The man in Princeton asks his way to Columbia. “If I were you,” he is unhelpfully told, “I wouldn’t start from Princeton.” Alas, if Princeton is where you are, you don’t have that choice; you have to start from there willy-nilly. That is how it is with many problems, especially those of a philosophical sort. And so it is with the problem of evil: we must start where we are, in the thick of it.

Because we are in the thick of it, it can seem obvious that the starting point for the problem of evil cannot be God and God’s omnipotent goodness—Augustine’s starting point. No, the starting point must be with the manifest evil there is in the world, and with the problem of how a good God, who could prevent it, does not do so. This is the approach of the Scottish philosopher David Hume. When it is a matter of where to start in addressing the problem of evil, it seems that the boot is on the skeptical foot.

There are of course categorical, non-skeptical views on either side as to where discussion about God and evil ends. Namely, either that, in the face of evident evils, a theistic answer cannot meet the conditions of certainty and proof, and must therefore be rejected; or, the evils done by human free agency being unavoidable, their occurrence can be no evidence against the goodness and power of God. But if we leave aside these categorical solutions, it can seem as though evil is a problem of fact for belief in God. It can seem as if faith in God can never convincingly explain away the world’s evil, let alone justify it, so that it is from those evils that we must start. It was thus that in the late eighteenth century, David Hume formulated the classic statement of what we now call the “problem of evil.” And that has been where, more or less, everyone has started ever since. “Epicurus’s old questions are yet unanswered,” Hume writes in Dialogues Concerning Natural Religion. “Is [God] willing to prevent evil, but not able? Then He is impotent. Is He able, but not willing? Then He is malevolent. Is He both able and willing? Whence then is evil?”

But skepticism, or at least agnosticism, should be allowed to cut the other way too. For if Hume is right and there is no defense of theism that could demonstrate consistency between the facts of evil and the power and goodness of God, it is also worth asking whether we are any better placed to formally demonstrate inconsistency between them. Maybe the provable absence of such a demonstration is all that theism needs. As far as I can see, in his Dialogues Concerning Natural Religion Hume makes no categorical case to the effect that God and evil are formally inconsistent, and his Epicurean questions, though skeptical, should not be understood as establishing some sort of proof of atheism on their own. This is because the drift of Hume’s Dialogues is meant to show how it is impossible to settle those questions either way. That of course would not be a reason why we shouldn’t “start from there”—that is, skeptically. Far from it. It may even lend credibility to the proposition that we should do so, if only in a dialectical spirit and with a view to showing that the matter is unresolvable. Such a conclusion would at least cohere with the generally Pyrrhonian form of skepticism that Hume so often declared himself to favor: as he puts it (in the mouth of Philo, the most Humean of the dialogue partners in Dialogues) “a total suspense of judgment is our only reasonable resource.”

So in this article, I want to put forward the case for starting where Hume starts, but to go a little further than he does to argue that the reasons for doing so are not just tactical, apologetic, and dialectical, but ought to be accepted by theists too—and for good theological reasons of their own. At no point will I attempt to answer directly the question of how God could allow moral evil of any kind, let alone allow the extent of moral evil there is; we cannot prove any answer to that question one way or the other. In short, the matter is demonstrably undecidable—you can prove that you couldn’t know the answer to it. That is about as far as anyone, theist or atheist, can get with the problem of evil. To the question “How could an all good and all powerful God allow evil?”, the only answer is that it is impossible to say. And it is here that Julian of Norwich comes into the picture.

God and natural evil: Hume’​s problem

How far may we go along with Hume? Every evil is a problem of some kind, if only the practical problem of how to cope with it. But I do not agree with Hume that every evil is a problem about God. For some evils we can take in our stride, there being no cause for theological, philosophical, or even moral alarm therein, even though they offend our sentiments. So we should start by taking the existence of such evils out of the debate.

