The jacket design of Thomas Kavanagh’s book depicts a naked Fortune led by the goddess Folly, amid a shower of stock certificates. Their chariot is pushed on by besotted investors and crushing others beneath its wheels. It is, in other words, a picture of the Mississippi “bubble” of 1719–1720, the disastrous speculative scheme launched in France by the financier John Law for investment in the colonies. But the picture could perhaps, with a little adjustment, be made to represent the “History of Ideas” led on by the false enchantress “The Enlightenment.” Was there ever such a spreader of chaos as that unfortunate term? One is thinking especially of its “The,” which crept in during the twentieth century (perhaps through a mistranslation of the “Die” in “Die Aufklärung“) and which has a fatal effect, making “The Enlightenment” a thing, a discrete event, a completed process or fully formed Weltanschauung. The phrase was no doubt coined in an effort to copy the dazzling success of Burckhardt’s “The Renaissance”; so it is important to see the basic reason why, unlike Burckhardt’s phrase, it is not acceptable. Eighteenth-century writers played endless intricate games with the term “enlightenment” (lumières, Aufklärung, etc.), games which it is very important for the historian to study. Thus he cannot afford to use this loaded term himself, as a neutral item in his professional vocabulary—any more than phrases like “phlogiston” or “the true church,” or “true nobility.”

I honestly think that it is this phrase which has given a skew to Enlightenment and the Shadows of Chance. It is in a way an attractive and rewarding book, but an extremely odd one. Its thesis is that the representation of chance and gambling in the French eighteenth-century novel is intimately connected with the rise and development of mathematical “probability theory.” One sits up at this intriguing proposition and looks forward with relish to finding it proved. But—here comes the strangeness—the book not only does not really demonstrate it, but addresses itself to proving something quite different.

The mathematics of probability, as is well known, was launched in its modern form by Pascal and Fermat, in the shape of an argument about the odds in gambling. It was then elaborated by Huygens and Jacques Bernoulli, was extended beyond the natural sciences by Laplace, and eventually, in the nineteenth century, it fathered the development of general statistical theory, which now governs so much of our thought and rules so much of our lives. Ian Hacking, whose book The Emergence of Probability is partly what inspired Kavanagh, has spoken of probability theory as the “great philosophical success-story” of the last three centuries. “Other philosophical ideas have waxed and waned and sometimes grown again, but probability has been monotone. It has waxed and waxed, shone and shone. It has been a success in metaphysics, epistemology, and pragmatics, to mention three of the classic philosophical fields.”

That there could be a mathematics of chance was, as even a non-mathematician can see, a momentous discovery. It meant the dethroning of the goddess Fortuna and her wheel and the end of reverence for hazard as a mysterious agency. The rational man was now invited to scorn a belief in chance as “the superstition of the vulgar” and to accept that the word “chance” was merely another name for our ignorance. Further, though Pascal was a devout theist and invoked probability theory as a support for the Christian “wager,” others saw the theory as a blood brother of necessitarianism, or what in the nineteenth century came to be called “determinism”—a view of the universe as in no need of a Providence and as running according to discoverable Newtonian laws.

There was, however, as should be remembered, room for a fair amount of confusion here. For the beauty of probability theory was not that it abolished chance, which is not possible, but that it sidestepped it. “Chance,” and the ignorance for which it is another name, remained (and remains) installed in the middle of our life; nor will it ever be possible for us to know whether a single tossed coin will fall heads or tails. What had been discovered, though the fact was not always grasped, was not a cure for human ignorance but a way of living with it. As for necessitarianism, it had in fact no close logical link with probability theory, but the two seemed to get on well together.

Kavanagh tells us that the original subject he set himself was “the questions of chance, gambling, and the aleatory” in the eighteenth-century novel. But then the thought occurred to him: Was it not significant that the rise of the novel coincided with the rise of probability theory? He concluded that it was and that “classical probability theory paralleled and in a very real sense sustained the emergence and hegemony of the modern novel.” The two were “part of a single shift in our understanding of the world and of how we represent our place within it.” Also he had a further thought: it was an important paradox of “the French Enlightenment,” though it had not been much discussed, that “while that period was massively fascinated with chance, gambling, and the aleatory in all their forms, its major figures suppressed that fascination in favor of ideals presupposing coherent systems of reason, law, or nature.”


Evidently the fact that “enlightened” writers such as Diderot came down heavily against gambling could be easily explained politically, as a move in the struggle between “third estate” and court. What was more expressive of a corrupt nobility, its enemies could argue, than a passion for the card table? (And moreover what more evil example to give the “People”?) But put this debate about gambling in a larger perspective, that of the history of probability theory, and—Kavanagh suggests—the matter can be seen in a different light. For what the success of probability theory led to was the dominance of the statistician. Thus in the long run its effect has been to enslave us to the average. To our twentieth-century eyes “the only conceivable ethics is one of what is probable within a context of large numbers.” The statisticians’ “average man” has triumphed.

