What Lucifer Wanted: Anselm, Aquinas, and Scotus on the Object of the First Evil Choice
What Lucifer Wanted: Anselm, Aquinas, and Scotus on the Object of the First Evil Choice
Abstract and Keywords
This paper discusses the views of three medieval thinkers—Anselm, Thomas Aquinas, and John Duns Scotus—about a specific aspect of the problem of evil, which can be dubbed ‘the Lucifer problem’. What was the object of the first evil choice? What could entice a perfectly rational agent placed in ideal circumstances into doing evil? Those thinkers agreed that Lucifer wanted to be happier, but while Anselm thought that that was something Lucifer could achieve by his natural powers, Aquinas held that it was not naturally possible for Lucifer to be happier, even though it was something he could obtain supernaturally. By contrast, Scotus posited that what Lucifer wanted was beyond what was logically possible, i.e. to be as happy as God (or to be God’s equal). An interesting consequence of Scotus’s hypothesis is that God could have done nothing to make Lucifer’s evil choice less likely.
1. The Lucifer problem
Suppose that God created the world entirely good and that evil made its first appearance through the free choice of one of God’s creatures—say, Lucifer.1 These two claims, which constitute the core of a long-standing way to explain why there is evil in a God-created world, have sometimes been thought to be inconsistent. Here is how a prominent critic of this tradition, John Hick, describes the problem:
The basic and inevitable criticism is that the idea of an unqualifiedly good creature committing sin is self-contradictory and unintelligible. If the angels are finitely perfect, then even though they are in some important sense free to sin they will never in fact do so. If they do sin we can only infer that they were not flawless—in which case their Maker must share the responsibility for their fall and the intended theodicy fails.2
(p.62) The claim is that, even though it is possible for an entirely or unqualifiedly good agent to be able to make an evil choice, actually making an evil choice is incompatible with being entirely or unqualifiedly good. The tacit piece of reasoning behind this claim is presumably that rational agents make evil choices only if they have a motive; but agents have a motive to do evil only if they are not entirely or unqualifiedly good, for it is assumed that to be motivated to do evil and to be entirely or unqualifiedly good are incompatible features. So if God created Lucifer entirely good, Lucifer had no motive to do evil and thus was not the sort of agent who would make an evil choice. By contrast, if Lucifer did make an evil choice, then he had a motive to do evil and thus it is not the case that God created Lucifer entirely good.
Striking as this argument is, one may object to the key assumption that to be motivated to do evil and to be entirely good are incompatible features. For it is not uncommon for morally good agents both to be motivated to make evil choices and actually to make evil choices. For example, they may not know that their choices are evil or they may believe that they are going to get something good out of what they would otherwise not choose. This objection, however, misses the point. For moral goodness is only one component of Lucifer’s overall goodness. The claim that God created Lucifer entirely good can be plausibly taken to entail that Lucifer was not only morally good, but also intellectually flawless and supremely happy, i.e. as happy as he could be. Because Lucifer was supremely happy, he had no reason to change his condition. Because Lucifer was intellectually flawless, he had no wrong belief about what was good for him—specifically, he did not have the wrong belief that he could be happier than he was. Therefore, Lucifer knew that he had no reason to change his condition. But Lucifer did choose to change his condition in some respect (e.g. by getting something that he did not have). Accordingly, his choice appears to be irrational. Now only agents who are flawed in some respect make irrational choices. Therefore, if Lucifer’s choice was irrational, Lucifer was not created entirely good and his choice was not the first occurrence of evil, contrary to what has been assumed.
So Hick’s criticism turns out to be potentially devastating for traditional attempts to explain the presence of evil in the world as the result of a created agent’s free choice. However, it is not accurate to claim, as Hick did immediately before the passage I have quoted, that the belief that evil was introduced by the fall of the angels was historically successful only because the thinkers who embraced it were “content to refrain from examining it.”3 In fact, that belief has been subjected to intense scrutiny at least from Augustine (p.63) onwards. Several medieval thinkers tried to solve the difficulty Hick mentions by considering what the object or goal of the first evil choice might have been. What did Lucifer want (and presumably fail) to achieve? Their hope was to identify an object such that a flawless agent might have plausibly and rationally wanted to attain it through a choice that should nevertheless be described as evil.
In the growing literature about medieval views on Lucifer’s fall, surprisingly little attention has been paid to the object of his choice. I have decided to focus on the views of three great medieval thinkers, i.e. Anselm, Thomas Aquinas, and John Duns Scotus, in order to identify some of the basic moves in this debate. As will emerge from what follows, there is a major divide between Anselm and Aquinas, on the one hand, and, on the other hand, Scotus. While both Anselm and Aquinas held that the first occurrence of evil could be explained only if God did not create Lucifer as supremely happy, Scotus held a view that did not commit him to that claim.
Those three authors’ treatments of Lucifer’s choice turn out to be of special philosophical interest in two different areas. First, from the point of view of moral psychology, Anselm, Aquinas, and Scotus inquired whether flawless rational agents can make evil choices without giving up their rationality. This is arguably one of the deepest problems concerning evildoing. Second, from the point of view of the philosophy of religion, those thinkers inquired whether God could have created the world in such a way that an evil choice, even though possible, would have been less attractive than it actually was—perhaps so unattractive that no rational agent would have actually made it. This is arguably one of the most challenging questions that a theodicy based on the notion of free will has to answer.
In what follows, I first consider Anselm’s claim that Lucifer’s choice did not exceed the limits of what was naturally possible for him. According to Anselm, Lucifer chose to obtain something that he could have obtained through the exercise of his natural capacities. Then, I turn to Aquinas’s view that Lucifer chose something beyond what was naturally possible for him, i.e. something that he could have obtained only through God’s supernatural intervention. Finally, I present Scotus’s view that the object of the first evil choice might have exceeded the limits of what is logically possible. Specifically, Lucifer may have wished to be God’s equal. In some way, it would be hard to think of an act of rebellion more radical than a wish not for an alternative state of affairs but for something literally inconceivable and unactualizable. Scotus’s suggestion is that, in a completely good world created by a supremely good God, evil could not have occurred unless one had considered what lay beyond the realm of what is logically possible.