For example: I know that others (including my wife) have different convictions than I do about the matter, but I have personally never had a theological problem with lions eating antelopes, though it is impossible not to feel sorry for the panicking beasts as they flee their predators in such wonderfully graceful leaps and bounds. Of course it is distressing that lions seem as unlikely as ever to get round to lying down with lambs, as Isaiah had hoped they would, but I cannot be troubled about God because they don’t. Lions lying down with lambs would of course be good news for lambs, but it would be terrible news for lions. Eating lambs goes with being a lion; being a lamb-eating machine is more or less what a lion is. And more generally nature seems to require a level of raw indifference in matters of tooth and claw. If there is to be variety and complexity in the natural world we know, including large carnivorous cats, the lambs, alas, are going to have to pay for it with their lives. “Did he who made the lamb make thee?” asks William Blake of the tiger burning bright. The question is rhetorical and the answer is yes: God did make tigers, and consistency would require of those who have a problem of this kind that they consider what alternative world they have in mind that doesn’t replace a problem for lambs being eaten with a problem for carnivores being starved for want of ovine nutrition.

It would appear to be the same with inanimate physical processes, for they sometimes impact unhappily, even tragically, upon human affairs. In the mid-eighteenth century, an earthquake in Portugal killed thirty thousand people, and Voltaire lost faith in God. More understandable would have been a loss of faith in human beings. It was they, after all, who had built Lisbon on a geological fault line and seemed willing to blame anyone or anything but their ignorance for the destructive outcome. We today have far less excuse for continuing to build San Francisco on the San Andreas fault line, and there seems to be something of a premodern and merely pagan superstition in supposing there would be a problem about God if someday soon San Francisco were to disappear forever down an immense sinkhole (for we do know now that the prospects are high that in due course it will).

And were it asked more generally why a good God who had alternatives available to him would create a world in which earthquakes are bound to happen, it is unclear what answer would meet the case either way. It would seem that in asking that sort of question about earthquakes we are asking about sets of physical processes governed by laws that originate at a point in time in the order of 1 to the power of -37 seconds after the Big Bang. So to require God to have created only an earthquake-free world, and to regret that he didn’t, is to regret too much, for to wish away earthquakes is to wish away the physical laws that govern the universe itself. There is no picking the good bits out of physics and leaving the bad, for that isn’t physics at all. For obviously if God makes a world in which there are going to be predictable outcomes, that will be because God wants us to be able to understand that world. But the world would become wholly incomprehensible to us if we could never know when physical laws were going to be suspended by God just to suit our particular preferences from time to time. There are those physical laws precisely so that, by getting to know them, we can learn to avoid building cities where earthquakes are bound to happen.

More challenging for some is the problem of physical pain. Hume, again, took the lead here. He seemed to think it obvious that a world in which no one suffers physical pain would be a better world than the one we have. He asked why, if God is good, he should have chosen an alternative so obviously the worse of the two. “It seems...plainly possible to carry on the business of life without any pain,” Philo says. Hume’s skeptical musings are rarely so thoughtless as when he speculates in this way, and there is a quick and sharp retort forthcoming from anyone suffering from that rare genetic disorder known as CIPA, the chronic inability to feel pain. Hume might be less convinced of the advantages of a pain-free life had he given a moment’s thought to the tragedy of a life threatened by scaldings in overheated bath water you cannot feel, by walking on a broken leg of which you are unaware, or walking in bare feet on broken glass without noticing, or having your hand in the middle of a flame without any painful sensation to tell you of your limb’s destruction. Then he might not have been so easily convinced that bodily pain is altogether a bad thing, and he would hardly think that, overall, he was much better off for the want of it.

But then, as if acknowledging that some pain has its purpose in animal life and conceding the general principle that some pain may be necessary, Hume presses the point: Why, he asks, so much pain? Why unbearable pain? Would not tolerable pain—or even some reduction in pleasure—serve the purpose of sending out the signals needed to warn of life-threatening courses of action? To which there is some sort of answer in the thought that pain cannot serve its purpose within the economy of human life if it occurs only at tolerable levels of mild discomfort. For, when tolerable, pain loses its point. It fails to do its job if it is less than too much, and it would be still less effective if it were replaced by a simple reduction in one sort of pleasure relative to others. Of course, it does not follow from this that we should not try to reduce the levels of pain that visit us. Of course we should, but only as far as it is safe to do so. A world in which analgesics were used to dull all pain to acceptable levels of discomfort would be a world in which, our bodies no longer serving with biological efficiency to warn us, we would always have to calculate how to avoid physically harmful forms of behavior. Pain makes for an immensely more efficient warning device than sluggish brainpower with its capacities for self-deception.

