Frankfurt and Beyond: Hierarchical Readings of Anselm's Analysis of Moral Choice with a New Test Case for his Concept of Freedom

The paper addresses criticism of Frankfurtian interpretations of Anselm's account of moral choice and defends a modified hierarchical model of the will which introduces, based on Anselm's theo-anthropology and Augustinian "beatitudinism," a third-order desire for happiness with God. In light of this, the issue of evil appears to result from a cognitive deficiency and not only from failure to order one's desires properly. The paper finally proposes, against the backdrop of universalism, a different test case for Anselm's concept of free will (i.e., not Satan's fall but the freedom humans may have on judgment day) and asks whether this choice might allow creatures to exercise freedom without the need for alternate possibilities. The paper thus aims to disentangle moral and theological-eschatological explications of Anselm's definition of freedom.

Introduction

Scholars like S. Tyvoll, K. Rogers and others have suggested "Frankfurtian" readings of Anselm's account of moral choice. They argue that Anselm—similar to 20th-century philosopher H. Frankfurt—thinks of a hierarchical relationship between the desires rational creatures are endowed with (i.e., the desires for benefit and justice), rather than seeing them as conflicting same-order desires. Rogers, in particular, has received a lot of criticism for her reading of Anselm.¹

Some critics (e.g. Williams, Visser) take issue with Rogers' "anachronistic" method of connecting "the work of a long-dead philosopher to philosophy in the present," something she openly admits (AoF, 1). Such reservations may be justified from a purely philological and historical perspective, but philosophy in its purest form is impossible without showing the relevance of ancient thought. Moreover, while textual interpretation should always strive for accuracy, it can still have philosophical value so long as it is guided by the eros to explore perennial questions, opens up new lines of thought, and produces philosophical insight, even if it at times strays from the author's original intent. As we shall see, this is not the case with Rogers, who provides sufficient textual evidence to support many of her claims.

Some critics focus on the debate between libertarianism and compatibilism and question the suitability of Satan's choice as an example that could be transferred to free choice in general (e.g., Ekenberg 2016a). But Anselm emphasizes the intrinsic connection of his three dialogues De veritate (DV), De libertate arbitrii (DLA), and De casu diaboli (DCD) and requests that they be "written together as a unit." The primary (negative) "test case" for his understanding of freedom, which he explains in DLA, is Satan's fall.² Rogers acknowledges that the focus on Satan may seem odd but highlights three analytic reasons Anselm has in presenting him as an "idealized example" (FaS, 217): Anselm is "interested only in morally significant choice and [is] deeply concerned to get to the bare metaphysics of free will. He . . . wants to examine a pure instance of choice, and he wants to put the central and most difficult puzzle in the starkest of terms: how could a being made perfectly good, with . . . no evil in the world to tempt him, possibly choose against the will of God" (AoF, 7).

Nonetheless, there remain certain difficulties in Rogers' account, and I shall address some of them in this paper. My main goal, however, is to clarify and defend a hierarchical reading of Anselm. I will do so by modifying Rogers' model of moral choice and suggest that we introduce a third-order desire, namely the desire for (true) Happiness (part 1). This third-order desire takes into account Anselm's theoanthropology and the fact that his view of the human being is influenced by what I call "Augustinian beatitudinism." In this context, I will have to contend with Rogers' claim that Anselm should be seen as an "Aristotelian" rather than a "Kantian" thinker: "Anselm has almost always been taken to espouse a sort of proto-Kantianism: the virtuous person is the one who chooses duty over the opposing options of self-interest and natural inclination" (AoF, 10). This claim is meant to support the hierarchical interpretation, but it seems to me that it is rooted in a misreading of Kant.

Following this, I turn to the question of evil and the failure to order one's desires toward benefit, justice, and Happiness properly (which, according to Anselm, is the cause of evildoing) and offer some critical reflections on Anselm's and Rogers' account of moral choice. I will suggest that evil is at bottom a cognitive issue. This then prompts me to apply the definition of freedom Anselm develops in DLA 3 to a different test case than Satan's fall, namely, to the freedom humans supposedly have on "judgment day" (part 2). I thus move beyond the scope of Anselm's own theology and use his concept of freedom within the context of universalism. I hope that this move may at the same time illustrate the theo-logic of this doctrine and emphasize the lasting relevance of Anselm's innovative reflections on freedom.

My discussion revolves around the cluster of issues having to do with the relationship between freedom and free choice, libertarianism and compatibilism, and moral and eschatological choice. It cannot be emphasized enough that all of Anselm's moral analyses are developed against the backdrop of his theology and eschatology. For him, moral choice is always about salvation and rarely about moral responsibility alone. The theological context, however, gives rise to certain problems. Anselm's definition of freedom is not primarily a moral but a theological concept which runs the risk of being reduced again to the moral choices finite creatures are expected to make against the backdrop of the idea of merit-based salvation. My proposal attempts to regain the high ground of his original definition and focus on a moment in which the kind of freedom Anselm has in mind may truly inform the free choice of creatures as well.

Part 1: N-Order Volitions and Theo-Anthropology

1.1 Some Preliminary Notes

It is generally agreed that Anselm "inherits two notions of freedom," of libertas and of liberum arbitrium, and combines them into one so that his single definition can stand for "such diverse realities as the freedom of self-perfection and the freedom of self-determination" (Kane, 305³), or freedom as a kind of virtue (self-perfection brings human excellence) and free choice. The "single definition" Kane refers to is Anselm's famous potestas servandem rectitudinem voluntatis propter ipsam rectitudinem (DLA 3), "the ability to preserve rectitude of the will for the sake of rectitude itself" (hereafter PSR²). Still, we need to explore with greater depth and precision what is being compressed into this single definition, that is to say, certain notions of freedom and free choice.

Although influenced by Augustine, Anselm rejects Augustine's concept of free will, i.e. "the ability to sin or not to sin." Anselm's reasons are theological: God and the good angels do not have that kind of freedom (DLA 1). Anselm therefore claims that the definition of freedom as such does not entail a "principle of alternate possibilities" (hereafter PAP). He does, however, say (most notably in DCD) that created rational natures need to have "open options" or real "a se choices" (Rogers) in order to prove and exercise their PSR². Rogers deserves credit then for showing that Anselm is not inconsistent (it seems, after all, that he is taking back his "distinctive definition" when it comes to created natures⁴) but presenting a proper application and explication of his definition and its entailments—one, moreover, in line with the more general idea that freedom is tested through free choice. Rogers spells this out repeatedly, but the idea is older. Kane, for instance, had stated that, "the form that freedom takes under the conditions of human existence is the ability either to sin or not to sin," so that "self-determination in a creature . . . is only possible if there are alternatives from which he can choose" (301f.).

It is noteworthy that Anselm doesn't simply consider alternative objects of one's desires or possibly opposing actions directed at those objects. Instead, he focuses on the subject (i.e., the agent's will and its motivating factors). This approach does indeed "prefigure" Frankfurt's analysis of the will but it is really in line with any motivational account of human acting. Anselm sees rational creatures endowed with two different desires which can, but do not have to, come into conflict: the desire for benefit (hereafter DB) and the desire for justice (hereafter DJ). Rogers holds that potential conflicts don't arise on the same order of desires (as in what she calls the "Kantian" reading of Anselm). They are not opposing desires but desires that, on principal at least, are complementary, belonging, that is, to different orders of the will. She thus suggests a "hierarchical" or "Frankfurtian model" to account for Anselm's view of moral choice and the will.

Anselm does not, of course, use Frankfurt's terminology, but his language clearly entails a hierarchical understanding of the will. Note, though, that the PSR²-definition itself also uses hierarchical language (if x is done for the sake of y, this implies that y is higher or more valuable than x), but this does not involve a hierarchical understanding of the relationship between DB and DJ; it simply indicates a hierarchy of goods and ends, a notion that can also be found in Aristotle (Nicomachean Ethics [NE] I 1.2). There is, however, plenty of other textual evidence for Rogers' interpretation, which she has presented by "unpacking this definition" (Rogers 2005, 251) and emphasizing its connection with the test case of free choice Anselm presents in DCD.⁵

My analysis will focus on passages from DCD (and, to a lesser extent, DC) but it will also draw on Anselm's discussion of freedom in DLA (mainly chapters 3 and 8). Note that the proposal for a hierarchical interpretation does not rely on a phrase from DLA 5 ("for everyone who wills wills his own willing") which has previously been used in this context, e.g. by Saarinen who calls it "Anselm's principle" (46). Goebel 2006 convincingly shows that this is merely an "emphatic, more elegant way to say that anyone who wills something just wills what he wills, and nothing else" (102). An actual choice is always a willful choice and only "improperly said to be done against the will" (DLA 5). But Goebel goes so far as to argue that Anselm thinks it is in general meaningless to postulate a second-order will (100), and he may have incorporated his criticism of Saarinen's reading of DLA 5 into his critique of FaS. It seems to me that this full-fledged rejection isn't entirely warranted; for Rogers' argument is mainly built around DCD and the two desires whose relationship Anselm discusses in that work. It is not merely about "a desire to will" (Goebel 2017, 127) and goes beyond the bare distinction between "Wunsch" and "Wille."⁶

What I have called the desire for benefit above (following Rogers, Tyvoll and others) is really Anselm's first notion of "happiness." Unfortunately, Anselm's terminology is not always consistent, a fact which may cause some irritation. In particular, it is not immediately clear that "benefit" and "happiness" are coextensive. Tyvoll, for instance, speaks of desires that are "beneficial, useful, or conducive to happiness," using these attributes interchangeably (157).⁷ It is, in fact, Anselm himself who uses benefit (commodus/commoditas) and happiness (beatitudo) interchangeably. He explains in DCD 12: "By happiness I don't mean happiness in accordance with justice but the happiness that everyone wills, even the unjust. After all, everyone wills his well-being," and then goes on to distinguish between "two goods and two evils that are contrary to them," one is justice (i.e. moral goodness), the contrary being evil injustice, and the other one "can be called the advantageous" or "beneficial" (literally what "brings commoditates"), the opposite being "the disadvantageous" (cf. DC 3,11−12). However, even the "happiness with justice" Anselm mentions in DCD 12 does not, in itself, denote the theo-anthropological notion of happiness as final human fulfillment which I will focus on later in this paper. Rather, it simply means "justly sought benefit." So the English word "desire for benefit" seems to be an appropriate translation for what Anselm describes as the will that seeks "both" commodum and beatitudinem, and I will reserve, as much as possible, the word "happiness" for the other notion of beatitudo that, I believe, we find in Anselm's writings as well.

1.2 Anselm and Hierarchical Accounts of the Will: An Overview

1. Although some critics hold that a Frankfurtian reading of Anselm's analysis of the will overcomplicates things, Rogers' view that, on the contrary, Anselm succeeds in presenting an acute analysis without the need for a "very complex process of desire formation" remains convincing (FaS, 261). It nevertheless seems necessary to point out and clarify certain issues and suggest some modifications to Rogers' account and, if her interpretation is correct, Anselm's account of moral choice as well. I shall first look at Rogers' account and compare it with Tyvoll's proposal on how to read Anselm's two desires in DCD and DC.

Actions spring from the will but are, as such, no part of it and should therefore not be included in a model that describes the hierarchical structure of the will. The will itself first contains first-order desires which are "desires to perform an action" (Tyvoll, 156) or "desires to do or not to do one thing or another" (Frankfurt 1971, 83).⁸ A person can have "numerous first-order desires at the same time" while she can only have one first-order volition at any given time, for this is defined as an "effective first-order desire, one that moves a person all the way to action" (Tyvoll, 156; Frankfurt 1971, 8). The introduction of second-order desires takes into account that we develop attitudes towards our desires; they are desires for a desire. But there is a difference between someone who "wants simply to have a certain desire or when he wants a certain desire to be his will" (Frankfurt 1971, 86). Note that contrary to Tyvoll's use of the term, Frankfurt reserves "volition" for the second order of desires (i.e. for those second-order desires that are effective and make a first-order desire effective or not). Effective first-order desires are simply called a person's "will" in the sense of providing "the motive in what [someone] actually does" (85) or tries to do.

In Anselm's view, the will yields volitions "through" inclinations (Tyvoll, 158). "Will" (voluntas) is here seen as instrument of the will. For Anselm says, we use the word in three different ways: as "faculty by which we will something (instrumentum volendi), inclination of that faculty, or desire, (affectio instrumenti) which moves it to will something . . . and the actual use of the will" (Rogers 2005, 252; cf. DC 3,11–13, DLA 7). In DC 3,11f., Anselm speaks of aptitudines which, he says, can also be called affectiones. These are the voluntas rectitudinis/iustitiae, or affectio ad [volendum] rectitudinem and the voluntas ad volendum commodum, or affectio commodi. They are inclinations or "occurrent desires" in the sense of "intention, volition or action-guiding plan" (FaS, 78). Tyvoll argues that they are "neither first-order desires nor first-order volitions" for "they are clearly not desires for specific actions" (158). He concludes that they are "desires of the second order" but with "different directions" (159). Rogers, however, sees a hierarchical relationship between these two so that one qualifies the other.

