[In the following essay, Kivy argues for the restoration of Candide's status as a text of philosophical significance.]

I

Voltaire's Candide is subtitled Optimism. It is about an impossibly naive young man who suffers incredible misfortune, while counselled by his teacher, Pangloss, to perceive the hidden benefits that this merely “apparent” misfortune and misery produce. Pangloss' speeches in this regard are well-larded with phrases and terms coined or made famous by Leibniz, and, so as not to leave the connection merely hinted at, the name of the philosopher himself is also invoked, as for example, where Voltaire has Pangloss say: “I still hold my original views, for I am still a philosopher. It would not be proper for me to recant, especially as Leibniz cannot be wrong; and besides, the pre-established harmony, together with the plenum and the materia subtilis, is the most beautiful thing in the world.”¹ Pangloss' various “explanations” and “arguments” are, of course, laughable and absurd. They are, on that account, the very soul of the book. And on that account, too, they, and the book, have been taken to be a devastating reductio ad absurdum of the Leibnizian optimism alluded to in the subtitle. One of the literary values of Candide, then, is supposed to be a philosophical one: its “refutation” of Leibniz; and it is to some considerable degree, at least, because of this philosophical value that Candide has maintained its reputation as a “classic” of modern European literature.

This traditional “textbook” view of Candide might be thought of as a conjunction of the following assertions:

• (1) Voltaire intended to “refute” Leibnizian optimism in Candide.
• (2) Pangloss faithfully represents Leibniz's position.
• (3) The combination of Candide's misadventures and Pangloss' laughable “arguments” constitutes a successful reductio ad absurdum of Leibniz's view.
• (4) The author's intention is realized.
• (5) The philosophical success of Candide is one, if not the chief, reason for placing a high value on the work.

But, clearly, nothing that happens or is said in the book succeeds in logically “refuting” the Leibnizian. So if Voltaire really did intend to “refute” Leibniz, he failed in his intention. Therefore, whatever its virtues, Candide is bad philosophy, and cannot be justly admired for any philosophical merit. In some crucial way, then, the “textbook” view of Candide must be defective.

In what way (or ways)? Erich Auerbach writes in Mimesis that “The novel Candide contains a polemic attack upon the metaphysical optimism of Leibniz' idea of the best of all possible worlds,” and adds: “… Voltaire in no way does justice to Leibniz' argument and in general to the idea of a metaphysical harmony of the universe, especially since so entertaining a piece as Voltaire's novel finds many more readers than the difficult essays of his philosophical opponents, which cannot be understood without serious study.”²

Two possible views are suggested by Auerbach's remarks. The first, what might be called a “frontal attack,” can be summarized as follows:

• (1) Voltaire intended to “refute” Leibnizian optimism in Candide.
• (2) Pangloss faithfully represents Leibniz's position.
• (3a) The combination of Candide's misadventures and Pangloss' laughable “arguments” does not constitute a successful reductio ad absurdum of Leibniz's view.
• (4a) The author's intention is not realized.
• (5a) The work is a philosophical failure, and, on that account, the philosophical content of Candide cannot be a reason for placing a high value on the work.

But a second view is also suggested, charging Voltaire not only with bad arguments but with serious misrepresentation of Leibnizian optimism. This “two-pronged” attack might go something like this:

• (1) Voltaire intended to “refute” Leibnizian optimism in Candide.
• (2a) Pangloss does not faithfully represent Leibniz's position.
• (3b) The combination of Candide's misadventures and Pangloss' laughable “arguments” constitutes a successful reductio ad absurdum of Pangloss' weak form of optimism, but not Leibniz's stronger and more sophisticated one.
• (4a) The author's intention is not realized.
• (5b) The philosophical success of Candide is a rather limited, trivial one, and cannot, on that account, be a reason for placing a very high value on the work.