None of these forms of evil—if indeed that is what they are—have any tendency to pose a problem of the kind that Hume thinks we are all forced to face. You can guarantee safety for lambs only on the condition of wimpishly vegetarian tigers and lions. You can have an earthquake-free cosmos only on the condition that there are no reliable physical laws to govern it. You can have a world free of physical pain only if it is also a world free of physical pleasure—in short, only if it is a world without nervous systems, which is to say, without bodies. Given the kind of world fit for bodies that we have, these pains are necessary evils where they are not necessary goods. And so it is hard to see why the existence of them is regarded as providing rational evidence against God. Indeed, they seem just as plausibly to be evidence for a providential benevolence within creation.

In any case, there is no need to bring God into the picture at this level, and it is no part of my argument that one should, since evolution will do as a perfectly good explanation for the emergence of the species we have, both lions and lambs, and for the fact that we animals all have diets disadvantageous to some other living species, and nervous systems that register pain. But if, like Hume and some fundamentalist Christians of our own time, you insist on bringing God into it one way or the other, the evidence from the natural world points at least as strongly against a skeptical conclusion as in favor of it. Ours seems to be just the sort of natural world you might expect a good and wise God would bring about, were God to bring about any world at all.

God and moral evil: Hume’s problem

But as for moral evil, for “evil done,” that would seem to be a very different matter. Here, at least, we might reasonably think that you have to start in Princeton—that is, with where we actually are. For here there really is a problem, and Hume gets halfway to an answer. He manages to show that there is no way of formally proving even the de facto consistency, nevermind the truth, of the three-way conjunction: God is all good, willing no evil; he is all powerful, hence able to prevent any evil; and yet there is evil. Hume says that the consistency of conjunction is logically possible, but that all the evidence of experience argues against it: “I will allow,” he writes:

that pain or misery in man is compatible with infinite power and goodness in the deity...[but] what are you advanced by all these concessions? A mere possible compatibility is not sufficient. You must prove these pure, unmixed and uncontrollable attributes from the present mixed and confused phenomena, and from these alone.

“A hopeful undertaking!” Philo adds ironically. “Here I triumph.” For it is not Philo’s job—or, we may suppose, Hume’s—to demonstrate formal inconsistency between theistic belief and moral experience. It is rather his purpose, he says, only to show that it is for the second of the three participants in the Dialogues, Cleanthes, who thinks he can pull off the theistic trick and solve the riddle of evil, to “tug the labouring oar, and to support [his] philosophical subtleties against the dictates of plain reason and experience.” Hume thinks he has shown that none of Cleanthes’s arguments hold. For Hume there just is a problem of evil—anyone who is not a philosopher with an ax to grind can see that—and Philo challenges Cleanthes to make the case for God that, in face of the manifest evils of life, seems quite counterintuitive. If he cannot do that, then like it or not he will have no choice but to start with Philo from Princeton. I am not sure this makes Hume an atheist. For certain it makes him a Pyrrhonian skeptic.

God and moral evil: Julian of Norwich’s problem

Let us leave Hume there for the moment, because I now want to draw attention to a surprisingly different time, place, and style of reflection on the problem of evil—that of the fourteenth-century English theologian Julian of Norwich. Unlike Hume she believes categorically in the existence of a good and all-powerful God. That said, she shares one thing with him: she is quite baffled at the quandary that is caused by the quantity and viciousness of sin. What’s more, just as Hume refuses to eliminate the problem by way of atheism, so Julian refuses to dissolve the problem by way of theology. She confesses that she does not know why a good and almighty God should have created a world in which there is evil. You might find the parallel between this fourteenth-century woman recluse and the worldly Scottish enlightenment skeptic to be surprising and unlikely. But setting Hume and Julian alongside one another may, I hope, shed some light on a distinction that is hard to grasp and often misunderstood. This is the distinction between a Pyrrhonian skeptic like Hume, for whom it cannot be known whether there is an answer to the problem of evil at all, and a theologian of the apophatic persuasion like Julian, for whom there is an answer, although it is unknowable. Both claim not to know, but their differences show that not-knowing can come in very different kinds. 

