Jump to ContentJump to Main Navigation
Embracing RiskThe Changing Culture of Insurance and Responsibility$

Tom Baker and Jonathan Simon

Print publication date: 2002

Print ISBN-13: 9780226035185

Published to Chicago Scholarship Online: March 2013

DOI: 10.7208/chicago/9780226035178.001.0001

Show Summary Details
Page of

PRINTED FROM CHICAGO SCHOLARSHIP ONLINE (www.chicago.universitypressscholarship.com). (c) Copyright University of Chicago Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in CHSO for personal use (for details see http://www.chicago.universitypressscholarship.com/page/privacy-policy). Subscriber: null; date: 21 June 2018

The Return of Descartes's Malicious Demon: An Outline of a Philosophy of Precaution

The Return of Descartes's Malicious Demon: An Outline of a Philosophy of Precaution

(p.273) Eleven The Return of Descartes's Malicious Demon: An Outline of a Philosophy of Precaution
Embracing Risk

François Ewald

Stephen Utz

University of Chicago Press

Abstract and Keywords

The new paradigm of security calls forth a new economy of rights and duties. While the language of risk, against a background of scientific expertise, used to be sufficient to describe all types of insecurity, the new paradigm sees uncertainty reappear in the light of even newer science. It bears witness to a deeply disturbed relationship with a science that is consulted less for the knowledge it offers than for the doubt it insinuates. Moral obligations are swallowed up in public ethics, and the principle of responsibility is seen as a reflection of a brand-new notion of precaution. The paradigm of responsibility is a paradigm of insurance—it assumes the logic of loss compensation. The paradigm of solidarity is also a paradigm of insurance, of universal and indeterminate insurance, of social and compulsory insurance. It is not so much concerned with voluntary and contractual forms of compensation as with the institution of pools of all kinds. The paradigm linked to the precautionary principle will undoubtedly remain a paradigm of insurance, but in a new shape that will have to integrate new cultural boundary conditions.

Keywords:   security, precaution, rights and duties, loss compensation, insurance, public ethics

We may be in the midst of a paradigm change concerning society's obligations for the physical security of its members and our shared political philosophy of security. The nineteenth century saw the dominance of a paradigm of responsibility. In the twentieth century this was fundamentally transformed: the prevailing paradigm was one of solidarity. Perhaps, at the beginning of a new century, we are seeing the birth of a new paradigm, one that has not yet found its true name, but whose arrival is presaged by various signs.

Where safety is concerned, rights and duties, legal and moral obligations are always issues; these fit together without necessarily overlapping. The paradigm of responsibility posits a certain economy of rights and duties in which the part played by moral obligations toward oneself and others is far greater than that of legal obligations. Implicit in the philosophy of liberalism, the paradigm of responsibility relies less on constraint than on freedom and individual will. Legal obligations with regard to others are summarized in the rule “do no harm to others.” Virtue holds an important place here under the twofold guise of prudence (for oneself) and charity (for others).

By contrast, the paradigm of solidarity, which is associated with the welfare state, considerably extends the role of legal obligations, which here tend to overlap with moral obligations. Solidarity is accompanied by the multiplication of social rights and by the recognition of a sort of general right to indemnity for every mishap in life. This paradigm belongs to an era of scientific and technical utopia, in which society may have the possibility of controlling itself, where knowledge may have an indefinite priority (p.274) over power. At its philosophical foundations, it is inseparable from the imperative of prevention: prevention of illnesses (with Pasteur's discoveries), prevention of crimes (with the system of social defense), prevention of accidents (with the sciences of safety), prevention of poverty and social insecurity (with social insurance).

The new paradigm of security calls forth a new economy of rights and duties. While the language of risk, against a background of scientific expertise, used to be sufficient to describe all types of insecurity, the new paradigm sees uncertainty reappear in the light of even newer science. It bears witness to a deeply disturbed relationship with a science that is consulted less for the knowledge it offers than for the doubt it insinuates. Moral obligations are swallowed up in public ethics, and the principle of responsibility is seen as a reflection of a brand-new notion of precaution.

The paradigm of responsibility is a paradigm of insurance—it assumes the logic of loss compensation. The paradigm of solidarity is also a paradigm of insurance, of universal and indeterminate insurance, of social and compulsory insurance. It is not so much concerned with voluntary and contractual forms of compensation as with the institution of pools of all kinds. The paradigm linked to the precautionary principle will undoubtedly remain a paradigm of insurance, but in a new shape that will have to integrate new cultural boundary conditions.


What characterizes the pattern of imposing liability and exacting damages that came to be known during the eighteenth and nineteenth centuries as “responsibility?”

A Political Strategy

We must first identify a broad public policy, an overall strategy of social control. Responsibility is based on the principle that “one person cannot transfer to another the burden of what happens to him.” It is in direct opposition to the principle of assistance. Adolphe Thiers, reasserting the liberal credo that had been constantly repeated for a half-century, wrote in 1850 that “[t]he fundamental principle of any society is that each man is responsible for meeting his own needs and those of his family, out of resources that are acquired or transmitted. Without this principle, all activity would come to a halt since, if man could rely on work other than his own to survive, he would gladly leave to others the tasks and difficulties of life.”

The formulation of this principle is linked to the advent of liberalism. It involved both making people provident, aware of the future, and preventing (p.275) them from living solely in the present. The principle of responsibility reflects a view of humanity's relationship with nature, according to which everything that happens to a person must be considered a sanction, good or bad. Responsible for myself, I would not even know how to attribute to someone else a causal role in the obstacles I face. The responsibility for these obstacles, even if other causal factors contribute to them, is mine. It is I who was unable to take a particular element into account; it is I who did not understand the laws of nature or was unable to use them. In any case, and always without exception, it is my fault. I am the only target for attribution. “To err is human,” as the saying goes. Accordingly, the principle of responsibility converts any mistake into a fault.

The principle of responsibility relies on a method of managing causality that makes it possible to devise self-regulation of conduct and activities. To the extent that one cannot lay one's own failures and sufferings off on someone else, failures and sufferings are able to become the indeterminate principle of their own remedy. The principle of responsibility, based on fault, thus serves as a universal converter of bad into good. But it is singularly demanding. Seen from this angle, security would not be a right, but merely a duty. For, on this philosophy, there is no room for the notion of victim. In this world, suffering a wrong gives you no right to anything (unless it results from the fault of someone else). And victims, whatever feelings of compassion and pity they may inspire, are always assumed to be the sole creators of their own destiny. The political principle of responsibility lies behind articles 1382 et seq. of the French Civil Code. It is precisely what these articles are intended to enforce. One might as well say that they were not intended to extend the reach of damages by increasing the possible number of responsible parties, but rather to limit them to the situation in which the injury suffered was due to the fault of another. A judge's task is to ensure that he maintains a delineation of fault that will uphold the sense of the general principle of responsibility.

The Virtues of Responsibility

The principle of responsibility and its legal sanctions aim to make man provident and prudent: provident as to the effect of fortune, prudent as to himself and the consequences of his actions. Fault by such logic is always a fault of prudence; it is the sanction for what one could or should have foreseen. A prudent and provident man has no excuse. To his own prudence he will not only owe his safety but also his capacity to associate with others to compensate for the effects of fate. If such a philosophy excludes any idea of insurance for responsibility, or even insurance for damages (long (p.276) considered immoral as though encouraging crime, and thus prohibited), it evokes, on the other hand, the idea of self-insurance, on a strictly voluntary basis.