The second of these attacks suggests a riposte: that Voltaire really had no intention of “refuting” Leibniz; that, as John Butt puts it, “Though he ridiculed Leibniz' terminology, … he attacked, not Leibniz' philosophy, but its popular perversions.”³ Thus, it is hardly fair, on this construal, to criticize Voltaire for not representing the philosophy of Leibniz accurately, and failing, therefore, to refute it, since he had no such intention in the first place. He was not, in other words, guilty of ignoratio elenchi. We arrive, then, if we accept this view, at a revised version of the “textbook” interpretation, which goes this way:

• (1a) Voltaire intended to “refute” a weak, perverted form of optimism in Candide.
• (2b) Pangloss faithfully represents this ersatz optimism.
• (3b) The combination of Candide's misadventures and Pangloss' laughable “arguments” constitutes a successful reductio ad absurdum of Pangloss' weak form of optimism, but not Leibniz's stronger and more sophisticated one.
• (4) The author's intention is realized.
• (5b) The philosophical success of Candide is a rather limited, trivial one, and cannot, on that account, be a reason for placing a very high value on the work.

Whatever textual evidence there might be for rejecting the revised “textbook” version, particularly its first premise (and I think there is sufficient), there is also what might be called a “sentimental” reason. We want to make the best of our well-loved Candide; we do not want it to be consigned to the scrap-heap of the philosophically trivial. What, after all, is left of the work, to justify its reputation as a “serious” classic, if it is divested of its philosophical reputation? Is it nothing more than an amusing diversion? This, of course, is not an argument, but merely a motive for finding one. What I wish to do is rescue Candide for philosophy, and restore that part of its reputation that rests thereupon. This requires walking tippy-toe between the “textbook” version, its revision, and such criticisms as Auerbach's. What is wanted is an interpretation that will preserve our traditional intuitions about the work: namely, that Voltaire succeeded in his intention, that Candide is “about” Leibnizian optimism and not some trivial kind, that Pangloss does fairly represent the Leibnizian, that its philosophical success is one of the chief reasons for placing a high literary value on the work. At the same time, a valid interpretation must, I think, acknowledge that the Leibnizian is not “refuted” by Candide.

What will be defended, then, is a fresh revision of the “textbook” interpretation which accepts unrevised its second, fourth and fifth premisses, and accepts in altered form the first and third. In this fresh revision, the “textbook” interpretation will preserve Candide's philosophical respectability, while admitting that some criticism of the “textbook” interpretation is well-taken (namely, part of Auerbach's), and some revision of the “textbook” interpretation required (but not Butt's). In order to accomplish this, we must now turn to the so-called “problem of evil,” and the Leibnizian solution. For the revision of the “textbook” interpretation I have in mind depends on a particular view of that solution: a view available to Voltaire but neglected by Candide's commentators, a view philosophically respectable and, perhaps, correct.

II

As is well known, the defining qualities of God seem incompatible with the existence of evil in the world. If God is omnipotent, he is able to prevent unmerited pain and suffering (and from now on I shall take “evil” and “unmerited pain and suffering” to be synonymous). If he is omniscient, he must know of their existence. If he is good, he must desire to prevent it. Yet evil is palpably present in the world. How can this be, given the omnipotence, omniscience, and moral perfection of the Deity?

To answer this question, Leibniz asks us to imagine Creation as the Deity's choice among competing “possible worlds,” that is, complexes of logically possible states of affairs. His choice is directed by his goodness; and so, of the logically possible worlds, he chooses that which, through his omniscience, he determines to be the best: thus, all things considered, the one with the least possible unmerited pain and suffering. This is how Leibniz puts it in the Monadology:

53. Now since there is an infinity of possible universes in the ideas of God, but only one can exist, there must be a sufficient reason for God's choice which determines him to one rather than another.

54. This reason can be found only in the fitness or in the degrees of perfection which these worlds contain, each possible one having a right to claim existence in the measure of the perfection which it enfolds.