In a short and dense exchange between student and teacher in Anselm’s dialogue On the fall of the devil, the student rejects the teacher’s suggestion that what Lucifer wanted was to be “inordinately like God.”4 The student’s argument is based on the premise that God is such that nothing like him can be thought. In other words, God is necessarily unique. The student further assumes that only somebody slow-witted could know what God is without knowing that he is necessarily unique. Lucifer, however, was not slow-witted (ita obtunsae mentis). Thus, since Lucifer must have known what God is (something that even an unbeliever knows, according to Anselm), he must also have known that God is necessarily unique. But a choice to be like God presupposes the belief that it is possible to be like God. And having that belief in turn entails having the belief that God is not necessarily unique. Therefore, Lucifer could not have wanted to be like God.5
The teacher accepts the student’s argument. He explains that the claim that Lucifer wanted to be like God should not be interpreted as meaning that Lucifer wanted to be “completely equal to God.” Rather, Lucifer wanted something that God did not want him to want. Accordingly, Lucifer wanted something by his own will, without subjecting his will to God’s will—he wanted what he wanted independently of whether God approved of his willing it or not. Now the teacher goes on to notice that it is God’s prerogative to will something by his own will without subjecting his will to a superior will, for no will is superior to God’s will. The implicit conclusion is that Lucifer wanted to be like God only in a qualified sense—not because the object of his pursuit was to be like God but because, by willing what he willed, he did not subject his will to God’s will. But this qualified equality with God was not the target of Lucifer’s choice. Rather, it was a by-product of his willing something that God did not want him to will.6
(p.65) What, then, did Lucifer want? The teacher admits, somewhat disappointingly, that he does not know. But teacher and student agree that the mysterious object of Lucifer’s choice must have satisfied four requirements. First, it must have been something advantageous for Lucifer, because it is assumed that rational agents only want what they believe is either just (i.e. something that God wants them to will) or advantageous for them. But it is impossible to make an evil choice by willing something just, because evil is nothing else than injustice, i.e. willing something against God’s will. Therefore, Lucifer must have willed something advantageous. Second, it must have been something that Lucifer was able to attain, presumably because rational agents do not intend to achieve something they know they cannot get. Third, it must have been something that Lucifer did not receive when he was created, because he had received all he had from God and so he could not have made an evil choice by willing what he already had. Fourth and finally, the object of Lucifer’s choice must have been something that God would have given to Lucifer if Lucifer had refrained from willing it.7 Accordingly, in the rest of the dialogue, Anselm refers to the object of the first evil choice merely as “something more (illud plus) that God did not want to give them [i.e. the angels] yet.”8
So Anselm can account for the possibility of the first evil choice only by positing that God did not create Lucifer and the other angels supremely happy. The reason Anselm gives for God’s decision is that he wanted the angels to deserve supreme happiness by their own merit, i.e. by not willing what God wanted them not to will. But Lucifer chose to be happier. Even though that choice was rational, it was nevertheless evil, because it was in conflict with God’s will.
There are two problematic aspects in Anselm’s view. First, there is something troubling in the claim that God decided to create his creatures less happy than they could naturally be. According to Anselm, God set a limit to the happiness his creatures enjoyed at the moment of their creation in order to allow them to prove themselves. Anselm’s God is a tempting God, who put his creatures to a test. Second, the claim that Lucifer’s choice to be happier was entirely natural but nonetheless evil is also worrisome. The reason why Lucifer’s choice was evil is not that it did not fit with Lucifer’s nature or needs but only that God did not want Lucifer to decide to be happier at that moment. Even though the responsibility for making an evil choice remains with Lucifer, that choice perfectly matched Lucifer’s nature. Not only did God put his creatures to a test, he also did nothing to make the test easier.
Aquinas agrees with Anselm’s two characteristic claims that Lucifer could not have willed to be unqualifiedly like God and that Lucifer and the other angels were not created supremely happy. But he gives new arguments in support of these claims. He also corrects Anselm in an important respect. According to Aquinas, it is not the case that Lucifer wanted something that he could have attained through the exercise of his natural capacities.9
Aquinas has two arguments to reject the view that Lucifer wanted to achieve equality with God “in an unqualified sense” or “absolutely.” Both arguments are grounded on Lucifer’s rationality.
First, Aquinas argues that for something to be equal to God involves a contradiction. Aquinas’s demonstration can be reconstructed in this way. Suppose that it is possible for something to be equal to God. Then it is possible for the type “God” to be instantiated (at least) twice. But this is logically impossible. Aquinas’s demonstration of the latter point is based on his characteristic view that God is subsisting existence itself (ipsum esse subsistens) and on the impossibility of there being two subsisting existences.10 Aquinas further argues that Lucifer cannot have been ignorant of this, because of his purely intellectual nature and flawless condition before the fall. But one may assume that no rational agent tries to achieve what she knows to entail something impossible. Therefore, Lucifer could not have willed to be equal to God absolutely, i.e. to be a god.11
Second, Aquinas argues that, even if it were possible for somebody to become equal to God in an unqualified sense, that agent would lose her nature, i.e. she would not be the same sort of thing and, consequently, she would not be the same individual. But any agent desires her own good and does not care for the good that would happen to her (so to speak) if she were to become a different individual. Again, because of his intellectual (p.67) nature, Lucifer could not have missed this point. Therefore, Lucifer could not have willed to be equal to God in an unqualified sense.12
By a similar argument, Aquinas also concludes that Lucifer could not have desired not to be subject to God, because Lucifer knew that any creature is ontologically dependent on God and as a consequence a creature’s not being subject to God entails its non-existence. But Lucifer could not will not to exist, because anybody who wills something wills something good for herself, and the satisfaction of that volition presupposes her existence.13
In these arguments, Aquinas presents Lucifer as a good metaphysician, well aware of subtle points concerning both his own identity and God’s nature.
So what about the object of Lucifer’s choice? It is here that Aquinas goes beyond Anselm. First, Aquinas claims that angels lacked nothing, for they were created perfect. In other words, Aquinas embraces the claim that Lucifer and the other angels were supremely happy at the moment of their creation. Second, however, Aquinas adds a qualification that puts him squarely back into Anselm’s fold. For he distinguishes between the natural and supernatural orders. Even though the angels were perfect and had no potentiality with regard to the order of nature, they did have a potentiality for a supernatural good, which they could obtain only through God’s aid. So Lucifer had no motive to make an evil choice concerning what he could achieve by his natural capacities, as he was perfectly happy in that respect. But Lucifer could have made an evil choice concerning something beyond what he could obtain by the exercise of his natural capacities, because in that respect he was not supremely happy.14
Even though it may look like an ad hoc move, the distinction between natural and supernatural orders and the corresponding distinction between a created agent’s natural and supernatural goals are actually key aspects of Aquinas’s thought. Suffice it to say that all rational creatures can reach a state of happiness by the exercise of their natural capacities. In addition to that, however, they also have a potentiality to experience a special kind of (p.68) intimacy with God, called the “beatific vision.” Even though rational creatures have a potentiality for this supernatural experience, at the enjoyment of which all their actions are ultimately directed, they cannot actualize that potentiality by the exercise of their natural capacities. Their inability depends not on God’s arbitrary decision but on the infinite gap between God and his creatures.15
Since ultimate happiness, even in the natural order, involves knowledge of God, at the moment of his creation an angel could achieve as much knowledge of God as he was naturally capable of. Specifically, he could know God by knowing himself as a faithful image of God.16 But such knowledge of God is not the union and intimacy with God that is reached in the beatific vision. Such a union and intimacy with God can only be granted by God’s grace. Although creatures can enjoy that union when God’s grace is given, they cannot achieve that state by the exercise of their natural capacities. So the beatific vision is something possible for rational creatures but beyond what is naturally possible for them.