Fault is a philosophical principle for attributing liability, which should have the merit of combining harmoniously not only the three functions of sanction, prevention, and compensation, but also of integrating ethics, law, and politics. Thus the great legal commentator Jean-Etienne Labbé could write at the end of the nineteenth century, at a time when the mechanism of responsibility was already deeply shaken, that “[r]esponsibility is the most perfect regulator of human actions”(1885). This model no longer corresponds to our experience, and yet it is difficult not to feel its coherence and even its proximity. Its persistence as a model, or as a regulating principle, in the face of contrary social facts is nothing new. Since the time when it was first instituted, it has been contested by the developments of industrialization. The form of providence supports the formulation of articles 1382-86 of the French Civil Code as they were interpreted until the end of the nineteenth century. The legal notion of fault echoes that of providence: one can only reproach somebody for what he or she should have known. As seen in the jurisprudence of article 1382 of the French Civil Code, the reference to the diligence of a “prudent head of family” or to certain standards of professional conduct indicates that one can only be held responsible to the extent of the available knowledge, which varies depending on the activity. Doubt, uncertainty, or suspicion cannot make one responsible. Such emotions are more likely attributed either to chance, or to the prudence that each person owes to himself in the conduct of his life.

Providence is the great virtue of the nineteenth century. It is the foundation of responsibility in the conventional sense of the term, and it prohibits blaming another for what happens to you (except in the case that it is due to the fault of another). Indeed, when the term was coined in the nineteenth century, “responsibility” did not designate, as it does today, a general principle of blaming another for mishaps, but precisely the opposite. Responsibility as providence consisted of being aware of the risks to which one was subject because of the need to face up to them on one's own initiative. The world of providence is one in which one must recognize his or her weakness and fragility, subject to incessant reversals of fortune; it is a world of chance events. It is an unbalanced world where one knows oneself to be vulnerable and scarcely hopes to use science and engineering (which are not readily available) in order to rebalance one's relationship with nature. One must instead rely on cunning, intelligence, and calculations of probabilities, which teach that accidents do not happen without laws and that these (p.277) laws are the foundations of opposed forces no one can face up to rationally. This is how insurance came to be promoted throughout the nineteenth century—as the institution of a rational providence. In the world of providence, one has no other resource than calculation and virtue. Faced with the accidents that one can hardly anticipate, the only possible tactic is to learn how to offset them. And insurance, which makes this possible, cannot be made compulsory, since that obligation would rule out the exercise of the virtue of providence.


The mechanism of responsibility was challenged, reformed, and replaced at the end of the nineteenth century, at least with respect to the coverage of certain types of events, by an arrangement based on solidarity. The major issue in all industrial societies was that of accidents at work and pensions. The mechanism of solidarity is not based on fault but on risk; its main instrument is insurance, rather than general legal rules for assigning liability.

Shift to Risk

The notion of risk, which makes its arrival in positive French Law in 1898 with the “professional risk” of the law on accidents at work, dated 9 April 1898, designates a way of envisaging the reparation which, without involving examination of the behavior of the worker or the boss, attributes them globally to the job and the firm. “All work has its risks; accidents are the sad but inevitable consequence of work itself” (Thoulon 1898). According to the principle of professional risk, the onus of accidents at work, whatever their cause, belongs to the firm, whether these accidents arise legally from a chance event or through the fault of the worker. Thus the result of the new law was to make the firm legally “responsible” not only for accidents resulting from personal imprudence or negligence, but also for those arising even when the firm has taken all precautions to avoid them, and those for which the worker is the cause (absent tortious intent on the worker's part).

The idea of professional risk thus divorces causality from attribution. The indifference of the latter in relation to the former is separated by the institution of a principle of attribution that no longer refers to the objective causality of damages. The rationality that underlies the notion of risk leads to a change in the legal allocation of damages, which is no longer in terms of “cause,” but of “distribution”—distribution between profits and expenses of firms, or more generally, a socially constructed distribution of expenses. (p.278) Risk only exists socially. Whether professional risk or social risk, it institutes a social contract between individuals.

The invention of professional risk and social risk consists of thinking through a principle of allocation that, freed from the old dependence on nature, finds its reference in a social relationship. This accounts for both the richness of the category and the difficulty of deploying it.

The idea of risk is linked with analyses of situations based on statistics and probabilities. Business risk, for example, characterizes a whole, the firm, which conserves its identity despite the variations that may affect its different parts. The regularity of risk is independent from the conduct of individuals. Their faults are factors of risk that do not affect the company's statistical reality. This is one of the principal benefits of the notion of responsibility: enabling the law to base itself on reparation of the actual fact of the accident, of the damage suffered, whatever the cause. While a principle of responsibility founded on an idea of cause implies a selective distribution of the costs, business risk, on the contrary, provides “solidarity.” Solidarity of the boss and the worker in the context of the firm recognizes a species of business risk that can put an end to the antagonism of capital and labor that the law of responsibility fueled. Social solidarity, more generally, allocates risk by shifting the cost of accidents to the firm, thereby creating a new balance between rich and poor, producers and consumers.

A New Social Contract

The new solidaristic doctrine of spreading costs was to be articulated in eloquent terms by the reforming jurists of the end of the nineteenth century. Let's hear what they have to say. First, Raymond Saleilles:

Modern life, more than ever, is a question of risks. Therefore, one takes action. An accident occurs, somebody must necessarily bear the consequences. This must be either the person who caused the accident or the victim. The issue is not of inflicting a penalty, but of knowing who must bear the loss, he who caused it or he who suffered it. Not the penal point of view but only the social point of view is at issue. Properly speaking, it is no longer a question of responsibility but a question of risks: who must bear them? Obviously, in reason and in justice, it must be he who in taking action has taken responsibility for the consequences of his deeds and his activity…. (1897:4)

This text is a testimony of the intellectual conversion that gave birth to the mechanism of solidarity. “Modern life,” says Saleilles, “is more than ever a question of risks” (id). This is an acknowledgment that the real world (p.279) is not one imagined by the drafters of the French Civil Code. Whatever diligence each person brings to his affairs, misfortune is not the exception but the rule. Injuries are “normal”—which does not mean they are inevitable. Social life is not naturally harmonious, but conflictual, prejudicial. When the “good head of family,” who served as reference in defining fault, takes action, pursues his business, does his work, he “naturally” causes, without wishing to do so, injury to other good heads of family. The outcome is necessarily that the injuries must be objectified as “accidents” and no longer be considered faults. These are risks.

Another consequence: the problem of responsibility is no longer that of determining “who was at fault,” but to whom the injury should be attributed, who should bear the loss caused by the injury. “Between two individuals, one of whom, even without being at fault, has caused a personal accident or a loss of property and the victim, who must bear the financial cost and, at the end of the day? On which asset-base must the final loss fall? That is the problem. The issue of fault has nothing to do with it” (Saleilles 1897:75).

Implicit in this approach is a problem of fairness in terms that are more economic than moral, causing the cost of damage to be borne by the victim or by another, in either case making one or the other “responsible.” Spreading risks means judging who must bear them, not in terms of the moral responsibility, which is irrelevant here, but in terms of a “social fairness” to be constructed. The source and foundation of responsibility are displaced from the individual onto society: one is responsible not because one is free by nature and could therefore have acted differently, but because society judges it “fair” to make you responsible, that is to cause you to bear the financial cost of the injury, whether you are actor or victim.

Let us now hear the version of L. Josserand:

The passer-by that I knocked down, the classmate that I injured, did not for their part commit any fault and yet today's doctrine declares them responsible for the accident, since it makes them bear the consequences. For, and this is an idea which, despite its evidence, is not sufficiently grasped, it is impossible to subtract the owner of the thing from the responsibility of injury without causing this same responsibility to be borne by the victim: Since the cause of the accident is unknown, some say, nobody must be responsible, since nobody is at fault. This reasoning is absolutely wrong: when an accident occurs, it is not possible for nobody to be responsible in the wide sense of the word, that is for nobody to bear the consequences of what happens: if the owner of the thing that caused the accident is not made to repair the damage, the victim (p.280) must of necessity be the one to bear it; the victim will incur the full responsibility for the accident, responsibility which is seen in the loss of life or health, without compensation. Whatever the solution adopted, there is always responsibility: the only question is that of knowing who must assume it. In a nutshell: is there responsibility each time there is final injury? (1897:107)

Can we be clearer on the transformations undergone by the notion of responsibility? Earlier, and always in common belief, responsibility was thought an attribute of human nature. Responsibility has changed its meaning. It is no longer the attribute of a subject, but rather a consequence of a social fact. Responsibility, in a way, has become a relationship without support.