55. And this is the cause for the existence of the best, which his wisdom causes God to know, his goodness makes him choose, and his power makes him produce.⁴

A word or two will be necessary here with regard to the “all things considered” clause inserted above. For, on Leibniz's view, the amount of unmerited pain and suffering in the world is not God's only moral consideration. There may be, in other words, possible worlds with less unmerited pain and suffering than our own. In particular, God thought it best to actualize that possible world with the fullest plenitude of being, “like … a learned author who includes the greatest number of subjects in the smallest possible volume.”⁵ Thus, we must assume, our world contains the least amount of unmerited pain and suffering commensurate with this fullness of being, and whatever else, in God's view, constitutes moral perfection. “It is true,” Leibniz writes in the Theodicy, “that one may imagine possible worlds without sin and without unhappiness, and one could make some like Utopian or Sevarambian romances; but these same worlds again would be very inferior to ours in goodness.”⁶ When we refer, therefore, in what follows, to the Leibnizian claim that this world contains less unmerited pain and suffering than any other possible, it must be understood always as the least amount consistent with all that goes into making this world the most perfect: the “all things considered” clause must be understood throughout.

Now Leibnizian optimism is frequently described as a position which solves the “problem of evil” by denying that evil exists. In a sense this is so, and in another it is not. Leibniz, after all, does not deny that pain and suffering exist. What he claims is that as little unmerited pain and suffering as possible exist, that God chose to actualize that possible world containing the least amount of evil. So one way of looking at the Leibnizian solution is not as a denial of the existence of evil but as a denial of God's omnipotence. There is something that God cannot do: he cannot do the logically impossible; logical necessity is a constraint on his power. God, then, has permitted evil in the world, but he is not culpable, since he has permitted the least logically possible; and to do better would have been to do the logically impossible, which is beyond his (or anyone's) power. Perhaps it is absurd to think of “impossibility” in the logical sense as a limitation of anyone's power, even God's. This appears to be the view of C. S. Lewis, in The Problem of Pain, where he writes that “It is no more possible for God than for the weakest of His creatures to carry out both of two mutually exclusive alternatives; not because His power meets an obstacle, but because nonsense remains nonsense even when we talk it about God.”⁷ For those, however, who think it intelligible to talk about the logically impossible being a constraint on the power of the Deity, it may seem plausible to represent the Leibnizian answer to the problem of evil in this way: as a denial of God's absolute omnipotence.

But if we mean by the world containing evil, that the world could have been better than it is, could, that is, have contained less unmerited pain and suffering, and if we mean to convey by could the logical possibility, then Leibniz is denying the existence of evil; for he is denying that there is a logically possible world, all things considered, better than our own. If there were, God would have chosen it rather than the one he did choose. We shall, therefore, henceforth represent Leibnizian optimism as denying the existence of evil and holding all purported cases of evil to be “apparent” evil only, in the sense outlined above, always with the caveat that this may not be the only way of representing the matter.

What now of the appearance of evil? Needless to say, we are not, like God, omniscient; and that is why for us, and not for God, there is apparent evil. God sees in every particular, why and how every instance of unmerited pain and suffering is the least possible, all things considered; but we cannot. “I cannot show you this in detail,” Leibniz says; “For can I know and can I present infinities to you and compare them together?”⁸ So we are in thrall to an illusion. How can we liberate ourselves? We can, at least, be given a paradigm explanation from our own everyday experience, which provides, as it were, a recipe—a recipe, however, that we cannot follow except as a kind of Gedanken-experiment. It is as if we were given the recipe for dispelling bent-stick illusions, “If you take the stick out of the water, it will look straight,” when we are never in a position actually to be able to pull sticks out of water.

The recipe for dispelling the illusion is taught us by our perfectly ordinary experience of things “working out”—of apparently unalloyed misfortunes revealing themselves in the event to be blessings in disguise. “We know … that often an evil brings forth a good whereto one would not have attained without that evil.”⁹ I have to make a flight to Seattle. I leave in plenty of time, but my taxi has a flat tire, and I miss my plane. Then I hear that the plane I missed has crashed, killing all on board. I now see my flat tire not as a misfortune but an apparent misfortune only, and, in reality, the most fortunate of accidents.