It is in the very analysis of this relationship that Rogers finds evidence of Anselm's hierarchical account of the will. Moral and immoral choices may be occasioned by objects (which may include acts, since the desire for x is the desire to do or get x, or do those acts that will lead to x) so that free choice is, on the surface, an object choice but it really is a value-judgment, which consists in the proper (or improper) ordering of desires, which incurs praise or blame and thus grounds moral responsibility.⁹ Anselm sees justice as superior to benefit and so DB needs to be "aligned" with DJ (AoF, 66ff.).

2. Generally speaking, it is not at all clear that this is the only conceivable relationship between benefit and justice since benefit is a relational concept requiring a genitive object ("whose benefit?") and one could think of different objects. For instance, justice itself could be the object, so that a choice is good when it is done "for the benefit of justice." Anselm, however, is clearly thinking of the agent's benefit, so that seeking one's benefit constitutes an integrative principle of all acting.

This view perhaps "prefigures" the motivational analyses of modern psychologists who focus on human self-interest (from Nietzsche to Maslow to advocates of "selfish genes"), but it really is nothing else than the old idea that human acting is also motivated by the "principle of pleasure and pain"; for all creatures seek advantage and try to avoid disadvantage. This is often called "hedonism," but the philosophical term rarely has anything to do with the vulgar hedonism it has come to be associated with in the popular understanding. Even the so-called hedonistic schools of philosophy which put pleasure (or happiness) at the center of their teachings (e.g. Aristippus, Epicurus) do not intend by this any form of egotistical selfishness or bodily pleasures but rather see it as an integrative principle springing from the experience of true self-knowledge (cf. Göbel 2002, 238–292). And a modern hedonist like J.S. Mill has to defend, in Chapter 2 of Utilitarianism, the principles of utility and happiness precisely by pointing out their integrative character. Mill emphasizes, for instance, that utility and happiness are identical, that pleasure is happiness, that it starts with the mere absence of pain, and that hedonism allows for different kinds and qualities of pleasures. Even more traditional approaches to anthropology and ethics turn out to be more integrative than one may first imagine, and philosophers like Aristotle,¹⁰ Augustine or Kant (to name just a few) consider "pleasure-seeking" as natural and valuable (although they dismiss some of its forms as unfitting and self-degrading for humans). Anselm shares the view that DB is "inseparable" from the human being (DC 3,12), for "for no one wills anything unless he thinks it is in some way beneficial for himself" (DCD 12). He even says that "by [DB] the will moves itself to other desires," since "no one . . . wills something, unless he first (prius) has a natural desire to avoid what is not beneficial or have what is beneficial" (ibid.).

3. This last quote may be seen as further supporting Tyvoll's reading which distinguishes between desires for certain actions (first order) and a second-order DB. Anselm's use of prius seems to suggest a relationship (a priority-relationship)—not, as Rogers holds, identity—between desires for particular actions and DB. If Tyvoll's view is sufficiently convincing and if at the same time Rogers is right in emphasizing the hierarchical relationship between DB and DJ, we might be forced into introducing a third order of desires at this point. This would call for a more complex structure of moral choices, since desires for certain actions (first order) would need to be aligned (in some morally relevant situations) with both DB (second order; this alignment would guarantee freedom in Frankfurt's sense) and DJ (third order; this alignment would guarantee justice and freedom in Anselm's sense).

However, Anselm's reference to the "prior" DB in DCD 12 is clearly different from Frankfurt's description of second-order desires: it is not a desire to have a certain desire¹¹; it is not "formed" (Frankfurt 1971, 83) in response to a first-order desire or a conflict of desires on the first order. Rather, it "precedes" any and all desires, and no "aligning" is needed with this higher-order desire for benefit. On the contrary, Anselm clearly holds that all desires are desires for benefit. Rogers' model therefore suggests that desires for specific actions, since they are all desires (plural) for benefit, can be subsumed under the concept of a "desire (singular) for benefit" (more precisely, they are all motivated by the underlying DB). At least for now, this allows us to preserve a two-order structure of the will and its desires. It is precisely the fact that created natures seek benefit in all their actions but justice only through some (only some desires for certain actions also fulfill DJ) that prompts Rogers' hierarchical understanding of the relationship between the two desires which in my view is the most convincing reading of the above quotes from DCD 12 and DC 3.

4. If DB is the integrative principle of all acting, justice could also, one might say, become subordinate to benefit. In other words, one may seek or do justice for one's personal benefit alone (e.g., to gain social status, one's peers' esteem or some other reward, or to avoid punishment by the law or from God in the afterlife). However, Anselm's view of the relationship between the two desires is quite clear: benefit is subordinate to justice. Anselm is concerned with just benefits or "justly sought benefits" and condemns any good deed done for wrong reasons as immoral (an analysis, by the way, that is fully in line with both Aristotle and Kant, insofar as they are both "virtue ethicists"). Goebel points out that "willing something and willing it for a certain reason [because one ought to will it] is all there is" to the supposed hierarchy of desires (2017, 127); Anselm's reference to DJ thus serves as nothing more but an adverbial qualifier. This claim resonates with Anselm's words in DCD 14: the creature is "not able to be happy unless he wills to be and wills it justly." But Anselm uses hierarchical language in prominent passages of his account of moral choice precisely because it allows him to capture more clearly (1) the value judgment underlying the adverbial qualification of the way in which one acts on one's DB and thus (2) the relationship between the two desires. DJ "limits excesses" of DB (DCD 14) and expresses a fundamental orientation of the human being towards justice.

Some of the skepticism towards hierarchical readings of Anselm stems from the fact that there are important differences between Anselm and Frankfurt, especially regarding value judgments and moral choices properly speaking.¹² But Rogers never claims that their views are in every respect identical. Rather, she has merely observed that Anselm, too, seems to distinguish between different orders of desires and that this distinction is by no means trivial when it comes to analyzing the operations of the will and furthermore is consistent with the way some contemporary philosophers speak on the matter.

5. A few words on the issue of moral judgment in Frankfurt and Anselm. Needless to say, Frankfurt incorporates judgment into his analysis of those desires for specific actions that Rogers subsumes under the concept of DB. The evaluative dimension is in fact the characteristic feature of Frankfurt's definition of freedom and personhood and prompts his hierarchical analysis of the will; "reflective self-evaluation" on a higher level is key for him (Frankfurt 1971, 83). This requires some degree of self-distance: "one must be able to step back from one's immediate desires, assess them, and identify with some over others" (89; cf. Rogers 2005, 250).

Now, Frankfurt and Tyvoll mainly discuss judgment and evaluation at a level "below" justice and morality properly speaking. They look at desires for specific actions and DB. I earlier said that there doesn't seem to be any "aligning" involved at this level, but this claim needs to be further qualified. Tyvoll's application of Frankfurt's theory to Anselm seems misguided in its focus on the relationship between desires for specific actions and DB (for all desires are desires for benefit); yet, when conflicts between desires for specific actions arise, it is the reference to or, rather, formation of a second-order desire (wanting to have a particular first-order desire) which decides the conflict, if, that is, the second-order desire is to be effective.

A decision may be made on account of the fact that desires are judged to be more or less beneficial (i.e. promise more benefit when turned into volitions and then actions) or based on judgments about false and proper benefit. Frankfurt and Tyvoll offer examples of addicts, which suggest that agents can, or cannot, choose long-term health over a momentary pleasure and short-term satisfaction (the desire to smoke may or may not be rendered ineffective by the second-order desire that one would rather desire not to smoke). Or it may simply be the case that a person prefers (i.e. desires more), in a given situation, benefit b over benefit a (e.g., "writing over reading": Tyvoll 156; cf. DC 3,11). Frankfurt's point is that "a person has freedom of will (i.e. their effective desires are free) in so far as they can control their desires" (Holton, 1; cf. Frankfurt 1971, 90) since it is the "peculiar characteristic of humans . . . to form . . . second-order desires" (Frankfurt 1971, 82). More precisely, Frankfurt sees it as "essential to being a person" (and not just a "wanton" who doesn't "care about his will") to have second-order volitions, rather than just "having second-order desires generally" (86), for the important feature of reflective self-evaluation, personhood and freedom is to be "concerned with the desirability of desires themselves" (87). It is, to be clear, a matter of free will, not just of free action. It is not significant "whether he is in a position to translate his first-order desires into action. That is the question of whether he is free to do as he pleases. . . Freedom of the will means that a person is free to want what he wants to want . . . or to have the will he wants. . . . It is in securing the conformity of his will to his second-order volitions . . . that a person exercises freedom of the will" (90). Frankfurt 1969 thus defines freedom simply as the harmony between first- and second-order desires and holds that this does not necessarily require a PAP. This looks very much like another surprising similarity between Frankfurt and Anselm, since Anselm rejects the Augustinian notion that a PAP (the ability to sin or not to sin) is essential to his definition of freedom (PSR²). Rogers, however, rejects this aspect of Frankfurt's account based on her "unpacking" of Anselm's definition and application to the case of creatures which, she insists, need open options as essential ingredients of true, libertarian freedom.

We will come back to this claim. It may suffice at this point to remind critics that the language of higher orders can simply be read as a new way to highlight the old philosophical idea that humans are self-reflective beings; their capacity to reason gives them their very dignity. Frankfurt and Anselm, as well as most other thinkers in the classical tradition of philosophical anthropology (including Aristotle, Augustine, and Kant) "suggest that we have a great deal of control over which desire we will entertain and allow to influence us" (FaS, 89).¹³

1.3 Anselm's Theo-Teleological Value System and the Structure of the Will

1. Above, I briefly considered a hierarchical model that combines Tyvoll's and Rogers' readings of Anselm and that would require us to introduce a third order of desires, so the question may arise if there could be any further higher-order desires, indeed, any indeterminate number of a potentially infinite series of orders (hence my reference to "n-order volitions" in the title of part 1). Although we have so far been able to preserve a dual structure, I will argue that there is another order of desires that we need to consider in Anselm. Nevertheless, there is a clear answer to the more general question at which point the series of orders will come to an end. Frankfurt himself has seen the problem of an infinite regress, and some critics argue that he merely suggests a pragmatic, arbitrary and unconvincing reason for stopping at the second order (cf. Holton, 2; Watson, 107; Goebel 2006, 100). Even Frankfurt himself initially says that "there is no theoretical limit to the length of the series of desires of higher and higher orders; nothing except common sense and, perhaps, a saving fatigue" prevents a person from stopping at a certain point (1971, 91). But is this really true in the end?

According to Frankfurt, "there are certain desires that we identify with and . . . free will consists in getting our desires in line with these" (Holton, 2), and "when a person identifies himself decisively with one of his first-order desires, this commitment 'resounds' throughout the potentially endless array of higher orders" (Frankfurt 1971, 91). It is, therefore, not at all arbitrary to stop at the second order if it is the order of those desires that we identify with, or if a solved conflict at the second order allows us to identify with a specific first-order desire. Note that second-order desires may also be in conflict, so what really matters is how they are turned into volitions, and thus, in turn, see to it that the corresponding first-order desire becomes effectively one's will. The choice which first-order desire a person wants to identify with is made at a higher level (second order). And if no second-order desire becomes effective (unresolved second-order conflict), a person turns into a "paralyzed, helpless bystander to the forces that move him" (91).¹⁴ While a "wanton addict may have no identify apart from his first-order desires," even the "unwilling addict identifies himself . . . through the formation of a second-order volition, with one rather than with the other of his conflicting first-order desires" (88).¹⁵

2. It seems to me that the emphasis Frankfurt places on self-identification is, generally speaking, very much in line with Anselm's focus on freedom as a kind of virtuous "self-perfection" (Kane) and self-choice or "self-creation" (Rogers¹⁶), and that this aspect may be overlooked by critics of a Frankfurtian interpretation. It remains, however, true that Frankfurt's motivational analysis of the will cannot fully account for the reasons why a person "decisively identifies" with a certain desire. This requires both an additional faculty of judgment (i.e., a faculty which is external to the will as such and allows us to "choose what we judge most desirable": AoF, 37) and also standards by which the decision is made to prefer desire a over desire b through turning second-order desire a into second-order volition a (making the desire effective), so that a person chooses to do a rather than b. Such a faculty of judgment, which also has access to external or internal standards for its choices, might be human reason or even emotions and natural sentiments.¹⁷

While Frankfurt's analysis certainly implies evaluation and judgment, it doesn't seem to refer to any particular value framework which would provide ultimate standards for someone's choice to identify with a certain desire and judge it as more beneficial or better than others. This has led Frankfurt's critics to ask if there is "anything special about higher order desires? Might they be trivial, or flippant (Consider my desire that I have cool desires). Why should we think of freedom as consisting in their efficacy?" (Holton, 2).