There is no injury where loss is solely individual. All injury is social. A distinction must be made between the injury suffered by an individual—this is an affair of chance or mischance—and the loss linked to the injury for which the attribution is always collective and social. In any case, society and the courts cause the cost of injury to be borne by someone, whether this is the victim or another: they spread the risks. The notion of spreading the risks implies that one conceives society not as an aggregation of individuals, who may be linked to each other by personal interest, but as a totality—the good and the bad of each individual dependent on everyone else. This is true whether one conceives of society as a totality in which no individual is any longer a third party in relation to others, or whether one conceives of society in accordance with the principle of solidarity. We move from an individualistic to a holistic perspective.

If injury is individual, it behooves society to spread the cost suitably, and this necessarily disrupts what nature or fate has decided, throwing a new light on the problem of justice. François Gény, another jurist from the end of the nineteenth century, prophesied the continuation of the transformations that he witnessed:

One can imagine a social ideal which, without claiming to halt the blows of fate or defy the decrees of providence, would aim to discover in the nature of things, placed by God Himself at our disposal, the means of sharing amongst all, in the form of an intelligently organized mutuality, the risks which incessantly threaten us as individuals. However, and without speaking of the almost insurmountable difficulties, of the prodigious organization required by the realization of such an ideal, one easily glimpses the utopia and the danger of a system, which, firstly, in order to remain loyal to its aim, should, by means of compensation, although necessarily imperfect, straighten out all inequalities, take into (p.281) consideration economic variations as well as material changes, share out the gains as well as the losses; which, secondly, however well-advised the application may be, will not fail to diminish the qualities of initiative, of diligence, of providence, which are among the conditions essential to the progress of humanity. (1902:817)

Here we have a description of the program of social security that took shape during the twentieth century, not only covering social risks (illness, old age, disability) but also in compensation for an ever more impressive number of accidental risks. It is a commonplace that, during the twentieth century, law and responsibility were transformed, bit by bit, into a legal system of compensation, based on risk and insurance. This is the case, on the one hand, for car accidents, but also for relief from natural disasters, for protection against assaults, and the consequences of crime and of injuries linked to defective products. Protection against sea pollution is also provided through these mechanisms, some of which must eventually be extended to land pollution. By way of responsibility, the twentieth century systematically thought compensation, so much so that today, in the early years of the twenty-first century, the victim of any sort of injury seems to be able to claim compensation, and the media are tempted to denounce the scandal as soon as they see what seems to be an unassimilated type of loss, as is still the case in France with medical accidents.

Prevention First

The paradigm of solidarity is not only a paradigm of compensation but also one of prevention. One forgets the extent that the paradigm has successfully distinguished the three problems of reparation, criminal sanction, and prevention. In fact, in the same way that compensation is no longer linked to the prudence or imprudence of individual conduct, prevention is objectified as a function now less of goodwill than of scientific know-how. The human being is no longer objectified as a free agent but rather as a link in a technical system in which faults are considered less as individual than organizational. We are all familiar with how the analysis of distinct systems has been swallowed up by this generalizing tendency.

Despite their separate spheres of application, compensation and prevention nevertheless derive from a single philosophical paradigm. The word prevention indeed took on its current meaning no longer a matter of reluctant reservation in the solidaristic environment as compulsory precaution, on the one hand, and as conduct bluntly designed to lower the probability of some unwanted happening on the other. In fact, the notion of prevention, applicable to Pasteur's discoveries on infection and their public (p.282) health consequences, as well as to the efforts of engineers to reduce the probability of mechanical accidents, assumes and accompanies the promotion of the notion of risk and—what comes down to the same thing—of measurable risk. Prevention (the vocabulary of which henceforth replaced that of providence) presupposes science, technical control, the idea of possible understanding, and objective measurement of risks. Thus the problem is no longer that of compensating for practically inescapable losses but of reducing the probability of their occurrence.

The nineteenth century's dream of security is tied to a scientific utopia ever more capable of controlling risks. While one cannot eliminate risks altogether (there is never zero risk), they will have been reduced sufficiently to be able to be dealt with collectively: accidents are the by-product, necessary although always more marginalized, of scientific and technical progress. These are special or abnormal risks, the responsibility for which should be spread over the community. Our concept of assured public health and safety involves prevention, the dream of an ever more complete reduction of risk. It is not, at least not on purpose, a measure designed to take on all human misfortunes.

Prevention is an attitude which, by principle, relies on trust in science and its know-how. It presupposes the adjustment of knowledge and power, an ever-possible control of power by knowledge. Its utopia, the asymptote of a knowledge always in the process of mastering techniques and practices derived from them, will eventually reach its limit. One cannot foresee what one does not know, even less what one cannot know.


The second paradigm of solidarity, however, may be unraveling before our eyes as a result of three major issues, which are also the three principal foci of contemporary problems of safety. These issues are medical accidents, particularly the aspect of serial risk revealed by transfusions, grafts, and transplants; global threats to the environment; and products liability of manufacturers (innovation risk).

The Precautionary Principle

These three problematic areas have two characteristics in common. First, they emerged at roughly the same time, during the 1980s. Second, they were experienced as an ordeal of unexpected vulnerability, unsuspected by individuals in developed societies, who were lulled by the promise of an ever-safer world. We are seeing the return of disasters, the insistence on individual and collective injuries of unequalled magnitude, at (p.283) least in peacetime. The nineteenth and twentieth centuries were obsessed with the problem of accidents (work-related or automotive); we are now rediscovering the existence of disaster, but with the difference that disasters are no longer, as before, attributed to God and Providence, but to human agency. It is in this deeply disrupted context that the precautionary principle now appears.

The category is henceforth part of positive law: international, European community, and municipal. First international: formulated at the time of agreements implemented, from the end of the 1980s, in order to combat the greenhouse effect and the hole in the ozone layer, in the context of the Brundtland Report on sustainable development (1987), the precautionary principle constitutes the tenth major principle recognized at the Rio Summit in 1992 (United Nations 1992). Here, precaution designates the attitude that consists of preventing serious or irreversible deterioration of the environment by a modification of the production, of the sale or use of products, of services or types of business, and this in accordance with a scientific and technical standard. Next the European community: the Maastricht Treaty specifies in article 130 R that “community policy on the environment” shall be based on the precautionary principle and on the principles that preventative action should be taken, that environmental damage should as a priority be rectified at source and that the polluter should pay. Finally, municipal law: since French parliamentaries have given a definition of the precautionary principle in the first article of the law, dated 2 February 1995, relating to the better protection of the environment, the precautionary principle, “in accordance with which the absence of certainty, taking into account the state of scientific and technical knowledge, must not postpone the adoption of effective and proportionate measures to prevent serious and irreversible damage to the environment at an economically acceptable cost,” is specified as the first principle that should take precedence in policies for the protection of the environment.

From these different texts it is immediately evident that the precautionary principle does not seem to have a universal meaning: its field of jurisdiction is limited in principle to the area of the environment. But, make no mistake, to the extent that this represents a certain type of decisionmaking in situations of uncertainty, the application of the principle will cause it to be exported beyond its original territory. We have already seen that the Council of State put the principle to work in extending state responsibility in the blood transfusion crisis.*

Otherwise, it appears that the precautionary principle does not target all risk situations but only those marked by two principal features: a context (p.284) of scientific uncertainty on the one hand and the possibility of serious and irreversible damage on the other. These two items are thoroughly problematic.