This should make us ask ourselves, Kavanagh suggests, whether something precious was not lost when the nobility were taught to be ashamed of gambling. In his view the modern mind is almost incapable of understanding what gambling meant in the culture of the ancien régime. According to the Duc de Saint-Simon, Louis XIV encouraged high-stakes gambling at court as a cunning move to ruin the nobility and immunize it as a political threat; but this, says Kavanagh, was a malicious myth. Gambling, and the visible contempt for money it expressed, was, on the contrary, an expression of the highest value conveyed in the word “noble.” The noble could afford to despise money because he had something better in which to ground his sense of personal identity:

a fixed, inherited sense of self that brought with it a freedom of thought and action far greater than that available to those whose social status depended entirely on acquired wealth and a compulsive mimicry of aristocratic conduct.

Gambling was for the noble a counterpart of valor on the battlefield; and, accordingly, whereas a noble was expected to bilk his tailor and his lady her dressmaker, a gambling debt was to be regarded as sacred.

What then of the novel and its “rise,” and what is their link to this account of probability theory and gambling? “Any representation of unalloyed chance,” writes Kavanagh, “strikes at the very raison d’être of the novel as a form.” Just as probability theory can say nothing about a single throw of a coin, so the novel can say nothing about an event with no before and after (a “real event,” as one might say, as opposed to a constructed and exemplary one). In order to develop, the novel had to forswear chance and aimless picaresque inconsequentiality. Like probability theory, it had to concern itself with “the law of large numbers” and the typical in human behavior; and this had political implications. It put it, so far as its “mainstream” is concerned, on the side of the “third estate.” The novel became a medium (I am foreshortening Kavanagh’s argument here, I hope not unfairly) in which the non-noble reader, the reader with no “fixed, inherited sense of self,” could learn how other people behaved, with the aim of following suit.

We had better pause here and consider a few objections. Can one swallow Kavanagh’s apologia for aristocratic gambling? Could one not equally well, indeed rather better, interpret the aristocratic passion for ruining oneself at cards or dice as expressing, not “freedom of thought and action,” but dreary conformism? Then, is it really right to speak of aristocratic gambling, as Kavanagh does, as a test of “prowess”? I think not. “Prowess” implies a competition in bravery and skill between human opponents, whereas gambling—in this resembling dueling—was essentially a competition with fate. Further, far from the ethos of gambling at the court of Louis XIV being almost beyond our imagining, it does not seem significantly different from what we read about in Balzac and Trollope. That gambling debts were supposed to be privileged and sacrosanct is one of the themes Trollope loves and glances at with semi-ironic relish.

The truth is, Kavanagh seems to have swallowed all the myths and self-serving propaganda of the nobility. That an ancient and “true” nobility, profoundly committed to the supreme importance of “land,” was at a given moment being challenged by a “rising bourgeoisie” basely committed to financial gain is a fiction that no sensible historian is likely to underwrite—for the good reason that it keeps turning up from century to century and from country to country. If we think of cross-Channel experience, the same idea was floated in England as early as the sixteenth century. It surfaced again during the civil wars, was all the rage in Tory circles during and after the War of the Spanish Succession, and was being pathetically voiced once more at the time of the Lloyd George budget. It is essential for the historian to study the workings of such perennial ploys, but he really must not start taking them literally.


But a larger objection obtrudes. Where is that “mainstream” fiction—that “novel of experience” as Kavanagh calls it—which rejects “chance” and wishes to impose the imperious rule of the “average”? He speaks as if it were already flourishing in the eighteenth century, but one cannot quite work out what writers he has in mind; and all the eighteenth-century novelists he actually discusses—Prévost, Voltaire, Vivant Denon, Crébillon fils, and Diderot—are, as he rightly shows eloquent celebrators of chance.

One cannot help feeling that in Kavanagh’s chronology a whole century has got mislaid. The “novel of experience” he is talking about is really a nineteenth-century phenomenon, and so, even more importantly, is that reign or tyranny of the statistical which probability theory eventually helped to usher in. Probability theory itself had already had a certain effect on civic life in the previous century. It influenced insurance, was thought to have a possible bearing on judicial decisions, and figured in the debates on inoculation. (Oddly, the great d’Alembert never wholeheartedly accepted probability theory, maintaining weirdly that it was physically impossible for a coin to come up heads a hundred times in succession; and partly for this reason he was lukewarm about inoculation, causing Diderot to attack him as both a bad mathematician and a bad citizen.) Still, it is hard to believe that anyone apart from mathematicians ever gave a thought to probability theory, and harder to believe that it transformed the ethos of a whole society. That the speculators in the Mississippi “bubble” were acting in the spirit of probability theory, or that Kant’s categorical imperative was an expression of the “law of large numbers”—these are ideas that hardly begin to persuade us. Kavanagh’s initial idea was a thoroughly beguiling one, but of the kind that one sometimes tries out and then has to let drop.