Now it is Aquinas’s contention that angels could make an evil choice just because they were not blessed, for rational agents can make evil choices only in the pursuit of some good they do not yet have. The object of Lucifer’s evil choice was precisely a missing good, i.e. the beatific vision. Lucifer simply wanted to be intimate with God, which he wasn’t yet. The problem was not what Lucifer wanted, which was in itself good—actually, the supreme good a creature could enjoy. Rather, the problem was the way Lucifer wanted what he wanted. He wanted to achieve beatitude without God’s help, i.e. not by grace. Thus, he wanted to see God by the exercise of his own natural capacities. Lucifer, however, never rebelled against his ontological dependence on God. Being a good metaphysician, he recognized that it would be impossible for him to exist if he did not depend on God:
Therefore, the devil’s first sin was that, to attain the supernatural happiness consisting of the complete vision of God, he did not elevate himself to God so as to desire with holy angels his ultimate perfection through God’s grace. Rather, he wanted to attain his ultimate perfection by the power of his own nature without God bestowing grace, although not without God acting on his nature.17 (Trans. Regan, 456)
(p.69) I call this explanation of the origin of evil the “delayed beatitude model.” An evil choice was a viable option for Lucifer only because his supernatural happiness (i.e. his union with God in the beatific vision) was delayed to a moment posterior to that of his creation.
It is worth stressing that Aquinas thought that the first occurrence of evil could not be explained by appealing to anything within the natural order. Rather, the appearance of evil can be explained only by considering the supernatural goal a rational creature can enjoy at God’s discretion. In this way, Aquinas manages to preserve God’s goodness. Unlike what is found in Anselm, God created Lucifer supremely happy, even though only with regard to the order of nature. Aquinas also manages to preserve Lucifer’s rationality. Similarly to what Anselm had posited, he holds that Lucifer chose more happiness, i.e. supernatural happiness.
It is also worth stressing that, for Aquinas, God could not have created the world in such a way that Lucifer could have achieved the union with God by nature and not by grace. In other words, God did not arbitrarily subtract from the happiness that Lucifer could have naturally enjoyed. What Aquinas calls a divine rule, i.e. that Lucifer could achieve supernatural happiness only by grace and not by his own forces, is due to the unavoidable gap between God and the world, not to God’s arbitrary decision. It is God’s prerogative to enjoy beatitude naturally. All rational creatures need God’s grace to reach that state, just because they are creatures.18
Since Lucifer wanted to achieve the union with God by his own forces, one can say that he wanted his own good inordinately and immoderately. This means that Lucifer wanted his own good in a way that was not commensurate to the limits of his nature.19 Lucifer’s faux pas was that, when he made his choice, he failed to consider the limits of his nature and his place in the world. According to Aquinas, this failure was cognitive but was not a genuine mistake. Lucifer did not entertain a wrong belief about his nature and his place in the world. Specifically, he did not think that he was able to be united with God without grace. Rather, Lucifer failed to consider that his nature was such that he could not be united with God without grace.20 Thus, his cognitive failure, which preceded his evil choice, (p.70) was not that he had a wrong occurrent belief but that he failed to have a true occurrent belief about the limits of his nature and capacities.21
Aquinas is adamant that Lucifer’s cognitive failure was not the sign of a natural flaw. By the exercise of his natural capacities, Lucifer could have had the right belief at the right time, i.e. he could have realized that he was unable to reach the union with God by the exercise of his natural capacities. That he did not form that belief was not in itself a fault. But it created the condition for a fault, namely the choice to pursue supernatural happiness by his own forces.22
In his account of Lucifer’s choice, Aquinas avoids commitment to the two questionable points I noted in Anselm. First, God sets no arbitrary limit to the natural happiness enjoyed by his creatures. That the enjoyment of supernatural happiness is beyond what rational creatures can reach by their natural capacities does not depend on God’s decision but on the gap between God and his creatures. Second, Lucifer’s fault was not that he chose to be happier against God’s will but that he did so without considering the limits of his nature (which even God could not have eliminated). That choice was an act of pride.
There is, however, a problematic point in Aquinas’s account as well. It is true that God could not have created the world in such a way that an evil choice would have been impossible, because the gap between God and his creatures, on which the difference between natural and supernatural happiness depends, is a necessary characteristic of the world. Nevertheless, it seems that God could have created the world in such a way that an evil choice, although possible, would have never been made by a rational agent. For God could have made Lucifer enjoy the beatific vision from the very moment of his creation. If Lucifer had been supernaturally happy at the moment of his creation, he would have had no motive to make an evil choice. Lucifer’s evil choice is explicable only if his beatitude was delayed. But God decided not to create Lucifer in a state of beatitude. So Lucifer could rationally make the evil choice to attain beatitude by his own forces. He did so, and evil entered the world. Aquinas justifies God’s decision by stressing that beatitude was not owed to the angels from the moment of their creation, because it was not part of their natures but the goal at which their natures were aimed.23 But the point of this claim is obscure. First, Aquinas himself admits that the beatific vision is the goal at which any rational creature’s nature is ultimately directed. Undeniably, Lucifer (p.71) was created naturally happy, but it is as if the rules of the game had been changed at the moment of his creation and a new goal was set for him—a goal distinct from that which he was naturally equipped to reach. Second, what is at issue is not so much what God, once he decided to create the world, owed to his creatures in order to make them naturally happy. Rather, what is at issue is whether God could have made things such that an evil choice would not have been made, independently of what he owed to his creatures out of strict justice. It is at best debatable whether Aquinas’s remark that God owed nothing to his creatures does anything to justify the creation of a world where it is not just possible but even likely for a rational agent to make an evil choice. There is something disturbing in the thought that the world could have been such that no rational creature in their right mind would have made an evil choice, because they would have had no motive to do so, but God decided not to create the world in that way.