In fact, the standard examples of smokers or addicts aren't really cases of moral choices properly speaking, or "morally significant freedom" in Anselm's sense (Rogers 2005, 250; Williams/Visser, 9); for they are not matters of justice, sin, and salvation (cf. DLA 6, DC 1,6 et al.). Anselm uses the very example of preserving one's health in DC 3,11 but solely as an example of benefit-seeking without justice. It is not immoral but morally irrelevant (amoral) in the proper sense of moral choices.¹⁸ Rogers notices that "a major difference between Frankfurt's analysis and Anselm's [lies in the fact that] Frankfurt explicitly states that the second-order desires, though necessary for personhood and free will, need not have any moral content" (2005, 259).

Rogers points out that Anselm avoids and solves the issue of a potential regress: "For Anselm, there is no threat of an infinite regress of orders of desire. God's will is the absolute standard of value and once one has chosen to select one's first-order desires against that standard, there is no higher value to appeal to" (ibid., 262). I would emphasize more clearly, though, that it is not Anselm's mere reference to God that puts an end to the order of desires but the fact that his value system has an absolute, external foundation (i.e., a natural order of things the agent knows, accepts, and even has a natural desire for).

3. The kind of criticism of Frankfurt that highlights the undistinctive character of second-order desires and, therefore, finds them superfluous or arbitrary cannot apply to Rogers' account of Anselm. Anselm, too, thinks of different orders of desires, but he clearly transcends the scope of Frankfurt's analysis which includes cases of reduced moral choice. Anselm is only interested in man's responsibility for full-fledged moral choices, and the introduction of a second-order desire (i.e., DJ) allows him to describe moral choice as hierarchical. A higher-order desire (or volition, since it is here assumed that it becomes effective) allows rational natures to make a moral judgement and apply their value system to the choice for a specific action and, in the case of a conflict, to the choice between desires for specific actions.

Goebel (2006, 101) invokes Leibniz' skepticism regarding a supposed "ability (or will) to will" and argues that a "reflective will" is meaningless. But the claim seems problematic if it implies that conflicts between desires are trivial. Doesn't Rogers capture a very common human experience when she describes a "torn condition" without which, she thinks, true freedom is impossible? Isn't she right in calling this a "decidedly conscious experience" (FaS, 91)? We are, in effect, torn between "open options" which result from our own, conflicting desires. Anselm clearly makes this point in DCD. Humans face the existential difficulty of dealing with the fact that they seem "condemned to freedom" (J.-P. Sartre) and hence to making responsible choices. Second-order desires allow us to be more than "bystanders to the forces that move us." We have already seen that, in Frankfurt's view, these second-order desires are still action-specific (i.e., they are desires for certain first-order desires which are desires for certain actions) and thus give reason and motivation for one's choice of acting. Furthermore, it is implied that this choice is made by certain standards (whether internal or external). It could therefore be argued that Rogers may simplify the role of second-order desires by subsuming, again, moral choices properly speaking (regarding conflicts between benefit and justice) under the (single) second-order "desire for justice" (this encompasses all conceivable second-order desires that first-order desires be effective for just reasons). Rogers nevertheless offers an accurate and illuminating account of moral choice and of Anselm's view of the two desires.

If we bear in mind this role of second-order desires, we could actually say that Anselm's claim that DB is prius to everything else (DCD 12) holds for all higher-order desires¹⁹: they are, as value standards which exist inside and outside the acting person, ontologically "prior" to other desires and choices, even though we may refer to them only "after" we experience a conflict of desires. The distinction between "priority in nature" and "priority in time" is typically Anselmian (cf. DV 2). So, if Anselm had used the language of numbered orders, he might have called desires for specific actions second-order desires and the higher-order desires by which we evaluate or motivate lower-order desires, or make them effective, first-order desires. Watson makes a similar observation in his critique of Frankfurt: "evaluations are prior and of the first order" (109). More precisely, second-order desires may be formed when a conflict arises, but a closer look at Anselm helps us understand that they are informed by something already present within the agent and, possibly, beyond his [not my] subjective value system.

4. Note, however, that Anselm's reference to God is not simply a theological claim that puts an end to any philosophical discussion; rather it is an integral part of his analysis of moral choice which, for Anselm, is always a theologically informed moral choice. This is most obvious in DLA 8 which takes us back to Anselm's understanding of freedom as PSR². This definition, of course, also serves as the foundation for Anselm's discussion of Satan's fall and for the doctrine of the two desires he introduces to address this test case in DCD. An analysis of DLA 8 provides even further support for the hierarchical reading of Anselm's structure of the will.

It is well known that Anselm's theory of truth and his concept of freedom are closely related. Anselm has a "double correspondence theory" of propositional truth (cf. DV 2). Statements are true when they correspond to reality or facts (this part is in line with the classic definition of truth as adaequatio intellectus et rei) but, more importantly, "when they correspond to the purpose of making statements" (Williams/Visser, 2). One determines the other, but the latter is essential; for truth and truthfulness are rooted in the teleological nature of things. And Anselm has a sort of double correspondence theory not only of truth but also of freedom because morality is rooted in ontology. Rectitudo (which is how truth is defined in DV 2) constitutes the crucial connection: everything, including human actions, is oriented towards the natural order of things. A being has rectitudo when it is as "it ought to be," "does what it received the power to do," and thus accomplishes its goal, "in keeping with the purpose for which it was made" (DV 2). While non-rational creatures like animals fulfill their purpose naturally and instinctively (cf. DV 12, DLA 2), humans (and angels) are meant to fulfill it through their particular nature (i.e., reason and free accordance with the purpose of things). Anselm's philosophy is teleological ("Aristotelian") in its emphasis on the being's orientation towards the natural order of things, which may take the form of ontological "correctness" or moral "uprightness" (both are common English translations of rectitudo²⁰). More precisely, Anselm's philosophy is theo-teleological since, as a theist and Christian believer in creation, he sees the natural order as God-given.

It is in the analysis of free will in DLA 8 (not in DLA 5, if Goebel's critique of Saarinen is correct) that Anselm first uses hierarchical language in a manner that reads almost like Frankfurt's "ability to want what one wants to want" (1971, 90). Recall that Anselm's "distinctive definition" of freedom does not in itself indicate an indeterminateness or "ambivalence with regard to good and evil" (Verweyen, 39) that is situated in the conflict zone of same-order or different-order desires but the capacity for justice or goodness, namely, the "capacity for preserving rectitude of the will for the sake of rectitude itself" (PSR²: DLA 3; cf. DCD 7), so that one wants to will what one ought to will ("for which one was given a will": DV 4) or wills "what God wills the will to will" (DLA 8; cf. DCD 4, DC 1,6). This language is almost identical with Frankfurt's, except that Frankfurt says "one" where Anselm uses the word "God." To be sure, Frankfurt singles out the human decision-making process and desires, just as does Anselm, but with Anselm we also find a reference to God's will. However, in DCD God's will gets translated into an innate desire rational creatures possess (i.e., DJ) thus shifting the consideration away from a purely theological one into that of a metaphysics of morals. Indeed, we can interpret the reference to God's will in DLA 8 as a reference to the higher-order desire that allows rational creatures to make truly moral value-judgments. Anselm refers to an evaluative dimension in the decision-making process and identifies its transcendent source. Since, for Anselm, this is God, freedom is understood in terms of self-perfection or self-fulfillment in accordance with the natural order of things, which is to say the order God intended for things.

Anselm is not merely suggesting a direct and simplified version of divine commandment theory. Rather, rational creatures have internal, natural access to, and take ownership of, divine law. Anselm's doctrine of the two desires which he believes God has implanted in rational minds demonstrates this. God's will is ingrained in human nature. But Anselm's analysis in DLA is also directed at the acting subject, so that the reference to God is indirect and has a cognitive (and motivating) function in the agent's choice-making process. Knowledge of what "God wills the will to will" (= what I ought to do) resides within the mind, namely, on that self-reflective higher level which Anselm in DCD calls DJ, so it really means that I am free (and just) when I can will what I want to will because I have brought lower-order desires in line with DJ, which is conformity with the natural order and, thus, with God's will. "Willing for sake of rectitude is the same as willing what God wills the will to will" (DLA 8). It has become part of my value system.

Note that such a view of ethics, as much as it relies on theology in Anselm's case, may not need God to ground morality and responsibility. An innate faculty, DJ and the natural access to the moral law it provides, as well as a faculty of judgment and application, may suffice. Such a view would be in line with later Scholastic teachings on natural law (in particular Thomas on the various types of law) but also with Kant's philosophy. In fact, this idea goes back to St. Paul's belief in a natural law and to the Stoic principle of a "natural order" which was adopted into Christian theology (as "divine order") and features prominently in Anselm's atonement theory (cf. Göbel 2015, 36–42).

5. So, Rogers' observation that Anselm's reference to God puts an end to any indeterminate, potentially infinite series of orders of the will is not just true because of the theological dimension Anselm introduces, but rather because God happens to be the absolute, external foundation of Anselm's internal value system. The hierarchy of desires comes to an end when the evaluative dimension of human decision-making finds values which serve as ultimate standards for the agent's judgments. They may have a theological foundation or merely human sources, or, as in Anselm, a combination of both.

In principle, however, Frankfurt's problem of a possibly infinite regress of higher orders doesn't need a return to Anselm to be solved, as Rogers seems to imply (2005, 250). In fact, a similar solution (although without reference to a theo-teleological foundation of value judgments) has long been reached within the tradition of Frankfurt's disciples and critics. Watson, for instance, has contrasted "mere wanting" with judging something as good (97.99.105). He also emphasizes the role of reason in making value judgments. Reason is not just a "faculty of determining what is true and what is false" but a real "source of motivation" and "original spring of action" (98f.) since "values provide reasons for action" (97). For Watson, "what one most wants" is not just, as in Frankfurt's analysis, someone's "strongest desire" but "what one most values" (100). Values are an essential part of a person's "motivational system" (97).²¹ This (old) insight doesn't need to abandon Frankfurt's hierarchical schema, though. Lewis thus identifies what we value with our second-order desires (115f.). If Anselm "prefigures" any of this, it is the combined or amended view of a Frankfurt-style hierarchical account of freedom, something that also accounts for the value-dimension of proper moral choice.

Critics of Frankfurtian interpretations of Anselm may have a point in maintaining that the idea of higher-order desires is meaningless if it remains within the sphere of a "superficial human wanting" (e.g. Goebel 2006), but this is precisely where Anselm's hierarchical model (as Rogers presents it) differs from and transcends the scope of Frankfurt's analysis. Watson states that "second-order desires are themselves simply desires" (108). Anselm's DJ on the other hand (although he calls it "desire" or "will") refers to a "valuation system" (105) which goes beyond mere desiring. Even more, its goal, justice, is not part of a subjective value system alone but has objective existence and a transcendent source in God who is the standard of all being and of all normativity. Emphasis on these differences between Anselm and Frankfurt may help reconcile Rogers with some of her critics. Talk of desire may underscore the subjective, human side of morality, but Anselm never loses sight of its objective foundation. In the just person, who is good and free, "what one wills the will to will" (because it is just) is "what God wills the will to will."

6. As much as I find the hierarchical account helpful and compatible with Anselm's approach, it is not clear to me how different this really is from Kant and "Kantian readings" of Anselm. Conflicts still arise, although they may arise "across orders" rather than on the same order of desires. But even that doesn't seem entirely true. There are still desires for specific actions that are mutually exclusive if the agent does his "duty" (in both Kant's and Anselm's sense) and aligns his DB with the DJ.

Rogers certainly acknowledges these conflicts; they are in fact essential in her emphasis on genuinely open options of acting without which created natures cannot prove their freedom: "We are not seriously torn in morally interesting ways unless we see mutually exclusive 'goods' to pursue" (FaS, 159; cf. AoF, 86.100). It even seems that she falls back on the language of same-order conflicts when she talks about "competing desires" and "two categories of good" (FaS, 92). But this is not the case. What seems like unclear language results from the very intricate way in which Anselm conceives moral choice. There is (1) no conflict per se between the two desires, but there are conflicts between options of choice. Yet, the "torn condition" does not (2) simply arise from something purely objective (i.e., from a conflict between options which merely present themselves externally to the acting person). Rather, the torn condition arises within the subject (the acting person's will) as a result of the desires humans are endowed with. This leads to a conflict between two ways to seek benefit: mere benefit and just (justly sought) benefit (when DJ qualifies DB). It is precisely the hierarchical schema that explains the difference and allows us to see that conflicts between same-order desires for specific actions really constitute cross-order conflicts between the (mere) DB and the (additional) DJ. In fact, there are enough passages in Rogers's works where she adds similar qualifiers. For example: "The torn condition arises when the agent desires justice on the one hand and, on the other, some conflicting 'mere' benefit which is inappropriate. But the desire for justice is a second-order desire about how to regulate our first-order desires. We are always motivated by what we believe will be of some benefit for us" (FaS, 118).