All we need now do is see every case of pain and suffering as a case like the flat tire, or a visit to the dentist, and we dispel the illusion of evil. But we can do this only in principle. We cannot actually trace out all of the consequences of the Lisbon earthquake that make the world in which it occurs the best possible: the one, all things considered, with the least amount of unmerited pain and suffering. Only God can do that. All we can do is begin the process by pointing out some of its beneficial consequences, and add that it is logically possible, at least, that when all of the consequences are weighed and measured, their benefits will outstrip the palpable evil of the event. That is to say, when the pain and suffering of the Lisbon earthquake, and whatever positive beneficial results it might have, are tallied up with all of the assets and debits of this world, and compared with all of the possible worlds which did or did not contain the Lisbon earthquake, it will be seen that our world, with the Lisbon earthquake, contains less evil, less unmerited pain and suffering, than any of these others.¹⁰ So, to get a little ahead of our story, when Candide explains optimism as “the passion for maintaining that all is right when all goes wrong with us …,”¹¹ we need only read “apparently wrong” for “wrong” to see that this depiction, in its way, is as accurate as it is funny. It must be the passion of the Leibnizian to dispel the “illusion” of evil with his paradigm, his recipe, just as it must be the passion of the Parmenidean to dispel the “illusion” of motion.

III

These are the bare bones of Leibnizian optimism, as I view it. Needless to say, to flesh it out fully, and do justice to its metaphysical niceties, lies beyond the scope of the present paper. What we must now ask ourselves is: Can the optimist be thwarted? Can his position be “refuted”? If by “refuted” we mean shown to be logically inconsistent, made the object of a formal reductio ad absurdum, the answer of Philo in Hume's Dialogues Concerning Natural Religion is “No”—and I suspect he is right. Leibnizian optimism does successfully, irrefutably show that one can consistently (although improbably) believe God is omnipotent, God is omniscient, God is good, in face of all the unmerited pain and suffering in the world. “I will allow,” Philo is forced to admit, “that pain or misery in man is compatible with infinite power and goodness in the Deity. …”¹² It is, as he phrases it, a “mere possible compatibility”; but as such it stands beyond logical refutation. How could we, after all, prove that, all things considered, this is not the world with the least amount possible of unmerited pain and suffering, since, ex hypothesi, we can never consider all things. Which is to say no more than that Leibnizian optimism might just possibly be true, as might the belief that Hitler is alive and well in Argentina.

On Philo's view, then, it would be pointless to try to “refute” optimism: a formal reductio is impossible. Is optimism, then, completely impregnable? By no means.

Let us suppose, for the sake of argument, a man I shall call “Pangloss.” How close he comes to being the character in Voltaire's book I will let the reader judge. Pangloss is certain that (i) God exists, (ii) God is omnipotent, (iii) God is omniscient, (iv) God is perfectly good. For him it is a necessity to embrace optimism in order to make his beliefs logically coherent. Since he is completely unwilling to give up (i), (ii), (iii), or (iv), he must accept optimism, no matter how unlikely it might seem in the face of pain, suffering, and world catastrophe. And logically his position is impregnable: we cannot charge him with inconsistency. Further, his belief in God's existence and nature is on a priori grounds: that is to say, he sees (i) through (iv) as the conclusion of a valid deductive argument with intuitively certain premisses. We cannot, therefore, do anything in the way of ratiocination to dislodge his optimism, short of beginning at the beginning, with his a priori proof (or in Leibniz's case, proofs) of God's existence and attributes, and showing the proof invalid or the premisses false. We can, however, get at Pangloss' optimism directly, without making this “end run,” in another way. We can, by bombarding him, so to speak, with the evil of the world, by, in other words, sticking his nose in it, force his belief in optimism closer and closer to the zero point, and, in so doing, put such a strain on his theological belief system that it may crack, thus impelling him to a reexamination of its a priori grounds. This is not “refutation” in the sense defined above: it is not, that is, a logical demonstration of falsity. So let us call it instead “confounding,” which will be, in this paper, a term of art.¹³