4. Scotus: a wish for the impossible
Scotus’s account of Lucifer’s choice is characterized by great caution. Without committing himself to just one explanation of what Lucifer actually chose, Scotus nevertheless defends the view that the object of Lucifer’s choice might have been to be God’s equal. He defends this claim by arguing that the first occurrence of evil might have been a wish for something logically impossible. In that way, he sets the discussion over the first occurrence of evil on new grounds.
Similarly to what Anselm and Aquinas had held, Scotus thinks that the general object of Lucifer’s choice was his own happiness.24 Like Aquinas, Scotus thinks that Lucifer’s choice was directed at his own supreme happiness. And again like Aquinas, Scotus holds that there is nothing evil (p.72) in the choice to be supremely happy. Such a choice was evil only because it was “immoderate,” i.e. unchecked by the power Scotus attributes to the will to direct itself towards what it ought to will according to one’s own nature.25
In his Parisian lectures on the Sentences, Scotus lists four possible ways Lucifer’s choice may have been immoderate. First, Lucifer may have willed his happiness more intensely than it was fitting. In turn, this may have occurred in two ways, either by loving happiness for oneself more than for God (i.e. by willing to be happier than God) or by loving God as an object of use, i.e. because the union with God would make one happy, rather than as an object of fruition, i.e. because God is good in himself. Second, Lucifer may have willed to be happy more than it was fitting for his own nature. Third, Lucifer may have willed to achieve a state of supreme happiness earlier than he ought to do. Fourth, Lucifer may have willed to achieve a state of supreme happiness without having to deserve it.26
Of the four ways listed by Scotus, the third and the fourth presuppose the delayed beatitude model. If Lucifer wanted to be supremely happy too early, then he was not supremely happy at the moment of his choice. Similarly, if his choice was to achieve supreme happiness without deserving it, the assumption is that God wanted Lucifer to receive supreme happiness as a reward for his merits and so did not give it to him at the moment of his creation.27 The first and the second way Lucifer’s choice may have been (p.73) immoderate, however, do not presuppose the delayed beatitude model. Suppose that Lucifer had enjoyed supernatural happiness since the moment of his creation. He could still have desired to be happier than God or have loved God in a utilitarian way. Similarly and even more interestingly, even if Lucifer had been as happy as he could be, he might have desired to be even happier, i.e. happier than it was fitting for his nature. In particular, Lucifer may have wished to be as happy as only God can be.28 In those scenarios, God’s decision to delay or not delay Lucifer’s beatitude is irrelevant to explain his evil choice.
The possibility that Lucifer may have willed to be as happy as only God can be, which Scotus mentions as the second possible explanation of Lucifer’s choice in his Reportatio, is particularly interesting, for it tallies with one of the most original aspects of Scotus’s account of Lucifer’s choice, namely his contention that Lucifer could have wanted to be God’s equal, in direct opposition to what both Anselm and Aquinas had claimed.29
In all three versions of his commentary on the Sentences, Scotus devotes an entire question to argue that Lucifer could have wanted to be God’s equal. Even though the basic elements of Scotus’s position remain the same, it is possible to detect some evolution in the relative weight Scotus gives to one of two strategies he uses to defend his position. Scotus consistently holds that neither strategy, if taken by itself, is sufficient to explain every aspect of Lucifer’s choice. But he changes his mind about which of the two strategies is fundamental and which adds some necessary but less important details to his solution.30
Scotus’s first strategy is to argue that, if an agent wills to do something or to be in a certain state, it is not necessarily the case that that agent has a belief that she can actually do that thing or be in that state. Specifically, Lucifer may have been willing to be God’s equal without having the false belief that he could be God’s equal. Scotus’s point is to show that Lucifer’s evil choice does not entail any cognitive failure on his part.
(p.74) Scotus’s argument is based on the traditional distinction between the so-called “love of friendship” (amor amicitiae) and “love of desire” (amor concupiscentiae). In Scotus’s rendition, these two kinds of loves are two kinds of acts of the will. The act of loving something by love of friendship involves a two-place relation between the lover and the object of love. For example, parents may love their children by love of friendship. By contrast, the act of loving something by love of desire involves a three-place relation between the lover, the first object of love (quod), and the second object of love (cui), i.e. that for the sake of which the first object is loved or desired. For example, a parent may desire a good education (first object) for her children (second object).
Now Scotus holds that anything whatsoever can be the object of either act of the will as long as that thing is apprehended as good. This has two interesting consequences. First, somebody can be both the subject and the object of both acts of the will. For example, I can love myself by love of friendship and I can desire for myself something by love of desire. Second, in the case of an act of love of desire, no fit is required between the object of desire and that for the sake of which that thing is desired. The intellect must apprehend each of the two objects as good before the will connects those two objects by an act of love of desire, but it does not have to judge that there is a match between those two objects. Rather, it is the will’s job to connect those two objects. The intellect does not have to follow suit. For example, if a parent loves her children by love of friendship and considers living forever a good thing, she may desire that her children live forever even though she does not have the false belief that living forever fits her children’s nature.
Scotus uses this model to explain how Lucifer could have made a choice for the impossible, i.e. to be God’s equal, without making any cognitive error. For Lucifer correctly cognized both himself and equality with God as good things. Accordingly, he could desire equality with God for himself. Since an act of love of desire does not require a corresponding complex act of the intellect judging that the two objects match each other, Lucifer did not have to form the belief that he could be God’s equal. Only if Lucifer had formed that belief would he have made a mistake, because, while it is true that equality with God is something good and so a possible object of desire for a creature, the belief that a creature can be God’s equal entails a contradiction.31
(p.75) Scotus’s second strategy is to argue that for an agent to will to do something or to be in a certain state is actually compatible with that agent’s having the belief that she cannot do that thing or be in that state. Specifically, Lucifer may have been willing to be God’s equal while being fully aware that that was impossible. Scotus’s argument is based on the distinction between two meanings of “choice” (electio). Let us call them “choice1” and “choice2”. A choice1 is any act of the will that follows the intellect’s act of full apprehension, i.e. an act of the will that is carried out neither in a state of ignorance nor in a state of emotional perturbation. A choice2 is a choice to do something, or, as Scotus says, an efficacious choice, i.e. an act of the will that follows the conclusion of a practical syllogism and as such is turned towards doing something. Even though Scotus does not say so explicitly, it seems that any choice2 is also a choice1. But Scotus is adamant that the reverse is not true. Not every choice that presupposes full apprehension is also a choice to do something. A choice1 may be a mere wish.