But at this point a word is also called for about that bugbear phrase “The Enlightenment.” Kavanagh writes, “The French Enlightenment began with the biggest and most devastating state-supported gamble Europe had ever seen.” He is referring to the story of John Law and the Mississippi company and says that “sacrificing everything to his passion for knowledge, Law became the harbinger of the reign of science to come.” This seems to put Law firmly on the side of “The Enlightenment” (and in the sense that he anticipated modern banking theory, the reference to “science” is fair enough). But then we read that it was because of John Law and the disaster of his speculation schemes in which thousands lost their investments that “the French Enlightenment most spectacularly elected chance and gambling as its scapegoat of choice.” That “The Enlightenment” should be both the sponsor and the arch-opponent of gambling is surely very awkward, and one of those tangles that the phrase is so good at creating—unnecessary tangles, so far as one can see, for if we forget about “The Enlightenment,” the fascinating story of John Law presents no problem to the understanding. Then, is a historian of ideas really justified in writing a sentence that begins: “The opposition between innervating credit and ennobling land left as a legacy of Law’s debacle mirrored and consolidated a number of parallel oppositions making up the ideology of nobility…”? What, within this swarm of mixed metaphors, is actually happening to what, one asks oneself? Since the things that a historian of ideas is studying are themselves figurative (“figures of thought,” in Foucault’s phrase), he has a duty to be scrupulous in his use of figures of speech.

Kavanagh, all the same, has made a real discovery: that, for their various reasons, some of the most original novelists of the French eighteenth century were preoccupied with chance, to the point of obsession. His pages on this theme are altogether suggestive and illuminating. From the time of Laplace it was more or less a general consensus that chance was a vulgar fiction and a mere index of our ignorance. There was, however, a solitary nineteenth-century dissenter, the mathematician Antoine Cournot. Cournot argued for the possibility of a chance event, defining it as one occurring at the intersection of two distinct and independent series of causes; and, as Kavanagh acutely points out, this is exactly the kind of event upon which Prévost’s Manon Lescaut continually turns, beginning with the moment when the seventeen-year-old Chevalier des Grieux meets Manon by accident just before she is to become a nun against her will.

Equally effectively he shows how Vivant Denon, in his brief novella Point de lendemain (1777)—I confess I had never heard of it—centers upon what it means for a person or an event to be sans conséquence. It is a story of what happens one night, of one delicious and (for the hero) totally unexpected seduction, all part of a complicated imbroglio that the hero never really comes to understand, but—this is its beauty—with neither antecedents nor consequences. A supremely chance event is torn out of the ordinary social round. Kavanagh’s reading makes subtle play out of the ambiguities of the phrase sans conséquence, extending them to Vivant Denon, who is, he says, both a “man of consequence” (he eventually became director general of all French museums) and, in view of his success in surviving so many convulsive political changes, “certainly and in the most profound sense of the word a man of no consequence.”

Kavanagh has no difficulty in showing the importance of Candide in the history of the novel as lying in “its unparalleled recognition of chance in all its forms,” a recognition which is implicit in the very texture of its prose, its dizzying speed, and the uncertainty about what terrible thing will have happened by the end of even a very short sentence. It was also an illuminating stroke to set Candide beside Voltaire’s later entry on “Destiny” in his Philosophical Dictionary. The peasant, says Voltaire, believes it has hailed on his field by chance. The philosopher knows that there is no such thing as chance and hail has from all eternity been destined to fall on that day. Yes, but, he continues, addressing an imaginary adversary, “I have from necessity the passion for writing this; and you have the passion for contradicting me; we are both equally fools, equally the playthings of destiny.” As Kavanagh rightly says, this exactly catches the pattern of the continuous dialogue in Candide.

Thomas Kavanagh is again very good on how, in the multiple story-telling in Diderot’s Jacques the Fatalist, chance and necessity keep changing places. He makes the admirable remark that “to recognize the cacophony of the world as an infinity of stories competing for our attention is to discover the reality of chance as a force presiding over the narration of those stories making us their own.” I think, though, he misses the point that this novel makes play both with fatalism and with its grayer rival, determinism. Fatalism is, as you might say, the fairy tale or Arabian Nights version of necessity, and in a certain sense it is almost the opposite of determinism. Unlike the latter, it is a version of predestination, asserting that future events are not determined by a sequence of particular causes but already written down in the Great Scroll and are likely, when they eventually unfold, to take us horribly by surprise. It is thus preeminently a philosophy for the storyteller, and Jacques, though a determinist of the school of Spinoza, reverts to fatalism when he wants to tell a story.

But then, “determinism,” which Kavanagh uses indistinguishably from “fatalism” as the antonym of “chance,” strikes one as an empty or misleading word, seeming as it does, as Hume says, “to imply something of force, and violence, and constraint, of which we are not sensible.” Laplace’s fantasy of a superior intelligence which, on the basis of the present state of the universe, would be able to discern the future as well as the past, has misled us: for all it could ever prove, even if it happened, is that somebody was particularly good at prediction—and we are all moderately good at it, or living would be impossible. Heisenberg’s discovery that certain phenomena are intrinsically unpredictable was of deep scientific significance but of very little philosophical significance. After all, we cannot even predict which way a tossed penny will fall.

This Issue

November 18, 1993