What I would like now to suggest is that we can best understand Candide by ascribing to Voltaire a view of Leibnizian optimism very much like Philo's in the Dialogues Concerning Natural Religion. Whether it is Hume's view is another question. But that a character in the Dialogues presents such a view shows, at least, that it was a view not unknown in the eighteenth century. That being the case, it would be no anachronism to impute it to Voltaire.¹⁴

But would the imputation be justified? There is a rather enticing passage in the Philosophical Dictionary—the article “Théiste”—which does tend to support the notion that Voltaire shares Philo's view. Here is the opening:

The theist is a person who is firmly persuaded of the existence of a Supreme Being, as good as he is powerful, who has created all extended beings, vegetative, sentient, and thinking; who perpetuates their species, who punishes crimes without cruelty, and repays virtuous actions with kindness.

The theist does not know how God punishes, how he bestows favors, how he pardons; for he has not the temerity to flatter himself that he knows how God acts; but he knows that God does act, and that he is just. The objections to [this] providence do not shake his belief in the least, for they are only great difficulties, and not proofs.¹⁵

This passage can be interpreted somewhat along the following lines. The theist believes that there is a God; that he is omnipotent and morally perfect. He believes, further, that God makes good the apparent inequities and injustices in the world: in other words, the theist is an optimist. And if the order in which Voltaire presents the theist's position has any significance, it is to suggest that the theist's optimism follows from—is a logical consequence of—his prior commitment to the existence of an omnipotent and morally perfect Deity.

But, Voltaire adds, there are serious objections to the notion of a Divine Providence that governs a morally perfect world. And these objections can be none other, it is safe to assume, than those embodied in the traditional problem of evil. Yet the aforementioned objections do not shake the theist's belief system, because they are only great difficulties (grandes difficultés), not demonstrations (preuves). That is to say, the palpable existence of unmerited pain and suffering in the world is not a formal reductio ad absurdum of theistic optimism, although it does present the theist with the not inconsiderable problem of accounting for the seeming moral imperfections of what he must believe is a perfect universe. The suggestion, then, is that optimism is a difficult position to hold, but not a logically contradictory one: exactly the view that Philo later is driven to in the Dialogues Concerning Natural Religion.

If, like Philo, Voltaire thought Leibnizian optimism logically impregnable, he clearly could have had no intention of logically “refuting” it in Candide; for he would have thought “refuting” it—i.e., proving it logically untenable—a logical impossibility. Nothing in the novel “implies” that Leibnizian optimism is false. But the reason for that is not, as Auerbach seems to suggest, a failure on Voltaire's part. It is Voltaire's view, I am arguing, that such a refutation is impossible. It is pointless, therefore, to blame him for not providing one: he never intended to.

What Voltaire did intend to do is to “overwhelm” the Leibnizian with the palpable pain and suffering in the world, in the hope of straining his belief system to the breaking point. Indeed, there is almost an echo of Voltaire's strategy in a speech of Demea's, from the same section of the Dialogues in which Philo's previously quoted remarks occur. “Were a stranger to drop, on a sudden, into this world,” Demea says,

I would show him, as a specimen of its ills, an hospital full of diseases, a prison crowded with malefactors and debtors, a field of battle strewed with carcases, a fleet foundering in the ocean, a nation languishing under tyranny, famine, or pestilence. To turn the gay side of life to him, and give him a notion of its pleasures; whither should I conduct him? to a ball, to an opera, to court? He might justly think, that I was only showing him a diversity of distress and sorrow.

That “stranger” is Candide, the fictional tabula rasa, as his name of course is meant to suggest. He is the instrument by which the “diversity of distress and sorrow” in the world is brought into sharp focus, and made the unavoidable object of the optimist's attention. In this way, Voltaire hoped, the premise that evil is illusory would be made to seem so unlikely as to lead, finally, to the optimist's rejection of the theological belief system which makes its acceptance necessary for him.

