambivalent relationship to PAP Ambivalent relationship to SD

Upon examination, a couple of puzzles emerge. First, why does SD have this ambivalent relationship with PAP? Why would PAP be required for SD in some cases (that of the angels and humans) and not in another (that of God)? Second, what is the precise relationship between SD and ADF? How can ADF and SD be related if ADF excludes PAP while SD sometimes requires it? [End Page 14]

In this paper, I argue that both puzzles can be traced back to cases of equivocation. First, I assert that SD is understood differently when discussing God and the angels. This shifting understanding of SD, moreover, importantly threatens to undermine the traditional theistic claim of God's excellence over the good angels. I then argue that Anselm conflates ADF with this somewhat equivocal understanding of SD. This conflation leads, in turn, to an equivocation on the notion of free choice. The conflation and subsequent equivocation on freedom of choice function to obscure the threat to God's excellence posed by the varying understandings of SD. And despite how often it may be conceded that freedom or SD mean something different in the case of God and the angels, it remains the case that God's excellence over the angels is, in some sense, called into question by the equivocation and obscured by the conflation.

Before I begin, I offer a few quick notes on my terminology. Recall that according to ADF, free choice is the "ability of preserving uprightness of will for its own sake." Since Anselm had previously defined justice as preserving uprightness of will for its own sake,5 ADF can be rephrased as the ability to uphold or preserve justice. I will thus sometimes discuss ADF in terms of upholding justice. Second, when I refer to PAP, I often have in mind the precise alternative possibility of being able to sin—precisely what Anselm denies is part of ADF. I will refer to this alternative possibility as PAPsin. It will also become clear that the notion of self-determination (SD) is equivalent to the notions of aseity, having a trait a se, and acting in a self-originating (sponte) manner. I will use these notions interchangeably.

Finally, I should comment on my charge that there exists equivocation in Anselm's account of SD. I will claim that SD is not understood univocally with regard to God and the angels; SD in the case of the angels is understood in a way slightly differently from how it is understood in the case of God. Given this lack of univocity, I state there is a degree of equivocation with regard to SD. This charge of equivocation does not rule out the possibility that there is some sort of analogy between God's SD and that of the good angels; there may indeed be an analogical relationship or family resemblance between these two notions of SD. It is not essential to my argument to discuss what "kind" of equivocation or analogy may be at play here. Similarly, my argument does not require that there be a "pure" equivocation with regard to SD. Instead, it is only necessary that I show that SD is [End Page 15] not understood completely univocally in both cases. I intend no more than this whenever I call Anselm's understanding of SD equivocal. The reader is thus advised not to get too "hung up" or troubled by my use of the term equivocation with regard to SD. Whether the various understandings of SD constitute equivocal as opposed to analogical (if such a distinction is legitimate) uses is ultimately beside the point. The point will be that the understandings of SD vary to some degree. This fact alone is enough to underlie the argument in this paper.

The Fall of the Devil

In De Casu Diaboli (DCD), Anselm offers a thought experiment in which God has created an angel prepared to will but not yet willing anything. Since an angel could not bring himself to will anything unless it had a disposition to will, God must grant to this angel a disposition (which he later calls an affectio).6 He then considers the hypothetical case in which the angel received only an affectio-for-happiness. Such an angel would be unable to will anything other than his greatest happiness. Furthermore, he would have no regard for the dictates of justice. As a result, this angel could not will justly. Paradoxically, the angel could not even be happy since one is not truly happy without a just will.7

Anselm then considers the case in which an angel receives only an affectio-for-uprightness. Such an angel would have to will that which he thought was just. But in willing such just acts, the angel would fail to be just because such willing would be necessitated. In Anselm's words, he would "not be able to will differently" than he does.8

Since having neither affectio alone is enough for the angel to be both just and happy, God granted all the angels both the affectio-for-happiness and the affectio-for-uprightness. But even doing this was not sufficient for the angels to will uprightness of will for its own sake. If both affectiones inclined the angels to will the same just action, then the angels' willings would once again be necessitated (since there [End Page 16] would exist no disposition in the angels' wills to will any differently). Their willing of justice would thus not have been for its own sake. As a result, God had to engineer a situation in which the affectio-for-happiness inclined the angels to will one way while the affectio-for-uprightness inclined them to will the other way. If an angel willed in accordance with its affectio-for-uprightness as opposed to in accordance with its affectio-for-happiness, the upright willing would be for its own sake and not due to either considerations of happiness or necessity.