Since practical syllogisms and deliberations are about things that agents believe to be able to achieve, only things that agents think are possible for them can be the objects of a choice2. By contrast, the object of a choice1 can be beyond the limits of what agents think they can attain. The object of a choice1 can even be logically impossible. To have a wish for the impossible poses no threat to one’s rationality as long as one judges the object of that wish to be impossible and as long as one does not decide to obtain it by a choice2.32
Scotus calls a choice1 that is not a choice2 a “wish” (velleitas) or “conditional volition” (velle sub condicione) as opposed to a genuine choice or simple volition. The idea is that wishes should be interpreted as the consequents of conditionals of the form “If such-and-such were the case, I would want for x to occur,” where the antecedent of that conditional may be either counter-factual or counter-possible. It is the nature of wishes that they can range not just beyond what is actual but even beyond what is logically possible.33
According to Scotus, wishes can be morally relevant. A mere wish can even be deemed worthy of the worst punishment or the best reward. In order for a volition to be morally good, it is both necessary and sufficient (p.76) that it is accompanied by the right circumstances, i.e. that it occurs to whom, when and how it ought to occur. But it is not necessary for that volition to be efficacious.34 For example, my wish that all wars be ended is not a choice to do something but is nevertheless morally good (and perfectly rational).
Scotus applies this model to Lucifer’s case. It would have been irrational of Lucifer to choose to be God’s equal by a choice2, i.e. to choose to become a god. But no threat is posed to Lucifer’s rationality if he merely wished to be God’s equal with full knowledge that to be God’s equal is something good but impossible for any creature. But even though it involved no irrationality on Lucifer’s part, that wish was nonetheless evil. Lucifer loved himself by love of friendship so much that he wished to be God’s equal, even though such a love should be reserved only to God:
With regard to the act of willing in the first way [i.e. as an efficacious volition], I say that the [evil] angel could not desire equality with God. In the second way [i.e. by a mere wish], he could, because he could love himself by as much love of friendship as that by which, according to right reason, he ought to have loved God. And nevertheless he could also have desired for himself as much good as he owed to God by love of desire, if one speaks of the act of the will that is called a “wish” (velleitas).35 (Trans. mine)
So while the first strategy identifies a class of two-object volitions, the second strategy distinguishes between wishes and choices to do something. Throughout his career, Scotus is consistent in regarding each of these two strategies as necessary to account for Lucifer’s choice. But in his first treatments he considers the first strategy as fundamental. Thus, the notion of two-object volitions is the core of Scotus’s solution in the question devoted to this topic in the first version of his Sentences commentary, the Lectura, as well as in its revision, the Ordinatio.36 In those two treatments, Scotus argues that Lucifer’s choice was a two-object act of love of desire. Only when dealing with the objection that there is no choice about (p.77) impossible things does Scotus introduce the distinction between wishes and efficacious choices and argues that, within the broader class of two-object volitions, Lucifer’s choice was a mere wish.37 By contrast, in later treatments of this topic, it is the second strategy, based on the distinction between a mere wish and a choice to do something, that Scotus regards as fundamental. Thus, in an addition to his Ordinatio, he remarks that such a distinction can be the ground for the affirmative answer to the question whether Lucifer wanted to be God’s equal.38 And in his late Paris course on the Sentences, the so-called Reportatio, Scotus actually makes the distinction between wish and efficacious volition the core of his solution. He now holds that acts of the will are fundamentally divided into wishes and choices turned towards actions. Each of those broad classes can in turn have one or two objects, as Scotus explains when responding to some objections.39
I think that Scotus has good reasons for this shift in emphasis. The two-object character of a certain kind of volition is certainly an important tool to solve the problem of the first evil volition, but it works only if the distinction between wish and efficacious volition is presupposed. Suppose that Lucifer wanted to be God’s equal by an efficacious choice, i.e. that he decided to become a god. The requirements for that choice being a two-object volition would be satisfied, but Lucifer would be the victim of a gross misapprehension, as he would believe that he could become God’s equal. Only if the two-object volition is a mere wish is Lucifer’s rationality preserved. The two-object structure of some acts of the will merely fleshes out the mechanics of Lucifer’s wish.
Before considering the philosophical cash-out of Scotus’s views on Lucifer’s evil volition as a wish for the impossible, let me briefly consider Scotus’s answers to Aquinas’s two arguments that Lucifer, being a rational agent, could not have willed the impossible.40
Aquinas’s first argument was based on the claims that for a creature to be God’s equal includes a contradiction and that Lucifer could not will what is contradictory. Scotus grants that for a creature to be God’s equal includes a contradiction. He also thinks that Lucifer knew perfectly well that that was the case. Nevertheless, Scotus trades on the claim that some acts of the will have two objects and that any two things can function as objects of those (p.78) acts as long as each one of them is believed to be good. He also uses his distinction between mere wish and efficacious choice. It is sufficient for Lucifer to apprehend equality with God as good in order for him to be able to wish to be like God. The will connects its two objects, i.e. Lucifer’s nature and God’s equality, by an act of love of desire. That act is a wish that does not require any prior intellectual judgment that Lucifer can be equal to God.41
Aquinas’s second argument was based on the view that for somebody to be God’s equal implies the loss of her identity and thereby the end of her existence. Scotus mentions a simplified version of this argument, but his answer can be used to address Aquinas’s more sophisticated version. Scotus proposes three ways to deal with this argument. First, he contends that it is possible for somebody to will her own non-existence, provided that that is not the main object of her volition but an unintended consequence of a prior volition. This is what happens any time somebody sins mortally, because when people sin mortally, they do not want to be subject to God, even though they can exist only if they are subject to God.42 Second, Scotus contends that it is possible to will the antecedent without willing the consequent, as when somebody wants to have a few drinks but not to get a hangover.43 Neither answer, however, is completely satisfactory if applied to Lucifer’s case. For it may indeed be conceivable for human beings not to think about the consequences of their own actions. But a purely rational creature in ideal circumstances such as Lucifer should know better. In his Reportatio, however, Scotus proposes a third answer. Somebody may know that there is a necessary connection between antecedent and consequent and wish that that connection did not obtain. For example, I may know perfectly well that after having a few drinks I will get a hangover. However, I may wish that the connection between drinking and getting a hangover did not obtain. Lucifer was in that situation. He knew that, if per impossibile his wish to be like God had been granted, he would have ceased to exist, but he wished that the necessary connection between those two claims had not obtained. Again, Lucifer’s rationality was not endangered because he was entertaining an impossible scenario with full knowledge that it was impossible.44
Scotus’s Lucifer is in stark contrast to Aquinas’s Lucifer. Aquinas’s Lucifer wants to be in God’s company. Scotus’s Lucifer wishes to be God’s equal. Aquinas’s Lucifer is engaged in a genuine choice—he fails (p.79) to consider that he is not able to reach the union with God by his own forces and then chooses accordingly to reach that goal. Scotus’s Lucifer is not engaged in any deliberation or genuine choice. He is a rebel only “in his mind,” with full knowledge that what he wishes is impossible. Nevertheless, he still wishes it. There is something of a romantic dreamer in Scotus’s Lucifer and there is something tragically futile in his choice.