What part of that belief system did Voltaire intend the Leibnizian to abandon? Not, clearly, the belief in God's existence. Voltaire was no atheist, if the Philosophical Dictionary is any indication; for it is full of enthusiastic endorsements of the “argument from design” for the existence of God.¹⁶ More likely, Voltaire inclined towards a denial of God's omnipotence—and this on explicitly stated Humean grounds, which can only lend support to a general Humean interpretation of Candide. The argument of Hume's, to which Voltaire refers, is to be found in Section 11 of the first Inquiry (“Of a Particular Providence and of a Future State”). In Voltaire's formulation: “The celebrated historian philosopher, David Hume, says in ‘Particular Providence’: ‘A weight of ten ounces is lifted in a balance by another weight; therefore this other weight is of more than ten ounces; but one can adduce no reason why it should weigh a hundred ounces.’”¹⁷ The conclusion of Hume's argument, which Voltaire puts to its intended theological use, is that “If the cause be known only by the effect, we never ought to ascribe to it any qualities beyond what are precisely requisite to produce the effect. …”¹⁸ We observe a world of immense size and complexity. From that we can conclude that the Creator of our world must be immensely powerful and capable of producing enormous complexity in his creations. But we cannot conclude that the Creator is infinite or omnipotent, since we do not know if our world is infinite or producible only by omnipotent power. “One must be very powerful,” Voltaire observes, “to have caused to be born spiders which spin webs to catch flies; but that is not to be omnipotent, infinitely powerful.”¹⁹ Thus there is no reason for us, in the face of palpable evil, to be Leibnizian optimists. “If the great Being had been infinitely powerful, there is no reason why He should not have made sentient animals infinitely happy; He has not done so, therefore He was not able.” So much then, on Voltaire's view, for the “problem of evil.” “This necessity [to which even God is hostage] settles all the difficulties and finishes all the disputes.” All is not well: “We say—‘All is the least bad that is possible.’”²⁰ That is to say, all is the least bad physically, empirically possible, given the physical, empirical infirmities of God, and given the natural necessities, the laws of nature, which neither God nor man can thwart.

We have established, then, that, on Voltaire's view, the Leibnizian is logically secure, but that pressure can be put on his optimism, for the purpose of shaking his belief in the omnipotence of God. On the strength of this, we are now in a position to present our revision of the first premise of the “textbook” interpretation of Candide; it is:

(1b) Voltaire intended to “confound” (but not “refute”) Leibnizian optimism in Candide.

IV

What, now, of Pangloss? Is he a fair literary version of the Leibnizian, given the humorous intent of his creator? In a recent article, Carolyn Korsmeyer has done much to restore Pangloss' good character as a disciple of the master; and I can confidently refer the reader there for further enlightenment.²¹ But for the present purposes, one word of caution ought to be enough to give Pangloss all the Leibnizian credentials he requires. We are not to look to Pangloss for a demonstration of optimism. The proof lies in the argument (or arguments) for the existence of God, and for the divine attributes of omnipotence, omniscience, and goodness that have preceded (we must imagine) the events befalling Pangloss in Candide. When he enters the stage in the first chapter, as teacher of “metaphysico-theologo-cosmolonigology,”²² he is already committed, on a priori grounds, to the theological premisses necessitating his optimism. His role thereafter is to provide what I earlier called the “recipe” for dispelling the illusion of evil that its denial requires. As Professor Korsmeyer correctly observes, “Pangloss's extended arguments generally have to do with the connectedness of events and with the view that evil in the world is illusory since it contributes to the good of the whole.”²³ In what is perhaps his most masterful Leibnizian disquisition, Pangloss argues that “There is a chain of events in this best of all possible worlds; for if you had not been turned out of a beautiful mansion at the point of a jackboot for the love of Lady Cunégonde, and if you had not been involved in the Inquisition, and had not wandered over America on foot, and had not struck the Baron with your sword, and lost all those sheep you brought from Eldorado, you would not be here eating candied fruit and pistachio nuts.”²⁴ Here is just the sort of “recipe” for dispelling the illusion of evil that optimism requires. But what this is frequently mistaken for, I think, is a parody of a proof, by the good effects of “evil” events, that optimism is true. And seen in this way it is, to be sure, a gross misrepresentation of Leibnizian optimism. Seen correctly, however, it is not a representation of the Leibnizian proof of optimism, but a necessary appendage: an instance of the paradigm argument for dealing with the “apparent” evil of unmerited pain and suffering.