A New Balance

What does “grave and irreversible injury” mean? Any bodily wound, much less death, must constitute grave and irreversible injury for the person who suffers it, such that no economic compensation can ever make him or her whole again. In fact, the precautionary principle does not so much focus on individual injury, such as may be caused by an accident, as on collective “catastrophic” injury. Still, it may be impossible to exclude from the principle's sphere of application, for example, a physician's lack of “precaution” in treating a single patient. Beyond this, it is useful to distinguish between the severity of injuries, which determines their nature and extent, and their irreversibility.

If an irreversible injury is always a serious one, the converse is not true. Irreversibility focuses on transformations of nature, the balance between humanity and the environment, the introduction of a long-lasting result in the process of life on earth. Irreversibility should doubtless be understood to relate to a “goal of lasting development that purports to satisfy developmental needs of present generations without compromising the ability of future generations to respond to their own needs” (art. 1 of the French Law of 2 February 1995). It is concerned with halting any event that might interrupt the course of natural history, those “revolutions” of which Cuvier spoke. It is concerned with ensuring the continuity of the future with the past. The precautionary principle is counter-revolutionary. It aims to restrict innovation to a framework of unbroken progress.

But the precautionary principle does not proceed solely from a consideration of the nature and amplitude of injury that seems possible from our standpoint today. For this principle to apply, it must change our understanding of injury itself. Injury partakes of the irreparable, the irremediable, the incompensable, the unpardonable, the nonprescriptive. The appearance of the precautionary principle is registered in the context of victims who are no longer satisfied with compensation, no matter how large, but who are only satisfied when those responsible are held criminally liable.

This is certainly a modification of the solidaristic equation that took for granted that a risk was acceptable as long as it was reparable or repaired. This was the principle of the workers compensation law of 1898 (Thoulon 1898), and it was certainly behind awards indemnifying victims' costs for investigation of those responsible for their injuries. One could take a risk (p.285) on the condition that one paid the price. This attitude, which presupposed that everything—gains as well as losses—had a price, was the principle of cost-benefit analysis, which legitimated entrepreneurial decisions. Was this not how the exercise of “objective responsibility” was justified, by internalizing “externalities” in the price of a product's production? The problems of decisionmaking, business activities, compensation, and responsibility henceforth became just economic problems. It was just a matter of adjusting the cost to the risk of the business enterprise.

The appearance of the precautionary principle no doubt marks the limit of this equation and its underlying logic. With the irreversible, we rediscover the irreparable. Not everything is a matter of economics. Not everything can be assessed a money equivalent. From now on, we no longer limit our attention to compensating, but are also concerned with preventing, forbidding, sanctioning, and punishing. A logic of responsibility returns alongside the solidaristic problematic of compensation.

The same idea underlies the apparently contradictory term zero risk. This involves not so much defining what an action without risk would be as asking what reason we have for running the risk—zero risk designates a risk that has a no price. Transformation, therefore, in our value systems and in our methods of valuation is beyond price. As such, it seems undeniable that the precautionary principle upsets certain postulates of an insurance-based society, which takes for granted that one can assign a price to anything, even the very notion of risk, the function of which is to attribute to a threat an objective value and price. In a way, the formula once used by Portalis to condemn life insurance again seems to fit: “People's lives,” he said, “have no price” (1989). Except that under the precautionary principle the idea applies not only to human lives but also to the lives of animals (via the protection of the diversity of species) and to nature itself.

Of course, this does not mean that, in a precautionary society, all injury suddenly becomes irreparable and the logic of insurance and solidarity disappears. First, precautionary logic is limited to serious and irreversible injury that, in principle, limits the scope of cognizable injuries. Second, precautionary logic is, above all, a logic of decision applicable in situations of uncertainty, including that of remedying injuries.

The second element of a precautionary situation bears on the “absence of certainty taking into account the scientific and technical knowledge of the time,” which is part of the French Law dated 2 February 1985. The formula defines the relationship between prevention and precaution: once there is certainty as to the consequences of an action, one remains with the logic of prevention, with all that this implies in conventional responsibility (p.286) terms. The notion of precaution concerns a situation in which only a relationship of possibility, eventuality, plausibility, or probability between a cause and its effect can be envisaged. The hypothesis does not concern exactly an unknown cause, but rather a probable or suspected cause. This reference to scientific uncertainty is particularly disturbing.

Uncertainty does not solely concern the relationship of causality between an act and its consequences, but also the reality of injury and the measure of the risk of such injury. The precautionary hypothesis puts us in the presence of a risk that is neither measurable nor assessable—that is, essentially a nonrisk. While the logic of insurance and solidarity had reduced uncertainty to risk, in order to make the former systematically assessable, the logic of precaution leads us once again to distinguish between risk and uncertainty. Precautionary logic does not cover risk (which is covered by prevention); it applies to what is uncertain—that is, to what one can apprehend without being able to assess. Do the changes we currently observe in weather patterns constitute an irreversible interruption in the history of climates? Or are they only a deviation, the likes of which the earth has already seen on many occasions? Should we continue to use blood transfusions once we suspect that the blood contains the hepatitis virus—A, B, or C—of which one is not even capable of assessing the potential danger?

The precautionary hypothesis focuses on the uncertainty of the relationship of causality between an action and its effects. This implies that from now on, along with what one can learn from science, in a context that is always relative, it will also be necessary to take into account what one can only imagine, suspect, presume, or fear. The precautionary principle invites one to consider the worst hypothesis (defined as the “serious and irreversible” consequence) in any business decision. The precautionary principle requires an active use of doubt, in the sense Descartes made canonical in his Meditations on First Philosophy. Before any action, I must not only ask myself what I need to know and what I need to master, but also what I do not know, what I dread or suspect. I must, out of precaution, imagine the worst possible, the consequence that an infinitely deceptive malicious demon could have slipped into the folds of an apparently innocent enterprise.

From this point of view, the formulations that one gives of the precautionary principle do not, at least superficially, avoid self-contradiction: anticipating limits in the attitude of prudence to be adopted, either in terms of scientific-technical criteria in accordance with the formula adopted at the Rio Summit, which speaks of measures to be taken “in accordance with a scientific and technical approach,” or in terms of more economic criteria if one follows the version of the French Law dated 2 January 1995, which (p.287) speaks of “effective and proportionate measures at an economically acceptable cost.” On the one hand, the decisionmaker is invited to expect the worst, the possible catastrophe, the irreparable; while, on the other, one finds that measures are to be taken only in a “scientific and technical” context—while, by principle, their validity is placed in suspense—or at “an economically acceptable cost”—just when one must envisage the nonassessable of the irreversible. This is no doubt explained by the fact that one wants to maintain a principle of economic and industrial development, which prohibits inaction in the face of uncertainty, at the same time as one seeks to limit as far as possible its harmful consequences. Hence the idea of “sustainable development.”

The precautionary hypothesis, finally, introduces uncertainty in both decisions and sanctions, both a priori and a posteriori. A priori: to anticipate danger or avoid injury, as is the prudential intent of the statutes. Precaution designates, first and foremost, a logic of decision to act or not to act, to undertake or not to undertake. It complicates this by introducing the dimension of uncertainty, based on the worst hypothesis. Strictly speaking, it only makes sense before the decision is taken. And yet, to the extent that one cannot fail to call it into play, in the sanctioning of responsibilities, once injury occurs the principle will find itself applied a posteriori. And one can see what this implies: that the sanction escapes the reference to available knowledge, a standard of determined knowledge, whether one is judged not only by what one should know but also by what one should have or might have suspected. Here, precautionary logic, even if it extends the field of subjective responsibilities because it focuses on the act of decision, does not strictly speaking enter into the former logic of fault. The latter presupposes the existence of knowledge, which is absent here because one is ex hypothesi uncertain.

