For anyone who has taken even a slight interest in tort law in legal systems other than our own, one of the striking features of French law is the absence of mechanisms with which to ring-fence liability. Where English and German law, to cite just them, cultivate numerous mechanisms capable of containing claims for remedy, our law advances, without any apparent qualms, towards a constant widening of the scope of the obligation to make good the harm caused to others. Recent examples of this trend are not wanting, at least in case law. France’s supreme civil court, the Cour de cassation regularly manifests its determination to relax the conditions for invoking liability, whether with regard to seminal acts, harm or causation.
This generosity towards plaintiffs bestows on French tort law an amplitude that is unknown to almost all foreign legal systems. For sure, this is not the only area in which France enjoys going it alone, but this Gallic oddity is still intriguing. First, it can be asked what the reasons are for this leaning towards compensation. Answering this question would involve adjoining to an examination of the specific technical features of French law a study of the other factors, which stringent positivism would characterise as extra-legal, that have contributed to this state of affairs and among which it can be surmised would figure prominently the concern to deal with all victims identically, a degree of hostility towards direct insurance and the influence of the ideas, unfortunately often a little distorted, of leading commentators, in the vanguard of which stands, of course, the scholar to whom these lines are dedicated. It would also be interesting to know what the economic consequences are of this greater extension of French law to civil liability, compared with foreign legal systems. Unfortunately, despite the development of the economic analysis of law, there are to date scarcely any works measuring the concrete economic effects of the application of civil law rules, in France at any rate.
This absence has as a consequence that it is difficult to know whether, from an economic point of view, the ease with which civil liability claims can be brought in French law is a good or a bad thing. Contemplated from other angles, though, this facility which many rejoice about may seem worrying. Might not going into overdrive on civil liability risk discouraging socially useful activities? True, the spectre of the ‘Americanisation of law’ should not be summoned up wrongly, but the Perruche ‘affair’ does give us food for thought, since the Cour de cassation’s choice to recognise a congenital disability as remediable harm for the child itself, which was not self-evident, prompted fear of sufficiently serious consequences for the activity of ultrasound practitioners for Parliament to pass legislation breaking with that case law. Moreover, through wanting to remedy pretty much anything and everything, is there not a risk of remedying less well? One might bite off more than one can chew, and by dint of having remedies for the most diverse forms of harm, the most essential interests of the person – those for which civil liability has been devised and developed – might receive less attention. Already, the relaxation of the conditions for bringing liability claims is boomeranging straight back at the victims themselves, as attested by the consequences of strict liability on the characterisation of the victim’s contributory negligence. Furthermore, with the multiplication of mass torts and the emergence of new types of remediable harm, which may prove substantial in scope, such as ecological harm, it is not certain that the apparently sacrosanct principle of French law, by virtue of which all types of harm must be fully remedied, will long remain tenable.
French law might therefore find itself constrained, sooner or later, to temper its remedial ardour. Most likely the legislature will have the opportunity, when the time comes, to step in to contain tort law and limit its extension. However, there is no need to wait until our backs are against the wall to think about means by which tort law could be bounded in a way that would be neither arbitrary nor unfair and that safeguards the main accomplishments of the development of this branch of law since the late 19th century.
Given the structure of French law, there are at least three levers that can be activated so as to control tort law: seminal acts, causation and harm. Each of these conditions for implementing liability may be defined more or less restrictively, with the consequences it entails for the obligation to make whole. Observation of foreign law suggests that a coherent policy for framing liability involves circumscribing as precisely as possible all of its conditions of application. In the limited context of this article, however, we shall restrict ourselves to the third term of the equation, namely harm.
The brilliant demonstration made of the progressive dissolution of harm in French law would itself alone vindicate that the efforts made to ring-fence liability should bear as a matter of priority on this so roughly-handled concept. Another reason for this choice, though, is that the centre of gravity of civil liability seems to have shifted from the conduct of the tortfeasor to the position of the victim. Consequently, it is not illogical to seek to bound the scope of liability by acting on what is now its primary component, both in terms of the genesis of the right to redress and in terms of theory, that is, harm.
Framing liability by the slant of harm presupposes that the concept is identified as precisely as can be. It matters in this respect that it be clearly accepted that harm is a legal concept and that the existence of suffering, shortcoming or impairment is not necessarily sufficient in itself to characterise harm as understood in tort law. Bounding the concept of harm in this way is not enough, though, to make it into an effective tool for framing liability. It must further be accepted that not all harm, even when recognised by law, can automatically be remedied by civil liability. In other words, not only does every wrong not necessarily constitute harm (section I), but moreover not all harm can necessarily be made good (section II).
1. NOT EVERY WRONG NECESSARILY CONSTITUTES HARM
Pinning down the concept of harm (préjudice) first entails clarifying its relationship with the concept of tort (dommage) (section A). That will make it possible subsequently to appraise whether the traditional – and often decried! – definition, which sees harm as detriment to a legally protected legitimate interest, is well-founded (section B).
