A Most Methodical Lover?

On Scotus's Arbitrary Creator(1)

Journal of the History of Philosophy 38 (2000): 169-202.

1. The Issue

Almost all interpreters of Scotus now agree in rejecting the old-fashioned charge that Scotus "thought God would act in a completely arbitrary way in his dealings with creatures."(2) As they point out, Scotus says in several places that God is ordinatissime volens--"a most methodical lover," as one translator has it.(3) Scotus also speaks of God as willing "most reasonably," and in one place he even says that God made all things "with right reason." Moreover, he devotes a whole question in the Ordinatio to arguing that there is justice in God. On the basis of such passages, interpreters conclude that Scotus's God, although he is of course perfectly free, always acts both reasonably and justly, never arbitrarily.

In this paper I shall examine these interpretations and the texts on which they are based. In every case, we shall find that interpreters have greatly overstated the constraints that God's rationality and justice impose on his willing. As a result, I shall have to admit that there is a good deal of truth in the charge that Scotus's God acts arbitrarily in some sense. But unlike those who have used the charge of arbitrariness to dismiss Scotus's views, I shall argue that the kind of arbitrariness Scotus recognizes in the divine will is nothing to worry about. Far from being an embarrassment, it is all of a piece with a most appealing picture of God and his relation to the created world.

First, however, I must delimit the scope of this paper, since there are really two distinct issues that arise in considering the degree of arbitrariness that must be attributed to Scotus's God. The first issue is what we might call the question of God's legislative rationality. Here the question is whether God's rationality constrains what he wills regarding the moral law. The mainstream interpretation holds that it does, and that the picture of Scotus as a radical voluntarist concerning the moral law--as someone who holds that God simply gets to make up whatever wacky rules he might come up with off the top of his head--fails to appreciate the rationality that Scotus ascribes to the divine will. The second issue, which we might call the question of God's creative rationality, is more general. Here the issue is whether God's rationality constrains the sort of world he can create and influences his dealings with his creatures. Again, the mainstream interpretation holds that it does; specifically, it holds that God must make his creatures in a fitting way and confer on them the perfections appropriate to them, and so not just any possible order of creation is in fact feasible for God as one who wills in a most reasonable way.

The two issues are not always kept separate in the secondary literature. Perhaps there is some justification for this, since many interpreters treat God's establishment of the moral law as a special case of his treatment of his creatures (specifically, of his human creatures), so that the question of God's legislative rationality turns out to be included in the question of God's creative rationality. But this assimilation of the two questions depends on a close association between the natural law and the exigencies of human nature, whereas Scotus takes great pains to repudiate any such association. Since I have discussed the question of God's legislative rationality at considerable length in another article, I here confine myself primarily to God's creative rationality.(4)

I shall take it as proved (for it is not in dispute) that Scotus's God cannot will contradictions. He cannot create square circles or married bachelors. And since Scotus is an essentialist, he takes God's inability to will contradictions to mean also that God cannot create beings of a certain kind without giving them the features that belong essentially to such beings.(5) Such constraints, however, are compatible with the greatest conceivable arbitrariness on God's part. For contradictions are not simply odd and exotic states of affairs that God is fortunately precluded from willing; they are not conceivable states of affairs at all. If God's inability to will contradictions is the only constraint on his will--that is, if God can will anything that does not include a contradiction--then he can will anything whatever, no matter how peculiar it might be.(6) So if we are to show that Scotus's God cannot act arbitrarily, we must identify some other constraints on his will. The two best candidates are divine justice and divine rationality; in section 2 I consider divine justice, and in section 3 divine rationality.

2. Divine Justice

a. The divine affectio iustitiae

One influential consideration regarding divine justice is Allan B. Wolter's insistence that the divine will, just like the human will, has an affectio iustitiae or affection for justice. Now the implications of this claim will depend very much on how one characterizes the affectio iustitiae--a matter to which I shall turn in a moment--but we must first ask whether the claim is even true. Certainly Wolter never cites any text in which Scotus ascribes an affectio iustitiae to God, and as far as I can find, there is no such text to be cited. In default of textual support, then, we need an argument to show that God has an affectio iustitiae. Wolter's argument(7) is that the affection for justice is a pure perfection, and since God possesses all pure perfections, he must possess an affection for justice.

The argument is obviously valid, and its second premise is beyond dispute, so the only question at issue is whether the affection for justice is a pure perfection. A pure perfection (perfectio simpliciter) is a perfection that does not imply defect or limitation of any sort. Scotus adopts the idea (although not the terminology) from chapter 15 of Anselm's Monologion. Anselm argues there that God "must be whatever it is in every respect better to be than not to be."(8) For example, it is in every respect better to be living than not to be living, so God must be living; but it is not in every respect better to be body than not to be body, so God is not body. Scotus takes Anselm's view a step further by insisting that the pure perfections are predicated univocally of God and creatures.(9)

Is the affectio iustitiae a pure perfection, or does possession of an affectio iustitiae imply some sort of limitation or deficiency? Scotus offers us more than one way to characterize the affection for justice, and commentators have offered still others, so there is really no way to answer the question other than to proceed through the various descriptions and show that, no matter which of them we choose, God turns out not have an affection for justice.

  • (1) The affection for justice is what provides a check on the affection for the advantageous.

  • Once again following Anselm, Scotus identified two affectiones or fundamental inclinations in the human and angelic will: the affection for the advantageous (affectio commodi) and the affection for justice (affectio iustitiae). The affection for the advantageous is always described as an inclination towards the agent's own happiness. For example, in Ordinatio 2, d. 6, q. 2, Scotus says, "If one were to imagine--along the lines of Anselm's thought experiment in On the Fall of the Devil--that there were an angel that had the affection for the advantageous and not the affection for justice . . . such an angel would not be able not to will advantageous things, or indeed not to will such things in the highest degree."(10) And earlier in that question he has pointed out that "the greatest advantageous thing is perfect happiness."(11) Similarly, in Ordinatio 2, d. 39, he says that a will with only the affection for the advantageous "would be nothing but the natural appetite of an intellectual nature, just as the appetite of a brute animal is the natural appetite of a sensitive nature."(12) And he makes it clear in many places that every creature's natural appetite is aimed at the perfection of that creature, and the perfection of an intellectual nature is happiness.(13) This appetite for happiness can, if left unchecked, be immoderate; there is such a thing as willing one's happiness too intensely, too impatiently, or in the wrong way (say, without taking the trouble to deserve it).(14) Hence, intellectual natures need some way of reining in this potentially immoderate appetite; such is the task of the affection for justice.

    If we understand the affection for justice in this way, it is clearly not a pure perfection. For possession of an affection for justice implies that the possessor has an appetite for happiness that is liable to get out of hand. God has no such appetite. For one thing, God necessarily possesses perfect happiness, and so it would make no sense to think of him as having an appetite for happiness in the first place. More to the point, however, it is simply not possible for God to love his happiness immoderately--and for two reasons. For one, God is incapable of sin. If the task of the affection for justice is to keep a possibly sinful will from falling into actual sin, God does not need one. For another, even if it did make sense to think of God as needing something to "keep" him from sinning, the particular sort of sin against which the affection for justice is supposed to guard--an immoderate love for one's own happiness--is one that is conceptually impossible for God to commit. God's happiness is an infinite good and therefore deserves to be loved with an infinite love. There just is no such thing as God's loving his own happiness too dearly, and so there is no place for an affection for justice that would prevent him from doing what cannot in any case be done.

  • (2) The affection for justice is that aspect of the human or angelic will in virtue of which it is free.

  • Scotus argued that a will that had only the affection for the advantageous would not be free. Just as the natural appetite of a sensitive nature follows deterministically upon sense cognition, so also the natural appetite of an intellectual nature would follow deterministically upon intellectual cognition. Freedom in a human or angelic will, then, must come from something else, the affectio iustitiae:

    Therefore, this affection for justice--which is the first controller of the affection for the advantageous, both insofar as the will need not actually desire that to which the affection for the advantageous inclines it, and insofar as it need not desire it in the highest degree (to the degree, I mean, to which the affection for the advantageous inclines it--this affection for justice, I say, is the innate liberty of the will, since it is the first controller of such an affection [for the advantageous].(15)
    Similar accounts of the relationship between the two affections are given at Ordinatio 2, d. 39, and 3, d. 26, where Scotus again argues that the will is free because it has an affection for justice that controls or moderates the affection for the advantageous.

    Scotus never applies this analysis to God's freedom, however, and it is not difficult to see why. For Scotus, (2) is tied up with (1): if we didn't have an affectio iustitiae, we wouldn't be able to put the brakes on our affectio commodi, and so we wouldn't be free. So once again, the possession of an affectio iustitiae implies a certain limitation or imperfection. It is not a pure perfection and therefore cannot be ascribed to God.

  • (3) The affection for justice is what disposes the will to act in accordance with right reason.

  • This is one of Wolter's descriptions.(16) He cites no text of Scotus in support of it, because there is none. But suppose there were. For a human being to will in accordance with right reason is, at least, for her to will in accordance with practical principles that do not themselves depend on her own choice. As we shall see at length below, Scotus denies that there are any such principles that constrain God's creative will. So there is not, strictly speaking, such a thing as right reason in God--Scotus says so explicitly.(17) And if there is no right reason in God, there is no affection for justice in God as described in (3).

  • (4) The affection for justice is what inclines the will to love something for its own sake, in accordance with its intrinsic worth.

  • Many interpreters argue that the affectio commodi inclines an agent to love objects for the agent's sake, whereas the affectio iustitiae inclines an agent to love objects for the objects' own sake. I have argued elsewhere against that reading, and I will not repeat the arguments here.(18) Oddly enough, even if those other interpreters are right and I am wrong, it still turns out that God has no affection for justice. For Scotus tells us that whatever God does, he does for God's own sake.(19) On this reading of the two affections, then, God would turn out to have an affectio commodi, not an affectio iustitiae.

  • (5) The affection for justice is what inclines someone to do justice (to do what is right).