Julian’s book of her “showings,” as she calls them, is an extended set of meditations on a central problem that besets her: she is profoundly troubled by her experience of evil, especially that consciously evil human behavior that she calls generically “sin.” And who would not be after reflecting on conditions in what must be the second nastiest century after our own, ravaged as it was by disease, death, war, poverty, malnutrition, starvation, and economic decline? Julian herself, at the age of eight or nine, had survived the Black Death, which in the space of two years took the lives of one third of the population of England. And in the face of her experience of the reality of evil—physical, social, and moral—she is told in her showings that God does not see sin, that for God sin is “no thing,” and that, contrary to all her own experience of human sinfulness, “all will be well, and all will be well, and every manner of thing will be well” (A Revelation of Love, chapter 27).

Julian is therefore confronted with a dilemma: the conjunction of her own intensely painful experience of sin (she says we experience it as “sharp pain”) and of the assurance that God does not notice sin compels her to seek some intellectual space within which the two conflicting propositions might be reconciled. You cannot sweep away the evil with some gesture toward the compensating goodness of God. Sin, she says, is real and inexplicable: it may be the source of—or may consist in—all sorts of illusions about ourselves, our fellow human beings, indeed about God. It may be the reason we fail to relate to others and ourselves as we should. But there is no sort of unreality in the fact of our failing to relate properly. The complex reality is that, on account of the world’s sin, unreality is the medium of our realities, of our actual relationships. This condition is the meaning of what Julian and Christians generally call the “fall.” And so the question that dominates her reflections is simple, and, remarkably, it is none other than Philo’s. Why, given a God who is omnipotent and all good, is there sin at all? For Julian as for Hume, it is a question that demands that she hold on to the dilemma without eliminating one of its horns. The omnipotent and unfailing love of God and the existence of sin are both undeniable. How, asks Julian, if we cannot deny either, can we assert both and hold the two in tension? There is sin. But why?

It is worth noting why this question seems so important to Julian and so odd to many of the philosophers and theologians of our own time. For the assumption equally widespread among both philosophers and theologians today is that the one thing for which you don’t need an explanation is that sin happens and is bound to. It seems to them, as it did to Hume, too obvious to be worth debating that if you create a world of free agents, where freedom must at least allow for the choice between good and evil actions, then necessarily some evil choices are going to be made. A world of completely free agents who never choose evil actions is certainly describable, but by strict logical necessity even an almighty God could not create one, since for God to cause a human world to be sinless would be for God to rob human choices of their freedom, it being assumed that no action of mine can be free if anything other than I is the cause of it. And that “anything other” includes God. So a world of sinless human beings is describable but uncreatable. In this view, Julian’s question “Why is there sin?” is redundant. Such, for example, is the view of the Calvinist philosopher Alvin Plantinga.

But Julian insists: God could indeed have created a world of human beings in which no one freely chooses to sin. That difference shows there is a vast spiritual, as well as intellectual, chasm between Plantinga, who evidently thinks that the human will could be free only if it occupied a space evacuated of the divine causal agency, and Julian, for whom, as for Augustine and Aquinas, our free choices are precisely where the presence of God’s agency is most evidently and directly working. You can see God acting directly in our free actions, for, precisely insofar as they are free, they are not subject to determination by natural causes, and so God’s agency is not mediated by them. Therefore God’s causality stands to my freedom only directly.

It is because Julian thinks God is the cause of our free choices that there is for her a real question as to why God did not create a world of free agents who freely choose not to sin. As Julian sees it, God could have done so. And so she tells us that

I saw that nothing stood in my way but sin. And I saw that this was so for us all generally, and I thought: if there had been no sin we should all have been clean and like to the Lord who made us. And thus...I often wondered why by the great foresight and wisdom of God sin had not been prevented. For then, I thought all should have been well.