Rogers is right in pointing out that, on principle, DB and DJ are not opposed to one another (AoF, 72, cf. DCD 14, DC 3,14). But some desires for specific actions are. In other words, some desires for benefit cannot be aligned with justice or DJ; for benefit needs to be sought in a way that is compatible with justice and the higher principles of morality. This, however, is not really different from Kant's view of moral autonomy which is also grounded in what we could call a "hierarchical" anthropology (this may be most obvious in Religion within the Bounds of Reason Alone²²). Rogers insists that commoditas is better translated with "benefit" than "advantage" to avoid the Kantian connotation of immoral selfishness (2005, 254); she claims that Kant "pits duty against self-interest and natural inclination" (AoF, 72) and that Kantian readings of Anselm fail to see the non-exclusive relationship between justice and benefit.

But Kant's account of moral choice, duty, and virtue is more intricate than this kind of dualism. He acknowledges that one source of human acting is sensibility, which he sees as a predisposition to the good that humans possess insofar as they are animals. He does call its principle "self-love," but this is not in itself evil. The "propensity" to evil is not a "natural property" precisely because "it can be imputed to the man, and consequently must consist in maxims of the elective will." Sensibility and self-love are natural, therefore not immoral but simply amoral. However, they can lead to evil if they are not controlled by reason, that is, if a person ignores his inherent access to the moral law, if, that is, "he reverses the moral order of the springs" of human acting. Evil does not result from "the distinction of the springs that he adopts into his maxim" but from a false "subordination" which makes "self-love and its inclinations the condition of obedience to the moral law," whereas morality and the natural order demand the exact opposite.²³ Evil results if self-love (as a sort of DB) is not aligned with practical reason and the moral law or "justice."²⁴ A closer look at Kant's hierarchical analysis of evil suggests that, despite Rogers' concerns, Anselm does indeed seem to prefigure not only Frankfurt but Kant as well, and that he may at the same time stand in an older tradition which goes back to Aristotle and others simply because they all share some fundamental insights into human decision-making.²⁵

One might argue that, for Anselm, seeking benefit is not just amoral but has genuine moral value. This, however, doesn't seem to be totally accurate. DB is an integrative principle of all acting and, as such, falls into the realm of what we might call today motivation analysis. It is only DJ (added to DB) that allows us to make morally significant choices and thus enter the realm of proper morality. DB is certainly part of the rectitudo of things (animals have rectitude), but it is not sufficient to ascribe praise, blame, justice, or to ground freedom and responsibility.²⁶

Kant acknowledges that self-love is a legitimate determining factor of the human will (although in itself amoral and only of moral concern if it conflicts with the moral law). Anselm's view that humans always seek benefit is similar, and he clearly sees that "mere benefit" can imply the kind of selfishness Kantian readings deem immoral. Conflict situations mark the difference and may turn an amoral desire into evil. At the heart of their moral analyses, both Anselm and Kant focus on a form of justice (moral goodness) that doesn't just result from desire but is freely chosen for its own sake.

Nevertheless, what Anselm calls DB will also motivate those actions that spring from DJ or, rather, actions springing from DB aligned with DJ. In other words, all choices retain their orientation towards benefit if this is indeed the underlying motive of all willing. Up to this point, DB may in fact have a wider extension, but it still remains of a lower order than DJ since, according to Anselm, the former needs to be aligned with the latter if one wants to act justly and freely; the reverse is never the case. The "will for justice does not aim at some object opposite to benefit. It operates at a higher level curtailing the excesses of the first-order desires for benefits" (FaS, 257).

It now seems to me, though, that actions springing from DJ (in addition to DB) provide some other kind of benefit (or happiness) which may not just fall under the benefit/happiness Anselm explicitly considers in his discussion of the two desires in DCD. We may, therefore, based on other passages in Anselm's work, need to think of a third-order desire to complement his model of the will and fully capture the meaning of this other conception of happiness.

1.4 Human Fulfillment and Theo-Anthropology: A Third Desire and Other Kind of Happiness?

1. Anselm's double correspondence morality entails respect for the natural order in one's moral choices and, thus, the agent's realization of his purpose and place within that order. Ethics is grounded in anthropology; the nature of the human being is part of God's creation and of his plan for the world. Morality provides happiness since it is completion of the human nature and gives the satisfaction of having achieved self-knowledge and self-realization. What I have called Anselm's (theo-)teleology above informs his view of the human being and thus extends to his ethics as well. In short, Anselm's analysis of morally right choices is grounded in his (theo-) teleological anthropology.

This is an idea that we have come to associate with a virtue ethics account of character development and personhood. Human flourishing is found in the fulfillment of our higher faculties. The end consists in the proper exercise of one's rational and moral abilities (Aristotle's telos-nature context of energeia and entelecheia). Rogers repeatedly elaborates on the similarities between Anselm and Aristotle (e.g. FaS, 33.38.77.118). On this account, the virtuous person will at the same time be a happy person because happiness is "activity of the soul in accordance with virtue" (NE I 7). At first glance, it may seem that happiness, as human fulfillment, trumps justice, but that is not actually the case. The Aristotelian notion of happiness is in fact in line with Anselm's hierarchical account which allows us to introduce the qualifier "true" or "proper" to preserve the subordination of benefit to justice Anselm upholds in DCD, and this is achieved precisely through the agent's value judgment which brings DB in line with DJ. This achieved justice then results in "true" benefit which can genuinely be called "happiness" (beatitudo). Rogers sees this, too: "Your over-arching life goal—the aim of your self-creation—is to flourish as a human agent" (FaS, 73). This is similar to Aristotle's virtue ethics in its emphasis on character-formation. Anselm shares the "view that your choices produce your character" (29.87).

Anselm defines his ethics of the good life anthropologically, informed, just like Aristotle, by an understanding of the purpose or proper function of the human being. Anselm's distinctive definition of freedom as PSR² aims at conformity between acting and being. Anselm thus "argues that as a property of a rational will freedom cannot be thought of abstractly but must be considered in relation to the view of man and purpose of the being who is said to possess it" (Kane, 300).

There is, however, an additional aspect to Anselm's ethics. For what is the function or purpose of the human being according to Anselm? It is twofold and thus transcends the Aristotelian (and any non-theological) account in that it extends from the realm of anthropology and ethics into the realm of theology.

(a) Man is, Anselm says, a "rational creature" (Monologion [= M] 68, cf. Cur Deus homo [= CDH] II 1). In line with a longstanding philosophical tradition, Anselm emphasizes the rational nature of the human being. This has relevance for his metaphysics of morals: the theo-teleology of being, which turns into a theo-teleology of acting when the rational agent does what God wills him to do, is not only something God or the natural order of things demand, but God or the natural order of things have made it such that humans desire justice and understand this demand by means of what is most distinctive of man, namely, reason (practical reason). So, for humans the natural order of things does not just consist in being what they are; they also need to use their reason and understand what they are ordained to do. Reason is that very faculty that allows them to order their desires properly.

(b) Furthermore, humans are created to "know and love God" whose "mirror and image" the human mind is (M 67ff.). We are asked to "imitate God" (M 68) and thus gain the "blessedness," "happiness," "salvation," or "happy immortality" we were made for (CDH Pr, I 10.19, II 1.10; cf. DC 3,13). This thought guides Anselm throughout his philosophical endeavors and takes the form of a personal confession in Proslogion (= P) 1: "I was created so that I might see you but I have not yet done what I was created to do" since "the image of God created in me is eroded by vices" and God will have "to renew and refashion" it so that Anselm can use his proper powers again. Like Aristotle, Anselm stresses the role of reason, but there is yet another dimension in man's orientation towards God which goes beyond a mere ethical fulfillment of the natural order and God's will (although he believes that the kind of happiness it results in can only be obtained through morality).

2. According to Anselm, a first goal of the human being is to imitate God through a proper use of one's powers of reason and will. It is "in being free that we are made in the image of God" (AoF, 113, cf. M 68, CDH II 10). These powers are in fact so great that humans are capable of "self-initiated action" (Williams/Visser, 8), and the emphasis that voluntarist accounts of Anselm's moral philosophy place on this essential ingredient of true freedom²⁷ is echoed by Rogers' focus on aseity.

For Anselm, the value of free will consists in the ability to imitate God, and the way in which we "actualize the image of God in ourselves" is through "aseity," or the ability to choose "from ourselves" between alternatives (Rogers 2005, 251; cf. DLA 14, DCD 12 et al.). More importantly, this choice determines our character and thus constitutes some form of self-creation. A se choices in morally relevant situations are all about "freely and responsibly forming our characters" (AoF, 9). The human being does not have self-creating powers in the fullest, divine sense of the word, and the tools we use come from God. But according to Rogers, Anselm insists that "self-creation is the whole point of freedom" (AoF, 12). Free choice and freedom in the sense of virtuous self-perfection belong together if our choices do indeed "help in the construction of our own characters" (FaS, 37). This is Anselm's point in DCD 18: the condition of a rational creature (i.e. being endowed with DB and DJ) is such that "he would be able to give justice to himself." Rogers holds that it is Anselm's goal to "carve out a genuine domain for created choice in the universe of traditional, classical theism," and he arrives at some sort of "reflective primary cause" which mirrors the "primary agency of God" (AoF, 19). Rogers maintains that Anselm thus departs from Augustine's "theist compatibilism" (31ff.) and insists on a libertarian interpretation of the choice between competing desires, which serves as a "measure of independence" or aseity (11.106). Desires "may be explainable in terms of preceding causes" (in our case, God) but "the winning out of one desire over another" is not (60). The phrasing is problematic: it seems to refer again to a conflict between the two desires, and it also seems to denote a certain passivity, as if the agent were nothing but a bystander in this conflict. That, however, is not what Rogers has in mind.

She speaks of "parsimonious agent-causation," acknowledging that for created beings "the space for aseity is small" (FaS, 81ff.). "All that is up to [man] is [to] hold onto God-given justice or throw it away" (AoF, 101). In DCD 2, Anselm refers to Jn 8:44 ("the devil did not hold on to the truth") and coins the term pervelle to describe the modest autonomy of rational creatures. What the agent does "from himself" is nothing but "hold on" to justice or "will it through to completion" when there are options of seeking mere benefit instead. He preserves rectitude justly (i.e., for the sake of rectitude itself: PSR²).²⁸ But this is still true aseity, and it is only possible for the created nature if one is confronted with "genuinely open options" (FaS, 9). It is here that the torn condition or PAP has its decisive role. Anselm modifies his general definition of freedom for creatures and exemplifies it with Satan's choice and the two desires of the will. Created freedom may only consist in letting justice prevail (or one's DJ when it is not compatible with the desire for a certain benefit), but it is "willed through" a conflict situation, and it is done by free choice. The desire comes from God (it is a "god-given motivation") but, under the conditions Anselm describes, the "winning out" becomes an instance of true "self-causation." Thus, it is "up to the creature" in a libertarian sense and therefore grounds responsibility (AoF, 60.72.118.153). In order for this to be possible God endowed rational creatures with both desires. Moreover, this choice is not determined in the kind of intellectualist sense that would exclude any form of voluntarism, since even if a person judges something as (1) beneficial and (2) just, he still needs to will it through and doesn't merely act on the rational insight alone.²⁹

It might, in any case, be more appropriate to speak of co-self-creation (rather than "self-creation") since Anselm clearly sees God as the one who provides everything except that choice itself in which the human agent holds on to justice. The role of creatures is, after all, ontologically limited; but the purpose of Anselmian free choice, as Rogers sees it, is clear: we "participate in our own creation by making ourselves better on our own" and thus the "created agent enhances his moral stature" (FaS, 24.156).

3. If rational beings are made in the image of God and are therefore "god-like," it would seem that final fulfillment of their being, or the highest degree of their powers, is achieved if and when they succeed in being like God. This would also provide them with supreme happiness, something that Anselm confirms in his discussion of DB in DCD 13 where he considers the (unsatisfying) hypothesis that the fallen angel could have received DB only and holds that "the higher he realizes happiness can be, the more he wills to be happy . . . therefore, he wills to be like God (vult esse similis deo)." But this cannot truly be the case. The angel must have had DJ as well to "curtail the excesses" of his desire for benefit/happiness (DCD 14), namely, to have what is not "fitting for him," to wish for things improperly or for benefits not meant for him. Thus, it is not "fitting" for a creature to (desire to) be like God.³⁰ It seems, however, that this self-restriction, which results from the proper ordering of desires (cf. DC 3,13), is not just a matter of curtailing DB, as Anselm says; rather, there is another kind of happiness/benefit involved which, in the final analysis, the self-restriction and self-control of human beings leads to.