To create this situation, God forbade the angels from willing a particular something that would, if willed, seemingly make them happy.9 The demands of justice would require them to follow God's command and not will it while their affectiones-for-happiness would incline them to disobey God and will their own happiness. It was then up to the angels to choose which inclination to will in accordance with.10 They could disregard considerations of justice and attempt to will their own happiness.11 Alternatively, they could forgo considerations of their own happiness and obey God. Doing so would be upright and it would not be for the sake of happiness; rather, it would be for the sake of uprightness of will itself.12 And as the story goes, the devil and some other angels willed their own happiness and thereby sinned while the rest did not sin by willing uprightly and obeying God.13 [End Page 17]

Self-Determination and PAP

In the story of the devil's fall, it seems the devil was presented with the alternative possibilities of sinning (i.e. willing happiness) and not sinning (i.e. willing uprightness of will). As a result, it is natural to assume the angels' freedom did indeed consist in having the alternative possibility of being able to sin (PAPsin). Many have argued, however, that Anselm is not concerned with freedom per se in presenting these alternative possibilities. To the contrary, it is often claimed that he is concerned with self-determination (SD). Visser and Williams go even further to claim "there is no appeal to PAP in this argument. Rather, the appeal is to the requirement that a free action have its origin in the agent rather than in some external cause."14 What Anselm is ultimately concerned with is that the angels be able to act "from themselves" and not out of any necessity. And PAPsin is required, in the angels' case, for such aseity.

Why would PAPsin be required for aseity? The answer is that nothing the angels possess is from themselves. Since they are creatures, everything they have is from God. The only way for them to have something truly from themselves, a se, is to allow them to make a choice among alternatives. In this way, they can have at least their uprightness, or lack thereof, from themselves. As Visser and Williams state, "alternative possibilities come into the picture as a kind of by-product . . . they just happen to be available, given the requirement that free action have its origin within the agent."15

When we turn to the question of SD with respect to God, however, the picture is different. God's SD does not require that "alternative possibilities come into the picture as a kind of by-product." In Cur Deus Homo II, ch. 5 Anselm is attempting to deny that God acted out of necessity when He provided a means of atonement. He states that "this necessity is nothing other than the immutability of His honorableness, which He has from Himself and not from another and which is therefore [End Page 18] improperly called necessity."16 Commenting on this passage, Visser and Williams write that "every upright divine action will be self-initiated; and for that very reason Anselm insists that no such action should be called "necessary." Divine aseity in fact guarantees that every action God performs is self-initiated. So all of God's actions are free, even if He never has alternative possibilities available to Him."17 Note that aseity in God's case does not require alternative possibilities. Since God is fully a se, all of God's actions would be self-determined. And since God cannot sin, then PAP is not required for God's self-determination. As a result, PAP is required for SD in the case of agents who are not already a se, but it is not required in the case of God.

The Problem with SD and PAP

I claim that SD's differential relationship with PAP means that SD is being used somewhat equivocally with regard to moral responsibility.18 Moreover, I hope to show that this equivocal use of SD implies that God, in some sense, had an unfair advantage with regard to being just in a self-determined way. There is thus a sense in which the good angels' SD for justice is more valuable than God's self-determination for justice.

Let us return to the angels' initial choice. In order for the angels to be truly self-determined, they had to be faced with the alternative possibility of sinning. Indeed, it could be said that they were tempted by sin, since it seemed to them as if they could have disregarded justice and willed their complete happiness without any negative consequences. The angels who resisted this temptation and obeyed God gave themselves (in a sense) justice. Those who gave into the temptation and disobeyed God made themselves unjust.19 Their self-determination arose from having true options. Those who determined themselves to be just did so by denying the temptation.

When it comes to God, God is fully a se. God Godself therefore determines God's own actions. But God is also supremely good a se and could not sin.20 When God acts, therefore, God could not act [End Page 19] unjustly; God can only act justly. But since God is fully a se, God's just acts are done a se. God thus self-determines Godself to act justly.

While on the one hand it is obvious that God's SD is different from the angels' SD since God is fully a se, it has not been pointed out (so far as I am aware) that there is a sense in which it was easier for God to be just a se than it was for the good angels. God never had to struggle or face any real temptation; there was never any true risk of failure in God's case of acting. All His actions must issue forth in justice. But for the angels, there was a real risk—so real that the devil and several other angels are putatively still suffering for having made the wrong choice. Seen in one light, one might even suspect the 'game' was rigged unfairly in God's favor. God did not have to worry about ever abandoning justice since He could not. And since he is completely a se, His inability to abandon justice counts as his determining himself to be just. But in the case of creatures, they could abandon justice. For creatures to determine themselves to be just, they had to face the real possibility and temptation of not being just. As a result, God seems to have possessed an unfair advantage with regard to determining Himself to justice. At the very least, God's determining Godself to be just is qualitatively different from the way in which the good angels determined themselves to be just. In other words, SD for uprightness of will is employed somewhat equivocally.

Another way to put this point is to appeal to John Hick's idea of "hard won virtues." While explaining his famed "Soul-Making Theodicy," Hick makes the point that God could not have created humans already virtuous and in the "likeness" of God. His reasoning is that "virtues which have been formed within the agent as a hard won deposit of his own right decisions in situations of challenge and temptation, are intrinsically more valuable than virtues created within him ready made and without any effort on his own part."21 Virtues that are earned in the face of temptation are more valuable than those that are acquired without struggle. Note, however, that the good angels' self-determined justice seems "hard won" in a way that God's self-determined justice is not. If Hick is right about the value of hard won [End Page 20] virtues, then it would seem that the good angels self-determining themselves to be just would be better than God's own non-hard won self-determination. In other words, God had it "easier," and God's self-determined justice is therefore not as good as that possessed by the good angels.