Verification of the observance of decisionmaking procedures takes on paramount importance, as should now be properly defined. Two contexts are to be taken into account: that where the injury is felt and that of the uncertainty of long-term causality. And one can hardly see how, under current law of responsibility, one could attribute to anyone an injury of unclear origin, except by employing new systems of causal analysis, vague logic, and other systems of probable causality, or by introducing a new law of proof, or by fixing responsibilities of principle to necessarily arbitrary foundations. The other context is that in which only with time does the causality of the injury become known and therefore assignable. What was only suspected now becomes progressively apparent. Under the old approach to responsibility, uncertainty of knowledge was innocence. The application of the precautionary (p.288) principle gives a very different result: the uncertainty is not an excuse, but rather a reason for greater caution. The problem of responsibility therefore arises in a somewhat unusual fashion: while one cannot say that the risk was totally unknown since one might have or should have suspected it, sanctioning such “responsibility” can only lead to a considerable restriction of entrepreneurial effort.

A new dimension in the problematics of safety under the precautionary hypothesis is time. The uncertainty of precaution resides to a great extent in the temporal span between cause and manifestation of harmful effect; the delay can be considerable. The precautionary hypothesis goes with an awareness of a dilation of time, of a new understanding of the duration in causality of human actions. This is in stark contrast to any analysis that acknowledges the possibility of an “accident.” Characterized precisely by the coincidence or the proximity of the cause and effect, an accident is conventionally defined by its sudden or instantaneous nature. Precautionary injury hardly has this nature; either time is needed for an injury to appear (cases of gradual pollution, climate changes), or it is only afterward that we understand that an effect resulted from a cause—a relationship that was, until then, unsuspected. The precautionary principle invites one to take account of considerably extended temporalities, which leads to the question of duration in the law of responsibility and of the very rationale of statutes of limitations. The principle specifies that “efficacious and proportionate” measures must be taken “without delay,” which presupposes that in a way it is already too late. What is hoped for is most often just avoiding the exacerbation of an already bad situation. The precautionary principle presupposes a new relationship with science and with knowledge. We knew all along that scientific knowledge was relative to a certain state of knowledge, that its validity was limited, that it behooved science to progress continually, and thus ceaselessly reform itself. But that is not exactly the case here. Within its domain of validity, science produces, if not definite certainties, at least an understanding of references, recognized by the scientific community.

The precautionary principle invites one to anticipate what one does not yet know, to take into account doubtful hypotheses and simple suspicions. It invites one to take the most far-fetched forecasts seriously, predictions by prophets, whether true or false. We should remember the accusation made from Heidelberg by forty Nobel Prize winners against their “colleagues” meeting in Rio, at the time of the World Summit. The precautionary principle returns us to an epistemology of the relativity of scientific knowledge.

The ethics of precaution are both sophistic and skeptical. They are sophistic (p.289) because they presuppose that any explanation is opposed by a countervailing explanation. To repeat the formula of Protagoras, man is the measure of all things, those he knows of as well as those he knows not (Plato 1991). Precautionary ethics are skeptical because one is invited to suspend judgment when faced with the assertions of science. For one must take all hypotheses into account, even and in particular the most dubious; one must be wide open to speculation, to the craziest imagined views. While the attitude of prevention presupposes a relationship with knowledge that guarantees the veracity of such knowledge, the precautionary hypothesis invites one to make the most deceptive malicious demon one's constant companion.

Marie-Angèle Sanson-Hermitte has already pointed out that the precautionary hypothesis compels one to take into account opinions that are acknowledged to be marginal and dissident within a scientific community (1996). Going into more depth, she calls attention to a disrupted relationship with science, such that we are now less interested in the confidence science provides than in the suspicions and doubts it can arouse both about what we know and what we do not know. With precaution, science becomes a principle of challenge. Science becomes a challenge to itself as well as to the many proofs that everyday life should not be the subject of permanent anxiety. Effectively, science today interests us less by producing new knowledge than introducing new doubts. Although we have scarcely recognized the change, certainty today is not procured so much by the conventional method of deduction as, rather like the Cartesian credo, by the logic of double negation: all that can be excluded is that anything should be excluded.

Development Risk

The extreme form of the precautionary idea is provided by the concept of risk of development that accentuates all the paradoxes and the difficulties. What is at stake is not suspicion but pure and simple ignorance. The question posed is whether one can permit a producer or a supplier, with objective responsibility for the consequences of product defects, to exonerate herself in the event that she could not have known the defect existed. Let us go further. The risk of development presupposes a product, affected by an undetectable and unforeseeable defect, which only becomes known after a certain span of time, and where the attribution to the product or to the producer presupposes a different level of science than that available to the producer. The most novel feature of the risk of development lies in the need to anticipate a transformation in knowledge, in awareness, in the very perception that transforms reality.

(p.290) The concept of risk of development is new because it presupposes that we entertain a new relationship with science and engineering. Until now, we lived under the illusion that science might control itself. No doubt prudence was needed, but any advance in knowledge reduced uncertainty. In the field of engineering, the engineer's sciences seemed to possess a certain kind of power of infinite self-control. As new dangers were introduced, new possibilities of controlling and reducing them were provided. We lived as if a balance between knowledge and power were continually possible. Risk of development has become an issue for us because we are now aware both of the dynamism of science and of the essential relativity of knowledge. Thus, contemporary epistemologists teach us that science progresses less by the accumulation and extension of knowledge than by “paradigm shift.” And we also move from one paradigm to another by the play of controversies and necessary conflicts, so that what is at stake is less the substitution of error for truth than the manipulation of power within the discipline. From this relativity of knowledge it can be deduced that, while scientific development always permits the discovery of new substances or new processes, it cannot do so without risk because no state of knowledge can claim a guaranteed mastery of itself. Science increases our powers and our capacities without reducing the uncertainty that it generates. We necessarily run the risk of the unforeseeable. In this rediscovered gap between power and knowledge, understanding and awareness, science and morals, we find the possibility and the necessity of a scientific ethic as well as the emergence of those previously unseen problems of decisionmaking and responsibility that we try to deal with by means of the precautionary principle.

In terms of responsibility, the novelty of the concept of risk of development lies in the fact that risk of development necessarily reveals itself after the fact, a posteriori. It does not take the form of risk objectively developed—when the risk becomes obvious, one has already left the context of precaution—but rather that of becoming aware that what one had believed to be of a certain nature was, in fact, of a different nature, foreign, even opposite. This aftereffect gives rise to a tragic or “dramatic” quality, as in the blood transfusion drama.

Now, the French legal tradition, at least since the Age of Enlightenment, was loath to accept these aftereffect problems. Since the Declaration of the Rights of Man and Citizen in 1789, the law of criminal responsibility has been founded on the nonretroactivity of the law. “Nobody can be punished except by virtue of a law established and promulgated prior to the offence and legally applied,” specifies article 8 of the Declaration. The same standard is repeated in article 2 of the French Civil Code: “The Law only deals (p.291) with the future; it cannot have a retroactive effect.” The conventional principle of civil liability, formulated by article 1382 of the Civil Code, which introduced fault, took for granted that one could only be responsible for what one knew about; one could not be found responsible without an awareness of wrongdoing. Until then, one was judged and sanctioned for what one was expected to know, for what one should have known, this being necessarily defined in the context of a certain state of science and knowledge.

The question posed by the concept of risk of development is new in the sense that it involves imposing a sort of conflict of laws over time. Can one be fair in judging an act other than in accordance with the elements that accompanied awareness of it? Is it not unjust to judge an act from the perspective of another state of awareness than that under which it was carried out? Is it fair, even for purposes of compensation, to appraise an act in accordance with suspicions and doubts that one is only capable of having after the event?