We have seen that Anselm does in fact insist on another purpose and meaning of human existence aside from being a rational being. This additional purpose is to "see" and "be with" God. This is clear from the emphasis Anselm places on the Christian notion of salvation (cf. CDH). He builds (e.g. in CDH II 1.4) on the Augustinian idea that humans have a "metaphysical nature" (cf. Fischer) which is both rational (they want to "know God" and understand last things) and affective (they "love God" and desire to be with him). This nature that is oriented towards God comes from God himself who "has made us and drawn us to himself," so that we may strive to seek, love, and understand him and, one day, dwell with him, since in his presence the restlessness that permeates the human condition (reason and will) will come to an end: "our heart is restless until it rests in you" (Confessiones I 1). It is this idea of man's God-given orientation towards God which I call Anselm's theo-anthropology, or, more precisely, his theo-teleological theo-anthropology.

Anselm's moral landscape is not so much, as Rogers holds, influenced by Aristotelian eudaemonism but rather by what I call Augustinian beatitudinism. The latter implies the former, but the purpose and goal of the human existence does not only consist in "flourishing by fulfilling one's natural purpose" of being rational and moral (FaS, 118). For Anselm, and any Christian thinker, the ontological and moral completion of the human nature (and, thus, of the natural order) in this life is not the final goal. Rather, it consists in being with God in the afterlife. Humans yearn for God's presence, but it is only actualized in heaven.³¹ Both eudaimonia and beatitudo mean "happiness," but the Latin term better captures this additional, transcendent, and theological dimension of Christian anthropology and ethics. Such happiness does not directly consist in doing what one is supposed to do, it is not simply the satisfaction one finds in self-knowledge and selfrealization, or self-determination and self-perfection; rather, it is a happiness granted by God as a reward for doing "what God wills the will to do" (DLA 8).³²

For Anselm, a moral choice is never a moral choice alone. The final purpose of all human acting is everything Anselm says about the nature and scope of freedom but in a more complete way, and justice (moral goodness) is the precondition rational creatures have to meet in order to receive the kind of happiness we have just described (DCD 13). "Justice is the requisite for our achieving the goal for which God made us: eternal happiness in His presence. Why is justice necessary? Because only by choosing justice can the rational creature 'polish' the image of God in himself and become a genuine imago dei, meriting the happiness for which he was made" (AoF, 56). As has already been pointed out: for Anselm, the question of genuine moral responsibility, i.e. the attribution of praise and blame to a human agent and his choices, has eschatological relevance; it is about sin and salvation, heaven or hell (cf. De conceptu virginali [DCV] 23, DLA 6.14, DC 3,9.12), so that "freedom is doing what one ought to do for the sake of salvation" (Luscombe 2006³³). The desire for this kind of happiness doesn't find its highest form in what Anselm describes as perfecting the desire for benefit/happiness he considers in DCD 13 (i.e., to be like God). So, shouldn't we rather conclude that good angels and humans who do what God wills them to do act from a desire that aims at being with God?

Anselm is certainly aware of this desire (e.g., M 70) as well as the fact that sin ("abandoning justice") affects one's happiness (Satan "took happiness away from himself": DCD 18). But he doesn't seem to consider it enough in his description of DB.³⁴ A hierarchical model of the will which is inspired by Anselm but aims at a complete account of moral choice—explaining, in particular, evil choices—will have to place sufficient emphasis on the longing for eternal happiness with God, and therefore I suggest modifying such models and introducing a third desire, the desire for Happiness (hereafter DH). The capital "H" distinguishes it from what I have so far called DB or, referring to Anselm's use of beatitudo, the "desire for benefit/happiness" (lowercase h). It seems to me that DH cannot simply be identified with DB (i.e., the happiness that is subordinate to justice) precisely because its goal (or highest degree and final fulfillment) is distinct from that of the "lower" desire which "merely" aims at likeness with God.³⁵

Although Anselm does not say this explicitly: evil springs in fact not only from disorder between DB and DJ but from confusion about the two forms of happiness.³⁶ Anselm does make it clear, though, that even the god-like aseity rational creatures have has a goal and context, namely, it fulfills the purpose of a rational being at a certain level (i.e., to imitate divine powers), but it still is a tool only for the sake of something else (i.e., to be with God). The final goal, and true freedom, is self-initiated action that leads to self-perfection through proper self-determination and thus merits the heavenly existence with God. We can therefore establish another hierarchical relationship between the two purposes and forms of happiness: aseity is a godlike ability; its goal, however, is not just to be like God but to be with God. It should be noted more clearly that beatitudo, not aseity, is the final perfection of the human being and reward for justice and freedom in Anselm's distinct sense of PSR². Taken out of context and deprived of its purpose, aseity can also lead away from God.

4. The introduction of a third DH into the hierarchical model of the will sheds new light on a puzzle that goes beyond an investigation into the nature and possibility of moral evil which Christian thinkers usually answer with some form of the free will defense. It is the question why humans (or angels) would ever have chosen evil in the first place. Anselm addresses both problems. But the supposed distinction between a desire for benefit/happiness and an additional DH calls for further reflection before we can turn to the problem of evil.

If justice leads to eternal happiness as a reward that is granted by God, one could argue that it ultimately serves another self-serving goal and brings about a personal benefit. But the goal is not selfish in the sense that both Anselm and Kant would deem immoral. It is nothing aimed at worldly advantage, nothing done at the expense of others. Choices that are just cannot come into conflict with DH, only with the DB Anselm discusses in DCD. The two are different precisely in that the latter needs to be aligned with DJ. If, however, a person does what he is meant to do, he perfects himself, is good to others, and merits God's favor in eternity. A good choice, in other words, manifests the harmony between anthropological, ethical, and theological determinations of the human existence.

On the one hand, the Happiness that results from justice and proper self-perfection is a transcendent quality. It therefore seems useful to see the corresponding desire in human beings as higher (belonging to a higher order) than DJ. On the other hand, DJ does not need to be aligned with this desire, insofar as it is not separate from justice and its theological context. One may thus be tempted to identify DH with DJ or see it, to use Aristotelian language, as a "consequent goal" of DJ. However, identifying the two desires wouldn't take into account that DH underpins—and may actually motivate—DJ. Accordingly, this is the ultimate reason why we should desire justice and have received DJ. It is precisely the fact that the desire to be with God motivates DJ which, I believe, is reason enough to keep the two separate. The relationship between different orders of desires should not be conceived of as merely a process of aligning and curtailing in a very narrow sense of these words. Frankfurt's emphasis on self-identification and the additional emphasis we have placed on higher-order desires and values as the appropriate instruments for the process of self-identification make this very clear. More precisely, DH motivates the agent to act on DJ and choose accordingly, and this is what the Frankfurtian model of the will describes as self-identification with a desire which then becomes effective and turns lower-order desires into volitions. DH is in this sense clearly a higher-order desire, higher, that is, than DJ and, thus, also higher than DB. So, a three-order structure of the will remains useful.

For similar reasons, it does not seem appropriate to subsume DH under DB. The latter, understood in the widest possible sense, could be conceived of as having two forms, with justice (or DJ) separating one from the other. I have pointed out that justice may be seen as the adverbial qualifier which allows human beings to seek benefit justly or to seek true benefit (i.e., Happiness). This means that the higher-order DJ has a qualifying effect on the lower-order DB, that is, by allowing a person to make a value judgment. But a two-order structure does not sufficiently explain the fact that we are, in turn, motivated by DH to act on DJ (the latter is "effected" by the former). DH does indeed "precede" (is prius to, cf. DCD 12) all other desires (including DJ), but it doesn't seem to be effective in all human choices, which is what Anselm says of DB. This additional difference underscores the need to distinguish the two.

In fact, it turns out that the supposed universality of DB, if it only appears under the form of the desire for "mere benefit" without justice, can only be asserted in a very superficial way. It may even be seen as self-contradictory. For DB without justice does not lead to proper benefit in the fullest meaning of the term (i.e., Happiness) but to its very opposite, meaning that those who perpetrate evil a se choices make themselves evil, abandon justice, and deprive themselves of the possibility to exercise PSR², the consequence of which is, according to Anselm, eternal unhappiness.

5. The distinction between two desires for happiness does not, however, solve the puzzle of evil choices, but it reconfirms something that should be commonplace in any theological ethics. Whatever the reason or direct motivation for an evil choice, one of its external yet necessary (albeit not sufficient) conditions is some kind of cognitive deficiency. This should not come as a surprise; for the created world is characterized by imperfection (Leibniz' "metaphysical" definition of evil is creaturely imperfection). Evil simply does not exist in the world of divine perfection. Some evil certainly results from irrational weakness of the will which leads a person to act against their best knowledge and give in to inappropriate desires.³⁷ Often, however, evil boils down to cognitive deficiency, the result being that the agent does not see his choice as evil. Anselm's insistence that rational creatures are not only endowed with DB but also with DJ aims at undermining this claim; the devil knows what God wants, his DJ gives him access to a natural moral law.

Rogers addresses some of these issues regarding the acting person's ignorance. Responsibility for character-determined choices can only be acknowledged if one is responsible for one's character ("tracing thesis"). Yet, what if a person does not understand that her choices determine her character? Rogers argues that the common principle of ethics that says that no one can be responsible for something done in ignorance does not apply here; for (1) humans ought to know that they build their characters by making choices, and (2) unawareness of the character-determining power of one's choices does not excuse evil choices (FaS, 217f.225).

Anselm addresses ignorance in DCD 23. He says that Satan could not have had full knowledge of all the consequences of his choice, namely, he couldn't have known for sure that he "would actually be punished," for otherwise he wouldn't have sinned. Satan does, however, know enough to be held responsible. "In Anselm's paradigm instance of choice, Satan recognizes that he could conform to the divine will and that, in choosing what he chooses, he violates a divine command" and deserves punishment. Anselm's example implies a distinction between "exculpatory and non-exculpatory ignorance" (FaS, 217.226). This analysis, however, only highlights the subordination of benefit to justice. More puzzling is the fact that the sinner (in the mere sense of evil-doer) seems to act against his own natural DH. The issue here is not so much that the sinner ignores DJ (evil clearly violates the principles of justice; this violation defines it as a moral term) or that his ignorance may have an effect on his responsibility. Anselm's theoanthropology sheds light on (or, rather, amplifies) the bigger puzzle of evil, namely, its motivation. Why exactly would anyone do evil if it ultimately leads to unhappiness, precisely, if he is fully aware that justly sought benefit would make him eternally happy?

Without the introduction of the third-order DH, this would be a flat contradiction of Anselm's claim that creatures always seek happiness (DB) or, rather, a self-contradiction in the very desire that according to DCD 14 is motivating Satan even in his evil choice. Keeping DB and DH separate can help avoid this issue, but it doesn't alleviate the cognitive dissonance at the heart of an evil choice.

Based on the separation between two desires for happiness, evil could simply be explained by saying that, while all humans are originally endowed with DB (which is always effective) and with DJ (which sinners "abandon"), some lack DH. But Anselm clearly thinks that all men were "made to enjoy God" (DC 3,13). If his theo-anthropology is correct, it has universal validity and all humans possess DH; they are all made to seek Happiness through justice, not "benefit simpliciter" (FaS, 82). It is important to note that DH cannot be mistaken for a merely moral imperative (like the demands of justice³⁸). It is a matter of self-fulfillment, not a (divine) commandment which could somehow be perceived as extrinsic and hegemonic, as somehow alien to myself. It is obvious, however, that this desire is not always effective in all human choices.

If their very nature commands humans to seek not only benefit but Happiness which results from justly sought benefit, evil choices are only possible in a state of ignorance which is, to be sure, not the root cause of evil but a condicio sine qua non. The person making an evil choice by abandoning justice, or valuing mere benefit higher than justice, may believe in some alleged advantage for himself (his DB remains effective) but he doesn't understand the full consequences of his choice, namely that his per-willing "mere benefit" will make himself unhappy in the long run. Evil thus seems to be possible due to a false value-judgment which results from limited knowledge and reasoning.³⁹ Christian theology is adamant about the fact that the cause of sin cannot be evil itself (evil is seen as the consequence, not the cause of the inappropriate choice Adam made), but all human choices are made under the condition of imperfection.⁴⁰ The imperfection of the created world is not a consequence of sin but precedes it; the world is "from," not "of God" (Augustine, De natura boni 10). This imperfection here is precisely a cognitive issue: it is the acting person's unawareness of his proper purpose as a human being and as God's creature, or unawareness—i.e., lack of direct knowledge—of the final consequences of a sinful choice for one's true Happiness (he only knows that his choice is unjust). Anselm describes this ignorance which co-determines Satan's paradigmatic choice and without which he would never have done evil and thereby make himself "unhappy" (DCD 22f.).

It thus turns out that we may want to take issue with the very context of morality and eschatology which is so characteristic of Anselm's moral philosophy. Sinners are often ignorant of the fact that their moral choice is, indirectly, (1) a choice for or against God and (2) a choice for or against their own nature. At the same time, Anselm (like most of his readers) focusses on moral choice, although his initial definition of freedom transcends the realm of morality. This, it seems to me, is problematic not only for non-theist ethicists but for theologians as well.