Christian theologians, especially those who seek the escape route between the horns of Hume’s dilemma that is known as the “free-will defense,” should give serious thought to Julian’s doubt here. One of the reasons Julian is too often presented as a cheerful and empty-headed goody-two-shoes in some pious Christian circles is that readers today fail to see how far she is from pursuing the escape route from the problem of sin available in the conventional “free-will” defense. For Julian to think—as Plantinga does—of God’s causality as excluding my free will, in the same way that your jogging my elbow would exclude my spilling the tea voluntarily, is to misrepresent the nature of divine causality in its relation to human freedom. It is just to this point, if no further, that one can observe the parallel between Hume and Julian: both resist that way out of the problem of evil that resorts to excising one of the dilemma’s horns. Unlike Plantinga, Julian sees no problem of consistency in maintaining that God could have created a world of free but sinless human beings; like Hume Julian will not let Plantinga get away with it. Julian therefore has the same problem as Hume: if God could have created a world without sin, should he not have done so, since such a world would have been the best possible world? And would you not suppose that a perfectly good and all-powerful God would, as a matter of course, make such, and only such, a world?

Is ours the best possible world?

At this point Julian is fully engaged in a sort of dialogical argument with the Lord about the matter—chapters 27–32 of her Long Text read like a sort of conversational disputed question in the academic mode of the medieval university—and it is a subtle one, requiring the sort of careful reading that it doesn’t always get in the secondary literature of our time. It is true that the Lord rebukes Julian for her anxieties about sin, but what she thereupon corrects herself for having thought is not that God could have created a world of free human agents who did not sin. Instead, she corrects herself for assuming that all would have been well only in such a sinless world. Her inclinations on that subject, until corrected by the Lord, were the same as Hume’s. They were also the same as those of the biblical Job. What the Lord tells her is that all is well in just this sinful world, and that puzzles her all the more: in recounting this episode in her “showings,” Julian admits she had to be persuaded how what she is told could be true. For plainly this world of sin could not be the best possible world. And why, she asks, would God create a lesser good when a better, indeed best, world was available for the creating. But here her answer is, sensibly enough, that the question is falsely put. That’s because nothing could answer the description “best possible world.” There could no more be a best-possible world than there could be a best-possible sonnet or string quartet. Here again she agrees with Hume, who, attributing the view that ours is the best possible world to the great German philosopher Leibnitz, deemed it to be “bold and paradoxical,” but otherwise would not entertain it.

Of course it might seem quite natural and obvious for a theist to respond that there must be a best-possible string quartet, perhaps imagining that there is the string quartet God would compose were he to compose one, and that God could not possibly compose a second-rate piece of music when he could create the unsurpassably good one. In fact, even were there such a piece of music as the best-possible string quartet, God could not be under any constraint to compose only it, because if God can compose only the best-possible string quartet, then there is something that Beethoven can do that God cannot. For while Beethoven could, and did, compose his String Quartet in C# Minor (Opus 131), it would follow on this account that God could not have composed it, since, good as it is, no one would say that Beethoven’s Opus 131 is the best-possible string quartet. There is at least possibly a better one in the cards, one even better than whatever quartets Beethoven himself might have composed had he lived longer.

As it is with string quartets, so it is with worlds. In agreement with Thomas Aquinas, Julian does not think this is the best possible world. Like any medieval Christian believer, she thinks that a world in which everybody is totally free but of such a mind that sinning cannot come into the picture any more is not only possible but actual, because God has created exactly that state of affairs in creating the heaven that is offered to us after death. There we are guaranteed for all eternity to be utterly sinless and to be utterly free in our not sinning. So it is not on account of her disagreement with Plantinga that Julian is rebuked, because the Lord evidently agrees with her, not Plantinga, about his having the power to create a sinless world of free agents. Rather, she is rebuked for having supposed that “all manner of things (would be) well” only in such a sinless world.

The “behovely”

What Julian is told beyond that is a puzzle. She is told that in this world with all its sin, with all its “sharp pain,” nothing is “amiss.” In this world, sin is “behovely.” Here again Plantinga comes into the picture for a very particular reason having to do with a mistranslation of the word “behovely” in the older Penguin modernization of Julian’s text by Clifton Wolters and even in the more scholarly Colledge and Walsh modernized version. “Sin is necessary” is how both these versions have it. This translation suggests that in a world of free agents there will necessarily be sin. That, however, is Plantinga’s idea. It is not Julian’s.