With risk of development we rediscover the face of destiny, but with a difference: in the Old World, destiny wore the face of the gods, while for us it now has always and necessarily a human face. The face of tragedy belongs to the world of technology, to those situations in which, because of transformations in knowledge and in the very nature of things, the consumer discovers a sort of retroactive revelation, the evil that strikes at him. On the one hand, he experiences disappointment and misplaced trust—“it wasn't what I thought, what I expected, what I was told, promised”—and on the other hand, the awareness that industry sought something that not only he didn't want, couldn't want, but that he had done everything to avoid: “I didn't do that, I didn't want that, I couldn't want that.” We now live in a new knowledge–power relationship and suffer the problems of responsibility that it poses, with an awareness that our societies are vulnerable to a new type of risk, a new awareness of tragedy.

The Return of Responsibility

In the now famous book by Hans Jonas, The Imperative of Responsibility (1984), one can find the philosophical bedrock of the precautionary principle. The importance of this book, written during the 1970s, lies in its highlighting the philosophical framework of contemporary ecological awareness. For Jonas, the history of humanity is marked by a clear division between the ancient and modern worlds. The human being's relationship with nature has reversed itself: while for a long time human being was in nature, as a part of nature, humans have now, thanks to science and (p.292) technology, gained control of nature. The balance of powers has shifted to the benefit of humans, who now possess the capacity to destroy themselves and the nature surrounding them. It is from this considerable increase in power that is born responsibility in the sense given to it by Jonas. Man must be aware of his power as unlimited, sovereign, in the sense of an infinite capacity to produce effects that he could not anticipate with certainty and in the sense that he has no other master than himself. This is Jonas's first thesis: the world of ecology is a world of the most extreme responsibilities, a world in which humanity no longer has nature as guide, a world of radical uncertainty, therefore also of necessary decision.

The powers of modern man confer upon him an infinite responsibility. His nature is revealed in fear, a feeling that reveals the power of his new capacities. Temporality, within which is situated his action, dilates to encompass the whole history of humanity, past and future, but it must be acknowledged that his powers are such that they threaten the existence of life itself. Contemporary man is becoming aware of himself in the feeling of anguish before the possibilities of annihilation that he bears in himself: for the first time, he discovers in himself the power to commit suicide as a species. Faced with this possibility, and in order to overcome his anguish, modern man is on a quest to find the rules of a morality that will limit his powers: the ethics of responsibility. His enormous power needs holding.

The ethics of responsibility contain risk and uncertainty to the extent that modern man must take account in his actions of both their long-term consequences and their possibility of sweeping along with them, at least in certain cases, the worst—the catastrophe. Instead of the categorical Kantian categorical imperative, there should be substituted an imperative adapted to the new type of human action: “Act so that the effects of your action are compatible with the permanence of an authentically humane life on earth” (Jonas 1984:5). For while we have the right to risk our own lives, we do not have the right to risk the survival of humanity. This imperative is the basis of the precautionary principle: it invites us to measure each possible action against the principle of the worst scenario. Morality becomes a sort of negative morality: it is not so much turned toward the positive quest for the best as toward the avoidance of the worst. The uncertainty of long-term prognostics confers the nature of a wager on human action, which leads to questions such as: do I have the right to endanger the interests of others in my wager?

Contrary to the supreme good and eternal timelessness, the aim of responsibility is the perishable, the vulnerable. The new obligation, born from threat, focuses on the ethics of conservation, of preservation, of impediment. (p.293) For we are living in an apocalyptic situation that presents as imminent a universal catastrophe due to the extreme power of our scientific, technical, and industrial civilization. It is necessary to “kiss the utopian ideal goodbye” and to denounce the psychological danger of the promise of prosperity. Given the warnings of the Rome Convention, Jonas was not afraid to affirm that “restriction, far less than growth, should become the watchword” (1984:190).

The appearance of a new word is never without significance. The nineteenth century invented “providence” and made it the highest of virtues. The twentieth century replaced “providence” with “prevention”—providence made compulsory for reasons of social security. Now we have “precaution.” These overarching concepts represent three attitudes toward uncertainty. Providence is linked to the notions of fate, chance and misfortune, and hazard. It involves incorporating the future into the present, but on an individual scale and with the idea of a possible mastery of the event. Insurance has long been presented as the science of providence. The logic, essentially liberal, of individual providence was, at the end of the nineteenth century, a victim of the Pasteur-Léon Bourgeois association: Pasteur's discovery of infection showed that the well-being of each person did not depend solely on his own conduct but also on that of his neighbor. From that time it became possible, for reasons of public health, to impose, when faced with risks, certain behaviors that one can call by the name, new at the time with this meaning, of “prevention.” Prevention is a rational approach to an evil that science can objectify and measure. Providence was contemporary with ignorance of the hazards of existence; prevention came with the certainties of science. It speaks the same language and beckons toward the reduction of risks and their probability. Prevention is a matter for experts who are confident in their knowledge. Precaution, such as we are seeing emerge today, focuses by contrast on uncertainty—the uncertainty of scientific knowledge itself.

The precautionary framework echoes ontological and epistemological conditions that are neither those that saw the emergence of providence, nor those which presuppose prevention. Precaution is not the result of an individualistic ontology such as providence. The threats that it involves are collective. It implies motivations that are not regional, but international. And it does not belong to the contemporary preventive dream of solidarism.

The recent formulation of the precautionary principle is doubtless linked to several factors. First of all, there is the awareness of the Utopian nature, relative and limited, of the attitude of prevention. We are, in fact, seeing the proof that our societies and their members are threatened by (p.294) risks that can be disastrous but are introduced by the very acts that sought to reduce these risks. The issue of medical accidents in the age of what Jean Hamburger has called the therapeutic revolution exemplifies this problematic. When our societies discover themselves, in accordance with the now-sacred term, to be “vulnerable,” they experience a certain inseparability of good and evil, that, within the paradigm of prevention were thought to have been indefinitely separated. We are also aware of the relativity of scientific knowledge and the necessity for an ethics of science, and finally of our overweening power over knowledge that is nevertheless at the foundation of that power. Modern science no longer offers certain knowledge, as can be seen, for example, from the proliferation of ethics committees. Science now seeks its own legislation. It never ceases doubting the effects of its own development. The building of nuclear power stations demanded strict measures of prevention. This was the task of the engineers. The blood transfusion drama put a halt, if not to scientific knowledge itself, at least to the relationship we had with it. Regrettably, those responsible had been able to claim that they were only able to make decisions based on scientific certainty and had not adopted a precautionary attitude.

Precaution starts when decisions must be made by reason of and in the context of scientific uncertainty. Decisions are therefore made not in a context of certainty, nor even of available knowledge, but of doubt, suspicion, premonition, foreboding, challenge, mistrust, fear, and anxiety. There is to some extent a risk beyond risk, of which we do not have, nor cannot have, the knowledge or the measure. The concept of risk of development is found within the limit of this new figure of prudence. We have seen, in fact, that once we are aware of the existence of risk of development we can no longer plead for industry to put into circulation products offering total quality, since the precautionary attitude affirms that this is not possible. Precaution finds its condition of possibility in a sort of hiatus and time-shift between the requirements of action and the certainty of knowledge. It enters into a new modality of the relationship between knowledge and power. The age of precaution is one which reformulates the Cartesian demand for methodical doubt. Precaution results from an ethic of the necessary decision in a context of uncertainty. The appearance of the precautionary principle is a sign of the profound philosophical and sociological transformations that characterize our fin-de-siécle.

Precaution and the Risk Society

In the mid-1980s, Ulrich Beck published Risk Society (1992), a widely read book in which he demonstrated how contemporary societies revolve (p.295) around the notion of risk. He could not have known then that his hypotheses would be borne out by the appearance and future development of the precautionary principle.