1.1. Tort and Harm
In its provisions on tort law, the French Civil Code does not use the term ‘préjudice’ (here harm) but that of ‘dommage’ (here tort), albeit without defining it. The two terms, however, were long held to be synonyms. At most some commentators observed that one spoke more readily of harm ‘préjudice’ in the event of a direct injury to a person and of tort ‘dommage’ in the event of interference with a thing. However, since the 1950s, one strand of doctrine, which is becoming ever more mainstream, has recommended recognising a conceptual distinction between tort and harm, a distinction that can supposedly be traced back to Roman Law. Even if there are still sometimes slight differences in approach among them, the observers who adhere to this distinction usually consider that tort (dommage) is defined as interference with a person, a thing or a situation. Harm (préjudice), for its part, consists in the consequences of this interference for the plaintiff. These consequences may be pecuniary (reduction in the value of an asset, loss of income, etc.) or non-pecuniary (physical or moral suffering, essentially).
The conceptual relevance of this distinction between tort and harm is seldom contested, but many authors mention it without dwelling on it, observing that it is absent from positive law and doubting its practical value. The simple fact that the distinction provides a better account of the actual structure of liability, at least in some instances, might, though, suffice to warrant it being adopted. The interference resulting from the tortious act is one thing, the consequences of that interference are another. The distinction between the interference and its consequences is besides made most explicitly in some legal systems, such as Swiss law, and lies just beneath the surface in others. It sometimes even crops out in French law. Several texts speak of remedying the tort (dommage) arising from interference with the person. True, these texts do not adopt the terminological distinction between tort (dommage) and harm (préjudice), but they do enshrine the conceptual distinction between the interference and its consequences. Traces of the distinction can be found in case law too, often rather cryptically but sometimes quite explicitly. And this is just in civil liability, because administrative law and even criminal law are more receptive to this distinction.
From a practical point of view, is there any real interest in dissociating harm from tort? De lege lata, the interest is probably not great, even if the distinction can be used to justify certain solutions that doctrine otherwise struggles to account for. De lege ferenda, however, the distinction has tremendous possibilities. Some need not be developed here, as they do not relate directly to the subject that concerns us. Others, by contrast, have a far closer connection with the problem of circumscribing liability. In particular, to leave greater scope for the distinction would probably be a means to remove certain ambiguities that presently beset the remedying of tort/harm and to emphasise the legal character of the concept of harm.
For one thing, case law tends to take a factual and concrete approach to tort/harm, which is expressed notably through the principle of the full remedy for the tort/harm and the principle of the final determination of its amount by the trial and appeal courts. For another thing, the Cour de cassation does not hesitate, sometimes, to abstractly determine the existence of tort/harm, and even its scope. Thus, the plaintiff in an action who is in a coma still has a right to compensation covering all ordinary counts of harm, including non-pecuniary harm, appraised objectively, notwithstanding the fact that he probably has no awareness of his condition and therefore incurs no harm from a subjective standpoint. In the same vein, the Cour de cassation has on some occasions decided that the infringement of a right necessarily causes harm to the right holder, even though it may be that, materially, the infringement has not entailed any actual adverse consequence for the right holder. It seems to me that there is unresolved tension in positive law between two contradictory trends, with tort/harm being sometimes thought of as a purely factual matter, the existence and extent of which should be gauged in concreto by the trial and appeal courts and sometimes as a legal concept that may be appreciated and evaluated objectively or abstractly by the Cour de cassation.
Endorsing the distinction proposed by doctrine between tort and harm might be one way of escaping this ambiguity. Tort would then be recognised as a purely factual component, the existence and nature of which would be matters to be determined exclusively by the trial and appeal courts. As for harm, it would be openly considered as constituting a legal concept, which would make it possible to explain that harm, in the legal sense, does not always coincide exactly with the actual detrimental consequences of the tort incurred. For in the same way as legal causation does not always map exactly to ‘factual’ causation, legal harm does not always map to ‘factual’ harm. The case law solutions cited earlier could then be reinterpreted as hypotheses in which the courts and tribunals infer the existence or even the scope of harm, as a legal reality on the basis of the finding of tort, a matter of fact, without necessarily matching the ambit of the harm to the detriment actually incurred and felt by the plaintiff.
Interpreting case law in this way obviously amounts to openly acknowledging departures from the principle of full remedy because in all these hypotheses, under cover of objectively determining the harm, one may in fact come up with compensation that exceeds the harm actually incurred or felt. This would not, though, be the first departure from the principle of full compensation. What is more, calling into question the principle of full remedy, the relevance of which has long been queried, might indeed be one of the beneficial effects of recognising the distinction between tort and harm. While remedy is conceivable with regard to tort – it is usually a matter of making good the interference, where possible, which is far from being always the case –, it is often meaningless with regard to harm, at least where this is non-pecuniary or emotional. In such instances, it is difficult to speak of anything other than compensation. Rather than clinging to the principle of full remedy, which is that much more debatable because, as seen, case law often appraises the harm in abstracto, regardless therefore of the actual loss felt by the plaintiff, it would be better to try to tell apart situations in which remedy for the tort is possible and situations in which only compensation for harm can be contemplated, for which full compensation may not be meaningful.