What is possibly the most perplexing aspect of Scotus’s account is that evil is introduced into the world through a mere wish. Both Anselm and Aquinas denied that that was the case. Anselm thought that Lucifer wanted something that he was actually able to attain. By contrast, Aquinas thought that Lucifer wanted something that he was not able to attain by the exercise of his natural capacities but that he could have obtained through God’s help. Saying that Lucifer wanted something impossible would pose an unacceptable threat to his rationality, according to both Anselm and Aquinas. By contrast, Scotus manages to reconcile Lucifer’s rationality with the claim that Lucifer wanted something impossible by arguing that Lucifer’s act was a mere wish. But isn’t it implausible to make what was possibly the most momentous act in the history of creation a mere wish? Wishes seem to be harmless. It is scary to think that evil may have been introduced into the world simply as the result of the wish for something known to be impossible.
With regard to this point, I would like to make two remarks. First, some wishes for the impossible are morally relevant and do tell much about the kind of persons we are. Think of some wishes about past events, such as the wishes associated with regrets or the lack thereof. Those wishes are morally relevant and some of them are not innocent. Just think about criminals’ repentance or lack of repentance about their past crimes. Likewise, Lucifer’s wish for something that did not merely go beyond what he could attain but even beyond any possible state of affairs that God could create looks like an act of radical rebellion. Even though Lucifer never had more than a mere wish and never took arms against God, that wish itself was as close as Lucifer could go to what may perhaps be described as an act of counter-creation—not merely the desire for an alternative order that God did not create but could have created, but the desire for something beyond the realm of what even God could have made.
Second, Scotus himself admits that Lucifer’s wish, no matter how bad, may not have been unforgivable. He suggests that Lucifer may have made many successive choices that took him down a slippery-slope. Lucifer may have started with a wish to be equal to God, which was motivated by excessive self-love. If at that point he had repented, he would still have been forgiven. But he did not repent. Rather, he had another wish. Since he realized that there was no room for two gods, he wished that God did not (p.80) exist. But to wish that God did not exist is an act of hatred against God. And that seems to be a rather serious fault, at least according to Scotus.45
Is there any further interest in Scotus’s view on the first evil choice or should it be regarded as a curiosity? I would like to conclude by indicating at least two possible reasons why Scotus’s position is worthy of serious consideration.
First, Scotus can explain the introduction of evil into the world without appealing to any supernatural event such as a supernatural union with God. The possibility of an evil choice follows from the existence of rational agents that are not God but may desire (quite rationally) to be like God. Specifically, Scotus does not need to postulate that rational creatures were not created supremely happy or that beatitude was delayed to account for the rationality of an evil choice.
Second, Scotus agrees with Aquinas that the possibility of an evil choice does not depend on a contingent decision made by God, such as the decision to create the angels less happy than they could have been with regard to their nature (as Anselm had contended). Rather, and again in agreement with Aquinas, Scotus thinks that the possibility of making an evil choice is an inevitable consequence of the gap between God and his creatures. Just as, according to Aquinas, no creature could be united with God by the exercise of their natural capacities, so, according to Scotus, no creature could be God’s equal. Since this does not depend on God’s arbitrary decision, God does not have any responsibility for the possibility of evil. But what about the actual occurrence of evil? Scotus’s and Aquinas’s accounts differ when one comes to consider whether God could have created the world in such a way that an evil choice, even though possible, would have been utterly unlikely. I have argued above that Aquinas held that nothing prevented God from giving beatitude to his creatures at the moment of their creation rather than later. Aquinas only stressed that, since God owes nothing to his creatures, God committed no injustice when he decided not to make his creatures blessed at the very moment of their creation. But it is compatible with Aquinas’s account that God could have created the world in such a way that no one in their right mind would have made an evil choice. This is not the case for Scotus. According to Scotus’s understanding of Lucifer’s choice, there is no possible scenario where it would have been irrational and therefore less likely for Lucifer to wish to be God’s equal. Given that both God and Lucifer exist, it is possible for Lucifer to have a wish to be God’s equal, because God is necessarily better and happier than Lucifer. Thus, God could have (p.81) done nothing to make it less likely for Lucifer to have that wish. Actually, God seems to have already done all that could be done to make evil less likely. According to Scotus’s hypothesis, Lucifer had to go beyond the realm of what is logically possible in order to introduce evil into the world. Since God’s power extends over all that is possible, Lucifer found his realm and autonomy in the only area left to him—the impossible.46
Adams, Marilyn McCord. “St. Anselm on Evil: De casu Diaboli,” Documenti e studi sulla tradizione filosofica medievale 3 (1992), 429–30.
Anselm. De casu diaboli, in F. S. Schmitt (ed.), Opera omnia 1 (Edinburgh: Thomas Nelson and Sons, 1946).
Bonino, S. T. (ed.). Surnaturel: A Controversy at the Heart of Twentieth-Century Thomistic Thought, tr. R. Williams (Ave Maria: Sapientia Press, 2009).
Courtès, C. “La peccabilité de l’ange chez saint Thomas,” Revue Thomiste 53 (1953), 133–63.
Forsyth, Neil. The Old Enemy: Satan and the Combat Myth (Princeton: Princeton University Press, 1987).
Hick, John. Evil and the God of Love (New York: Harper & Row, 1966).
Hoffmann, Tobias. “Aquinas and Intellectual Determinism: The Test Case of Angelic Sin,” Archiv für Geschichte der Philosophie 89 (2007), 122–56.
—— “Duns Scotus’s Action Theory in the Context of His Angelology,” in L. Honnefelder et al. (eds.), John Duns Scotus 1308–2008: Investigations into His Philosophy (Münster: Aschendorff, 2011), 403–20.
—— “Theories of Angelic Sin from Aquinas to Ockham,” in T. Hoffmann (ed.), A Companion to Angels in Medieval Philosophy (Leiden: Brill, 2012), 283–316.
Isidore. Sentences, in Patrologia Latina 83 (Paris: Vivès, 1862).
John Duns Scotus. Opera omnia (Paris: L. Vivès, 1891–95).
—— Opera Omnia, ed. C. Balić et al. (Vatican City: Typis Polyglottis Vaticanis, 1950–).