One might respond by claiming that Hick is simply emphasizing the value of SD in general when he lauds the value of hard won virtues. By contrasting hard won virtues with those "created within [the agent] ready made and without any effort on his own part," Hick is simply emphasizing the importance of the agent self-determining herself to be virtuous. Under this interpretation, the value of hard won virtues is not that they are "hard won" per se, but rather that they have their origin in the agent and are not given to the agent from without. As a result, Hick's concept of hard won virtues could not be used to support my claim that God's SD is "easier" and thus not as valuable as that of the good angels.

For what it is worth, this is not necessarily the rationale Hick himself employs. Instead, Hick seems to emphasize the effort. He writes,

this principle expresses a basic value-judgment, which cannot be established by argument but which one can only present, in the hope that it will be as morally plausible, and indeed compelling, to others as to oneself. It is, to repeat, the judgement [sic] that a moral goodness which exists as the agent's initial given nature, without ever having been chosen by him in the face of temptations to the contrary, is intrinsically less valuable than a moral goodness which has been built up through the agent's own responsible choices through time in the face of alternative possibilities.22

The emphasis seems to be on the fact that hard won virtues are valuable because the agent had to "fight" for them. Ultimately, what Hick himself thinks is not directly relevant. Instead, I can simply follow Hick's approach by presenting a principle which, I presume, may be compelling to others—namely, that a just will that is due to one's own choices in the face of alternative possibilities and the real potential of failing is more valuable than one that is due to one's own nature, even if that own nature is (as in the case of God) completely a se.23 [End Page 21]

Consider the act of winning Wimbledon. It seems like it would be a "bigger deal" and "better" for me to win Wimbledon than for God to win it. After all, God is God! God is omnipotent and defeating all the tennis foes at Wimbledon would be no problem. Such a win for God would be about as satisfying as beating a 3rd grade basketball team would be for the NBA champion Golden State Warriors. There thus seems to be a real sense in which it would be "better" for me to win Wimbledon than for God. My winning Wimbledon would be "hard won" in a way God's winning Wimbledon would not. But what holds for Wimbledon should hold for having uprightness of will. God is completely good a se and is incapable of acting unjustly. Given God's completely a se nature, he does not need to struggle to have uprightness of will. Having such uprightness would be automatic for him. For God to win the "uprightness of will" game (so to speak) should be no more satisfying for God than for Him to win Wimbledon. It is, in a sense, too easy given God's a se nature. God's self-determined upright will is thus different from and, in some sense, not as good as the self-determined upright will of the good angels.

One might resist this line of thought by arguing that the "hard won-ness" of justice is irrelevant to determining which case of self-determination is better. Such a response might proceed by insisting instead that simply having a virtue or quality a se is itself a good. It is good because it is better to give a quality to oneself than to receive it from another. And the more one is "from oneself"—i.e., the more a se one is, the better. Since God is completely a se, His self-determination for justice is (in contrast to my charge above) better.

But once again, note that God is a se in a way different from how the good angels are a se. God cannot be said to have "given" any of His qualities to Himself. Instead, God came "ready made" with all God's qualities.24 In the words of Wes Morriston, there is a real sense [End Page 22] in which "God is simply 'stuck' with his moral nature, he is not responsible either for it or for what follows from it."25 Given God's nature, God's moral rightness was unavoidable.26

Consider Morriston's "Bonnie Chance" thought-experiment. He asks us to suppose a finite person, "Bonnie Chance," who comes "into existence purely by chance and without any cause whatsoever" and whose "nature prevents her from ever choosing what she sees to be less than the best." Bonnie's actions would ultimately arise from herself and from "no external chain of causes." In this sense, her actions could be considered a se. But despite this a se nature of her actions, Morriston denies we would call her free. Why? Because "she is 'stuck' with a nature that prevents her from ever going wrong."27 Although Morriston phrases his discussion more in terms of freedom than self-determination per se, the example is helpful. If one is, in some sense, "stuck" with a nature, then actions issuing forth from that nature should not be positively credited to the agent. But God is similarly "stuck" with His own nature. By the same token, therefore, God's moral actions should not necessarily count for Him as "free" and worthy of moral credit.28 At the very least, God's self-determining Godself to be just seems less meritorious than the good angels (who were not "stuck" with justice) determining themselves to be just. [End Page 23]

Even the staunchest defenders of Anselm, when commenting on the topic of God's aseity, inadvertently lend support to the claim that God's SD for justice is different from that of the good angels due to God's "stuck-ness." For example, Katherin Rogers writes that God's absolute a se existence means "there is no need in the divine will for some apparatus by which it may, in Anselm's words, 'give itself justice.'" God's a se justice thus arose in a different way from the a se justice of the good angels. The good angels did have to give it to themselves, whereas God's nature entailed that God "just had" it; God's acting justly was inevitable. Rogers continues by stating that God's "inevitably doing the best in no way conflicts with His power to keep rightness of will for its own sake."29 Even if this is true, the basic point remains unaltered: the self-determined nature of God's justice is different from the self-determined nature of the good angels' justice. God did not have to "do" anything to give Himself justice whereas the good angels did. It therefore seems that the good angels' just natures are self-determined in a way that is, in some sense, different from and better than how God's is self-determined.