  • This description is certainly correct as regards human and angelic wills, but Scotus never applies it to the divine will. For (5) is closely tied to both (1) and (2),(20) and since God cannot have an affection for justice as described in (1) or (2), it is hard to see how he could have an affection for justice as described in (5). Scotus does indeed talk about something in the divine will that prompts it to do justice, but it is not an affection for justice; it is simply divine justice. Once that fact is understood, the crucial question for the purposes of this paper is whether, according to Scotus, divine justice constrains God's creative will. I shall now argue that it does not.

    b. Divine justice proper

    As Wolter rightly observes, if justice is a matter of paying back what one owes, then strictly speaking there is no justice in God with respect to creatures; he cannot fail to give them what he owes them, since he owes them nothing at all. To put it another way, justice means giving everything its due; but nothing is due to creatures from God, so we cannot properly speak of God's justice in relation to his creatures.(21)

    We would be too hasty, however, if we were to draw what might seem the obvious conclusion, namely, that God will be acting justly no matter how he treats his creatures. For Wolter argues that God is not utterly debt-free. While he owes nothing to his creatures, he does owe something to himself. He must do justice to his own perfect goodness, to his infinite love and generosity. So when we say that God acts justly in his dealings with his creatures, we cannot mean that he gives them their due, but that he gives himself his due. As Wolter puts it:

    the justice which obliges God to be true to his own goodness does modify his creative act in regard to whatever he chooses to create. . . . In a word, if he actually creates something, God must be true to himself and to his infinitely good and perfect nature. . . . He owes it to himself that whatever he chooses to create will have a beauty and natural goodness about it.(22)
    By way of illustration, consider the following analogy. Suppose I am a rather high-minded, Platonistically-inclined philosopher. My sole allegiance is to philosophy; as I see it, my only obligation is to seek out and make known the truth with all the fervor I can command. So I have no obligations to my students as such. Would it then follow that I would be acting in conformity with my one obligation no matter how I treated my students? Obviously not. For I would surely find that some ways of treating my students would prevent me from making known the truth as effectively as possible. Classroom theatrics and a lively, colloquial delivery might do the job better than scholarly austerity and elegant diction, for example; and gentleness and patience might turn out to be more effective than contemptuous rejoinders to ill-informed questions. And so I would then have an obligation to be a gentle classroom ham. But notice, this obligation derives entirely from my first and paramount obligation to philosophy. My passion for philosophy modifies my treatment of my students, even though my students do not of themselves give rise to any obligation on my part.

    In the same way, God's sole allegiance is to himself. But not just any old way of treating his creatures will do justice to himself. Consequently, God's justice to himself will "modify his creative act," even though no creature can of itself give rise to any obligation on God's part. Creatures are "secondary objects" of divine justice, just as my students are secondary objects of my passion for philosophy.

    This specific analogy is my own, but something very much like it(23) is clearly at work in Wolter's understanding of divine justice. Now even before we turn to the texts of Scotus, we surely have good reason to doubt whether any such analogy can really represent Scotus's views. In my analogy, my derivative obligations arose because fulfilling them was a necessary means for me to fulfill my one "real" obligation. It is very difficult to see how any way of treating creatures could be a necessary means for God to fulfill his one real obligation. Cannot God love himself perfectly well no matter how he treats his creatures? Wolter says that God would not be "true to himself" if he withheld "beauty and natural goodness" from his creation.(24) But unless we are already smuggling in the idea that creatures somehow deserve beauty and natural goodness, it is hard to see why God's withholding those things would constitute a breach of duty either to creatures or to himself.

    I can illustrate this point by putting a slight spin on my original analogy. This time, it is still true that my students deserve nothing from me, but now we will say that no matter how I treat my students, I can still meet my sole obligation to seek out and make known the truth. Can we somehow salvage my derivative obligations towards my students? I do not see how. Similarly, then, if creatures deserve nothing from God, and if God necessarily loves and enjoys his own goodness, it is hard to see how we can salvage the derivative obligations towards creatures that Wolter wants to insist on.

    The question, then, is which version of the analogy accurately represents Scotus's views on divine justice. Does God's one non-derivative obligation give rise to derivative obligations to creatures, or can God do his duty no matter how he treats his creatures? This question is decisively answered by the texts of Scotus, to which we should now turn. Scotus's discussion of divine justice is found in Ordinatio 4, d. 46, which comprises four connected questions: (1) Is there justice in God?; (2) Is there mercy in God?; (3) Are justice and mercy distinguished in God?; and (4) Do God's justice and mercy concur in his punishment of the wicked? I shall concentrate on question 1, which is the primary source for both Wolter's reading and my own.

    In question 1 Scotus asks whether there is justice in God. He begins by considering Anselm's and Aristotle's definitions of justice. Anselm defines justice as "rectitude of the will preserved for its own sake." This definition applies to justice in its general sense, in which all right conduct is a matter of justice. If we want to define justice as a particular virtue distinct from the others, as Aristotle does in Book Five of the Nicomachean Ethics, we can simply add Aristotle's qualification that justice is "with respect to another."(25)

    Scotus then goes on to expound the various subdivisions of justice identified by other thinkers. Justice is first subdivided into (I) legal justice, or rectitude with respect to a general law, and (II) particular justice, or rectitude with respect to some particular thing pertaining to that law. Particular justice is in turn divided into (IIA) justice with respect to what is strictly speaking other and (IIB) justice with respect to oneself as if with respect to another.(26) Justice with respect to what is strictly speaking other is in turn divided into (IIA1) commutative justice and (IIA2) distributive justice. (I) is in God because the law "God is to be loved" precedes any determination of the divine will. (IIB) is therefore also in God; it consists in his willing particular things that befit his own goodness. (IIA1) is seen in God's acts of rewarding and punishing, while (IIA2) is seen in his giving to various natures their due perfections.

    Scotus then simplifies the discussion by settling on a twofold distinction. (I) and (IIB) amount to practically the same thing, Scotus says,(27) and he brings them together under the heading of God's "first justice." (IIA1) and (IIA2) both belong to God's "second justice." God's first justice is "the rectitude of his will in its ordering towards what befits the divine will"; his second justice is "the rectitude of his will in its ordering towards what is required by that which is in the creature."(28)

    So here we have the claim Wolter wants to make: that there is a sense in which God can be said to be just towards his creatures. God's "second justice" requires him to give to creatures what their natures demand. Wolter admits that the claims of this second justice are not absolute in the way that those of the first justice are. Nonetheless, he insists that on the whole and for the most part God is obligated to give his creatures what their natures demand. It would not be just for God wantonly to deny heat to fire or heaviness to earth. He can, however, withhold what is due to one part of creation in the interests of doing justice to creation as a whole.

    So far we have three main claims:

    (1) There is in God not only a first justice, which has to do with how he treats himself, but also a second justice, which has to do with how he treats creatures.

    (2) God can go beyond his second justice, but not beyond his first justice.

    (3) God can withhold what is due to a part of creation, but only when doing so is necessary for him to give what is due to the whole of creation.

    Wolter's attributing (1) to Scotus rests on an embarrassing and elementary mistake. Scotus sets out these distinctions only to reject them. Immediately after the passage we have been considering, he continues: "Nonetheless, without bothering to disprove these distinctions, in answer to the question I say more briefly that in God there is only one justice, one in reality and in ratio."(29) That one justice is God's justice to himself; there is no second justice at all. Scotus will go on to speak about secondary acts and secondary objects of divine justice, but from here on out--where he is at last setting out his own views rather than expounding the views of others--there is no further mention of "secunda iustitia." Indeed, Scotus takes care to emphasize a few lines down that God's acts regarding himself and those regarding his creatures do not involve "notionally distinct justices, as it were" (quasi distinctae iustitiae ratione), and he speaks of God's justice to himself as God's "one and only justice" (ista unica iustitia).

    The attribution of (2) is, if anything, even more puzzling. Scotus does indeed consider the argument that God can go beyond his second justice but not his first justice. Anselm, after all, had said in the Proslogion, "When you punish the wicked, it is just because it accords with their merits." This would seem to be a clear case of God's acting in accordance with his second justice; he is giving his creatures their due. But Anselm immediately goes on to say, "When you spare the wicked, this is just, not because it accords with their merits, but because it accords with your will and goodness." So here we would have a case of God's going beyond his second justice by not giving his wicked creatures the punishment that is their due, but respecting his first justice by giving himself his due.

    So far so good; but Scotus then offers two arguments against the view that God can go beyond his second justice but not his first--that is, against the view that Wolter ascribes to him. And Scotus endorses the second of those arguments in these words: "The second argument appears to prove manifestly that God's second justice can incline the divine will to anything to which his first justice inclines it. . . . And thus there will be no distinction between these [two] on the basis of his being able to go beyond it or his not being able to go beyond it.(30) Clearly, then, Scotus rejects (2).

    According to (3), God can withhold what is due to a part of creation, but only when doing so is necessary for him to give what is due to the whole of creation. Wolter bases this notion on his reading of a passage that comes a few paragraphs later in question 1.(31) I will set out Wolter's translation in parallel columns with the Latin text, for reasons that will quickly become clear.
    [D]ico quod sicut in istis politiis legislator respicit in se simpliciter iustum quod est rectum boni publici, secundum quid respicit alia recta partialia, semper quidem in proportione ad istud rectum et ideo in quibusdam casibus rectum est non servare leges iustas, respicientes ista recta partialia, quando scilicet observatio earum vergeret indetrimentum [sic] iusti publici, scilicet bene esse reipublicae, ita Deus simpliciter determinatur ad iustum publicum, non communitate aggregationis, sicut est in civitate, sed communitate eminentis continentiae, quod est rectum condecens bonitatem suam. Omne autem aliud rectum est particulare et [nunc hoc iustum,] nunc illud iustum secundum quod ordinatur vel convenit huic recto. 

    Dic[o] ergo quod potest velle Petrum damnari et recte velle, qui[a] illud particulare iustum APetrum salvari--non necessario

    I say the legislator in matters of state regards something as simply just if it is right for the public good, whereas he regards other, partial rights always in the qualified sense that they do not militate against this unqualified right of the community at large, and therefore in certain cases he sees it is right not to observe just laws concerning these partial rights, namely, when their observation would be detrimental to what is just publicly, namely, to what is in the best interests of the state. In a similar fashion God is determined to do what is just publicly as something right and becoming to his goodness, and to do this not for a group that is just an aggregation of citizens, but rather for a community whose members are knit together in a far more excellent way. But everything other than what is right for this community is only a partial right that may be just in this case but not in that, depending on how it is ordered to or in harmony with that more basic right. 

    I say therefore that God could will that Peter be damned and be right to will such, because this particular instance of what is just,

    The reader who takes the time to compare Wolter's English to the Latin will be struck by how persistently Wolter has translated his view into a text that actually says something quite different. There are two crucial points Wolter wants to make by using this analogy: God can withhold justice in specific instances when doing so is necessary for the good of creation as a whole, and God's regard for the good of creation as a whole is a matter of his doing justice to his own goodness. The translation makes both these points; the Latin makes neither of them.

    To see why this is, we must first get clear on the analogy that Scotus is developing. One side of the analogy is quite clear. Legislators must have regard to what is just or right for the whole (the rectum or iustum publicum), but in light of the common good they may sometimes need to go beyond what is just or right for individual citizens (recta partialia or particularia). The other side of the analogy is less clear. The analogate for recta partialia is of course the "demands" of creatures: that fire be hot, that Peter be saved, and so on. But what corresponds to the iustum publicum? In Wolter's translation the iustum publicum is the good of creation as a whole; but in the Latin it is the good of God himself. Compare Wolter's translation (on the left) with a more literal rendering:
    In a similar fashion God is determined to do what is just publicly as something right and becoming to his goodness, and to do this not for a group that is just an aggregation of citizens, but rather for a community whose members are knit together in a far more excellent way. But everything other than what is right for this community is only a partial right that may be just in this case but not in that, depending on how it is ordered to or in harmony with that more basic right. In the same way God is unqualifiedly determined to the iustum publicum--not for the community as an aggregation, as in [the case of] a city, but for the community as superordinate containment--which is the right that befits his own goodness; whereas everything else that is right is a particular [right], [so that] now this is just and now that, depending on its being ordered to or in harmony with this [public] right.