Premise (2) of the “textbook” interpretation, to the effect that Pangloss fairly represents the Leibnizian, can be accepted, then, without revision, with the understanding that we do not confuse the recipe for dispelling the illusion of evil with the Leibnizian proof of optimism. It is the former that Voltaire puts in Pangloss' mouth, not the latter, which, indeed, never appears in Candide at all; for that proof amounts to the argument for God's existence and the divine attributes of omnipotence, omniscience, and perfection.

The revision of the third premise of the “textbook” interpretation follows directly from the revision of the first. Voltaire intended to “confound” rather than “refute” optimism. And the realization of this intention is perfectly compatible with the failure to provide a logical reductio ad absurdum, something which Voltaire, if I am right, thought impossible. Thus the third premise reads:

(3c) The combination of Candide's misadventures and Pangloss' laughable arguments does not constitute a reductio ad absurdum of Leibniz's view but a devastatingly successful “confounding” of it, which is all that is possible, given its logical impregnability.

In what way is Candide a successful “confounding”? The key is given by Candide himself: “The terrified Candide stood weltering in blood and trembling with fear and confusion. ‘If this is the best of all possible worlds,’ he said to himself, ‘what can the rest be like?’”²⁵ In other words, when we are brought face to face with the immensity of pain and suffering in this world, that there is no logically possible world that has less, no logically possible world the least little bit better, becomes well nigh incredible: not logically impossible, but so remotely possible as to approach probability zero. We are asked to believe that almighty power, all-seeing intellect, and holy will cannot have made a world more perfect in any respect than the charnel-house in which we live. This is the whopper the Leibnizian must accept, as Candide so forcefully impresses upon us.

We are now, finally, in a position to justify the fourth and fifth assertions of the “textbook” version. Voltaire's intention has been realized in Candide: it is a successful “confounding” of Leibnizian optimism, and so this, its philosophical success, can be a reason for placing a high value on the work, if philosophical success is a literary value in the first place (and I have no reason to believe it is not). The lack of “rational argument” in Candide, far from being a philosophical failure, is, on the contrary, the result of a sound philosophical insight: that optimism is logically unassailable, and formal refutation a misguided endeavor. What Voltaire intended, instead, was to “confound” the optimist by putting his belief system under stress. Neither optimism, nor pessimism, for that matter, can be “refuted” by confrontation with the world. Indeed, the optimist's “recipe” for dispelling the “illusion” of evil is a double-edged sword, serving equally well as a “recipe” for dispelling the “illusion” of goodness, since, just as it is logically possible that every cloud has a silver lining, it is also logically possible that every silver lining has a cloud.

V

Candide is a philosophical success because it does not attempt the impossible, but accomplishes the possible supremely well, at least so far as humor can do, which is not to say that a walk through Auschwitz or a week in a cancer ward cannot do more to crush the spirit of optimism. Yet, in conclusion, it might be well to observe that there is more to be said for comedy as a weapon against optimism than simply that it relieves us of the terrible burden of staring pain and misery full in the face. Ironically, it was said for the Enlightenment by (among others) one of its most confirmed optimists, as well as one of Voltaire's favorite targets, the Third Earl of Shaftesbury. “Truth,” he wrote in his defense of the freedom of wit and humor, “… may bear all lights; and one of those principal lights, or natural mediums, by which things are to be viewed, in order to a thorough recognition, is ridicule itself, or that manner of proof by which we discern whatever is liable to just raillery in any subject.”²⁶ No one understood better than Voltaire what ridicule could do to test the optimist's creed. More important, no one knew better what it could not—and that, I have been arguing, is the signal virtue of Candide most often perversely seen as its fault.²⁷