Part 2: Judgment Day: A Different Test Case for Anselm's Concept of Freedom

2.1 Morality and Eschatology: Evil as an Indirect Choice of God and Self

1. Questioning Rogers' "Aristotelian" reading of Anselm, Visser emphasizes that he "is not an Aristotelian with respect to character formation" (Visser, 412); habituation doesn't seem to have a role in his account of good and bad. Indeed, Anselm picks Satan's fall as his paradigmatic example when it comes to freely choosing between good and evil, which amounts to a once-and-forever "single primal" choice, and Rogers' analysis of "character-formation" traces act-responsibility back to one decisive moment that determines a person's character "forever" (FaS, 123.120 et al.).⁴¹ This is consistent with what Anselm says in DC 3,4: "having thrown away justice," a person can never do anything good again ("unless God grants her the grace to do otherwise"). It is in this sense that she is now determined to be evil and "can only will badly and stupidly" (FaS, 119).

Evil choices are unfree because they don't meet the PSR²-definition and result from a rational being's loss of control over its own desires, but they are also determined because the evil person cannot be not evil since the loss of DJ is unrecoverable for the sinner and he therefore lacks the strength of will to persevere in times of temptation and steadfastly hold on to justice. Ekenberg (2016, 64) therefore cites Goebel who concludes that "Anselm is a voluntarist only with respect to good angels and human beings, whereas he is an intellectualist with respect to the morally bad." More precisely, Goebel's point is that Anselm's "theory of acting is voluntarist and indeterminist in the case of the morally good person but intellectualist and determinist in the case of the morally bad person" (2006, 94). The evil person can, at most, be called free "in a very different sense" (105), for Anselm claims in DLA 3–4 that he retains the freedom described as PSR² but cannot use it anymore because he has abandoned justice and the motivation and strength to be just (cf. DC 3,13). It is indeed questionable whether it is even meaningful to call such a person free. It is, Goebel claims, nothing but the "appearance of responsibility" since the evil person doesn't have any alternatives of acting properly speaking (118), and he points out that early readers of Anselm, like Hugo of Amiens, part ways with him on this point. Hugo "accepts Anselm's definition of freedom as the "ability to act in a morally good way" but teaches that the evil person doesn't possess it anymore" (119).⁴² However, Goebel (2017, 126) also underlines "that Anselm consistently refuses to attribute a free will (voluntas libera) to the unjust." Anselm's understanding of the relationship between freedom and free choice (and free will) is even more intricate than is frequently seen. "All he does attribute to those who are not morally good is an (idle) capacity to keep justice . . . a capacity he calls the "freedom of decision" (libertas arbitrii)" (ibid).⁴³

This part of Goebel's analysis seems nevertheless reconcilable with Rogers. He emphasizes that Anselm's discussion of the power to preserve rectitude and resist temptation in DLA 5–7 only refers to good people (i.e., people with an upright will) (2006, 106⁴⁴). This may be true but doesn't preclude that the exemplary account of evil choices Anselm offers in DCD (on which Rogers focusses) is still rooted in the logic of DLA's definition of freedom as PSR² and its application to creaturely existence. Anselm himself points this out in his preface to the three dialogues when he says, "what I wrote about the bad angels was the heart of the question, whereas what I said about the good angels [in DCD] was a side issue." More importantly, the choices of evil people are determined, but that is so due to their own self-determining primal choice. Libertarianism does not hold that nothing in the world is determined; it simply says that some things are determined due to choices that are in themselves indetermined. Rogers points out that Anselm assumes the same for sinners who have lost the ability to be non-evil by their own doing. Of true interest to us are not the individual—inevitably evil—choices evil people make, but how that first, free evil choice is possible ("first" in a temporal and metaphysical sense). This is precisely the question Anselm and Rogers seek to address.⁴⁵

I emphasized, however, in the previous chapter, the role of a cognitive deficiency even in that choice. I contend that this condition in fact limits the freedom of choice. Ignorance is not a necessary condition of freedom (as Anselm and Rogers hold, referring, to be sure, to the moral freedom of creatures) and actually undermines Anselm's notion of "meritorious freedom" (Rogers 2002, 71); it is only a necessary condition of evil.⁴⁶ Ignorance and the randomness that comes with it are impediments for a truly free choice in the sense of PSR². Under the conditions of imperfection, it seems unreasonably and unfairly difficult for some to exercise this kind of freedom; it does not seem possible as a general human quality.

2. The issue of self-determination is even more pressing because of the eschatological relevance it has in Anselm: forming one's character through one choice also determines one's afterlife. Moral freedom is not just a matter of "co-self-creation" for rational agents but also of co-self-redemption as opposed to co-self-damnation.⁴⁷ "Anselm . . . proposes that a single choice in the past might cast the die for all eternity" (AoF, 84). Humans earn the happiness that is beatitude through their moral performance.

This focus on a single moment in which man loses rectitude sounds like an account of original sin (Adam's fall), and it certainly is at times (in DLA, DCD, DC, DCV). But it also serves as an account of individual responsibility for either sinning or holding on to justice. We don't just live in a post-lapserian world but also after the restoration⁴⁸ that has already happened in Christ. Christians, Anselm points out, participate in it through baptism (DCV 17, DC 3,8). Anselm doesn't believe that they are all saved⁴⁹; he merely lays out the foundations for the ambivalent state of the human situation where good (through God's original creation and Christ's redeeming sacrifice) and evil (through Adam's choice) are possible. Anselm addresses both the pre-lapserian will and the issue of subsequent grace. DLA and DC, in particular, go beyond Adam and Satan's fall and give an account of any a se choice one may and must make in order to prove one's individual justice after the human nature has been restored in Christ. Rogers' point is that it is still just one choice that counts.⁵⁰

But this one moral choice of self-determination or co-self-creation with eschatological relevance in an individual's life is preceded (1) by Adam's choice and (2) by God's gracious choice to send his son and thus restore humankind to its original dignity (cf. CDH, DCV 22). I suggest that (3) the individual's moral choice Anselm and Rogers focus on, if it is at all relevant, may be followed (or replaced) by another choice (4) which is independent of it or may correct and reverse it.⁵¹

The broader theological question is this: are we content with the idea that this world of cognitive and moral ambivalence is the arena in which the unconditionally loving father-God Jesus Christ proclaims submits each individual to a "test of freedom" so that we make that one character-determining choice (act from inordinate desires: DCD 4.7) in a state of unawareness regarding the consequences of our choice not just for DB and for the demands of justice but for our own purpose and DH? Are we satisfied with the idea that the lack of direct knowledge regarding the implications of our choice lets us abandon justice and lose the ability to exercise the "virtue" of freedom and self-perfection, and that God will then judge us accordingly and deny us that greater happiness of self-fulfillment for which he made us? I for one am not and in light of this will therefore suggest an alternative scenario.

3. Let it be understood that it is not Anselm's distinctive definition of freedom as PSR² that gives rise to theological problems. It is the other kind of freedom that is also present in Anselm's thought, namely, the free choice which in the case of creatures, he says, requires a PAP. This takes us back to the claim that PSR² combines the two notions of freedom and free choice into "one single definition" (Kane, 305). We have seen that Rogers' "unpacking" does indeed show that it is meant to cover the entire range of implications the idea of freedom has, at least in moral contexts.⁵² But I wonder how convincing the need for a PAP actually is if Anselm's distinctive definition really sets the standard for any discussion of freedom. As mentioned above, it was Frankfurt again who questioned the need of "alternate possibilities for moral responsibility," and Rogers vehemently rejects this aspect of his account of freedom.

She points out that Frankfurt seems to confuse freedom of action with freedom of choice, external with internal freedom (AoF, 78ff.),⁵³ and focusses on the knowledge required for free choice. In CDH II 5, Anselm gives the example of a monk who "voluntarily keeps his vows" and stays in the monastery; for this choice to be praiseworthy it is irrelevant if external circumstances actually allow him to carry out the opposite action (leave the monastery). It doesn't matter if he knows that he would be forced to stay should he actually try to leave. Rogers concludes that Anselm "explicitly rejects a PAP if the principle is about overt deeds . . . but when it comes to created choices, Anselm insists that there must be options" (FaS, 155).

My concern with the PAP is different from Frankfurt's. Specifically, it stems from the underlying theological issue of Anselm's insistence (which most of his readers and most theologians in general seem to agree with) that rational creatures have to prove their PSR² in the ambivalent state that characterizes the finite world after the fall, a world in which both good and evil are equally possible while at the same time the indirect, theological and theo-anthropological consequences of evil choices are clouded by ignorance. The moral-eschatological assessment of the human being which Anselm proposes reintroduces a certain imbalance between the two kinds of freedom in favor of the traditional view: freedom is, after all, to sin or not to sin, that is, the self-determined choice that makes the individual just or unjust. If we place much emphasis on the moment (and its irreversible consequences) in which justice is lost, our attention necessarily shifts from Anselm's distinctive definition of freedom (PSR²) to the Augustinian understanding as choice between good and evil. At the heart of Anselm's moral inquiry, at least on Rogers' account, stands not freedom (PSR²) but the freedom (free choice) to give up the ability to use one's freedom by abandoning justice and becoming a "slave to sin" (DLA 2.10ff.) and to one's unrestricted DB (DC 3,11.13).

Should we not return to Anselm's actual innovation, that is, the understanding of freedom as PSR²? Can we find a test case which allows creatures, too, to prove their PSR² without the need for a PAP? A change of perspective, away from the tracing problem that defines act-responsibility by looking into a person's past and towards a "future" choice may actually help avoid the problems raised so far. The test case I have in mind is what the Christian theological tradition calls "judgment day" and the choice or freedom that humans may have before God. I shall call this "judgment day freedom" (hereafter JDF, which, to be precise, in the following often refers to the free choice humans exercise on judgment day), and I suggest we identify it with PSR². It seems to me that the truly self-realizing choice Anselm has in mind—freedom as virtuous self-perfection—is only possible at that point.⁵⁴

4. It may sound peculiar that God's judgment should offer human agents a free choice. This notion, however, has become a commonplace in theology; it goes back to St. Paul's concept of a natural moral law which even the gentiles "know in their hearts" (Gal 5:11 and Rom 2:14–16⁵⁵). Luther, 53 maintains that people will actually be judged "by their own thoughts" and God will simply "assent to your own judgment about yourself and confirm it. . . . According to the witness of your thoughts and of your conscience, you therefore deserve either heaven or hell." Ratzinger (1978, 176) has expressed a similar thought: sinners are not "turned away by God" but "turn away from God."⁵⁶ Ratzinger sees this choice (which God will respect) as the ultimate fulfillment of human freedom, but I don't think it necessarily requires a PAP (i.e., the possibility of turning away from God).

It is against this conceptual backdrop that I shall place Anselm's definition of freedom in the context of the doctrine of universal salvation and reflect on JDF. This is where PSR² may have its proper place, not in the analysis of a moral choice which has indirect eschatological effects. If Anselm carries the idea of "choosing God" and one's true self-fulfillment into his analysis of the will, and if "judgment" is all about a direct choice of God, both these ideas would require that such a choice not be made under the conditions of a clouded judgment but once the truth is fully "revealed" and the "veil of ignorance lifted" (cf. 2 Cor 3:14).

Furthermore, employing Anselm's PSR² allows us to preserve the idea that the human person does in fact act freely on that "day" and makes a truly self-determined and self-determining choice. We can thus solve one of the biggest issues Christian theologians have with the doctrine of universal salvation: it seems to do away with free choice, for everyone is saved anyways, regardless of their moral choices and regardless even of that final choice for or against God. Indeed, universalism has gained a certain popularity among Catholic Church leaders as well.⁵⁷ The conflict between the concepts of universal reconciliation on one hand and freedom on the other hand only arises if freedom is understood in the traditional way as the ability to choose between good and evil. With Anselm's PSR² there may be a chance to preserve human freedom against the backdrop of universalism.

2.2 Judgment Day and Anselm's Notion of Freedom

1. Understanding freedom not primarily as a choice between good and evil, but as free choice of the good, as a free and liberating assent to the order of things according to the creator's plan, allows us to gain a new understanding of the relationship between salvation and freedom, religion and morality.

Anselm emphasizes that the freedom for the good, whose use had been lost through sin, has become possible and desirable again because of the salvation Christ has brought. This is obvious in his moral works, as we have seen, though Anselm develops the theological background most prominently in CDH: through Christ, the human being has been "restored to its proper dignity" (CDH II 21). This is true for moral, character-determining choices, but it is even more true for the afterlife-determining choice of JDF. Jesus himself is unequivocal about his role on judgment day: "And I, when I am lifted up from the earth, will draw all people to myself" (Jn 12:32). Recall that the divine judgment has little to do with the human notion of retributive justice (cf. Rm 3:21ff, Gal 3:6ff); it is rather the notion of "making just" (justi-fication). The way in which this justification is brought about gives humans a role that may qualify as some sort of "parsimonious agent-causation." It is not something God simply imposes on his creatures; on the contrary, they participate in it actively, through their freedom, without, however, falling into any form of Pelagianism. Paul's and Luther's understanding of God's judgment is key in my proposal for using Anselm's PSR² in the context of universal reconciliation. If it is true that humans should once be asked to judge themselves, I assume that, when they are "face to face" (1 Cor 13:12, cf. M 70) with the loving God whose love "trumps everything," in the presence of Christ who is love, indeed everybody will "call on him," as Paul says with that final act of salvation in mind (Rm 10:12), rather than choosing eternal separation from God.