In effect, the principle of precaution rounds out the agenda of the “risk society” in several ways. First of all, the principle of precaution appears as one of the primary instruments of “reflexive modernization” which, as Beck demonstrates, characterizes postmodernism. Through the notion of precaution, modern society thinks about its problems and questions its basic assumptions. More than ever, modern society finds itself cut off from the natural world, removed as it were from the scientific and technological project on which it is based. Precaution appears when scientific expertise comes up against its own limitations and forces the politician to make sovereign decisions, alone and without recourse to others. It can perhaps be said that, with the principle of precaution, the analysis made by Beck within a specific national context has become an international preoccupation. Indeed, the true sphere in which the principle of precaution applies is international law—environmental law, trade law, and government liability. Along with the principle of precaution, the notion of reflexive modernization has become a central preoccupation of the international community.

Through the notion of precaution, the experience of risk takes on three additional dimensions that build on the descriptions provided in Risk Society. First, the power dimension. The issues of liability and risk are linked to the phenomenal technological capabilities we have achieved. The industrial “will to power”—to borrow Nietzsche's famous formulation—is apparent in the proliferation of highly efficient technical systems. As this century comes to a close, industrial power is such that it no longer only causes accidents, but also catastrophes. We experience the industrial promise as fraught with the threat of disaster. When the multinational firms that manufacture genetically modified organisms (GMOs) state that, thanks to their efforts, world hunger will soon be eradicated and that environmental problems linked to intensive farming are already a thing of the past, the news sets off alarm bells. This is because such exploits are in fact tremendous displays of strength. Modern industrial power is a superpower of such might that the horizon of our accountability necessarily extends over the very long term. We are responsible for future generations. The Stoics, in an attempt to free man from worry, suggested that he draw a distinction between that which depends on us and that which does not, it being understood that we are only accountable for those things over which we have control. The problem with industrial power is that it appears that we no longer lack control over anything. Hence the boundless nature of liability (p.296) that is so worrisome. It leaves no room for innocence, as we can see from the rarefaction in law of the notions of the fortuitous event and force majeure (or act of God). However, this sense of boundlessness is merely the reflection of our awareness of contemporary industrial power.

But as Beck points out, risk is also a social relationship. The notion of precaution also raises the issue of power relationships in a global society during an era of globalization. Specifically, the power of the industrial society is exercised through relations of power, which are fundamentally asymmetrical. Modern technologies create dependencies, not equality. The more technologically developed societies become, the less they seem capable of being governed by the contractual model. The asymmetrical relationship between employer and employee is epitomized by the notion of the employment contract, which organizes the employee's subordination and the apportioning liability for professional risks. The relationship between producer and consumer is also asymmetrical, as is the relationship between the professional and the layman, underpinning the right to consume. Today, questions of liability turn essentially on these asymmetries. It is this asymmetrical dimension, and the feeling of dependency to which it gives rise, that lie at the heart of precaution. Risk is not only a danger, it is a social relationship. It is the relationship between those who have technological power and those who benefit or perhaps suffer from it.

In addition to power in the relationship with nature and power in relationships between human beings, there is the third dimension of the experience of precaution: that of harm suffered. In the industrial society, it is believed that activity and business are impossible without risk. Risk is considered normal; it is not contested in and of itself. The only question is how to organize the apportioning of risk. We can't even conceive of the idea that the acceptability of an activity or business could hinge on its posing no risk for others, provided that we do not hold liable only those who suffer harm due to the risk, and that those who put others at risk be held responsible.

With precaution, we are witnessing a remarkable change in this schema. The problem is no longer so much to multiply the responsibility for risk and to organize the solvency of those who are liable through insurance, but rather to prevent certain risks from being taken. Not only is prevention taking precedence over compensation, we are also trying to anticipate and prevent risks whose existence has not been proven. There are two major reasons for this: one is that the nature of damages has changed from mere individual accidents to catastrophes, and the other is that there has been a reevaluation of the cost of risk. A perfect illustration of the new scale on which risk is (p.297) measured: during World War I, a general could send three hundred thousand soldiers to their death in waves of fifteen, as in the battle of the Chemin des Dames. Today, the only acceptable war is the “zero risk” war. This is a peculiar transmutation of values. In the traditional cost-benefit trade-off, it was enough that the advantages sufficiently outweighed the risks for us to feel justified in taking action and thereby accepting the residual portion of the risk taken. Today, we tend to measure the risk on the basis of this residual portion: what is worth sacrificing for this? Are those who are unlucky enough to be among the victims of so much less value than the others? This is the method of valuation that lies behind the zero risk problematic.


  1. 1. If it appears that precaution results from an appropriate epistemologico-legal system, than it is worthwhile to distinguish providence and prevention, for precaution introduces a world that is neither that of responsibility nor of solidarity; it should be remembered that these three formations, far from being incompatible, are in fact complementary. We are not concerned with three worlds that succeed each other over time, each replacing another, but rather with three attitudes with regard to uncertainty, assessed and developed at three moments in time. They already existed before receiving the formalizations that have been seen over the last two centuries. We are concerned with three faces of prudence, to the extent that we interpret this notion in the sense that Aristotle gave it of behavior in the face of uncertainty. These three attitudes are not the only possible attitudes of prudence. Each has its field of competence and area of validity. For this reason, it is essential not to confuse them and to respect their spheres of influence and jurisdiction: while providence is always necessary, it is inadequate in order to confront these “global threats” that themselves relativize the attitude of prevention. Precaution condemns neither providence nor prevention. It only introduces another level of preoccupation in the conduct of humans in certain situations of uncertainty.

  2. 2. Precaution, properly understood, reintroduces the true, sovereign decision in public policy and in the practices of responsibility. Conventional law on responsibility does not so much sanction a decision well or badly taken as it does the disregard of available knowledge. The logics of fault and prevention presuppose that, in the spheres they govern, it is always possible to articulate a standard of conduct that everyone must observe. One incurs one's responsibility as soon as one does not respect the practical consequences of available knowledge, a standard that makes possible the formulation of sanctioned obligations. Precaution, which resituates us in a context (p.298) of uncertainty, reintroduces a pure logic of decision. And the rationality of the decision can no longer satisfy itself with the conventional cost-benefit balance, which is in principle unknown or at least dubious. It derives from logics that are, according to risk economists, irreducible to conventional utility functions. This has a twofold consequence: the decision still belongs to the politician rather than to the expert and is the result more of an ethic, of the respect of certain procedures, than of a morality linked to the application of a preexisting framework. It does not follow that scientific expertise is useless, but that it will not release the politician from the sovereignty of his or her decision.

  3. 3. Precaution, to a certain extent, brings us out of the age of insurance companies. It creates a world in which, in principle, compensation no longer has meaning, because the only rational attitude is to avoid the occurrence of a threat with irreversible consequences. Precaution is an attitude of protection rather than compensation. Concurrent with the growth of the theme of precaution, one can observe a great prudence from insurers as to the possibility of covering the corresponding losses. This is understandable: one is in the order of the unlimited (from the point of view of amounts), not only of the indefinite but also the undefinable (from the point of view of risk). But precaution does not itself disqualify the need for compensation once the threats at issue have consequences for individuals. From this point of view, precautionary logic accentuates the dissociation between responsibility and compensation that already characterized the face of solidarity. If the cause of injury is no longer revealed by, properly speaking, a logic of risk, the same can be said of compensation. Precautionary logic does, however, impose the construction of new outlines: contractual definition of the risks covered either by a limitation of the duration of claims periods, by fixing limits of coverage or by changing the size of ranges of mutualization. In this respect, the mechanism implemented in France in 1982 for the coverage of natural disasters provides an interesting model. It makes prevention into a political function; it institutes a mechanism for solidary compensation, at two levels, anticipating coverage if necessary from the state in the event of a superdisaster (Magnan 1995).