Notes

1. Voltaire, Candide, trans. John Butt (Harmondsworth: Penguin Books, 1976), p. 136.

2. Erich Auerbach, Mimesis, trans. Willard Trask (Garden City, New York: Doubleday Anchor Books, 1957), pp. 358-60.

3. In the introduction to his translation of Candide, p. 8.

4. Gottfried Wilhelm von Leibniz, Philosophical Papers and Letters, trans. and ed. Leroy E. Loemker (Chicago: University of Chicago Press, 1956), II, p. 1053. The Monadology, of course, like most of Leibniz's philosophical works, was not published in his lifetime, or in the period immediately following; but the same views are expressed in the Theodicy, which was. I have chosen the passage from the Monadology simply because it presents in a small compass all the relevant points.

5. Discourse on Metaphysics, Loemker, I, pp. 468-69. The Discourse, like the Monadology, was not published in Leibniz's lifetime, and, indeed, was only first printed in 1846.

6. Theodicy: Essays on the Goodness of God, the Freedom of Man, and the Origin of Evil, ed. Austin Farrer, trans. E. M. Huggard (London: Routledge and Kegan Paul, 1951), p. 129.

7. C. S. Lewis, The Problem of Pain (New York: Macmillan, 1962), p. 28. Cf. Lloyd Reinhardt, “Metaphysical Possibility,” Mind 87 (1978): “when we come … to things of a logical sort …, it is misleading to talk of anything of a logical sort preventing something” (p. 217).

8. Theodicy, p. 129.

9. Ibid.

10. Strictly speaking, on Leibniz's view no other possible world but ours could contain the Lisbon earthquake, since the Lisbon earthquake, the event that occurred in our world, is defined by its relations to all the other events in our world. There could, of course, be events very like it in other possible worlds.

11. Candide, p. 86.

12. Part X (text of the 2nd ed., 1779).

13. I am grateful to Ann Watts for suggesting this term.

14. Needless to say, I am not claiming that the text of Hume's Dialogues influenced Voltaire, as Candide predates it.

15. Voltaire, Dictionnaire philosophique, ed. Etiemble (Paris: Garnier, 1967), p. 399. I am grateful to my wife, Lindley, for helping me with this translation.

16. As, for example: “Certainly the world is an admirable machine: therefore, there is somewhere or other in the world an admirable intelligence. This is an old argument, but none the worse for that.” Dictionnaire philosophique, p. 460.

17. Voltaire's Philosophical Dictionary, no translator credited (Cleveland and New York: World Publishing, 1943), p. 240.

18. David Hume, An Inquiry Concerning Human Understanding, Section 11.

19. Voltaire's Philosophical Dictionary, p. 243.

20. Ibid., pp. 243-44.

21. Carolyn Korsmeyer, “Is Pangloss Leibniz?,” Philosophy and Literature 1 (1977): 201-208.

22. Candide, p. 20.

23. “Is Pangloss Leibniz?,” p. 203.

24. Candide, p. 144.

25. Ibid., p. 37.

26. Anthony, Earl of Shaftesbury, “Sensus Communis: An Essay on the Freedom of Wit and Humour,” Characteristics of Men, Manners, Opinions, Times, ed. John M. Robertson (New York: Bobbs-Merrill, 1964), I, p. 44.

27. I am grateful to my colleagues in the English Department, Irwin Primer and Ann Watts, for many helpful suggestions both as to style and content. They are not responsible, of course, for my few instances of stubborn refusal to be guided by surer hands.