    I have translated "communitate eminentis continentiae" by "community as superordinate containment" because I wanted an English phrase that was as inscrutable as Scotus's Latin. Whatever exactly those words mean, they certainly do not mean "for a community whose members are knit together in a far more excellent way." Wolter's translation unequivocally identifies the community with the universe as a whole, but Scotus's phrase leaves one wondering exactly what the "community" in question is--no minor point of puzzlement, since it is the good of that community, whatever it turns out to be, that determines what the iustum publicum is, and by reference to which the violation of particularia recta can be justified.

    Fortunately, Scotus does not leave us in suspense. In the next paragraph he tells us exactly what the iustum publicum is. Again, it is instructive to compare Wolter's translation with a literal rendering:
    I say therefore that God could will that Peter be damned and be right to will such, because this particular instance of what is just, viz., "Peter is saved," is not necessarily required for what is just for the community in the sense that its opposite could not also be ordered to that same end, namely, what is just for the community as befitting divine goodness. I therefore say that God can will, and will rightly, that Peter be damned, because this particular instance of justice--Peter's being saved--is not necessarily required for the iustum publicum in such a way that its opposite cannot also be ordered to that very same [iustum publicum], namely, to the befitting of the divine goodness.

    Here Scotus puts "the befitting of the divine goodness" in apposition to "iustum publicum"; they are one and the same thing. Wolter, by contrast, puts "what is just for the community as befitting divine goodness" in apposition to "iustum publicum," as if the iustum publicum involved what suits the good of the community rather than, as Scotus says, the good of God himself. Those additional words in Wolter's translation completely change the meaning of the text, and they correspond to nothing in the Latin.

    There is, in short, nothing in this passage about the good of creation as a whole. There are the goods of particular creatures, which correspond to the goods of individual citizens; and there is the good of God himself, which corresponds to the iustum publicum. Where Wolter has Scotus saying that God can withhold justice in specific instances when this is necessary for the good of creation as a whole, Scotus actually says that God can do so whenever this would not violate what he owes himself. And when is that? Always. For Scotus tells us that the divine goodness is an end for which no "means" is indispensable.(32) And as he puts it a little earlier in this question, "There is nothing in the divine will in virtue of which it is inclined to any secondary object in such a way that it cannot be inclined justly to the opposite of that object. For just as it can without contradiction will the opposite, in the same way it can will [the opposite] justly. For otherwise it could will something absolutely and not justly, which is absurd."(33)

    There is still more in Wolter's analysis that we need to examine. Wolter says that "the justice which obliges God to be true to his own goodness does modify his creative act in regard to whatever he chooses to create."(34) By this he must mean--at the very least--that God's justice somehow makes a difference in God's creative act. That is, God's justice rules out certain creative acts that God could perform if (per impossibile) he were not just. But this is precisely what Scotus denies, notwithstanding his use of the word 'modify' (modificare) in this context. In order to get the full force of Scotus's words, we need to look in some detail at the argument.

    Once again Scotus looks to an analogy to help expound his view, but this time the analogy is not between God and a creature, but rather between two different aspects of God's relation to creatures. God is related to creatures by both intellect and will; he both knows them and chooses them. Both the divine intellect and the divine will, however, have as their first object the divine essence. God first of all knows and loves himself, and only secondarily does he know and love creatures. In the case of the divine intellect, Scotus notes three important facts. First, God's knowledge of creatures in no way depends on the creatures themselves; they are intelligible because he understands them, rather than vice versa. Second, God's act of understanding is necessary with respect to its secondary objects, creatures, just as it is necessary with respect to its primary object, himself. Third, the act by which God knows creatures is notionally different from, but really the same as, the act by which he knows himself.(35)

    Of these three facts, only the first applies to God's willing of creatures. God's willing to create one thing rather than another does not depend on anything about that creature; it is entirely a matter for God's free choice. The second clearly does not apply: God necessarily loves himself, but he does not necessarily will any particular creature. And because the second fact does not apply, the third cannot apply either; since God's volition of creatures is free, we cannot speak of notionally different acts in the divine will as we could in the divine intellect:

    Now if we should wish to distinguish what is really one act into many notionally different acts, as before we distinguished what is really one intellection into many notionally different ones insofar as it goes forth over many secondary objects, I say that there are not, as it were, notionally distinct justices with respect to them; nor is there [numerically] one justice [that applies to both God and creatures], whether distinct in some way or indistinct. For a habit inclines naturally, and so it inclines determinately to one. Consequently, a potency that is habituated by that habit cannot tend to the opposite. But there is nothing in the divine will in virtue of which it is inclined to any secondary object in such a way that it cannot be inclined justly to the opposite of that object. For just as it can without contradiction will the opposite, in the same way it can will [the opposite] justly. For otherwise it could will something absolutely and not justly, which is absurd.(36)
    We must not overlook how uncompromising Scotus's argument is here, and how decisively it speaks against Wolter's view that God's justice makes a difference in how he creates. Justice is a habit, and it is of the nature of habits to rule out certain courses of action. That is what Scotus means by saying that a habit "inclines determinately to one": for at least some pairs of contradictory actions, one who possesses the habit will be decisively inclined towards one action and away from the other. Now there is no problem in saying that God has such a habit when it comes to himself. God's justice requires him to love himself and decisively prevents him from failing to love himself. But we cannot say that God has any such habit regarding creatures. For if he did, his justice would decisively prevent him from willing certain things regarding his creatures. Since in fact God can will anything regarding his creatures that does not involve a contradiction, we know that there is no such habit in God. Notice how Wolter turns Scotus's argument on its head. Wolter says that there is justice in God with respect to his creatures; therefore, God's justice makes a difference in how he creates. Scotus argues that God's justice cannot make any difference in how he creates; therefore, there can be no justice in God with respect to his creatures.

    Although God's justice makes no difference to his creative act, Scotus is still willing to say that God's justice "modifies" his volitions with respect to creatures. He simply takes care to empty the word modificare of its usual sense:

    Nonetheless, it could be said that [God's] one and only justice, which does not incline [the divine will] determinately except to its first act [viz., the act of loving his own goodness], modifies the secondary acts [viz., acts of loving contingent things], although it does not modify any of them necessarily, in such a way that it could not modify the opposite [act]. Nor does it precede the will, as it were, inclining it after the manner of nature to some secondary act; rather, the will first determines itself with regard to each secondary object. And this act is modified by that first justice in virtue of the fact that that act is in harmony with the will to which it is conformed as if the rectitude inclining it in this way were the first justice itself.(37)
    While you would expect the word modificare to mean (as Wolter thinks it does) that God's justice makes a difference in his creative act, Scotus here explains that it has no such implication. We can say "God's act of willing contingent thing x was modified by his first justice," but we are not entitled to conclude that God's act of willing the complement of contingent thing x would not also have been modified by his first justice, if that is how God had chosen to act instead. In short, whatever God can will regarding contingent things will be "modified" by his first justice in this attenuated sense. And since God can will anything whatsoever, it follows that every possible order of creation would, if willed, conform to God's first justice in the only sense that Scotus allows for that expression.

    One final look at my high-minded philosopher will help illustrate what Scotus is saying here. Suppose once again that my only obligation is to seek out and make known the truth. Suppose further that I can carry out this obligation no matter how I treat my students. Strictly speaking, then, I have no duties to my students; I cannot be unjust to them, since I owe them nothing. Scotus is now saying that there is still a sense, albeit an attenuated one, in which my treatment of my students is "modified" by my love for the truth. The will by which I seek the truth is the very same as the will by which I treat my students with gentleness. We can therefore say that my treatment of my students is in harmony with my seeking the truth, or at least that it is not in disharmony. Of course, the same could be said if I instead chose to treat my students with frosty disdain. The will by which I seek the truth would then be the very same as the will by which I treat my students coldly. We would still be able to say that my treatment of my students is in harmony, or at least not in disharmony, with my seeking the truth--precisely because my seeking the truth is compatible with absolutely any treatment of my students whatsoever.

    So does it make any sense at all to talk about justice in creation? Scotus says, "In a second way, what is just is said to be in a creature from the correspondence of one created thing to another; e.g., it is just on the part of a creature for fire to be hot, water cold, fire above, earth below, and so on, since this created nature demands that as something corresponding to it."(38) But we must not say that God's justice requires him to bring about these "just" states of affairs. The primary act of God's will, his loving himself, has nothing to do with creatures one way or the other; and no secondary act of God's will is determinately required by divine justice.(39) (As we shall see later, however, divine rationality does impose certain constraints on the secondary acts of God's will.)

    In standard Scholastic fashion, the question closes with Scotus's reply to three arguments offered at the beginning of the question. The third of these had argued that there is no justice in God, since justice inclines one to pay back what one owes, and God owes no one anything.(40) Scotus's reply to this argument is another passage on which Wolter relies heavily, and it will again be useful to compare Wolter's translation (on the left) to a more literal translation(41):
    To the third, I say that God is no debtor in any unqualified sense save with respect to his own goodness, namely, that he love it. But where creatures are concerned he is debtor rather to his generosity, in the sense that he gives creatures what their nature demands, which exigency in them is set down as something just, a kind of secondary object of this justice, as it were. But in truth nothing outside of God can be said to be definitely just without this added qualification. In an unqualified sense where a creature is concerned, God is just only in relationship to his first justice, namely, because such a creature has been actually willed by the divine will. To the third [objection] I say that he is a debtor, in an unqualified sense, only to his own goodness, that he love it. To creatures, however, he is a debtor in virtue of his generosity, that he communicate to them what their nature requires. This requirement is set down as something just in them, as a secondary object of [God's] justice. But in truth nothing external to God is determinately just except in a certain respect, viz., with the qualification "so far as it is on the part of a creature." The unqualifiedly just is only that which is related to the first justice, i.e., because it is actually willed by the divine will.