Why is it so egregiously wrong to translate “behovely” as “necessary”? Because it involves a misunderstanding of the nature of the theological predicament that Julian feels constrained to address. For, as we have seen, “necessary” is exactly what Julian thinks sin is not. That is why, believing that things could have been otherwise, she has a problem Plantinga does not have. She needs to take an entirely different theological tack.

“Necessary” won’t do as a translation of “behovely” because “necessary” is a term forming a joint in a linear, inferential sequence: if this is the case, then that necessarily follows. Understanding sin as “necessary” would appear to be attractive to those for whom, maintaining that there could not be a creatable world without sin, a philosophical solution to the problem of evil is thereby made available. Because evil is necessary in any world that God can actually create, God can’t help but that there is sin. But Julian thinks there is no such solution, and her approach to the problem of evil evokes a quite different vocabulary of explanation. Just as the logician’s “if p, then q” is inferentially linear, so Julian’s “it is behovely that there is sin” is narratival. “Behovely” is the connective tissue of a storyline, not of a syllogism.

As a rough translation into modern English, Julian’s “behovely” means something like “fitting,” or “befitting,” implying that there is something that the behovely fits with and gets its sense from. Perhaps one could also translate it as “appropriate.” Or, as I would prefer, one could translate “sin is behovely” as “sin is just so.” To get closest to a distinctly medieval meaning of “behovely,” the best way to translate it is not with a modern English word, but rather with a medieval Latin term of theological art—namely, conveniens. One way of getting a grip on the non-logical character of the term is from a standard medieval question: “Was the Incarnation necessary?”—to which the received answer, from the time of Anselm’s late-eleventh-century treatise known as Why did God become Man? (Cur Deus Homo), was twofold. Absolutely speaking, the Incarnation was not necessary. God, after all, was under no necessity of nature, nor under any obligation in justice, to do anything at all about the sinful predicament of creatures; and, if he were to do anything about it, many possibilities of relieving that predicament were available to God other than the Incarnation—and all of them at lower cost. Yet the second person of the Trinity, the Word, was made flesh and dwelt among us. Why?

Duns Scotus said early in the fourteenth century that God became man not as a response required by an unanticipated event, nor as a solution to any kind of problem, at least not principally. Scotus thought that God just fancied the idea of becoming man regardless of whether Adam sinned or not. As Proverbs puts it, it was Wisdom’s “delight to be with the children of men” (8:31). On the other hand, if God was under no constraint of necessity to become man, neither was the Incarnation a mere whim on the Father’s part to send his Son into the world, to preach and suffer and die for the world’s sake. If he was neither under any necessity to do so nor merely indulging his power to do it, then the question why God chose to set in motion just those particular events needs to be understood other than in terms of either logical or natural necessity. For the question “Why did such and such happen?” we need instead a vocabulary that is closer to how one explains an event’s occurrence within a particular narrative—because you are explaining how the narrative makes sense of just that happening. And, as to the Incarnation, the term of art that, after Anselm, the medieval theologians used in answering the question “Why did God become man?” was that it was conveniens. It was not that it just so happened, as if by accident or on a whim. Nor was it a sort of Plan B, things having gone so terribly wrong with Plan A in the Garden of Eden. But though it wasn’t a necessity imposed on God, it was indeed “just right” that God should do it, it was conveniens—“behovely,” or perhaps “just the thing,” a Godlike thing to settle on such an over-the-top solution. And you can get to see just how right it was if you can get the hang of the story it fits within. In short, conveniens in Latin, or “behovely” in Middle English, are terms descriptive not of how logical and linear sequences are formed, but rather of how narratival and spiral sequences are formed. It is a term descriptive of what is just right about a good story, such that if you are to understand why an event happened you need some access to the whole story in which it takes place.