As additional support for my claim that God had it "easier" with respect to SD, consider an argument made by Theodore Guleserian. Guleserian is concerned with the "property PW—perfection of the will," a property sufficiently similar to Anselm's notion of uprightness of will that we can use it as a surrogate. After contrasting having PW essentially from having PW freely, he notes that that the traditional theist (Anselm included) must be "committed to hold that God has essential PW" instead of "free PW" since "the property of having essential PW is a greater excellence than the property of having free [i.e. contingent] PW."30 Translated into Anselmian terms, the traditional theist holds it would be better for God to have uprightness of [End Page 24] will fully a se instead of contingently (e.g. after a struggle where God could have forsaken uprightness of will). So far, so good! Anselm would agree. But Guleserian then notes that if having essential PW is a greater excellence than having free, contingent PW for God, then this "should be true for any moral agent who can have essential PW." Since an omnipotent God, moreover, could have created "excellent nonfree creatures having essential PW," those are the creatures God should have created instead of creatures with contingent, free PW (e.g. the good angels).31 Obviously, those are not the kinds of creatures God created. But if God did not create those kinds of creatures, then it must be the case that the traditional view is wrong. It is not better to have PW essentially. Instead, it is better to have PW freely.

Guleserian has basically run a reductio on the claim that having essential PW is better for God. Since God did not create creatures with essential PW, it must be the case that free, contingent PW is better than essential PW. And this is precisely the point I have been making in this section of the paper. The good angels' PW (i.e. uprightness of will) would seem to be better than God's PW (i.e. uprightness of will) since the good angels' PW is contingent while God's is essential. Guleserian himself uses these arguments to reject "the traditional view that God is essentially morally perfect in favor of the nonAnselmian view that God is morally perfect by his own free (i.e. libertarian) choice" since, as he has argued, free PW is better.32 If the Anselmian is not willing to abandon the view that God is essentially morally perfect, then the alternative is the one I have been making: God's PW (i.e. uprightness of will) is in some sense less valuable than that possessed by the good angels who putatively achieved their uprightness of will through a libertarian choice. While both God and the angels may be self-determined in their uprightness of will, there is a strong case to be made that the self-determination of the good angels is more valuable.

In response, it might be claimed that essential PW is better overall (thus the reason God has it), but that creating angels with essential PW would have nonetheless been worse than creating them with free PW. In other words, one might try to deny that essential PW is a greater excellence "for any moral agent who can have essential PW." In order to make this defense, one might claim that God's essential PW arises fully a se whereas an angel with essential PW could possess it only from [End Page 25] another. In order for the angel to be truly good and a se and autonomous with regard to its PW, it would need to possess PW freely.

Guleserian addresses this objection in terms of autonomy and argues that it is not as decisive as may initially seem. He notes that a God who has essential PW a se would still not be completely "autonomous in the sense claimed by the traditionalist" and would not be independent from "determining external factors."33 His basic point is that the actions of a God with essential PW would sometimes be determined by the actions of others. Perfection of will would entail that God would have to react in certain ways to the actions of certain individuals and, in that sense, God would "be subject to the will of another." If God promised, for example, a certain Luke forgiveness upon repentance and Luke did indeed repent, "God's volition is in this case determined by the external fact of Luke's act." Guleserian's point is that if autonomy is important (as the objector is asserting), complete autonomy will not be secured for God by claiming God has PW essentially (or, in our terms, by claiming that God's uprightness of will follows from His completely a se nature): "If complete autonomy is important one can find it only in a being that is fully free in the libertarian sense."34 If Guleserian is correct, then even an emphasis on autonomy ends up supporting the claim that choosing among alternatives (as the good angels did in deciding not to sin by preserving uprightness of will) is better than preserving uprightness of will due to one's own nature—even if that nature is a se and essential. Ultimately, whether Guleserian's specific arguments are successful or not is somewhat beside the point. What is not beside the point is that there is a real sense in which the angel's SD for justice in the face of alternative possibilities is different from, and seemingly more laudable than, God's SD for justice in the absence of alternative possibilities.

What Does Freedom Have To Do With SD?

We have seen why Anselm claims SD sometimes requires PAPsin. The moral responsibility of the angels depends upon their being self-determined, and PAPsin is required for their SD. But if SD is the ground for moral responsibility, and if SD's differential relationship to PAP can be explained as we have just done, then why did Anselm feel the need to offer a separate definition of freedom? Why should freedom [End Page 26] not have simply been defined as "the ability to act in a self-determined way"? After all, we have already seen several quotations above in which SD cannot be spoken of without referencing freedom in some sense or other (the two quotations from Visser and Williams under "Self-Determination and PAP"). And freedom, moreover, has typically been associated with moral responsibility. Given these facts, why would Anselm not simply define freedom in terms of self-determination instead of offering ADF and thereby avoid the counterintuitive claim that PAPsin is not part of freedom.