    Note first that Wolter says that God is a debtor to his generosity, whereas Scotus says that God is a debtor in virtue of his generosity. The difference is crucial. As we have seen, Wolter wishes to argue that God owes it to himself to make his creatures good, so that God's generosity to creatures is itself a matter of justice. The expression "a debtor to his generosity" would support that interpretation. But what the text actually says is not that God owes it to his own generosity to make his creatures good, but that it is solely a matter of generosity for God to do so. That is, God is not being just either to himself or to creatures in making them good. All he owes to himself is to love himself, which he can do perfectly well without making creatures in any particular way. And he owes nothing at all to creatures, so when he confers suitable perfections on them, it is not justice but generosity that prompts him to do so.(42)

    For similar reasons, Wolter mistranslates the last two sentences of the passage. The phrase "God is just only in relation to his first justice" also tends to support Wolter's view that in conferring perfections on creatures God is being just, not really to his creatures, but to himself. No such phrase appears in the Latin. Scotus does not say that God is just in relation to his first justice, but that that which is unqualifiedly just is so in virtue of its relation to the first justice. And that relation is given simply in virtue of the the fact that God wills it: "The unqualifiedly just is only that which is related to the first justice, i.e., because it is actually willed by the divine will." There is simply no grammatical way to get either Wolter's translation or his interpretation out of Scotus's words.

    There is one last misinterpretation of this passage that requires extended treatment. As we have seen several times, Wolter insists that creatures come in as a secondary object of God's justice, and he often relies on this very passage. For Scotus says, "This requirement is set down [ponitur] as something just in [creatures], as a secondary object of [God's] justice." Wolter therefore claims that God's generosity to creatures is a matter of justice. But there are three arguments that decisively rule out this interpretation.

    First, Wolter fails to appreciate Scotus's use of the word 'ponitur'. Scotus often uses 'ponitur', like 'dicitur', to introduce a view that he rejects.(43) That this is what he is doing here is made clear by the contrast between 'ponitur' and 'tamen secundum veritatem'. The sense of the passage is therefore as follows: "This requirement is set down--by others, whose views I do not endorse--as a secondary object of divine justice; but the truth of the matter is that nothing in creatures is determinately just unless God wills it."

    Second, not only is this reading the only one that makes sense of the antithesis between 'ponitur' and 'tamen secundum veritatem', it is the only one that coheres with what Scotus said above in the body of his discussion of divine justice. Scotus has argued that if there were justice in God, his creative options would be limited; and since his creative options are in fact unlimited, there is no justice in God with respect to creatures. He has also argued that we can speak of God's justice as "modifying" his creative choices only in a highly attenuated sense. Why, then, would he here endorse a view that is totally at odds with his main conclusions? If Wolter's reading is right, certain states of affairs involving creatures are just; God would be unjust if he failed to bring them about. But that is exactly what Scotus denies in the body of the question. So Wolter's reading has Scotus contradicting in the reply to an objection the claims for which he argued at length in the body of the question.

    Third, Wolter's reading also renders this passage itself internally contradictory. Scotus does indeed say here that God's justice determinately requires him to love his own goodness, but he also says that nothing external to God is determinately just unless it is actually willed by God. From these two claims it follows that nothing external to God is necessary for God to love his own goodness. Wolter's reading requires us to say that Scotus flatly contradicts himself in the space of two sentences. If we reject this charge, as we surely should, we must say that for Scotus God's justice towards himself places no constraints on his creative act.

    3. Divine Rationality

    a. God's orderly and rational willing

    Since considerations of divine justice do nothing to quiet the suspicion that Scotus's God acts arbitrarily, we should now turn to what Scotus says about divine rationality. It is here that some interpreters appeal to passages in which Scotus speaks of God as willing ordinatissime or rationabilissime.(44) For example:
    Another characteristic of the divine will is that it wills everything most reasonably and in a most orderly and correct way. . . . The Subtle Doctor has indeed been blamed by certain authors for defending a notion of divine will that denies any rational element, as though he thought God would act in a completely arbitrary way in his dealings with creatures, including man and the natural moral law. A simple look at some of the pertinent texts will convince the reader that nothing could be farther from the truth.(45)
    Scotus in fact recognizes two different senses of what it means to will in a rational or orderly way. Since the first sense is always associated with the word rationabiliter, whereas the second is usually, although not always, associated with the word ordinate, I shall identify the first sense as rational willing and the second as orderly willing. Let us look first at his conception of rational willing. Scotus says, "God causes or can cause all things--not irrationally, therefore rationally. Therefore, he has a ratio according to which he forms them."(46) This ratio is a divine idea, an eternal blueprint in the mind of God. God wills rationally insofar as he wills in accordance with such an idea. Now the mind of God has ideas of every possible creature. So on this conception of what it means to will rationally, God will be acting rationally no matter what he creates. To act irrationally would mean to act blindly, and God could do that only if he had, quite literally, no idea of what he was creating. A necessarily omniscient God cannot act otherwise than rationally.

    Scotus's conception of orderly willing is somewhat fuller.(47) To will in an orderly way is to adopt an end and then to will the means to that end (i) in the order of their proximity to that end and (ii) for the sake of that end. Since God wills in a most orderly way,(48) there is one obvious constraint on his willing: having willed an end, God must go on to will the necessary means to that end. But this constraint turns out not to amount to much, since there is only one end that God cannot fail to will: himself. And Scotus maintains that nothing contingent--whether angels or human beings or unicorns or the Fifth Commandment--is necessary for God to will his own goodness.(49) So God's willing ordinately does not require him to will any of those things.

    The fact that orderly willing understood in this way is consistent with a great deal of arbitrariness on God's part is made quite clear when we consider the context in which Scotus typically talks about orderly willing. In all four of the passages cited in note 48, Scotus is talking about predestination. To take one representative passage:

    Everyone who wills reasonably first wills an end; second, that which immediately attains the end; and third, other things that are more remotely ordered to attaining the end. Therefore, since God wills most reasonably. . . he first wills the end. . . . Second, he wills those things that are immediately ordered to it, namely, by predestining the elect, who immediately attain it. . . . Third, however, he wills those things that are necessary to attaining this end, namely, the goods of grace. Fourth, on account of [the elect] he wills other things that are more remote: for example, this sensible world, first of all, that it might serve them.(50)
    From this we can see that the concept of willing ordinately cannot be intended to imply any significant constraint on God's dealings with his creatures. For Scotus insists that even though God must love his own goodness, he need not create human beings at all, much less predestine any of them; and still less must he predestine any particular human being as opposed to some other. The fact that Scotus keeps raising the concept of willing ordinately in the context of predestination--where Scotus himself admits that we are in the region of the mysterious and inexplicable(51)--is decisive evidence that this concept is in no way intended to mitigate Scotus's voluntarism. God wills in a most orderly way, but what he wills in that orderly way is very much up to him.

    b. Two models of rationality

    So far we have found nothing that might count as a genuine restraint on God's creative activity. But there is another approach we can take in order to get a clearer picture of just how rational Scotus's God has to be. We can look at familiar models of rationality and see how far they might apply to God as Scotus understands him. In this section I wish to look at two such models. The first perhaps embodies the notion of rationality that is most ingrained in our ordinary non-philosophical thinking: instrumental rationality. In this sense, I act rationally when I act in a way that is apt to help me fulfill my desires or attain my goals--whatever those desires and goals may happen to be--and irrationally when I act in a way that is apt to frustrate my desires or keep me from reaching my goals.

    Scotus's conception of orderly willing is clearly a version of this model. God's action is rational in part because he first wills an end, and then wills other things for the sake of that end. But as I remarked earlier, there is only one end God must adopt--himself. He necessarily loves himself and wills his own blessedness, but any other, subordinate, ends he might adopt are purely contingent. Since God wills in an orderly way, he will not fail to will whatever means are necessary to achieve those ends. To this extent, those who have objected to the picture of God's creative activity as arbitrary have been justified; God will not frustrate his own freely chosen subordinate ends.

    But lest we exaggerate the extent to which such instrumental rationality limits God's creative action, Scotus insists on two crucial points. First, the subordinate ends God can choose to adopt are all contingently related to his only ultimate end, which is himself. That is, they are related to the ultimate end only because God chooses to make them so.(52) So, for example, when Scotus says that God wills the salvation of the elect for the sake of the ultimate end, he must mean that this for-the-sake-of relation between the elect and God exists only because God brings it about. Second, God's ultimate end is something he will possess no matter what else he wills. For God is a necessary being, and necessarily enjoys perfect blessedness. So no matter what God chooses, he will have everything he wants, since he will have himself.

    Given these considerations, our first model of rationality turns out to apply to God only in a very limited way. God will always act rationally in the sense that he will not frustrate his own attainment of his subordinate ends. If God wants a world in which unicorns thrive, he will create all the conditions that conduce to the flourishing of unicorns. But God's wanting such a world does not seem to be the sort of thing about which this model has anything to say. For God's desire to create a world that is hospitable to unicorns need not in turn be related to any further end that God intends to realize by creating such a world. Nor does this model imply that if God creates unicorns, he must will (and consequently provide for) their flourishing. As far as this model of rationality goes, God could rationally will that there be systematically frustrated unicorns, so long as the frustration of unicorns is not incompatible with some other goal that he freely adopts.(53) This model says only that if God adopts the flourishing of unicorns as an end, he must will whatever is necessary for unicorns to flourish.

    A second model to which one might appeal is more Thomistic. Aquinas frequently speaks of morally good actions as being "in accordance with right reason" and of morally bad actions as being "contrary to reason." Without attempting more than the very sketchiest discussion of what this means, we can say that for Aquinas there are certain ways of acting that are properly called "reasonable" independently of the judgment of any human agent. When an agent correctly grasps what is reasonable and acts in accordance with that grasp, that act is reasonable and hence morally good. This is, of course, a theory about the rationality of human agents, but it is not difficult to see how it might be extended to the divine agent. One might wish to say that certain possible states of affairs are such that God would act irrationally in creating them. If so, then God can be described as acting according to right reason (in an analogous sense of that expression) when he creates a world of which no such state of affairs is a part. Interpreters who dread the charge of arbitrariness usually seem to have some such view in mind.

    There are two serious problems with the attempt to pin such a view on Scotus. First, I have said that we would be able to speak of God as acting in accordance with right reason only in an analogous sense. It is not easy to determine just what that analogous sense would amount to. As Aquinas never tires of reminding us, the starting points of human practical reasoning are ends, which are in the practical realm what first principles are in the theoretical realm. An action will therefore be reasonable insofar as it bears the proper relation to our end, the human good.(54) No action of God's can be described as reasonable in quite this sense. For no created state of affairs is necessary for God to enjoy his end (and indeed, we cannot even talk about God as having an end without doing grave injustice to his sovereign perfection(55)). It is not even easy to see what it would mean to say that some possible state of affairs bears (or lacks) a proper relation to the divine perfection.(56)

    So we first have the problem of figuring out what it would mean to apply a Thomistic model of rationality to God. Even if we solve this problem, however, we face a more serious difficulty, which is that Scotus does not accept that model of rationality anyway. You will never find Scotus describing an action as "reasonable" or "contrary to reason" in the Thomistic sense. One reason for this is his libertarian conception of choice, according to which human beings have some control not merely over their actions but also over what will count as reasons for action.(57) Another reason is that Scotus repudiates a teleological conception of morality. An action cannot be described as "reasonable" or "contrary to reason" simply because of its fit or lack of fit with our natural end. For Scotus, right actions are right simply because God has freely and contingently commanded them, and wrong actions are wrong simply because God has freely and contingently forbidden them.(58) We might, I suppose, call it reasonable to act as God commands, but since God's commandments have their ultimate source in divine will rather than in divine reason, such a designation looks like a slipshod extension of the word 'reasonable'. You will not find Scotus employing it in that sense.