2. I believe that my notion of JDF fulfills Anselm's definition of freedom as PSR² and that it is largely compatible with everything he says about human and angelic choice in DCD and DC. There is, however, one passage that seems hard to reconcile with my proposal and casts doubt on the claim that JDF, as a "choice for God," is actually free.⁵⁸

The passage, DCD 22–23 on Satan's ignorance, has already been mentioned, but we need to look at it more closely now. Although Satan clearly knew that "he deserved (just) punishment if he sinned" because he knew what he ought to do (he still possesses the DJ), Anselm says that he cannot have foreknown for sure that he "would actually be punished." The reasons Anselm considers include the fact that Satan may think he gets away with sinning, especially since there had been no precedence of divine punishment, or that God may first want to punish him but may subsequently change his mind due to his merciful nature, not wanting "to damn his own creature."⁵⁹ Anselm's point is that, had Satan refrained from his sin because of this knowledge, he would have remained steadfast, and thus good, but not free. If he had had certain knowledge (seeing the final consequence of his choice), "as one who both willed and had happiness, he could not have spontaneously willed that which would make him wretched." He would not have had a real choice: he "would not have been just for not willing for what he ought not, since he would not have been able to will it." And refraining from sinning "out of fear of punishment" does not count as being just. Anselm therefore holds that Satan couldn't have known for certain that God would punish him; he sins freely, thus abandons justice and so deserves his just punishment. This kind of ignorance is indeed a "necessary facet" in Anselm's moral account "of angelic freedom" (AoF, 96).

My sketch of the choice a person makes on judgment day, in the presence of Christ, looks structurally similar to this hypothetical case of Satan having "before his eye" (cf. DLA 9) the real punishment that will inevitably follow as a consequence from his sinful choice. The question therefore is: can we be called free (and just) if we choose to be with God on judgment day because we see the reward in front of us? I think we can.

3. First, in the hypothetical case Anselm considers, Satan sees the consequences of a sinful choice. A bad choice, however, doesn't complete one's being and provide that other kind of happiness humans and angels were created for. There is an important difference between punishment for an evil choice and self-fulfillment through a good choice, namely, the choice to preserve rectitude for the sake of rectitude alone which, again, is not only the "per-willing" of justice over benefit through situations in which mere benefit seems attractive but also the per-willing of Happiness (the effective third-order DH) which fulfills the very nature of one's being. There is an asymmetry between good and evil choices if the good end of a free choice is grounded in the nature of the being which makes the choice.

It seems in fact that Anselm sees a similar asymmetry between human choices and Satan's case. In DLA 9, he gives the example of a human agent who is aware of the indirect consequences his choice has in the long run (i.e., on the plane of eternity). Anselm discusses the case of someone who is threatened with torture and death if he "preserves the truth," and this person chooses to give us his uprightness in order to save his life (cf. DLA 5f.). Anselm's point is that he still makes a free choice. Although he didn't choose to be in that situation, he is not compelled to value his life higher (will it more) than the truth in that crucial moment of moral relevance. Anselm says: If the will "saw right before its eyes the eternal glory that it would immediately attain if it preserved the truth, and the torments of hell to which it would be given without delay if it lied, there is no question that it would quickly be found to have strength enough to preserve the truth" (DLA 9).

The moral assessment of the situation may sound rather harsh, almost cynical (although it is, again, quite Kantian in the radical view of what constitutes morality), and we may disagree with Anselm's account of the justice of the unconditionally loving father-God, but that is not the point here. Anselm seems to be implying that the person's choice would still have been "spontaneous" (sponte, i.e., free and a se, not compelled by an external force) even if he had possessed secure knowledge of God's ensuing judgment and if it had motivated him, through both DJ and DH, at the time he made his choice, thus siding with eternal glory over the preservation of his physical existence.

This brings us back to the cognitive issue of evil. Rogers acknowledges the exculpatory quality of "non-culpable ignorance" but insists that the "epistemic requirement" for moral responsibility is fulfilled in Satan's case because he "understands what he is doing and what the consequences might be" (FaS, 226). This is true, but it doesn't mean that seeing "the eternal glory before one's eyes" doesn't fulfill the epistemic requirement and fulfills it even more fully, thus enabling a person to make an even "freer" choice (cf. DLA 1.9). The consequences Satan is indeed aware of include the fact that his choice goes against the demands of justice, but the most important outcome of his choice is unclear to him (i.e., the indirect consequence of factual divine punishment and thus the fact that his choice will prevent him from achieving Happiness). Satan hopes to get away with his sin, but his reflection seems limited to the context of benefit and justice. He doesn't want to miss out on his Happiness; in fact, he doesn't even consider it. "No-one can will to plunge themselves into misery" (FaS, 227; cf. DCD 22–24). But it is most obviously an effective third-order DH that could prevent this from happening. Satan's evil choice proves that he lacks sufficient knowledge to make a good choice.

This is not to say that his choice is unfree; we saw earlier that he knows enough to make a free choice. But I suggest to reverse Rogers' argument (cf. FaS, 227): ignorance is the condition under which created natures normally make their (moral) choices, but it is not essential in making a se choices and required for freedom proper, which Anselm describes as PSR². If a created nature cannot see all the consequences of her choices, freedom as self-fulfillment is not possible, only a choice between good (justly sought benefit) and evil (mere benefit). Yet, freedom as self-fulfillment is, as we have seen, the ultimate meaning and goal of PSR² due to Anselm's concept of rectitudo which is informed by his theo-anthropology.

Both ignorance and open options are conditions Anselm and Rogers regard as necessary ingredients of creaturely freedom, but that is because freedom (PSR²) is reduced to free choice again (PAP), a step which may seem warranted when taking into consideration the conditions of human reality where freedom (inner freedom) seems to exist in moral choices alone. It is a different issue, however, if we look at another moment and the kind of choice that transcends the finite beings' existence on earth, at the transition from this world to the other world (JDF).⁶⁰

4. The fact that the ultimate reason of Satan's choice ("but why?" asks the student in DCD 27) remains mysterious even at the end of the dialogue does not mean that there is no reason for his choice. Anselm insists that "in saying that he abandoned [justice] by willing what he ought not, I indicate clearly both why and how he abandoned it. He abandoned justice because he willed what he ought not to will, and he abandoned it by willing what he ought not to will." This latter will, however, was not "preceded by any other cause except that he was able to will." He willed it "simply because he willed it." Rogers sees this answer as fitting with a libertarian account of choice. But the teacher's answer only refers to "efficient causes" of creaturely acting, as Anselm himself is quick to say. Satan's choice can still be explained in ways that seem sufficient to answer the student's question: Anselm refers to reasons already given, and that includes the Augustinian and Genesis account of original sin, namely desiring, in an "excessive," "inordinate" or "perverted" way "to be like God" (sin of pride: Gen 3:5; Augustine, Confessiones II and De libero arbitrio III 25.74–76; DCD 4). Satan feels the "desire to exalt himself in an inappropriate imitation of God" (AoF, 48); the actual good he desires can remain undefined.⁶¹ Anselm's reference to earlier parts of the dialogue recalls the disorder between DB and DJ. Putting benefit above justice (or letting himself be driven by desires that seek benefit without seeking justice), Satan disregards the purpose of aseity as a tool for justice and true Happiness, seeking in its place unlimited power. He wants to do everything and anything a se.

There is a difference between the motivational explanation of evil choice and the ontological explanation of the possibility of (moral) evil. On a libertarian account of freedom, only the choice itself can cause evil; there are no preceding factors. Anselm refuses to follow Augustine in "ascribing explanatory power to the nothingness" God created the world from⁶² and accepts the theologically problematic notion that "there are events in the world that are not caused by God but by a created being" which thus mirrors God's aseity (AoF, 12). This may seem to be a mystery, especially against the backdrop of the modern scientific paradigm which aims for a complete explanation of the world and all events in it with natural causes alone; but, as Rogers says, it is a "mystery [that] ought to be expected" (AoF, 105), for it is the mystery of creative power.⁶³

This account of evil does not, however, affect my hypothesis that, on judgment day, a final cause for one's acting—the "sight" and effect of God and his love, or the "attracting" presence of Christ—does not prevent true freedom in the sense of PSR² as long as its efficient cause is the agent's will and the proper order of its desires.

5. One may still speak of a last "judgment" in the traditional sense of an eschatological PAP, heaven or hell, understood as the hypothetical possibility of ultimate separation from God which results when humans turn away from him and reject his love and communion with him in heaven. But in the light of the unlimited, unconditional love of God, this possibility, we may hope, can barely become reality. Karl Rahner chose a similar approach in his attempt to come to terms with the notion of hell: he didn't abandon the idea but said of hell that it was empty (Schumacher, 214). Or one could think of a combination of two theological models of judgment day: an open choice for some, who freely choose to be with God, while others may be compelled to choose God by his intervening grace if and when they fail to make this choice on their own. This hypothesis, which takes inspiration from Anselm's own example of the unwilling monk (CDH II 5), does not diminish the freedom of those who choose God, although unbeknownst to them, actions resulting from a different choice (i.e., to turn away from God for all eternity) couldn't become real.

Proponents of universalism could also give up the notion of any freedom on judgment day and refer to the irresistibility of grace. This doctrine and its various strands are controversial in the context of any reward model or "commercial theory" (L. Berkhof) of salvation precisely because it seems to diminish or abolish human freedom.⁶⁴ But theologians from Thomas Aquinas to Karl Barth are convinced that the eschatological moment makes it almost impossible to "escape God." We are constrained to accept this revelation, "constrained not by the compulsion of force but by the compelling power of divine love exerted in our favor. We could evade force but we cannot evade love" (Barth, 387).

However, the effects of God's restoring grace on judgment day can be understood in a slightly different and more intricate way which is in line with Anselm's PSR² and allows us to hold on to human freedom in what I would call the most decisive moment in a person's existence.

6. Anselm is concerned about grace without merit, honor, order, and justice (DC 3,9 et al.). This is also why, in CDH I 12, he rejects the notion of atonement sola misericordia, a mere intervention of God's grace, and argues for God's mediating (re-conciliating) action which—through incarnation, Christ's obedience unto death, resurrection and victory over death, devil and sin—has restored the human condition (II 21). But it is precisely this idea that comes into effect when we use his concept of freedom to speculate about judgment day. Recall that Anselm distinguishes (in DLA 3–4, cf. DCD 12) between the mere possession of a capacity (PSR²) and its use. The possibility of the latter depends, among others, on external conditions; it is, therefore, a "mediated" reality, but it is nevertheless "self-determined" or "self-initiated action" (Kane, Williams).

On closer inspection, and considering Anselm's analogy with eyesight (I may not be able to see a mountain due to darkness or clouds or other impediments, but I still retain the power of vision), the act of justification on judgment day brings us back to the cognitive issue of "clouded judgment" in human choices: the all-outshining presence of Christ/divine love is like the light that dissolves the clouds or "lifts the veil" (2 Cor 3:14) and enables humans again to exercise the power of seeing, or that proper freedom for the good whose active capacity was lost through sin. The "light of faith" is, Pope Francis' encyclical Lumen fidei (no. 34) reminds us, "incarnate in Christ," but it is not only present during our life; rather we may hope, with Paul, that its presence will shine and enlighten us even more on judgment day as well. Thus, what God does for man in his salvific grace is indeed a final "restoration of men to their original dignity" (CDH II 21), and moral goodness (justice) is not the condition for, but a possible consequence of salvation. The idea of universal reconciliation does not deprive humans of their last and greatest freedom but they are, on that "last day," finally enabled to use their PSR² (again) and say yes to God.

This is not forced on them in a way that would preclude freedom by Anselm's standard of self-initiated action; it does not fall under the conditions that limit free will: "compelled by external force" or "bribed by some extraneous reward" (DV 12). It remains a human choice which is but motivated by the presence of Christ. "Irresistibility of grace" is, as I see it, a figurative way of speaking; judgment day grace doesn't overpower man but empowers him (again) to do what he ought (was meant) to do. Anselm may take issue with Augustine's compatibilist belief that, although "grace doesn't destroy the human will," it still "changes . . . and . . . helps it" (De gratia et libero arbitrio 41; AoF, 34). But the only grace-effected "change" my account of JDF implies regards the external conditions under which the creature is asked to make his choice: the veil of ignorance regarding the consequences of one's choices has been lifted through Christ's presence. The choice itself is still up to the individual. It is at this moment of final reconciliation that God's grace and the freedom humans were given by God work hand in hand.