Anselm originally offers ADF because he wants a definition of freedom that would apply to God and the good angels as well as humans. But a definition in terms of SD, as Anselm understood it,35 should have also accomplished this aim! Why offer instead a definition of freedom that appeals to preserving uprightness of will for its own sake? While a case could certainly be made that aspects of ADF relate to self-determination, there appears to be no need to invoke the concept of "uprightness of will" if his true focus was on the fact that one's actions be self-determined. To appeal to "uprightness of will" seems to steer the definition off onto a direction tangential from his main concern.36

I want to suggest that part of the answer lies in the fact that Anselm (no doubt inadvertently) conflates ADF with his somewhat equivocal notion of SD. This conflation has the effect of causing him to equivocate between the two definitions of free choice (freedom defined as ADF and freedom defined as self-determination) when discussing liberum arbitrium. Moreover (and this is the key part!), the conflation and equivocation have the overall effect of helping Anselm gloss over the problems with SD and PAPsin discussed above. In other words, Anselm's appeal to ADF helps obscure the fact that the SD of the good angels is in some sense better than the SD of God since the good angels faced PAPsin. Let us begin by exploring the claim that Anselm conflates SD and ADF. [End Page 27]

Conflation & Equivocation: ADF, SD and Freedom of Choice

Since SD grounds moral responsibility and ADF is phrased in such a distinctive way, one may reasonably think that freedom is not related to moral responsibility in Anselm's system. Indeed, Stan Tyvoll has gone so far as to state that "the two concepts—what free will is and what makes me ultimately responsible—are separate and distinct."37 But if freedom of choice has really been severed from moral responsibility, Anselm's action theory should count as a shocking and unexpected break from the philosophical tradition.38 The simple rejoinder is that Anselm's emphasis on aseity should not imply that freedom has been thus severed from moral responsibility. Instead, one might assert that aseity and freedom go hand in hand; to act with aseity, one must act freely. This close connection between freedom and aseity can be observed even in Tyvoll's article. He prefaces the comment above by stating that the distinction is between "what the essence of free will is and the way in which someone is responsible for exercising it," thus expressing a desire to maintain some connection between free will and moral responsibility.39

This close connection between aseity and freedom is implicitly, if not explicitly, made by most who comment on this part of Anselm's action theory and by Anselm himself. For example, Visser and Williams cannot avoid alluding to freedom and free actions while commenting on Anselm's appeal to aseity. They state that "a free action [must] have its origin in the agent" and that there is "the requirement that a free action have its origin within the agent."40 As we have also seen, they state, "Divine aseity in fact guarantees that every action God performs is self-initiated. So all of God's actions are free, even if He never has alternative possibilities available to Him."41 Note the close and apparently [End Page 28] interchangeable nature among aseity, self-initiation, and free action. Katherin Rogers does the same. While discussing DCD in a section entitled "Aseity," she writes, "the free choice must originate in the agent himself."42 In her most recent book she again assumes the close connection between aseity and freedom. In the section "Aseity and Alternative Possibilities," she asserts that "the most basic requirement for the morally significant freedom which can ground moral responsibility is . . . that one's choices come from oneself, they are a se."43

The important point to note is that Anselm's concept of SD is often discussed in the secondary literature with reference to freedom. Freedom of choice becomes conflated with SD and the terms are used, if not fully interchangeably, as intricately related. In these instances, freedom of choice is primarily understood not as preserving uprightness of will, but rather as a surrogate for SD. It is this conflation of freedom with SD that leads to an imperceptible equivocation on the notion of freedom of choice—sometimes understood as SD, and sometimes understood as ADF.

This type of conflation can be observed in Anselm's own writings. In DLA ch. 2, Anselm twice states that the devil's sin was through his freedom of choice.44 Anselm himself thus makes an explicit connection between freedom of choice (liberum arbitrium) and the devil's sin (which resulted from his self-determining himself for PAPsin). In that same chapter, Anselm goes on to state:

They sinned through their choice, which was free; but they did not sin through that in virtue of which it was free, that is, through the power by which it was able not to sin and not to be a slave to sin. Instead, they sinned through that power they had for sinning.45

Commenting on this passage, Visser and Williams claim that Anselm is simply saying "that they sinned by an exercise of that power" (italics mine). But when Anselm claims that neither sinned "through [End Page 29] that in virtue of which [their choice] was free," he is (according to Visser and Williams) "emphasizing the teleological nature of freedom; full-fledged freedom of choice is the power for self-initiated action for some good end."46 Their idea seems to be this: freedom of choice per se is teleological in nature; its purpose is to allow the agent to uphold justice. This is why its official definition excludes PAP. But freedom of choice brings along with it, so to speak, the side-benefit (or rather, side-drawback?) of being able to choose to sin by failing to uphold justice. As a result, freedom can be used to sin without the ability to sin being part of the definition of freedom.