    Since he has no room for the Thomistic sense of 'reasonable', it is no surprise to find that Scotus also has an idiosyncratic conception of what "right reason" means. For him, an action is in accordance with right reason when the agent has correctly discerned what it means to follow the moral law in a given set of circumstances--in other words, when the agent has correctly discerned what is required in order to do God's will in those circumstances. This peculiar usage obviously exacerbates our earlier problem about applying a conception of right reason to divine action. If our acting in accordance with right reason simply means our acting as God wills us to act, then what would it mean to say that God acts in accordance with right reason? It would seem to mean nothing more than that God acts however he wills to act.

    c. Right reason in God

    Such a use of "right reason" as applied to God is so peculiar that one would hesitate, out of simple interpretive charity, to ascribe it to anyone--except that Scotus explicitly speaks in just that way. The point is important, since at least two interpreters have attempted to apply the Thomistic model of rationality to Scotus's God on the basis of a passage in which Scotus speaks of God as acting in accordance with right reason.(59) Wolter, for example, writes, "in the examined report of his late Paris lectures [Scotus] says expressly, paraphrasing St. Augustine: 'Whatever God made, you know that God has made it with right reason'."(60) If we read "right reason" in its Thomistic sense, it does appear that Scotus is here acknowledging some sort of restriction on God's creative activity. When we look at this sentence in its context, however, we find that Scotus is using "right reason" in the peculiar sense I have already ascribed to him.

    In this passage Scotus is replying to an argument that appears in Augustine's De libero arbitrio III.5. Augustine's argument is worth quoting at some length:

    Therefore, it is possible for something to exist in the universe that you do not conceive with your reason, but it is not possible for something that you conceive by right reason not to exist. For you cannot conceive anything better in creation that has slipped the mind of the Creator. Indeed, the human soul is naturally connected with the divine reasons on which it depends. When it says "It would be better to make this than that," if what it says is true, and it sees what it is saying, then it sees that truth in the reasons to which it is connected. If, therefore, it knows by right reason that God ought to have made something, let it believe that God has in fact done so, even if it does not see the thing among those that God has made.(61)
    Note that even Augustine does not quite say that God acts in accordance with right reason. It is our reason, not God's, that is described as right. Leaving that quibble aside, however, we can see that Augustine here endorses something like what I have been calling the Thomistic model of rationality. There are certain standards of rationality, "divine reasons," and God's creative act must conform to those standards. Moreover, we can (at least to some extent) discern what those standards are. If, per impossibile, we were to find that God had failed to live up to them, we would have legitimate grounds for complaint. But in fact we can rest assured that God has amply met the standards of reason, even if the evidence that he has done so eludes us.

    Since Scotus here adopts Augustine's language about right reason, Wolter attributes Augustine's view to Scotus. But if we read Scotus's response as a whole, we find that Augustine's language is the only part of the view that Scotus adopts. Here is the whole text of the response:

    To the other [argument] I say that "whatever occurs to you as better according to right reason, you may be sure that God has made it"--one must say that nothing is unqualifiedly better according to right reason than insofar as it is willed by God. And so those other things that would be better if they had been made are not in fact better than existing things. Hence, the authoritative passage [from Augustine] means nothing more than this: whatever God made, you may be sure he made it in accordance with right reason. For as it says in the Psalm, "all things whatsoever that he willed, he made"--blessed be his will.(62)
    So just as we should have expected, Scotus holds that God's acting in accordance with right reason simply means his acting however he wills to act. In direct opposition to Augustine, Scotus insists that God's rationality imposes no constraint on his creative act, since independently of that act there are no standards to which God must conform if he is to act reasonably.

    He expresses the same view, and puts a similarly non-Thomistic gloss on the expression 'recta ratio', in his discussion in the Ordinatio of God's knowledge of producible things. The question at issue is whether God's knowledge of possible creatures counts as practical knowledge. Practical knowledge, according to Scotus, is cognition that precedes action and dictates whether that action is to be done or not. Accordingly, God's knowledge of producible things is not practical knowledge. For "if, before the act of the divine will, the divine intellect could have such cognition, it would have it merely naturally and necessarily. . . . Therefore, he would necessarily know that such-and-such was to be done, and then his will--to which his intellect would present this [possible action]--would not be able not to will it."(63) And if that were true, God's freedom in creation would disappear.

    In order to sustain his denial of practical knowledge in God, however, Scotus must respond to the two initial arguments for the opposite view. It is the first that is of concern to us now, since in responding to it Scotus carefully delimits the notion of right reason as it applies to God. The initial argument quoted Augustine as referring to "the art of the wise God." Since art is a practical habit (the argument continues), there must be practical knowledge in God.(64) Scotus replies:

    To the authoritative passage from Augustine I say that art is "a productive habit with true reason" (from Book 6 of the Ethics). Now if this definition of 'art' is understood in its fullest sense, 'right reason' is understood as being what directs or rectifies the power whose job it is to act according to art. But art is a "habit with true reason" in a diluted sense when it is merely a habit that apprehends the rectitude of things to be done, and not a habit that directs or rectifies in things to be done. One can concede that there is art in God in this second sense. For, given the determination of his will with respect to any of the things that are to be done, his intellect knows the order of things to be done. In that case, there is right reason in God--that is, [reason] that knows rectitude--but not right reason in the sense that it directs the active power.(65)
    Right reason in the first sense is the one that I have been associating with the Thomistic model; it is also the one that Augustine employed in the argument Scotus alluded to in the Reportatio examinata. In this sense, right reason is what lays down the law for actions; it prescribes what is to be done. Scotus will not admit that there is any such thing in God directing his creative will. For apart from that creative will, it is not the case that one thing is better than another (as the Reportatio examinata puts it) or that one thing rather than another is to be done (as the Ordinatio puts it). So in him there can only be right reason in a diluted sense. God's knowledge of what is better, or of what is to be done, follows rather than precedes his will. "Right reason" in this attenuated sense does not direct God's decision; it merely reports it. Scotus could hardly be more explicit in rejecting the Thomistic model of rationality as applied to God.

    d. Divine artistry

    Scotus's mention of art in the passage just cited merits further discussion. For of all the strategies used by interpreters of Scotus to make Scotus's God look more rational and less arbitrary, perhaps none has been more abused, and twisted further from Scotus's real views, than the comparison of God to an artist. The following passage from Wolter nicely exemplifies the use to which this analogy has been put:
    It has often been said of a fine artist or master craftsman that he cannot turn out a product badly done. I think this might serve as the model for what Scotus is saying equivalently. For God is obviously the most perfect of artists, a craftsman like no other. He owes it to himself that whatever he chooses to create, will have a beauty and natural goodness about it.(66)
    If we spin out the comparison in somewhat more detail, we can see how beautifully apt an analogy it is for reinforcing the view that God's rationality contrains his creative choice. If I am a good composer, my mastery of the art of music does not require me to write a symphony rather than a song, to prefer duple meter to triple or major keys to minor. I have the freedom of the creative artist to make what I please. What I cannot do, though, is make bad music. Whatever I write will be genuinely musical; it will be excellent after its kind. Similarly, God's creative rationality does not require him to create one possible world in preference to another. He has the freedom of the creative artist to make what he pleases. But he cannot turn out shoddy work. Whatever he creates will be genuinely beautiful and good; it will be excellent after its kind.

    This, as I said, is a beautifully apt analogy for its purpose; but unfortunately it is entirely an invention of interpreters. Scotus does occasionally compare God to an artist, to be sure, but he never puts the analogy to such a use. As Scotus understands it, the point of the comparison is that artists do not work blindly. They first form an idea of the work they are going to execute. And when the Creator sets out to work, he too has a stock of ideas on which he draws. We find Scotus deploying the artist analogy in order to establish a catalogue of divine ideas. He uses it to make three points:

    1. Scotus wishes to argue against the view that God has practical ideas only of those things he will actually make, whereas he has merely speculative ideas of everything else. In response he points out that an artist's having the habit of art does not depend on whether he chooses to exercise that habit. Similarly, God has practical ideas with respect to every possible creature, even those he chooses not to create.(67)

    2. An artist that produces a whole and every part of that whole must have a distinct cognition of the parts. God therefore has distinct ideas of the parts in every whole he creates; in particular, he has ideas of genera distinct from his ideas of species.(68)

    3. A created artist works on a pre-existing subject or matter and therefore need not distinctly cognize everything in his finished product. If I make a box that floats, I may not understand why it floats; I made the box, but I am not responsible for the buoyancy of the wood out of which I made it. God, however, produces, and so must distinctly cognize, everything in his creatures.(69)

    Such are the uses to which Scotus puts the artist analogy. It helps him clarify what God must know about his creatures, but it says nothing at all about what he can choose concerning them. As far as we are told here, God's artistry is related indifferently to what he does make and what he does not make. We therefore cannot appeal to the notion of artistry in order to explain why God makes some things rather than others, or why he makes them in one way rather than in some other way. The use of the analogy to argue for restraints on God's creative choice is utterly without a basis in Scotus's writings.

    To make matters worse, that use of the analogy also has implications that Scotus is at pains to reject. Consider the creaturely analogate, the "fine artist or master craftsman" whose work cannot be anything short of excellent. The analogy would lose its point unless there were some independent standard of artistic excellence by which the artist's work would have to be judged. (We need not hold, of course, that such standards are identical across cultures or eternally fixed, merely that they are not of the artist's own making.) There must be some rules or principles of art, conformity to which is necessary for a work of art to be truly artistic and hence to be a worthy product of a masterly artisan. So to say that "God cannot turn out a product badly done" is to imply that there are standards of evaluation independent of God's choice. After all, if there were no such standards, anything God created could count as "artistic"; God could simply decree that the standards of artistry were whatever they needed to be in order for his creation to measure up.