Responsibility, in the strict sense of the word, corresponds to providence and solidarity to prevention. Categories that were simultaneously moral, political, and economic, responsibility and solidarity described a form of social contract, a method of distribution of rights and obligations in society. The word is missing that would correspond to them within the precautionary paradigm. Maybe one can, while waiting for a better locution, speak of safety, inasmuch as the term describes, as in nuclear terms, a particularly (p.299) strict safety requirement. And, in the long term, precaution demonstrates a sort of inflexion in our attitude with regard to risk. Fear, from which our societies had perhaps, in accordance with the progressive utopia of the twentieth century, believed they could free themselves, returns in a new form. Solidarity had almost made us riskophiles; now we are almost riskophobes, individually and collectively, and will likely remain so for some time.

The emergence of precaution accompanies the crisis of progress, a certain overvaluation of the past in relation to the future, the desire to limit the destructive effects of time, and perhaps also a new suspicion aroused on the human species and the rationality of its development. This occurs with the proviso that what is concerned is freedom, enterprise, innovation, their unwanted consequences, in the long term, on the scale of the species. Precaution is a way of raising, considerably, the price of innovation. The alert has been given by a succession of affairs in the sectors of the environment and public health. Damage, serious and irreversible, has been caused that now appears could have been avoided by the observation of a precautionary attitude. It is behind the crisis of confidence that means that the spirit of enterprise, of creation, and of innovation, even in this period of unemployment, is no longer valued as before, with a certain privilege being given to abstention. In the age of precaution, the value of enterprise depends less on the well-being that it procures than on the urgency and the degree of necessity for products that it makes available. We wondered, in a previous age, whether it was better to have a big enterprise or a small enterprise; now what counts is frugality. Paradoxically, at the dawn of the twenty-first century, value is no longer in abundance, but in scarcity. So be it; but if precaution is no doubt necessary, one can also fear the consequences of too great a precaution. Precaution demands regulation.

As we become aware that certain activities demand a precautionary attitude, and that precaution is being discovered as a principle of responsibility, we see the formulation of precautionary legislation. In fact, precaution may be the worst or the best of principles. As the saying goes: “When in doubt, do nothing.” The danger of precaution is that it may result in inaction. There is another saying: “Too many cooks spoil the broth.” The precautions to take in the context of a technologically developed society are no doubt necessary, but, they must remain “reasonable” to prevent them from leading to an exhaustion of innovation and therefore to a revolutionary change in society with even more unfortunate consequences. This is the difficult tight rope that we must tread in pursuit of the idea of sustainable development.

(p.300) Notes

An earlier version of this essay was published as “The Return of the Crafty Genius: An Outline of a Philosophy of Precaution,” Connecticut Insurance Law Journal 6, no. 1 (1999): 47–80, and is reprinted by permission of the Connecticut Insurance Law Journal.

(*) Editors' Note: France's tainted blood crisis received only cursory attention in the United States, and it is therefore helpful to include a brief explanation. In June 1985, Michel Garretta, then-director the of French National Blood Transfusion Center, ordered the normal distribution of nonheated blood products. Garretta decided that the widely accepted and American-pioneered heat treatment procedure that inactivated the AIDS virus in blood was unnecessary. This order was in direct conflict with the practices of other nations, including the United States and Canada. As a result of this order, French patients received untreated and untested blood for six months. In 1992, it was reported that about one-half of France's four thousand hemophiliac population had contracted the virus and more than 250 had already succumbed to it. A battle of who was to blame ensued—scientists pointed the finger at the politicians and the politicians at the scientists. And there were also those who claimed they were blameless. Georgina Dufoix, at that time the Social Affairs Minister, stated paradoxically that she was “responsible but not guilty.” As one author noted, “[N]one individually were responsible; the system was assigned responsibility. And now they all claim that they were not at fault. The hemophiliacs are dead because they were not seen. As public technocracies spread, perhaps this is the sort of defendant we had better get used to.” Wall St. Journal 1992; see generally Feldman 2000.


Bibliography references:

Beck, Urlich. 1992. Risk Society: Towards a New Modernity. Translated by Mark Ritter. Beverly Hills: Sage.

Cuvier, Baron Georges. [1863] 1969. The Animal Kingdom. Reprint, New York: Kraus.

Descartes, Rene. Meditations on First Philosophy. [1641] 1992. Edited and translated by George Heffernan. South Bend, Ind.: University of Notre Dame Press.

Feldman, Eric A. 2000. “Blood Justice: Courts, Conflict and Compensation in Japan, France and the United States.” Law and Society Review 34:651–702.

Gény, Francois. 1902. “Risque et responsabilite.” Revue Trimestrielle de droit civil. Paris: Sirey.

Jonas, Hans. 1984. The Imperative of Responsibility. Translated by Hans Jonas, in collaboration with David Herr. Chicago: University of Chicago Press.

Josserand, Louis. 1897. Responsabilite du fait des choses inanimees. Paris: Libr. nouv. de droit et de jurisprudence, A. Rousseau.

Labbé, Jean Etienne. 1885. Paris: Sirey.

Mangan, Serge. 1995. “Catastrophe Insurance System in France.” Geneva Papers on Risk and Insurance 20:474–80.

Plato. [n.d.] 1991. Protagoras. Translated by C. C. W. Taylor. Oxford: Clarendon Press.

(p.301) Portalis, Jeane-Etienne Marie. 1989. In Naissance du Code Civil, an X11—1800–1804. Paris: Flammarion.

Saleilles, Raymond. 1897. Les Accidents de travail et la responsabilite civile. Paris: A. Rousseau.

Sanson-Hermitte, Marie-Angèle. 1996. Le sang et le droit: essai sur la transfusion sanguine. Paris: Editions du Seil.

Thiers, Adolphe. 1850. Rapport Au Nom De La Commission de l̓assistance et de la prevoyance publiques 6. Paris: Paulin, Lheureux.

Thoulon, Edouard. 1898. La Responsabilite des accidents don't les ouvriers sont victimes dans leur travail. Paris: Libr. nouv. de droit et de jurisprudence, A. Rousseau.

United Nations. 1992. U.N. Conference on Environment and Development, Rio de Janeiro. The Rio Declaration on Environment and Development, Principle 10. The Earth Summit, AGENDA 21, rio declaration, forest principles. New York.

Wall St. Journal. 1992. “Murder by Bureaucracy.” Wall St. J., 5 August. A14. (p.302)


An earlier version of this essay was published as “The Return of the Crafty Genius: An Outline of a Philosophy of Precaution,” Connecticut Insurance Law Journal 6, no. 1 (1999): 47–80, and is reprinted by permission of the Connecticut Insurance Law Journal.

(*) Editors' Note: France's tainted blood crisis received only cursory attention in the United States, and it is therefore helpful to include a brief explanation. In June 1985, Michel Garretta, then-director the of French National Blood Transfusion Center, ordered the normal distribution of nonheated blood products. Garretta decided that the widely accepted and American-pioneered heat treatment procedure that inactivated the AIDS virus in blood was unnecessary. This order was in direct conflict with the practices of other nations, including the United States and Canada. As a result of this order, French patients received untreated and untested blood for six months. In 1992, it was reported that about one-half of France's four thousand hemophiliac population had contracted the virus and more than 250 had already succumbed to it. A battle of who was to blame ensued—scientists pointed the finger at the politicians and the politicians at the scientists. And there were also those who claimed they were blameless. Georgina Dufoix, at that time the Social Affairs Minister, stated paradoxically that she was “responsible but not guilty.” As one author noted, “[N]one individually were responsible; the system was assigned responsibility. And now they all claim that they were not at fault. The hemophiliacs are dead because they were not seen. As public technocracies spread, perhaps this is the sort of defendant we had better get used to.” Wall St. Journal 1992; see generally Feldman 2000.