If this interpretation is correct, Anselm is equivocating with regard to freedom of choice. Freedom of choice (properly understood) does not entail the ability to sin. But freedom can, according to Anselm, nonetheless be "exercised" so as to commit a sin. But if it can be so exercised, it does seem to entail the ability to sin after all! As such, freedom of the will would seem both to preclude and include an ability to sin—an equivocal position. To say freedom of choice does not include the possibility to sin when "properly" understood but that it can nonetheless be that through which an agent can sin seems to be a dubious case of wordsmithing at best. At any rate, it constitutes some degree of equivocation. And this equivocation arises precisely because freedom of choice is conflated with the somewhat equivocal notion of SD according to which PAPsin is sometimes required.47

Stated differently, the equivocation occurs by means of freedom's conflation with SD. Since SD admits of two slightly different meanings and different relationships to PAP and since freedom is conflated with SD, freedom comes to have two slightly different meanings and different relationships to PAP. Sometimes freedom of choice/SD means the type of freedom/SD possessed by the angels in their initial state (one that involves PAPsin), and sometimes freedom of choice/SD means ADF and thus precludes PAPsin. But since both meanings are grouped under the concept of "freedom of choice," the change of meaning easily goes unnoticed. One can then mistakenly presume that ADF and its preclusion of PAPsin is a definition of freedom that covers all cases and thereby gloss over the problems pointed out in this paper. [End Page 30]

Obscuring The Fact That God Had It "Easier"

We hereby come to my final point. The conflation of freedom with SD and the resultant equivocal nature of freedom of choice allows Anselm to obscure the fact focused on in the first part of the paper, namely that God had it "easier" with regard to SD for justice. Allow me to explain. As we have just seen, it is natural to conflate SD with freedom of choice. In fact, this tendency to conflate the two notions is so strong that at one point Anselm even predicates aseity of God's own freedom of choice itself: "One type of freedom of choice is a se."48 Tomas Ekenberg even claims Anselm "clearly and explicitly makes free choice, liberum arbitrium, a condition for moral responsibility."49 As we have seen, he does not; Anselm makes SD the condition for moral responsibility. But Ekenberg's misattribution is understandable given the way SD and freedom of choice are so often conflated.

Once this conflation is established, the Anselmian can become imperceptive of the fact that (as I have argued) God's excellence may be somewhat diminished compared to that of the good angels. How? Even though the good angels had it "tougher" with regard to SD for justice, the Anselmian can remind herself that SD is (due to the conflation) the same as free choice (call this "free choice1"). It is then a small step for her to say that free choice (call this "free choice2") is ADF. And according to ADF, the good angels' freedom and God's freedom are the same. Thus, it appears that both God and the good angels acted out of the same type of freedom of choice when determining themselves. Since they supposedly acted through the same type of freedom, there is no rationale upon which to claim the good-angels' SD for justice is somewhat superior that of God's. According to this line of thought, both of the SDs arose from ADF which is the same for both. But it is essential to note what has happened here. "Free choice1" is different from "free choice2." The former is related to the understanding of SD that includes PAPsin whereas the second is related to ADF, an understanding which precludes PAPsin. As a result of this conflation, the Anselmian thinks SD and ADF are identical concepts. She thus fails to recognize the equivocal nature of SD and the fact that this equivocal nature of SD threatens an aspect of God's preeminence over that of the good angels. [End Page 31]

In summary, I have argued that SD is not a univocal notion for Anselm. A proper understanding of Anselm's concept of SD would lead to the conclusion that the angels' SD for justice is somewhat superior to that of God's. I have then argued that this conclusion is somewhat obscured by the fact that Anselm's concept of freedom of choice is often conflated with SD and itself becomes equivocally understood. I now conclude by conceding that Anselm himself, while insisting that there is a "common definition," is nonetheless willing to divide "free choice" into different types.50 But simply admitting that there are different understandings of free choice does not undercut my argument in this paper. If anything, this admission makes it all-the-more puzzling that Anselm did not recognize the threat of equivocation with regard to these concepts in the first place. [End Page 32]

Michael Barnwell
Niagara University

Footnotes

1. "Libertatem arbitrii non puto esse potentiam peccandi et non peccandi" (De Libertate Arbitrii (DLA), ch. 1). All Latin quotations are from the standard S. Anselmi Cantuariensis Archiepiscopi Opera Omnia, Ad fidem codicum recensuit Franciscus Selesius Schmitt (Stuttgart–Bad Cannstatt: Friedrich Frommann Verlag, 1968). Unless otherwise noted, translations are mine.

2. illa libertas arbitrii est potestas servandi rectitudinem voluntatis propter ipsam rectitudinem (DLA, ch. 3).