    If artistry involves rules or principles independent of the artist's choice, God is no artist, according to Scotus. As we saw earlier, Scotus says that art, properly speaking, directs the will of the artist; since there is no practical habit in God directing his creative choice, there is no art in God.(70) Indeed, God is free not only in his particular actions, but even in his choice of practical principles. God, in other words, gets to make up the rules:

    In terms of a distinction between instants of nature: in the first, [the divine intellect] apprehends every possible operation(71)--those that are principles of possible operations, just like particular possible operations. And in the second, it offers all these to the will, which from among all of them--both practical principles and particular possible operations--accepts [only] some.(72)
    Note the word 'all', which is repeated with unmistakable emphasis. The divine intellect presents the divine will with all possible operations and practical principles, and the will in turn chooses from among all of them. If Wolter were right about the artist analogy, there would be some possible operations and practical principles that would not be available for the will to choose, since the divine intellect would apprehend them as "inartistic," so to speak. That is precisely what art does: it is a habit of the practical intellect that forestalls certain choices.(73) No such habit is in God; what he knows does not determine what he does.(74) "Hence, when the divine intellect, before an act of the will, apprehends the proposition 'x is to be done', it apprehends it as neutral, just as when I apprehend the proposition 'There is an even number of stars'."(75)

    To sum up, then: God's pre-volitional knowledge is permissive rather than restrictive. It shows him what he can create, and says nothing at all about what he cannot or should not create (except insofar as it concerns what is logically self-contradictory). It takes in the entire field of possibility, and it does so without distinction. God is no respecter of possibles.

    So if God is an artist, he is one who not only creates new works of art but also determines the rules that govern his own artistry; for "no law is right except insofar as it is given force by the acceptance of the divine will."(76) Such a situation is by no means as difficult to imagine as one might think. Wolter's use of the artist analogy encourages us to think of artists or artisans who work within a well-defined art, where the standards are already set and the artist is content to work within those boundaries to produce what any well-informed and attentive person will recognize as art. Certainly some artists work in that way. But there are other artists whose creations are new in a more radical way. These are the artists who invent new forms, pioneer new techniques, introduce new idioms that were, perhaps, not even implicit in the art as it had been practiced before. Their work will not commend itself so readily to every well-informed person, since part of what it is to appreciate this radically new art is to unlearn, or at least agree to suspend one's adherence to, the principles by which one judges the work. If I try to listen to Schönberg in the same way I listen to Bach, I will be as bewildered as someone who watches a cricket match and insists on regarding it as a deviant and perverse attempt at baseball. Schönberg's music is (we shall agree) no less musical than Bach's, just as cricket is no less a sport than baseball; but the rules are quite different.

    Scotus's God, I am suggesting, is more like Schönberg than like Bach. He consults no treatise on heavenly harmony before scoring the music of the spheres; there is no celestial Strunk & White for him to follow in writing out his plan for the universe. That the analogy is an imperfect one, I freely admit; but the imperfections all point to an even more radical creative freedom for God than any created artist could hope to achieve. Even a Schönberg emerges out of a history on which he draws, but God is truly the origin of his creation. Schönberg's conceptual raw materials had to be gathered from somewhere, but God creates out of nothing.

    As Scotus sees it, then, the creedal affirmation that God is the Maker of all that is, visible and invisible, includes the claim that God creates all contingent practical principles. For if they are not among the 'invisibilia' that God is said to create, God's freedom is restricted without good cause, his sovereignty is imperilled, and his omnipotence is degraded. I have suggested that our experience of radical innovation in the arts gives us a way to begin to understand this claim--to see, that is, that Scotus's claim is not nonsensical, and that, however it may appear at first glance, it is not altogether outside our mundane experience. Why, then, do we find ourselves so reluctant to ascribe such artistry to God?

    In replying to this question, Scotus would have to argue in something like the following way: What we must not forget is that now--thanks to God's creative will--certain practical principles are in fact true, and others are false. So when we try to imagine how God might have created otherwise, we cannot, because we invariably judge alternative possible orders of creation by the only standards of evaluation we have: the standards God actually--but contingentl--decided to create. We are like the baseball fan who cannot see cricket as anything but botched baseball, like characters in A Midsummer Night's Dream who would not be able to imagine that the mind from which they sprang could just as easily write tragedies as comedies. From our limited perspective, hedged about by an actual order that excludes a host of unimaginable possibilities, we cannot effectively think ourselves into the position that God occupied when creating. But though we cannot fully grasp that position, what we know about divine sovereignty tells us that God must indeed have been in it:

    I therefore say that not only can God act otherwise than as he has ordained in some particular ordering, but he can even act ordinately otherwise than as he has ordained in his global ordering--in other words, [otherwise than] according to the laws of justice. For both what is beyond that ordering and what is contrary to that ordering could have been ordinately brought about by God by his absolute power.(77)

    4. Conclusion

    I conclude that neither divine justice nor divine rationality imposes any substantive constraints on God's creative will. God may be a most methodical lover once he decides what to love, but that decision seems to be just about as mysterious a process as any other case of falling in love. Love, it has been said, is akin to madness; and (if I may be allowed the word 'madness' in such a context) perhaps we are to conclude from Scotus's account that God's love is no less madness for having some method in it.

    1. Wherever possible I cite texts of Scotus from the Vatican critical edition (Vatican City: Typis Polyglottis Vaticanis, 1950-; indicated in the notes by a 'V'). Elsewhere I give references to the Wadding edition (Lyon, 1639; reprinted Hildesheim: Georg Olms Verlagsbuchhandlung, 1968; indicated in the notes by a 'W'). I have edited texts from the Wadding edition on the basis of the following Codices: A (Ordinatio 2-4: Assisi, bibl. comm., cod. 137), Q (Ordinatio 2-4: Paris, bibl. nat., cod. lat. 15854), S (Ordinatio 2-3: Vatican City, bibl. apost., cod. vat. lat. 883), Z (Ordinatio 2-3: Paris, bibl. nat., cod. lat. 31141), and P (Ordinatio 2: Paris, bibl. nat., cod. lat. 15360). I am grateful to Fr Luka Modric, President of the Scotistic Commission, for recommending these manuscripts. At the suggestion of Fr Allan B.Wolter, I have also consulted Codex M (Oxford, Merton College, cod. 66) for the text of Ordinatio 4, d. 46, q. 1. Where I am criticizing Wolter's translations, however, I reproduce his own editions as they appear in Duns Scotus on the Will and Morality (Washington, DC: The Catholic University of America Press, 1986).

    Work on this paper was made possible in part by the generous support of the Vice President for Research Discretionary Fund, an International Travel Grant, and an Old Gold Fellowship from the University of Iowa. I am grateful to Marilyn McCord Adams, Richard Cross, Peter King, Gareth Matthews, Scott Ragland, Allan B. Wolter, and two anonymous referees for this journal for their comments on earlier versions of this paper.

    2. Bernardino M. Bonansea, Man and His Approach to God in John Duns Scotus (Lanham, MD: University Press of America, 1983), 190.

    3. Allan B. Wolter, OFM, in his Introduction to Will and Morality.

    4. See "The Unmitigated Scotus," Archiv für Geschichte der Philosophie 80 (1998): 162-181. In the notes to the present paper I point out certain parallels between God's legislative rationality and his creative rationality.

    5. To be sure, Scotus sometimes says puzzling things, as when he claims that God could create fire and make it cold (Reportatio 4A, d. 46, q. 4, n. 10 [W 11.2:878a]: "Potest enim Deus secundum iustitiam suam agere quod terra sit sursum et ignis deorsum; et potest facere secundum oppositum actum, faciendo ignem frigidum, etc."). Richard Cross has persuaded me that the evidence for Scotus's essentialism is too strong to founder on such rogue statements as this, and that the best reading is that Scotus just holds an odd view about the essence of fire, taking heat to be merely an accidental modification of fire.

    6. Similarly, God's legislative rationality prohibits him from dispensing from certain necessary moral precepts but allows him complete discretion in establishing contingent moral precepts. See "The Unmitigated Scotus," 170-180, and "Reason, Morality, and Voluntarism in Duns Scotus: A Pseudo-Problem Dissolved," The Modern Schoolman 74 (1997): 84-93.

    7.  Stated somewhat obliquely in "Native Freedom of the Will as a Key to the Ethics of Scotus," reprinted in The Philosophical Theology of John Duns Scotus, Marilyn McCord Adams, ed. (Ithaca, NY: Cornell University Press, 1990), 158, and more explicitly in Will and Morality, 14.

    8.  Anselm, Monologion and Proslogion with the Replies of Gaunilo and Anselm, Thomas Williams, trans. (Indianapolis: Hackett Publishing Company, 1996), 29.

    9.  Anselm himself rejects univocal predication. In chapter 27 of the Monologion he writes, "And since [God] alone among all natures has from himself whatever existence he has, without the help of any other nature, is he not uniquely whatever he is, having nothing in common with his creatures? Accordingly, if any word is ever applied to him in common with others, it must undoubtedly be understood to have a very different meaning" (Ibid., p. 46). But Scotus argues--rightly, it seems to me--that Anselm's teaching about the pure perfections will collapse (peribit) unless there is univocal predication. Scotus makes this argument at Ordinatio 1, d. 3, pars 3, q. 1, nn. 38-40 (V 3:25-27).

    10. n. 8 (W 6.1:539): "Si enim intelligeretur--secundum illam fictionem Anselmi De casu diaboli--quod esset angelus habens affectionem commodi et non iustitiae . . . talis angelus non posset non velle commoda, nec etiam non summe velle talia."

    11.  n. 5 (W 6.1:537): "maximum autem commodum est beatitudo perfecta."

    12.  n. 5 (W 6.2:1021): "praecise tantum esset appetitus naturalis naturae intellectualis sicut appetitus bruti est appetitus naturalis naturae sensitivae."

    13. See especially Ordinatio 2, d. 39; 3, d. 17; and 4, d. 49, q. 10.

    14. Ordinatio 2, d. 6, q. 2, n. 9 (W 6.1:540).

    15. Ordinatio 2, d. 6, q. 2, n. 8 (W 6.1:540): "Illa igitur affectio iustitiae, quae est prima moderatrix affectionis commodi, et quantum ad hoc quod non oportet voluntatem actu appetere illud ad quod inclinat affectio commodi, et quantum ad hoc quod non oportet eam summe appetere, quantum scilicet ad illud inclinat affectio commodi: illa, inquam, affectio iustitiae est libertas innata voluntatis, quia ipsa est prima moderatrix affectionis talis."

    16.  In "Native Freedom" Wolter says, "This native liberty or root freedom of the will . . . is a positive bias or inclination to love things . . . as right reason dictates" (152). And, although he never states the equivalence outright there, the association between right reason and the affectio iustitiae is one of the dominant themes of his Introduction to Will and Morality.

    17. See section 3c, below. It is true that there are some practical principles that do not depend on God's will, namely those that are necessarily true; but these do not affect his creative activity. If someone wanted to ascribe right reason to God on the grounds that God always follows those norms, and from that to conclude that he has an affectio iustitiae, my only objection would be that Scotus never speaks in that way. That is, Scotus never says that the affectio iustitiae is what prompts the will to act in accordance with right reason, and he never thinks of recta ratio as involving conformity to necessary truths. I could have no objections to the substance of the view, since this pseudo-affectio iustitiae would imply no limitations at all on God's creative choice.