Sin is behovely

The vocabulary of the conveniens or the “behovely” is, then, that of connective tissue for the spiral of a narrative that accumulates meanings as it goes along and, as it accumulates those meanings, progressively demands the readjustment of the narrative curve. And no doubt some such notion of the “behovely” would provide an understanding of Julian’s theology were we considering only her description of the Incarnation itself in those terms. For her, the coming of Christ could not have been anticipated had it not been prophesied. We needed to be told because no necessity entailed it. But when the Incarnation happens, everything becomes clear: we have a new hermeneutic of the Old Law, obscure poetry that seemed to mean one thing becomes a hermeneutical key for everything—one has only to think of how differently the prophecies of Isaiah or Hoseah read when they are read, after the fact, in light of the Incarnation. The role of the term “behovely” in Julian’s theology would be perfectly clear and uncontroversial if she had merely been proposing the medieval commonplace that the Incarnation was in this way behovely.

But this is not all she says. What Julian calls behovely is “sin.” So much for Julian as the cheerfully upbeat dispenser of piously optimistic nostrums. The problem of credibility that Julian’s theology presents us with deserves to be faced squarely. Here we have a great theologian of the Christian Church telling us that sin is behovely. She tells us this not because she is naïve about the world’s evil but because, knowing the world’s evil for what it is, she believes that it follows from core Christian beliefs about God’s power and love; that sin so “fits” with the divine plan that nothing can be “amiss.” And this can seem implausible, even scandalously so. For we must suppose Julian’s theology to entail that behovely—and so not amiss—was the bureaucratic, cold efficiency with which the murder of 6 million Jews was planned and executed; behovely, the ideologically motivated mass exterminations of the Pol Pot regime; behovely, the frenzied pogroms of Rwanda and the mass rapes of Syria; behovely, the betrayals of every adulterous spouse; behovely, every lie told in breach of trust; behovely, every sexual abuse of a child; behovely, every rich person’s denial of food to the hungry. If these are not amiss, it would seem that nothing could be amiss. It would seem that Julian’s response to the problem of evil is simply to deny, a priori and in the face of the overwhelming weight of evidence to the contrary, that there is any possible evil that could be a problem for belief in God.

It has to be admitted that at this point in the argument it is all too easy for the contemporary theologian to lose his or her nerve—and that is why the free-will defense of Plantinga and others can seem to be the only way out for believers. You take Plantinga’s line in order not to be stuck with Julian’s and because you feel weighed down by the heavy burden of human evil. If all you need to say is that evils of such incalculable extent and intensity can’t be helped, that they wouldn’t be sins if they weren’t freely done, and that our not sinning would not be free if God had prevented it, then it’s all our fault and you cannot blame God for sin. In this view, a world without sin would be without humans, occupied only by automata preprogrammed by God. A sinless world is impossible given freedom, and without freedom there are no human beings.

For Julian, such a conception of God and human freedom is not all right. Significantly, Julian’s position is closer to that of most atheists of our times than it is to that of most contemporary Christian theologians, and I think it is closest of all to the skeptical Hume. In The Nature of Necessity (1974), Plantinga writes, “It was beyond the power of God himself to create a world containing moral good but no moral evil.” Julian disagrees. God could have created such a world and did not. Then there is John Mackie, an atheist precisely on account of believing Julian to be right and Plantinga wrong: “God was not…faced with the choice between making innocent automata and making beings who, in acting freely, would sometimes go wrong: there was open to him the obviously better possibility of making beings who would act freely but always go right. Clearly [God’s] failure to avail himself of this possibility is inconsistent with his being both omnipotent and wholly good” (“Evil and Omnipotence,” Mind, 63, no. 254) And here again is Julian, the fourteenth-century anchoress, sorely tempted to agree with Mackie: “And I thought that if sin had never existed, we should all have been pure and like himself, as God made us; and so in my folly I had often wondered before now, why, in his great forseeing wisdom, God had not prevented the beginning of sin; for then, I thought, all would have been well.” What is clear, then, is that Julian would rather have Mackie’s problem, with or without a solution, than be forced to conclude as Plantinga does, because the price of his conclusion would have to be paid in the currency of what a later age called “deism”—the doctrine of a God whose presence is expelled from that part of his creation that is most in his image and likeness: the human freedom of choice. Once again, Hume agrees: deism would be a way out, but Julian will have none of it, any more than will Demea in the Dialogues. Why, Demea asks, “should man pretend to an exemption from the lot of all other animals?”