3. DLA, ch. 1.

4. The specific options the angels had to choose between and what constituted the sinful choice are described below.

5. De Veritate, ch. 12.

6. In DCD, these dispositions are called "wills." He thus refers to a will-forhappiness (voluntas beatitudinis) and a will-for-uprightness (voluntas rectitudinis) in DCD chs. 12–14. When he later writes De Concordia Praescientiae et Praedestinationis et Gratiae Dei Cum Libero Arbitrio, (DC), he refers to these two "wills" affectiones, or dispositions (DC III, ch. 11). It is clear that Anselm intends the "wills" of DCD chs. 12–14 to be these affectiones.

7. DCD ch. 13.

8. Ita hic si vellet convenientia non idcirco esset iusta voluntas, quoniam sic hoc accepisset ut non posset aliter velle (DCD ch. 14).

9. Anselm does not specify what that thing may have been. For a discussion of what the devil might have been willing, see Giorgio Pini, "What Lucifer Wanted: Anselm, Aquinas, and Scotus on the Object of the First Evil Choice," Oxford Studies in Medieval Philosophy 1 (2013): 61–82.

10. There is some dispute in the literature over how to understand willing in accordance with one affectio or another. See, for example, pp. 66ff. of Katherin A. Rogers, Anselm on Freedom (Oxford & New York: Oxford University Press, 2008) and Thomas Williams, "Review of Katherin Rogers, Anselm on Freedom," Notre Dame Philosophical Reviews 2009, no. 2 (2009).

11. I say "attempt" to will it because, as it turns out, willing their own happiness did not make them happy. Instead, they were punished for disobeying God. It was, however, required that the angels did not know they would be punished. If they had known, their affectiones-for-happiness would have also inclined them to obey God just as their affectiones-for-uprightness did. As a result, the consequent willing justly by obeying God would have been necessary instead of for its own sake. It was thus necessary that the angels suffer some, in the words of Marilyn McCord Adams, "necessary ignorance" (see Marilyn McCord Adams, "St. Anselm on Evil: De Casu Diaboli," Documenti E Studi Sulla Tradizione Filosofica Medievale 3 (1992): 439–49). For Anselm's discussion, see DCD, ch. 23.

12. DCD, chs. 22–24.

13. For my arguments that this account of the devil's fall is incoherent, see the following: Michael Barnwell, "The 'Harder Problem' of the Devil's Fall Is Still a Problem: A Reply to Wood," Religious Studies 53, no. 4 (2017): 521–43; Michael Barnwell, The Problem of Negligent Omissions: Medieval Action Theories to the Rescue (Brill, 2010), ch. 3; Michael Barnwell, "Why Can't the Devil Get a Second Chance? A Hidden Contradiction in Anselm's Account of the Devil's Fall," The Saint Anselm Journal (Fall, 2017).

14. Sandra Visser and Thomas Williams, "Anselm's Account of Freedom," in The Cambridge Companion to Anselm (Cambridge: Cambridge University Press, 2004), 187. Similar points are made in Rogers, Anselm on Freedom, 76–78; Katherin A. Rogers, Freedom and Self Creation: Anselmian Libertarianism (Oxford: Oxford University Press, 2015), 154–57.

15. Visser and Williams, "Anselm's Account of Freedom," 187.

16. My identification of this passage's relevance and the translation is due to Visser and Williams, 193.

17. Visser and Williams, 193–94.

18. Please see my comments at the beginning of this paper regarding my use of the term "equivocal."

19. DCD, ch. 18.

20. See, for example, Monologion, ch. 1.

21. John Hick, "An Irenaean Theodicy," in Encountering Evil, ed. S.T. Davis (Atlanta: Knox Press, 1981), 39–68. The details of Hick's soul-making theodicy need not concern us here. The interested reader can find Hick's classic expression of this theodicy in John Hick, Evil and the God of Love, 2nd ed. (Basingstoke: Pal-grave MacMillan, 2010). I provide a succinct summary in Michael Barnwell, "Soul-Making Theodicy and Compatibilism: New Problems and a New Interpretation," International Journal for Philosophy of Religion 82, no. 1 (August 1, 2017): 29–46.

22. Hick, "An Irenaean Theodicy," 44, italics mine.

23. In the block quotation, Hick is primarily focused on contrasting the case of a creature given virtues against that in which a creature is not given virtues; he is not contrasting either case with that of a God who has these virtues and His being completely a se. It is thus admittedly a bit difficult to ascertain Hick's own view on the matter presently under discussion. To my ear, the tenor of his discussion seems focused on the fact that the value of hard won virtues is derived from the fact that the agent had to "fight" for them. As I have already pointed out above, however, Hick's own view is ultimately irrelevant to the argument being made in the paper. It is being used to illuminate my claim that a just will gained as a result of facing alternative possibilities is in some sense different from (and better than) a just will that is determined to be so due to the completely a se nature of the Being possessing it.

24. Of course, the use of the term "came" is misleading, since God would be eternal. Despite this inaptness, the basic point I am making should be clear. Also, note that in contrast to the case of God, there is a sense in which the good angels "gave" justice to themselves (DCD, ch. 18).