King, Peter. “Scotus’s Rejection of Anselm: The Two-Wills Theory,” in L. Honnefelder et al. (eds.), John Duns Scotus 1308–2008: Investigations into His Philosophy (Münster: Aschendorff, 2011), 359–78.
—— “Angelic Sin in Augustine and Anselm,” in T. Hoffmann (ed.), A Companion to Angels in Medieval Philosophy (Leiden: Brill, 2012), 261–81.
Montano, Edward J. The Sin of Angels: Some Aspects of the Teaching of St. Thomas (Washington: The Catholic University of America Press, 1955).
Robiglio, Andrea A. L’impossibile volere: Tommaso d’Aquino, i tomisti e la volontà (Milan: Vita e Pensiero, 2002).
Thomas Aquinas. Opera omnia (= ed. Leonina) (Rome: Commissio Leonina, 1882–).
—— Scriptum super libros Sententiarum, ed. P. Mandonnet (Paris: Lethielleux, 1929).
—— Summa theologiae (Cinisello Balsamo: Edizioni san Paolo, 1988).
—— On Evil, tr. R. Regan (Oxford: Oxford University Press, 2003).
Williams, Thomas. “The Libertarian Foundations of Scotus’s Moral Philosophy,” The Thomist 62 (1998), 193–215.
—— “The Unmitigated Scotus,” Archiv für Geschichte der Philosophie 80 (1998), 162–81.
(1) The claim that evil was introduced into the world by a rebel angel called “Lucifer” is based on an idiosyncratic interpretation of a few biblical passages, first and foremost among which is Isaiah 14:12–14. See Neil Forsyth, The Old Enemy: Satan and the Combat Myth (Princeton: Princeton University Press, 1987), 134–9. Whereas most details of Lucifer’s story will not be relevant to what follows, the two aspects to be retained are that God created Lucifer entirely good and that Lucifer’s evil choice was unprecedented. For an excellent treatment of the main philosophical difficulties connected with the first evil choice, see Scott MacDonald, “Primal Sin,” in G. B. Matthews (ed.), The Augustinian Tradition (Berkeley: University of California Press, 1998), 110–39.
(2) John Hick, Evil and the God of Love (New York: Harper & Row, 1966), 68–9. As Hick notices, this criticism is already found in Schleiermacher.
(3) Hick, Evil and the God of Love, 68.
(4) The traditional claim that Lucifer wanted to be God’s equal was based on Isaiah 14:14: “I will ascend above the height of the clouds, I will be like the most High.” See also Ezekiel 18:1–9 and the similar promise made by the serpent to Adam and Eve in Genesis 3:5, “you shall be as Gods.” (I quote from the King James Version.)
(5) Anselm, De casu diaboli 4 (ed. Schmitt, 241). For a slightly different interpretation of this passage, see Marilyn McCord Adams, “St. Anselm on Evil: De casu Diaboli,” Documenti e studi sulla tradizione filosofica medievale 3 (1992), 429–30.
(6) Anselm, De casu diaboli 4 (ed. Schmitt, 242). For a recent treatment of Anselm’s account of Lucifer’s fall, see Peter King, “Scotus’s Rejection of Anselm: The Two-Wills Theory,” in L. Honnefelder et al. (eds.), John Duns Scotus 1308–2008: Investigations into His Philosophy (Münster: Aschendorff, 2011), 359–65; and King, “Angelic Sin in Augustine and Anselm,” in T. Hoffmann (ed.), A Companion to Angels in Medieval Philosophy (Leiden: Brill, 2012), 273–81.
(7) Anselm, De casu diaboli 6 (ed. Schmitt, 241–3).
(8) Anselm, De casu diaboli 6 (ed. Schmitt, 243).
(9) I base my analysis on Aquinas’s Questions on evil. See Aquinas, Quaestiones disputatae de malo (= DM) 16.3. For parallel treatments, see Summa theologiae (= ST) I, 63.3; Scriptum super libros Sententiarum 184.108.40.206. See Tobias Hoffmann, “Aquinas and Intellectual Determinism: The Test Case of Angelic Sin,” Archiv für Geschichte der Philosophie 89 (2007), 122–56; and Hoffmann, “Theories of Angelic Sin from Aquinas to Ockham,” in A Companion to Angels, 286–9. Specifically on the object of Lucifer’s choice, see Edward J. Montano, The Sin of Angels: Some Aspects of the Teaching of St. Thomas (Washington: The Catholic University of America Press, 1955), 115–59.
(10) Here Aquinas is giving a condensed version of an argument he states in the De ente et essentia 4 (ed. Leonina 43: 376–7).
(11) DM 16.3 (ed. Leonina 23: 293b).
(12) DM 16.3 (ed. Leonina 23: 293b–294a).
(13) DM 16.3 (ed. Leonina 23: 294a).
(14) DM 16.3 (ed. Leonina 23: 294a). See also ST I 62.1. On the angels’ perfect happiness and impeccability in the natural order, see C. Courtès, “La peccabilité de l’ange chez saint Thomas,” Revue Thomiste 53 (1953), 133–63. It should be mentioned that the issue of angels’ impeccability is closely connected to the relationship between what is natural and what is supernatural, which is a very controversial theme among some Thomistic scholars. For some background, see S.-T. Bonino (ed.), Surnaturel: A Controversy at the Heart of Twentieth-Century Thomistic Thought, tr. R. Williams (Ave Maria: Sapientia Press, 2009).
(15) ST I 12.4.
(16) ST I 56.3.
(17) DM 16.3 (ed. Leonina 23: 394a). The English translation is from Thomas Aquinas, On Evil, translated by R. Regan, edited with an introduction and notes by B. Davies (Oxford: Oxford University Press, 2003).
(18) ST I 12.4 and 62.4. On the divine rule establishing that all creatures need grace in order to attain a state of supernatural happiness (i.e. the beatific vision) and on Lucifer’s failure to subject himself to that rule, see DM 16.2 ad 1 (ed. Leonina 23: 289b); DM 16.2 ad 7 (ed. Leonina 23: 290b); DM 16.3 ad 1 (ed. Leonina 23: 294b). See Hoffmann, “Aquinas and Intellectual Determinism,” 134.
(19) DM 16.2 ad 4 (ed. Leonina 23: 289b–290a).
(20) As Aquinas argues in ST I 12.4, this is not a revealed truth; rather, it can be known by natural reason.
(21) DM 16.2 ad 4 (ed. Leonina 23: 290a).
(22) DM 1.3 (ed. Leonina 23: 16a). See Hoffmann, “Aquinas and Intellectual Determinism,” 140–1.
(23) ST I 62.1.