So does Julian have a solution? She makes it every bit as clear as Hume does that neither she nor anyone else is in possession of one. But there is a difference between the two, and it is all the difference in the world. Hume does not say that there is no solution to the problem of evil. He says only that whether there is or is not cannot be known, at least not by “natural reason.” Julian, by contrast, knows that there is a solution, for “all manner of thing will be well,” but she also knows that we do not and cannot know how that could be so. Therefore, how we are to understand Julian’s skepticism in distinction from Hume’s depends on how we are to understand her when at the very end of her Long Text—after perhaps many more than twenty years of turning the problem of sin over and around—she tells us that, though “this book was begun by God’s gift and his grace,” it “is not yet performed, as to my sighte.”

Clearly, Julian does not mean by this that her text is unfinished. On the contrary, the statement that her work is incomplete is clearly intended as the appropriate and responsible conclusion to a theological treatise as carefully constructed as any in the fourteenth century. Nor does she mean that, though some theological progress has been made in the Long Text that she had not been able to make in her earlier, much shorter version, perhaps a third attempt at it might yield a text that is finally “performed,” or complete—a text that once and for all answered the question of “How could a good God allow sin?” in the sense that John Milton seemed to think was required of a theologically satisfactory response. Milton thought he could construct a narrative of paradise lost and regained that is able to “justifie the ways of God to men” (Paradise Lost). Julian’s text refuses completeness in Milton’s sense, in which the occurrence and pervasiveness of sin is said to have been theologically “justified,” for she thought no such justification of God’s ways is possible for us in our time. In not providing one, Julian’s text does not fail of completeness. In what sense, then, is her text not yet “performed”?

It is the “not yet” that matters here, the provisional. What is provisional is Julian’s theological refusal of both the logical completeness of Plantinga, which would purport to demonstrate the formal consistency of an infinite love’s creating just this sinful world, and the narrative completeness claimed by Milton, which would purport to finish the story that “justifies the wayes of God to men.” For all her doubts about how it could be true, Julian accepts that “sin is behovely.” But I think she knows enough about how the logic of the behovely works—as narratives do—to understand that we could see how sin is behovely only if we were in possession of the complete narrative that makes sense of it, and we are not. All we possess is but a narrative fragment, a torn-off corner of the manuscript of salvation history, and it tells Julian of nothing but the paradox of an innocent man judicially executed for a reason he too begs to know of, though he dies, as we will, the reason why denied us all. Somehow, Julian knows that the meaning of sin, its character as behovely, lies in that incomplete narrative of the Cross that is at the heart of her showings, a narrative whose incompleteness is necessary, for “not yet” belongs to the nature of human existence in time. It is thus that Julian’s apophatic theology moves far away from the skeptical Pyrrhonian world of David Hume.

Julian, then, cannot complete her “book,” for incompleteness is in the nature of the narrative spiral itself. It is not just sin’s being behovely that is being told by that narrative, it is we who are being told by it. For Julian, this includes her own two attempts to grapple with that fragment of the narrative that is shown to her. In that fragment is the meaning of sin, but it is hidden from her—it is, she says, a “great secret” the meaning of which is withheld from us. Julian knows that her attempts to lay hold of the complete story are themselves but episodes within it. As the postmodernists of our times would have it, she is being read by the narratives she believes herself to be writing; she is being told by the narrative she twice attempts to tell. What would answer Julian’s question “Why is there sin?” is the narrative completed, her “book…performed.” And that cannot be done within history. For the completed narrative is, literally, the end of the story. And that, Julian knows, is the beatific vision, the price of which is death.

In the meantime, there is but the meantime, the “not yet.” And Julian does her theology obedient to the temporality in which neither understanding nor living can yet be “completed.” Julian is the theologian she is because she knows that all theological writing submits to a necessary condition of incompleteness, and like Hume she refuses an easygoing and peremptory ultimacy. For writing that is pretentiously “finished” is not theological; it is parody. It is Jeremy Bentham’s “nonsense upon stilts,” the ridiculous parading as the sublime. Julian’s theology is truly spiral. It begins and ends where unending begins, as T. S. Eliot says. And maybe Eliot did get it right. At any rate, he got it from his reading of Julian.