25. Wes Morriston, "Is God Free? Reply to Wierenga," Faith and Philosophy 23, no. 1 (2006): 93–98.

26. I am purposely avoiding using the term "necessary" to refer to the way God's moral uprightness must follow from God's given nature. The reason I am doing so is to avoid the objection that since God's actions flow from God's own nature, they cannot properly be called "necessary" (cf. Cur Deus Homo II, ch. 5). What label we give to God's action is somewhat beside the point. The main point is that God's willing uprightly is determined by His nature, which was just "given." His moral uprightness was inevitable.

27. Morriston, "Is God Free? Reply to Wierenga," 96.

28. Morriston does concede that God's simplicity, if true, might constitute a relevant difference between the case of God and Bonnie. See Morriston, 98. And Anselm, of course, does believe God is simple (cf. Monologion, ch. 17). The defender of Anselm, therefore, might want to resist the appeal to Bonnie. But even if God is simple, I think the Bonnie Chance example is instructive in bringing out my main point. Given that God is simple and identical to God's moral nature, it still remains the fact that God is "stuck" with it; God could not change His nature. In fact, simplicity would even entail that the Moral Nature (i.e. God) "just is." And since the Moral Nature "just is," then God (which is identical to Moral Nature) would 'automatically' be just. And this moral nature that is "automatic" would still, in some sense, be "easier" and qualitatively different from a moral nature in an angel that is not "automatic" and is not "just is."

29. Rogers, Anselm on Freedom, 85.

30. Theodore Guleserian, "Divine Freedom and the Problem of Evil," Faith and Philosophy 17, no. 3 (2000): 350. Guleserian defines free PW as "the property that x has just in case, for every situation S and act A, whenever x believes that the moral law requires that anyone who is both in S and has the power to do A in S ought to do A, x freely wills to do A." Important in understanding the distinction between free and essential PW is the fact that x having "free PW entails that x has PW as a contingent (i.e., nonessential) property." It is because of this consideration that Guleserian reasonably assumes the traditional theist would claim God has essential PW, defined as "the property of having PW essentially." Moreover, God cannot have PW both freely and essentially. This is "because the freedom of one who has free PW requires that this person freely wills to do a wrong act in some possible world" (Guleserian, 350).

31. Guleserian, "Divine Freedom and the Problem of Evil," 354.

32. Guleserian, 360.

33. Guleserian, 356.

34. Guleserian, 357.

35. If SD does have a slightly different meaning in different cases (as I have been arguing), using it in a definition would not be legitimate. But Anselm presumably did not regard SD as equivocal while discussing it in his works on the will.

36. The close reader of Anselm will suggest that this definition derives from his discussion of rectitudo in De Veritate (esp. ch. 12), a work meant to be read in conjunction with DLA and DCD. This may explain why Anselm offered ADF but does not philosophically justify why ADF is to be preferred over a definition in terms of self-determination.

37. Stan R. Tyvoll, "Anselm's Definition of Free Will: A Hierarchical Interpretation," American Catholic Philosophical Quarterly 80, no. 2 (2006): 170.

38. Interestingly, Thomas Williams has recently argued that Anselm is a quiet radical with regard to his understanding of the ontological origin of the nature of choices. See Thomas Williams, "Anselm's Quiet Radicalism," British Journal for the History of Philosophy 24, no. 1 (2016): 1–20.

39. Tyvoll, "Anselm's Definition of Free Will: A Hierarchical Interpretation," 170.

40. Visser and Williams, "Anselm's Account of Freedom," 187, italics mine. While these quotations are very similar, they are indeed two separate quotations from the same page. Visser and Williams thus tie aseity to free action twice within a few paragraphs.

41. Visser and Williams, 193–94, italics mine.

42. Rogers, Anselm on Freedom, 77, italics mine.

43. Rogers, Freedom and Self Creation, 155, italics mine.

44. Visser and Williams make this point. See Visser and Williams, "Anselm's Account of Freedom," 183.

45. Peccavit autem per arbitrium suum quod erat liberum; sed non per hoc unde liberum erat, id est per potestatem qua poterat non peccare et peccato non servire, sed per potestatem quam habebat peccandi, qua nec ad non peccandi libertatem iuvabatur nec ad peccandi servitutem cogebatur (DLA, ch. 2). The English translation is that made by Visser and Williams. See Visser and Williams, 183.

46. Visser and Williams, 184.

47. Note moreover that this apparent equivocation cannot be solved by asserting there are two different "freedoms" here. Anselm has in mind the same freedom. Indeed, Visser and Williams even claim that the sin occurs "by an exercise of that power [i.e., liberum arbitrium]" (Visser and Williams, 184, italics mine). There is no indication that there are two different faculties or powers involved.

48. Libertas arbitrii alia est a se" (DLA, ch. 14).

49. Tomas Ekenberg, "Voluntary Action and Rational Sin in Anselm of Canterbury," British Journal for the History of Philosophy 24, no. 2 (2016): 215–30, p. 216. Interestingly, Ekenberg goes on to claim that Anselm denies that voluntariness is required for moral responsibility.

50. DLA, ch. 14.

Share