    18. See "How Scotus Separates Morality from Happiness," American Catholic Philosophical Quarterly 69 (1995): 425-445.

    19. Ordinatio 2, d. 37, q. 1, n. 2 (W 6.2:980): "quidquid Deus facit, propter seipsum facit (Omnia propter seipsum operatus est altissimus)."

    20. See "How Scotus Separates Morality from Happiness" for a discussion of some of the intricate connections Scotus saw between morality (5), freedom (2), and the transcending of the natural appetite (1).

    21. Scotus argues thus at Ordinatio 4, d. 46, q. 1, nn. 1, 12 (W 10:238, 253-254).

    22.  Will and Morality, 18-19 (emphasis mine).

    23.  Any analogy that has the following formal structure will do just as well. S has non-derivative obligations only with respect to x; S will not meet those obligations with respect to x if S behaves in ways W1 . . . Wn towards y (where x is not equal to y). So S has obligations with respect to y (namely, not to treat y in ways W1 . . . Wn ), but those obligations derive entirely from S's obligations with respect to x. Wolter himself prefers to compare God to an artist, and it might be wondered why I have substituted my analogy for his. There are two reasons. First, Wolter never spells out his analogy in the way that he needs to in order to make his point, and it is not altogether clear that his analogy can be expounded in such a way as to exhibit the required form. But let us suppose that it can; there is still a second reason for preferring my analogy. If Wolter's analogy has (or can be made to have) the required form, it will be vulnerable to the same objections that I am about to raise against my own; but his analogy brings in additional complications as well. These complications cannot be investigated adequately until later in this paper, and so it seemed better to substitute an analogy that is tailor-made for the present discussion.

    24.  Will and Morality, 19.

    25. Ordinatio 4, d. 46, q. 1, n. 2 (W 10:238): "Hic primo de iustitiae definitione. Rationem generalissimam iustitiae ponit Anselmus, De veritate, cap. 12, quod 'iustitia est rectitudo voluntatis propter se servata'. Haec ratio specificatur per iustitiam secundum quod de ea tractat Aristoteles, 5. Ethicorum, qui addit ultra hanc rationem hoc, quod est 'esse ad alterum'."

    26. If I give you what you deserve, this is an instance of (IIA). If I give myself what I deserve, this is an instance of (IIB), because we are in a sense considering me as if I were two persons, one of whom "owes" the other a certain sort of treatment. Scotus's point is simply that the definition of justice as involving rectitude "with respect to another" should not be taken to imply that we cannot have obligations to ourselves, since for certain purposes we must think of ourselves as if we were "another."

    27. n. 3 (W 10:238): "Et ista duo membra, scilicet iustitia legalis, et particularis ad se quasi ad alterum, in Deo quasi idem sunt: quia rectitudo voluntatis divinae respectu suae bonitatis."

    28. n. 5 (W 10:241): "Sic ergo in genere tota illa definitio iustitiae, eo modo quo potest ad Deum pertinere, potest reduci ad duo membra; ut primo modo dicatur iustitia, rectitudo voluntatis in ordine ad condecentiam voluntatis divinae: alio modo rectitudo voluntatis in ordine ad exigentiam eius quod est in creatura." These two sorts of justice are first designated "prima iustitia" and "secunda iustitia" a few lines down.

    29.  n. 7 (W 10:252): "Non improbando distinctiones, dico tamen brevius ad quaestionem, quod in Deo non est nisi unica iustitia, re et ratione." "Non improbando distinctiones" must mean "without bothering to argue against these distinctions" rather than "not by way of disparaging these distinctions" (as Wolter has it), since if the earlier distinctions were rightly drawn, there would be more than one justice in God, in ratio if not in reality, and that is just what Scotus here denies.

    30.  n. 6 (W 10:241): "Sed secunda ratio videtur evidenter concludere quod ad quidquid prima iustitia inclinat, iustitia secunda inclinare potest voluntatem divinam. . . . et ita non erit distinctio istarum in hoc, quod est posse agere praeter eam et non posse agere praeter eam."

    31. n. 11 (W 10:253; Will and Morality 250-253).

    32. n. 11 (W 10:253): "Est enim illud finis quidem nullum ens ad finem determinate necessario requirens."

    33.n. 8 (W 10:252): "Sed ad nullum obiectum secundarium ita determinate inclinatur voluntas divina per aliquid in ipsa quod sibi repugnet iuste inclinari ad oppositum illius, quia sicut sine contradictione potest oppositum velle, ita potest iuste velle; alioquin posset velle absolute et non iuste, quod est inconveniens."

    34. Will and Morality, 18-19.

    35. I isolate these three points of comparison from the text cited in the next note, and from the following passage, which immediately precedes that text: "Sic etiam habet unum actum rei et rationis ad quem determinate inclinat illa iustitia quae est respectu suae bonitatis; sed iste actus ex consequenti respicit multa obiecta secundaria, et eo modo quod dictum est in primo libro, quod intellectus divinus praeter hoc quod habet unum obiectum primum et unum actum, respicit multa obiecta secundaria. Sed in hoc est differentia hic et ibi, quia intellectus obiecta secundaria necessario respicit: hic autem voluntas obiecta secundaria tantum contingenter. Et ideo non solum non dependet ab eis actus hic, sicut nec ibi, sed nec necessario determinatur ad illa hic, sicut determinatur ibi" (n. 7).

    36. n. 8 (W 10:252): "Si autem velimus distinguere unum actum re in multos rationis, sicut ibi distinguitur una intellectio rei in multas ratione, ut transit super multa obiecta secundaria, dico quod respectu illorum non sunt quasi distinctae iustitiae ratione, sed nec una qualitercumque distincta vel indistincta; quia habitus inclinat per modum naturae, et ita determinate ad unum, ut per hoc repugnet potentiae habituatae per habitum illum tendere in oppositum. Sed ad nullum obiectum secundarium ita determinate inclinatur voluntas divina per aliquid in ipsa quod sibi repugnet iuste inclinari ad oppositum illius, quia sicut sine contradictione potest oppositum velle, ita potest iuste velle; alioquin posset velle absolute et non iuste, quod est inconveniens."

    37.  n. 8 (W 10:252): "Tamen posset dici quod illa unica iustitia, quae non inclinat determinate, nisi ad primum actum, modificat actus secundarios: licet nullum eorum necessario, quin posset modificare oppositum. Nec quasi praevenit voluntatem inclinando eam per modum naturae ad aliquem actum secundarium: sed voluntas primo determinat seipsam ad quodcumque obiectum secundarium. Et ex hoc ille actus est modificatus ab illa prima iustitia, quia consonus voluntati, cui adaequatur, quasi pro rectitudine inclinante iustitia prima." Note that this passage states, albeit in different language, the same view that was expressed in the text from the Reportatio cited in note 5, where Scotus explicitly links the view to the claim that God could justly make fire cold. So it seems quite clear that if God had chosen to make fire cold, that act would, in the language of the Ordinatio, have been "modified" by his justice, just as his actual act of making fire hot was modified by his justice; or, to use the language of the Reportatio, either of those acts would be "in accordance with" divine justice.

    38. n. 9 (W 10:252): "Secundo modo iustum in creatura dicitur esse ex correspondentia unius creati ad aliud, sicut iustum est ex parte creaturae ignem esse calidum, aquam frigidum, ignem sursum, terram deorsum, et huiusmodi, quia natura ista creata hoc exigit tanquam sibi correspondens."

    39. Ibid.: "Sed ad istud iustum non determinat iustitia divina intrinseca prima ut est respectu primi actus isto modo, quia ille actus non respicit illud obiectum; neque respectu actus secundarii, quia ut respicit istud obiectum secundarium non est illa iustitia determinate inclinans, ut dictum est."

    40. n. 1 (W 10:238): "Praeterea, iustitia inclinat ad reddendum debitum. Deus nulli est debitor. Ergo."

    41. n. 12 (W 10:253-254; Will and Morality 252-255): "Ad tertium dico quod non simpliciter est debitor nisi bonitati suae, ut diligat eam; creaturis autem est debitor ex liberalitate sua, ut communicet eis quod natura sua exigit, quae exigentia in eis ponitur quoddam iustum, quasi secundarium obiectum illius iustitiae; tamen secundum veritatem nihil est determinate iustum et extra Deum nisi secundum quid, scilicet cum hac modificatione, quantum est ex parte creaturae, sed simpliciter iustum tantummodo est relatum ad primam iustitiam, quia scilicet actualiter volitum a divina voluntate."

    42. So, as Richard Cross rightly observes concerning this passage, "the claim is not that God is essentially generous, but that the term 'debtor' is being used metaphorically" ("Duns Scotus on Goodness, Justice, and What God Can Do," Journal of Theological Studies 48 (1997): 48-76, p. 63). In the Reportatio examinata, d. 44, q. 2, Scotus again exploits the difference between justice and generosity, stating explicitly that no divine perfection requires God to give creatures their characteristic perfections: ". . . dico quod non est invidia in Deo si non fecit omnia ita bona sicut potuit. Invidia enim est in subtrahendo ab aliquo bonum quod est sibi debitum. Deus autem non est debitor alicuius creaturae quantum ad aliquam perfectionem in eo, quia mere liberaliter omnia facit."

    43.We see another instance of this usage earlier in the question, where Scotus is discussing the view that God can go beyond his second justice but not his first. He says, "Et tanta ponitur membrorum istorum distinctio, quod Deus contra primam operari non potest, nec praeter eam, sed praeter secundam potest operari, licet non universaliter, quia non potest damnare iustum nec beatum" (n. 5, W 10:241). As we have seen, Scotus goes on to reject this distinction; hence his use of 'ponitur' to introduce it in the first place. (The Wadding edition has 'est' for 'ponitur', but all three manuscripts I have consulted--A, M, and Q--agree on 'ponitur'. The text given here is that of Codex A.)

    44. See especially Fidelis Schwendinger, OFM, "Die Metaphysik des Sittlichen nach Johannes Duns Scotus," 3. Teil, 2. Hälfte, Wissenschaft und Weisheit 3 (1936): 93-119; Bonansea, 190-191; and Wolter, Will and Morality, 9, 17, 19-20, 55, 57.

    45. Bonansea, 190.

    46. Ordinatio 1, d. 35, n. 39 (V 6:261): "Deus omnia causat vel causare potest, -- non irrationabiliter, ergo rationabiliter; ergo habet rationem secundum quam format." Similar remarks about the priority of cognition to volition are made at Ordinatio prologue, pars 5, q. 1-2; 3, d. 32, n. 6; 4, d. 46, q. 1, n. 10; and (with the analogy to a craftsman who conceives his work before he executes it) Reportatio 1A, d. 36, qq. 3-4. The role of the divine ideas and the significance of the artist analogy will be discussed in section 3d.