(24) Strictly speaking, Scotus holds that Lucifer’s choice for something that would make him happier is Lucifer’s second volition. That act is preceded (logically but possibly even temporally) by Lucifer’s act of loving himself excessively, i.e. as only God should be loved. While the first act is an act of love of friendship directed at himself, the second act is an act of love of desire directed at the enjoyment of happiness for himself (on the distinction between these two kinds of acts, see below). Here I will take into account only Lucifer’s second act, directed at what would make him happier. Only the second act explains how the first act of self-love is excessive and therefore evil. See Lectura (= Lect.) 2.6.2, nn. 25–6 (ed. Vat. 18: 377); Ordinatio (= Ord.) 2.6.2, nn. 34–36 (ed. Vat. 8: 39–41); Reportatio (= Rep.) 2.6.2, nn. 4–5 (ed. Vat. 22: 618b–19b). In the absence of a critical edition, I make reference to the text of the second book of Scotus’s Reportatio as printed in the Vivès edition. I have checked it against Tobias Hoffmann’s transcription from mss. Oxford, Merton College, 61, and Oxford, Balliol College, 205. I thank Tobias Hoffmann for generously making his transcription available to me.
(25) Scotus takes over from Anselm the view that the will has two powers or inclinations, a first power directed at the agent’s happiness (the affectio commodi) and a second power checking the will’s exercise of its first power (the affectio iustitiae). See Lect. 2.6.2, n. 36 (ed. Vat. 18: 381); Ord. 2.6.2, nn. 49–51 (ed. Vat. 8: 48–51); Rep. 2.6.2, n. 9 (ed. Vivès: 621a–b). On Scotus’s version of Anselm’s doctrine of the two powers or inclinations of the will, see Thomas Williams, “The Unmitigated Scotus,” Archiv für Geschichte der Philosophie 80 (1998), 162–81; Williams, “The Libertarian Foundations of Scotus’s Moral Philosophy,” The Thomist 62 (1998), 193–215; and most recently King, “Scotus’s Rejection of Anselm,” 365–77. For an overview of Scotus’s position on angelic sin, see Tobias Hoffmann, “Duns Scotus’s Action Theory in the Context of His Angelology,” in John Duns Scotus 1308–2008, 403–20.
(26) Scotus, Rep. 2.6.2, n. 10 (ed. Vivès 22: 622b–623a). Neither in the Lectura nor in the Ordinatio does Scotus mention what in the Reportatio he gives as the second way. Moreover, in the Lectura, he subdivides the first way into two and expresses a preference for what in the Reportatio he lists as the second aspect of the first way. Finally, in the Ordinatio Scotus says that Lucifer’s will to be happy may have been immoderate in still other ways that he does not care to mention. See Lect. 2.6.2, nn. 37–8 (ed. Vat. 18: 381–2); Ord. 2.6.2, nn. 52–3 (ed. Vat. 8: 51–2).
(27) Scotus accepts the distinction between natural and supernatural happiness, which is presupposed by the delayed beatitude model. See Lect. 2.4–5.1–2, n. 17 (ed. Vat. 18: 361); Ord. 2.4–5.1–2, n. 16 (ed. Vat. 8: 7). He also holds that God, as a matter of fact, did delay beatitude, even though he could have given it to his creatures at the moment of their creation. See Ord. 2.4–5.1–2, n. 48 (ed. Vat. 8: 22). The controversial issue is whether God’s decision to delay beatitude plays any role in explaining the origin of evil because it made room for Lucifer’s evil choice, as Aquinas had maintained.
(28) Since happiness is the state of ultimate perfection of a certain essence, where all the capacities of that essence are exercised in the best way, it follows that the better a certain essence is, the happier it can be. So God is happier than any creature can ever be.
(29) The literal interpretation of the claim that Lucifer wanted to be “like God” had been commonly discarded after Anselm’s criticism. Before Anselm, however, Isidore of Seville had endorsed the view that the devil thought he was God’s equal and even God’s superior. See Isidore’s Sentences 1.10, in Patrologia Latina 83 (Paris: Vivès, 1862), 555b.
(30) Scotus, Lect. 2.6.1 (ed. Vat. 18: 371–4); Ord. 2.6.1 (ed. Vat. 8: 25–35); Rep. 2.6.1 (ed. Vivès 22: 614–17).
(31) Ord. 2.6.1, nn. 9–10 (ed. Vat. 8: 27–8). See also Lect. 2.6.1, n. 10 (ed. Vat. 18: 373); Rep. 2.6.1, n. 6 (ed. Vivès 22: 616a). In the Reportatio, Scotus calls the will a “comparative power” (vis collativa). My interpretation of this expressions is slightly different from the one proposed by Hoffmann. Hoffmann thinks that the claim that the will is a collative nature entails that the will forms its own judgment. By contrast, I think that Scotus is calling attention to a class of volitional acts having two objects. I do not think that any “judgment of the will” is involved in those acts. See Hoffmann, “Duns Scotus’s Action Theory,” 408–12.
(34) Rep. 2.6.1, n. 5 (ed. Vivès 22: 615b). For Aquinas’s denial that Lucifer’s choice can be a mere wish for the impossible because such a wish is not morally relevant, see DM 6.3 ad 9 (ed. Leonina 23, 294b–295a). It should be noticed that the Leonine text has Aquinas saying just the opposite, i.e. that a wish (velleitas) is not morally relevant. But the Leonine editors have deleted a non before est voluntas at 295, line 293. That deletion has no textual support, for the non is contained in the entire manuscript tradition, as evidenced in the critical apparatus and in the Étude critique, ed. Leonine 23, 58*. As a consequence, I think that the non should be reintegrated and the Leonine text should be corrected. On the notion of wishing and conditional willing in Aquinas, see Andrea Robiglio, L’impossibile volere: Tommaso d’Aquino, i tomisti e la volontà (Milan: Vita e Pensiero, 2002).
(38) Ord. 2.6.1, n. 16 (ed. Vat. 8: 30). As the Vatican editors note, the manuscripts make it clear that this paragraph must be considered as a later addition made by Scotus to the text of the Ordinatio.
(40) Scotus takes into account four arguments against his position. Two of these arguments correspond to those Aquinas had given in DM 16.3 and ST I 63.3.
(45) Ord. 2.6.2, n. 78 (ed. Vat. 8: 65–6).
(46) I am grateful to the participants in the 2012 Cornell Summer Colloquium in Medieval Philosophy and an anonymous referee for many useful comments that led me to revise earlier versions of this paper. I would also like to thank the audience of a Scotus conference held at Oxford in 2008, where I first tackled this topic.