Besides, there is, as it were, a "reward" involved for this choice (i.e., salvation) but it is not "extraneous;" rather, it is the fulfillment of man's innermost being, purpose, and dignity. This activation of one's "inherent freedom" (Saarinen, 45) translates into a voluntary act that chooses the good because the rational creature finally sees clearly again the natural order and telos of things, and acts from this understanding.⁶⁵

7. We may even be able to take this a step further and thus address the lingering issue whether, and in what sense, the judgment day choice can be called free. In human matters, necessity (unless it is consequent necessity⁶⁶) seems to prevent freedom, but God always acts with necessity and is still called free. The ontology behind this is: "God exists a se, and therefore all of his choices are self-caused . . . he is free without having options" (FaS, 266; cf. DLA 14, DCD 1, CDH II 10). Necessity, according to Anselm's theo-ontology, doesn't preclude freedom if it is the inner necessity that describes the natural explication of a being's essence, or essential (self-)perfection (CDH II 5.10.17, DV 8, DLA 8, DCD 12, DC 1,2–3, DCV 19, P 7 et al.). Anselm is of course speaking about God who is necessarily, yet freely, always all that he is. The same idea, however, also applies, to some extent at least, to man's JDF.⁶⁷

It is here that the conflict of desires, whether on the same order (as in Kantian readings of Anselm) or across orders (as in Frankfurtian readings), finds its final solution, for the goals of both desires Anselm describes in DCD coincide in the kind of happiness that the third-order desire I have suggested aims at, namely, Happiness attained through justice. So, Happiness is thus, after all, not merely a "consequent goal" or an additional goal that would supersede justice, but it is fulfillment of the rightness (rectitudo) of the natural order and, thus, of Anselm's teleological theo-anthropology. All three desires of the will are now in harmonious order, and the created agent gains the complete happiness of "blessed immortality" (CDH Pr) through his choice. Man has not become god (he is still a creature); his freedom, however, at least in this one choice in which, directly and explicitly, he says yes to God, has been perfected in its proper essence which is divine or, as Anselm puts it, the "same for God and man" (DLA 1). PSR² is indeed, for creatures, a "potential perfection" (Goebel 2006, 117), but it can now become reality (cf. DLA 14). The kind of universalism which adopts Luther's reading of Rm 2:14ff specifies that JDF allows everyone to exercise it, even "under conditions of creaturely existence" (cf. Goebel 2006, 116), although not (anymore) under the purported necessity of making a "moral choice for God."

We saw earlier that Anselm holds there is only one definition of freedom and that he has theological reasons for rejecting Augustine's view of free choice. He thus arrives at his distinctive PSR²-definition of free will thinking of the divine being; it really is a theo-logical definition first, although it is then applied to created natures as well and "reduced" to moral choices (a moral PAP). The human choice, however, which I have described as JDF fulfills and perfects freedom in the sense of virtuous self-determination.

This also means that JDF may not require open options and thus fulfills the PSR²-definition without the need for an additional PAP. We have found a case which truly brings the two notions of freedom together: it is the "full-fledged freedom of choice [which] is the power of self-initiated action for some good end" (Williams/Visser, 8). It is, I think, this "full-fledged freedom" we are talking about when we reflect on judgment day following the theo-logic of unconditional love and universal salvation. Rogers no doubt succeeds in showing that Anselm sees a logical connection between PSR² and PAP; there is no inconsistency or undecidedness in his thinking, as if Anselm might "sometimes admit a freedom to choose between alternative possibilities but still comes back to the view that freedom is not being able to do other than as one does because freedom is doing what one ought to do for the sake of salvation (and loss of freedom lies in doing what one ought not to do)" (Luscombe 2006). It is nevertheless true that this latter view constitutes freedom at its fullest (and good people have it "inseparably after their death:" DLA 14), but Anselm continues also to identify it with the moral choice rational creatures are expected to make during their lifetime.

8. It should be abundantly clear by now: a possible objection that the above account of JDF does not fall under the category of moral choice strictly speaking is moot since that is exactly the point I am making here. Universalism reminds us that we should keep separate what Anselm (and others) merge together, often confusing the two: morality and eschatology. Created beings that have freedom participate in a divine quality; God is the epitome of freedom, but his freedom is not "morally significant freedom" (AoF, 14f.). PSR² is, as such, not a definition of morally significant freedom, and JDF is not a moral choice that has eschatological relevance. It does not demand an implicit, indirect choice of God and one's own happiness in a state of "obstructed sight," under the conditions of ambivalence, imperfection and cognitive deficiency. Rather, the matter at hand now is a direct choice of God, and its consequences are clear.

There is no torn condition anymore, no conflict of desires, no choice between selfish benefit and justly sought benefit. Or more precisely, a hypothetical PAP is, in fact, still conceivable but only on account of a radical voluntarism-turned-atheism which a pre-modern thinker would hardly have thought of: a "Nietzschean atheist," who turns Anselm's desire "to be like God" into the desire "to be God" and in his "will to power" cannot "bear not to be a god" if "there were any gods" (Zarathustra II 24), may consciously choose to turn away from God even when he sees God before his eyes and realizes that he will miss out on his higher fulfillment as a metaphysical being.⁶⁸ On principle, however, the choice to perfect and fulfill one's very nature doesn't require the polarized thinking of moral and metaphysical alternatives (i.e., an "either/or" between heaven and hell). JDF allows us to simply act on our DH.

The example of the Nietzschean atheist also shows that even though JDF may be motivated by one sole desire (although DH, not either DB or DJ), it is not a fully determined choice and unfree in the moral sense Anselm and Rogers discuss with created natures in mind (i.e., contrary to what Anselm says in DCD about the hypothetical case of someone endowed only with DB or DJ ⁶⁹). For the agent acts "on one God-given motivation," but he still has to give his rational assent and "will it through" to the end. The act of pervelle doesn't require a torn condition per se. Even the non-moral, metaphysical self-choice of JDF may thus, in a way, fulfill the criteria of libertarian freedom and Anselmian aseity, that is, the "choice originates in the agent" and he "may have done otherwise" (AoF, 60). The question simply is what this choice consists in; can't it just be a Stoic willing of something that would happen anyways, a conscious "let-it-happen" or "per-will it"?

This calls into question a common form of dichotomy between (indeterminist) voluntarism and (determinist) intellectualism in two ways: (1) What is good may have been "decided by God" (Ekenberg 2016, 70), but the acceptance of Anselm's value framework itself (i.e., the Christian theology-morality context discussed in part 1) can be seen as a truly free decision of the will. (2) The truth, once intellectually recognized (through reason and judgment), still requires some act of the will, a personal assent, to make it effective in one's choices. This may take the form of "belief" which is "to think with assent."⁷⁰ The idea seems unfounded that voluntarism needs to be anti-intellectualist if reason and will are oriented towards the truth; that, to be free and a "robust power and independent faculty," the will needs to be able to go against "reason's dictates" (ibid, 60⁷¹), especially if it is DH that motivates the choice. Reason here serves a desire; this desire, however, is not opposed to reason and its judgment that doing x is good and therefore ought to be done. Reason informed by DH recognizes choices that are not only just but also beneficial (having established the proper order between DB and DJ) and are therefore not opposed to choices that would bring mere benefit. I can only act in spite of this insight if I act "in spite of myself."

The way in which Rogers uses pervelle in her account of Anselm's moral philosophy can actually support the contrary argument I am making here: according to Rogers, Anselm's view of moral choice is libertarian because the agent causes an a se choice, which is "absolutely up to him," by "per-willing one God-given desire over another." She takes it to mean "to will through to the point of intention" (FaS, 32). Is this not what Frankfurt sees as turning desire into volition, making it effective in determining one's actions? Rogers explicitly rejects Frankfurt's other claim that freedom simply consists in "harmony between first- and second-order desires" (90) because for her, libertarian freedom requires a PAP⁷²: "this per-willing occurs in a situation in which the agent is torn between genuinely open alternative possibilities" and desires two incompatible moral "objects" so that the per-willing of one entails overriding the desire for the other. Rogers' main point, however, is that aseity requires no effective principle other than the acting person herself who "causes the non-determined choice . . . simply by per-willing (following through with), and that means without the introduction of any new sorts of causes" (32) and without any "additional causal power beyond the motivating factors preceding the choice" (96). Rogers' critique of Frankfurt may be justified with moral choices in mind, but JDF, not being a moral choice, may not require a PAP. The freedom of rational creatures is mainly a matter of judgment regarding God's natural order of things. Under the conditions of JDF, Anselm's PSR² can fully coincide with Frankfurt's non-binary self-identification. It is now possible simply as an act of rational assent to something that is recognized as self-fulfilling. It is a matter of self-identification with (and acting on) DH. No other "causal power" is required.

In short, the human being freely perfects the inner necessity that lies in his nature and seeks actualization. A choice that fulfills one's purpose is, in some sense, both necessary and free if freedom consists in the power to "become what one is." For humans this means that they let DJ determine and control DB and thus fulfill DH. This is PSR², and the conditions of JDF enable humans to practice it (again) and preserve justice, the rectitude of being, without the need for a moral PAP and indirect choice for or against God and self-fulfillment. According to Anselm, the main "point of freedom is to choose justice" (AoF, 61), and that is indeed, as I have tried to argue, starting from Paul's and Luther's view of divine judgment, what humans are asked to do and able to do on judgment day.

To be sure, created beings will never have absolute freedom; they are "thrown into being" (M. Heidegger) and have received, from God or nature, all the tools they use in making their choices (i.e., the faculties of reason and will) (DCD 1 et al.). The libertarianism Anselm shares with Rogers is still different from any view that maintains human responsibility but sees God's grace at work not only in providing those tools but also in the act of choosing good over bad.⁷³ Anselm and Rogers hold that the choice itself offers some small space for creaturely freedom⁷⁴ and that this account applies to the various choices mentioned in this paper, namely, the choice that Adam makes for all humankind (which is reversed in Christ) and the individual's self-determination of his moral character. This is the realm of moral choice which is characterized by ambivalence and which also yields ambivalent results. Some make evil choices, others good ones.⁷⁵ I assumed the same kind of freedom for the "final choice" an individual has to make (JDF). Its additional, external condition (God's direct presence) does not change or limit human aseity. It is simply God's way of restoring, for the individual, the original capacity for goodness and Happiness that he had endowed humans with in the first place.

In order to further unpack the claim that this is true freedom, I qualify my above statement that self-choices are both necessary and free. The apparent contradiction may dissolve if we compare human and divine freedom again. The necessity of those choices is not a mere imperative that lacks logical necessity; it is not just a necessary condition if one wants to fulfill one's being. If the mind is, as Anselm says, a "mirror of God," we may apply what he says about creation to the sphere of human acting. God does not "have to" create but he nevertheless did so simply because his nature is love (cf. CDH I 18) and he always acts in accordance with his being. In a similar way, humans may not, strictly speaking, "have to" use their a se powers to choose, on judgment day, God and good (hypothetical PAP), but we may hope that they nevertheless will do; for they were "made to seek and love God," and are then able to see and live out their proper nature again (the hypothetical PAP is thus suspended).

2.3 Conclusion

I have stressed the importance of looking beyond the moral context Anselm himself focusses on in DLA and DCD and apply his distinctive PSR²-definition of freedom to the eschatological moment of eternal self-choice (JDF). Anselm helps us understand that it is not the potential choice to "turn away" from God that constitutes an integral part of the highest degree of human freedom; rather, it is the restored ability to turn towards God that fulfills the theo-teleology of the human nature and its freedom.

Note that my proposal does not touch on the more general questions of theodicy Anselm addresses: Why were humans not directly saved? Why is there all the hideous evil in the world? And I do not call into question the significance of human freedom in moral choices and its role in causing moral evil which may indeed be seen as the "price for the great good that is freedom" (AoF, 91).⁷⁶ The idea of universal salvation simply offers an alternative to the rather pessimistic thought that the evil humans cause in this world should continue in the other world and that the realm of divine eternity is equally divided into a dualism of good (heaven) and bad (hell). Universalism certainly has numerous practical—ethical and theological—consequences, and here is not the place to discuss them further. Suffice it to say that universalism may require Christians to be "over-human" and deal with something that seems outright "counterintuitive,"⁷⁷ namely, that we may have to separate morality from eschatology and dissolve the Anselmian nexus between eternal happiness and merit. Radical universalism does away with the idea of merit (be it moral performance or mere faith) as a relevant factor in God's judgment (be it seen in the traditional way or with Paul and Luther). Yet, everything the all-loving God of the Christian Gospels does is "paradoxical" (G. Theissen) and may therefore seem "counterintuitive" but only, I must stress, by all-too-human standards, if, that is, human definitions of retributive justice, fairness and morality are used to do theology and describe the divine judgment. There is, in fact, a theo-logic to the paradoxical unity of love and omnipotence, of highest moral standards and absolute forgiveness, which characterizes God's being and his attitude towards mankind. The gospel of love is not illogical, and its logic is not only accessible to the human heart but also, by and large, to reason. It is to the credit of Anselm that he understood this well.⁷⁸