Because I have reread Candide more than once I have become convinced that this is how things stand: those who cannot see beyond their nosesthe Senegalese on the street corner, the contemporaries who have died, the two burglaries in their own apartment, the leukemia of the neighbors' daughter in the house by a transmission towerare all like Candide, while Pangloss represents statistics. In Voltaire's day statistics were less efficient and less invasive, but there was Providence or Progress, there was Faith or Reason; and Providence and Progress, according to which this is the best of all possible worlds, stood opposed to the personal or family life of individuals overcome by suffering, tragedy, fear, and violence. If Pangloss were living today, he would be a professor of statistics. This confirms the irresistible tendency of statistics to optimism. Because figures, even sad ones, are consoling. If they promise happiness and progress, they regard me. If they point to misfortune, pain, and death, they do not regard me. They are averages, precisely, hence abstractions, hence they always leave room for an exception for someonethat is, for me. Even statistics about Africa, which today provide a sort of baseline, an antiworld for our reassurance, are at last "progressive" in that the numbers for demographic increase are higher than those for exterminations by human hand or natural means. But if this is so, we need to reread Candide. That is, we need to return to individuals, to persons who are not informed and reeducated by a political pedagogy that puts them in step with statistics and its rational deductions. The limit of individual life is not a defect, it is life itself. A measureless disproportion arises as the distance between the realm of direct experience and the greater world is reduced and fritters away. (The myth implied in the word globalization is one sign of this short circuit.) But statisticsprobabilityought to continue to service the life of individuals, not vice versa. If that does not happen, and the separation between the two continues to grow, the life of individuals will be condemned to some sort of superstition, fanaticism, or shame. Everyone becomes richer: I am ashamed of being poorer. Everyone gets thinner: I am ashamed of being fat. Or, to offer another possibility, the Vatican and the scientists (the majority of them, or the most famous ones) declare that electromagnetic waves are innocuous: my daughter has leukemia, so I will go to Lourdes or to pray to Padre Pio. (It should be unnecessary to remark that even the scientists, the politicians, or the bishops have a personal life in large part independent of their professional obligations. They smoke, and they are just as afraid of dying as their grocer, if not more.) We need to negotiate a mutually honorable compromise between factual information and those who command it, on the one hand, and the lives of individuals, on the other.

Some kinds of science dismiss highly improbable phenomena as irrelevant. Average quantities apply to phenomena that are in equilibrium or stationary: in nonlinear phenomena the researcher's perception becomes, once again, highly important. In real life, however, the relevance of a phenomenon increases with its improbability. Even the politicians, according to whom individuals must be raised to the level of the statistical average, would not dream of challenging the existence of lotteriesfar from it! But if I buy a lottery ticket, I have one chance inwho knows? a hundred thousand? a million?of winning. The improbability of winning the lottery is enormous. But it happens. Someone wins it. This is enough to make people buy a ticket. This is a decision more absurd and irrational than becoming alarmed about one's safety despite the data on a decline in crimes. If I am not mistaken, Pascal, who was not just a great mathematician or a great saint, was interested in the question. In gambling, the lower the frequency, the higher the investment. (If this were not so, who would ever bet on an overwhelming improbability like the existence of God, and of a good God to boot?) Gambling has two faces, gain and danger. Alarm about personal safety is an inverted lottery. You can explain to me, numbers in hand, that the probability that I will be attacked on the street by a Romanian is so low as to be statistically insignificant, but the stakes are so high that I venture into the street as if setting off for the front. I would go so far as to say that there is in this disposition, officially considered a superstition or a weakness, something like an anthropocentric compensation. Certain moral doctrines invoke the infinite as a corrective measure against an excessive human security. The infinitely great and the infinitely small, the infinitely remote in past or future time, put me back in my place. But I continue to reduce billions of years and billions of stars to my own measure and make myself the norm. In today's world it is not the too big and the too small, the too far off in the past or the future, that overwhelm me or weigh on my comprehension: galaxies or nanoparticles, they are as concrete as I am and are my daily experience. The large numbers in statistical samples and statistical averages are far from me and incommensurable with me because they are abstractions. Even figures for world population are an abstraction: most people's estimates would be off a billion or two, and they would shrug their shoulders. The problem of politics (of all politics, hence of even the best sort) is to find a reasonable way to liberate itself from the average abstractions and return to the singularity of lived lives, including the lived life of the great communities. This is no mean task, to be sure. Perhaps something of the sort happened in France recently, almost by chance, in the electoral success of myriads of local, spontaneous organizations.

     In a recent book entitled Le Meilleur des mondes possibles: Mathématiques et destinée, Ivar Ekeland attempts to rescue Maupertuis and the principle of minimal action from Voltaire's riddance of them and Leibnitz from automatically being identified with Pangloss. I am unable to evaluate Ekeland's conclusions, but I am struck by the imperceptible shift from the question of optimism and pessimisman archaism: the last to use the terms was an Italian prisoner who spoke of the pessimism of reason and the optimism of willto the question of optimization, in biology, in economics, and regarding the destiny of the Earth. We can improve the world a bit, perhaps, although there exists no providence, natural or transcendent, and we can do so thanks to the fact that we have become capable of destroying itthe world, that is. We have lost God, but we have in our hands a thread that leads, at its other end, to the end of the world.

There is the question of what Italians call il cammino dell'ubriaco (the drunkard's walk) and use to express "random walk" or Brownian movement. If that drunkardby chance and with equal, 1/2 probabilitytook one step to the left and one step to the right for every fraction of time (one step per second, for example), his average speed would be zero. If his average speed were assumed as a significant dimension, our drunkard would not move from place. But in fact he does move. Which means that in this case, average speed has no meaning in physics or dynamics and we need to select another dimension, one that departs from the mean, to describe his motion. I do not know if I have understood the situation correctly, but there is certainly a strong temptation to use the drunkard who moves but stays in one place, thus escaping averages, as a description of progress.

To read the entirety of this unedited article, please proceed to Rough Cut