    47. For a more detailed discussion, along with citations of a number of relevant texts, I refer the reader to "The Unmitigated Scotus." What follows here is a summary of those remarks.

    48. Ordinatio 1, d. 41, n. 41; 2, d. 20, q. 2, n. 2; 3, d. 7, q. 3, n. 3; 3, d. 32, n. 6.

    49. On this see in particular the discussion of the Decalogue in Ordinatio 3, d. 37, and the discussion of divine justice in Ordinatio 4, d. 46, q. 1.

    50.  Ordinatio 3, d. 32, n. 6 (W 7.2:692-93): "omnis rationabiliter volens, primo vult finem, et secundo illud quod immediate attingit finem, et tertio alia quae remotius sunt ordinata ad attingendum finem. Cum igitur Deus rationabilissime velit. . . primo vult finem. . . . Secundo vult illa quae immediate ordinantur in ipsum, praedestinando scilicet electos, qui scilicet immediate attingunt eum. . . . Tertio autem vult illa quae sunt necessaria ad attingendum hunc finem, scilicet bona gratiae. Quarto vult propter istos alia quae sunt remotiora, puta hunc mundum sensibilem, pro aliis, ut serviat eis."

    51.  On this see especially the lengthy discussion in Ordinatio 1, d. 41.

    52. What Scotus says here about God's creative rationality has a close parallel in what he says about God's legislative rationality. Just as the subordinate ends of creation are related only contingently to the ultimate end of creation, which is God himself, so also particular moral laws are related only contingently to the ultimate end of human beings. In Ordinatio 2, d. 37, q. 1, n. 8, Scotus claims that in every mortal sin there is aversio, meaning that the will is related inordinately to something that is necessary for the end. He then asks, "Whence is this necessary?" His answer is characteristic: "from the divine will commanding that it be observed 'if you will to enter into life', not from some practical syllogism--for here one ought not investigate the moral teaching of the philosophers, but rather the commandments of God in Scripture. . . . In this way, aversio is nothing other than an inordinateness of the will with respect to something ordered to the end by divine commandment, with respect to which it ought to be ordered." (W 6.2:984: "aversio est communis omni peccato mortali, quia in omni tali peccato voluntas inordinate se habet respectu alicuius necessarii ad finem. Unde est illud necessarium? A voluntate divina praecipiente illud observari si vis ad vitam ingredi, non ab ali[qu]o syllogismo practico (hic enim non oportet quaerere moralem doctrinam philosophorum sed praecepta Dei in Scriptura). . . . et hoc modo nihil aliud est aversio nisi inordinatio voluntatis circa aliquid ordinatum ad finem ex praecepto divino, circa quod deberet ordinari.") The fact that Scotus is here speaking in theological terms ("mortal sin") rather than in purely philosophical terms obviously complicates the matter. But his remarks elsewhere, in contexts where there is no reference to the order of merit, confirm the suggestion made here that behavior in accordance with the commandments of the second table of the Decalogue is related to our end only because God has made it to be so. So even granted that God creates beings with our nature--which is to say, beings with our end--he need not impose on them the obligations that he in fact chose to impose on us.

    53. As we saw earlier, God would also not be violating justice in the strict sense if he were to set up the world in this way. See the texts cited in notes 37-39.

    54. Then doesn't the Thomistic conception collapse into the first model? It does not, and for a variety of reasons. First, in the instrumental model, any desire or goal the agent happens to have will count as an end; in the Thomistic model, not every end will serve to make action for the sake of that end fully rational. Second, in the Thomistic model, there is no suggestion that actions and their objects can be related only instrumentally to our end. Third, on the instrumental model, all evaluation is agent-relative. We can call an agent rational only with reference to those desires or goals that the agent happens to have. On the Thomistic model, however, evaluation can in a sense happen from outside the perspective of the agent. It makes sense to talk about the agent's real end, even if that end is something that has never yet entered the agent's mind. From this third consideration follows a fourth. In the instrumental model, to say that an agent is rational all-things-considered will be at best a judgment about coherence--the suitability of means to ends and the coherent relation among various ends. In the Thomistic model, however, to say that an agent is rational all-things-considered is also to make a judgment about correspondence--the fact that what the agent desires and pursues is really, and not just apparently, a good for that agent.

    55. Hence, the closest Scotus comes to applying this model to God involves talking about God as being an end rather than having an end. In Ordinatio 2, d. 37, q. 1, n. 2, Scotus says, "Whatever God does, he does for his own sake ("The Most High has worked all things for his own sake"); but he does it out of the most perfect charity, which he himself is. Therefore, such an act is perfectly ordered, both in virtue of its end and in virtue of its operative principle." (W 6.2:980: "quidquid Deus facit, propter seipsum facit (Omnia propter seipsum operatus est altissimus), sed a charitate perfectissima, quae ipse est, facit. Ergo talis actus est ordinatissimus, tam ex fine quam ex principio operativo.") Note that whatever God might do will be perfectly ordered in this sense.

    56. The most obvious solution to this problem is to detach the notion of right reason from any teleological association, to make of it the sort of notion that would be more at home in the moral philosophy of the Cambridge Platonists than in that of Thomas Aquinas. To say that God acts in accordance with right reason would then mean simply that God acts according to certain objective standards that limit which possible worlds he can create. Unfortunately, Scotus cannot say such a thing, as will soon become clear.

    57. I discuss this conception in "The Libertarian Foundations of Scotus's Moral Philosophy," The Thomist 62 (1998): 193-215.

    58. There is a small class of exceptions to this general rule, which we need not go into here. See Ordinatio 3, d. 37, nn. 5-7.

    59. Wolter, Will and Morality, 19, and Cross, "Duns Scotus on Goodness, Justice, and What God Can Do." Cross differs from Wolter in that he holds that Scotus's God can act contrary to right reason, but even that claim implies the Thomistic view that there are independent standards of reasonableness for divine action.

    60. Will and Morality, 19, citing Reportatio 1A, d. 44, q. 2. Wolter goes on to identify this claim with the view that God is ordinatissime volens.

    61. Augustine, On Free Choice of the Will, Thomas Williams, trans. (Indianapolis: Hackett Publishing Company, 1993), 80.

    62. Reportatio 1A, d. 44, q. 2: "Ad aliud dico quod 'quicquid recta ratione tibi melius occurrit, hoc scias Deum fecisse': verum est quod nihil est melius simpliciter recta ratione quam inquantum volitum a Deo, et ideo alia quae si fierent essent meliora, non sunt modo meliora entibus. Unde auctoritas nihil plus vult dicere nisi quod 'quicquid Deus fecit, hoc scias cum recta ratione fecisse'. 'Omnia enim quaecumque voluit fecit', in Psalmo--cuius voluntas sit benedicta."

    63. Ordinatio 1, d. 38, n. 6 (V 6:305): "si ante actum voluntatis divinae posset intellectus divinus aliquam talem cognitionem habere, haberet eam mere naturaliter et necessario . . . de necessitate ergo cognosceret hoc esse faciendum, et tunc voluntas -- cui hoc offerret -- non posset non velle istud." Note that Scotus is denying that God has practical knowledge of possible creatures (factibilia). God does have practical knowledge that concerns himself, most notably that he is to be loved (Ordinatio prologue, pars 5, q. 1-2, n. 324; cf. n. 333). Since, as we have seen, God's obligation to love himself does not bear in any way on his creative act, I will not discuss this sort of practical knowledge here. It does, however, affect God's legislative act: it grounds the distinction between the first table of the Decalogue, which enjoins the love of God and therefore contains necessary precepts from which God cannot dispense, and the second table, which pertains to creatures and therefore contains contingent precepts from which God can dispense. On the distinction between the two tables, see "The Unmitigated Scotus," 170-180.

    64. n. 2 (V 6:303): 'Contra: VI Trinitatis cap. ultimo: 'Ars sapientis Dei' etc.; ars est habitus practicus; ergo etc."

    65. n. 11 (V 6:307-308): "Ad auctoritatem Augustini dico quod ars est 'habitus cum vera ratione factivus' (ex VI Ethicorum): et prout complete accipitur definitio artis, intelligitur 'ratio recta', hoc est directiva sive rectificans potentiam cuius est operari secundum artem; deminute autem est ars 'habitus cum vera ratione', quando est tantum habitus apprehensivus rectitudinis agendorum et non habitus directivus sive rectificans in agendis. Hoc secundo modo potest concedi ars in Deo, quia posita determinatione voluntatis eius respectu quorumcumque operandorum, intellectus eius cognoscit istum ordinem operandorum: et tunc est ibi 'ratio recta', hoc est cognoscens rectitudinem, -- non tamen 'ratio recta', hoc est directiva ipsius potentiae operantis."

    66. "Native Freedom," 158; cf. Will and Morality, 19.

    67. Reportatio 1A, d. 36, q. 3-4, nn. 11, 17-18, 57. The paragraph numbers refer to the critical edition forthcoming in Medioevo. I am grateful to Timothy B. Noone for allowing me to make use of this edition.

    68. Ibid., nn. 24-25.

    69. Ibid., nn. 26-27, 51.

    70. Ordinatio 1, d. 38, n. 11, quoted in note 65.

    71. 'Possible operation' here renders 'operabile' and is meant to serve as a shorthand for the ungainly "what can be done or made."

    72.  Ibid., n. 10 (V 6:307): "sed distinguendo de instantibus naturae, in primo apprehendit quodcumque operabile (ita illa quae sunt principia operabilium, sicut operabilia particularia), et in secundo offert omnia ista voluntati (quorum omnium aliqua acceptat, tam principiorum practicorum quam particularium operabilium)."

    73. Again, see the text quoted in note 65.

    74. At Ordinatio 1, d. 38, n. 5, Scotus makes explicit that God's action is not constrained by any knowledge of (i) practical principles, (ii) practical conclusions, or (iii) terms that entail anything in categories (i) or (ii).

    75. Lectura 1, d. 39, q. 1-5, n. 44 (V 17:493): "Unde quando intellectus divinus apprehendit 'hoc esse faciendum' ante voluntatis actum, apprehendit ut neutram, sicut cum apprehendo 'astra esse paria'." This is perhaps not an altogether satisfactory analogy, since the number of stars actually is either even or odd--I just don't know which. But God cannot be ill-informed in this way, so if he apprehends a proposition as neutral, it really is neither true nor false.

    76. Ordinatio 1, d. 44, n. 8 (V 6:366): "nulla lex est recta nisi quatenus a voluntate divina acceptante est statuta."

    77. Ordinatio 1, d. 44, n. 10 (V 6:367): "Dico ergo quod Deus non solum potest agere aliter quam ordinatum est ordine particulari, sed aliter quam ordinatum est ordine universali--sive secundum leges iustitiae--potest ordinate agere, quia tam illa quae sunt praeter illum ordinem, quam illa quae sunt contra ordinem illum, possent a Deo ordinate fieri potentia absoluta."