Challenging the principle of full remedy would furthermore have the great merit of putting an end to the idea, which reassertion of the principle tends to sustain, that civil liability is designed to make good any and all interference that might befall those in the frame for liability and to erase all consequences of accidents that might happen to anyone. For it is wrong to believe that this is so. Civil liability law does indeed seek, where it can, to restore victims to the position they would have been in had the fact in cause not occurred, but that presupposes that there was a remediable harm recognised as such by law. Now, not every infringement of a person’s interests necessarily constitutes harm. Harm is a legal concept and law is authorised to decide what falls within the category of harm and what does not. To recognise the distinction between tort and harm would be one way of asserting clearly the legal character of the concept of harm. This, in turn, would justify a definition of harm that selects from among the instances of interference or suffering that a plaintiff may evoke, those for which compensation or remedy may be claimed in civil liability law. It is towards such a definition that it is now time for us to turn.
1.2. Harm, Interference with a Legally Protected Legitimate Interest
Distinguishing tort (dommage) from harm (préjudice) amounts to seeing in the latter the negative consequences that the former may have for the claimant. That, however, does not provide a substantive definition of harm. Harm was long perceived more or less implicitly as a loss with respect to some condition or earlier situation. This conception, though, is not readily compatible with the development of non-pecuniary or moral harm. It has also been radically challenged by the Perruche decision, which saw a harm in a congenital disability. Another approach, long prevalent in Germany and having influenced other legal systems, is to define harm as a measurable reduction in assets. It has never really resonated in France, though, and barely deserves closer scrutiny insofar as it proves incapable of taking satisfactory account of non-pecuniary harm. Rather than as a loss or as a pecuniary difference, it is better to think of harm as interference with an interest. Contrary to appearances, this definition is not to be conflated with the definition that sees in harm a loss or deterioration compared with some earlier state of affairs. To take just one example, the interest may be the achievement or obtaining of some condition, in which case the failure to achieve that condition might be an interference with that interest, without there being any loss or deterioration for all that. Thus, if it is accepted that living without disability is an interest that is recognised by law, being born with a congenital disability may constitute a harm, even though there is no loss compared with some earlier state of affairs.
Defining harm as interference with an interest, apart from having the merit of being straightforward, clearly illuminates the very reason for being of civil liability, which is to protect certain interests (indirectly by compensating or remedying infringement of them). However, not all the interests that a claimant may invoke are intended to be covered in this way. The interest interfered with must still be recognised by law and judged by law as worthy of protection. In other words, the recognition of a harm implies a value judgement made by the legal system. It is because law has this capacity to select from among interests that may be interfered with those that are worth protecting that harm is and must be a fully legal notion, capable as such of being defined and framed by law.
This necessity to distinguish among interests that might be interfered with those that law wishes to protect and those to which it does not intend to afford its protection compels us to refine the definition of harm just proposed. However, it is not a matter of laying down a priori a list of interests the interference with which would constitute a harm. This would be contrary to the spirit and tradition of French law. It is better to resort to a general formula. The most satisfactory seems to me to be that which sees in harm the interference with a legally protected legitimate interest.
Admittedly, the formula is a sulphurous one, having been used by case law in the middle of the last century to prevent remedying the harm incurred by an unmarried woman in the event of her partner’s death. It seems to me, though, to deserve better than the opprobrium many have heaped on it. For sure, if one likens the requirement of the interference with a legally protected legitimate interest to that of interference with a subjective right, on the basis of the definition Jhering gave of such a right, it may appear excessively stringent. If, however, we leave aside the concept of subjective right, I see nothing shocking in the law demanding, to receive a plaintiff’s action in liability, that the claimant argues for interference with a legally protected legitimate interest. On the contrary, ‘legally protected’ underscores this essential aspect, already mentioned, which is that the law can and must select among all interests the protection of which may be called for by litigants. Only those interests that law deems legitimate must be protected by means of civil liability. The question then, obviously, is how can one tell what a legitimate interest is. Before attempting an answer, it should be recalled why the requirement of a legitimate interest has been deemed problematical.
It has been criticised first of all on the ground that it would create confusion between the basic conditions of success for an action in liability and the general condition of admissibility of actions in justice laid down by article 31 of the Code of Civil Procedure by which ‘an action may be brought by anyone with a legitimate interest in a claim succeeding or being dismissed’. While it is true that the use of the same term may give rise to ambiguity, it must be seen, however, that the interest mentioned by the Code is not identical to the interest the interference with which constitutes the harm. This infringed interest, by definition, pre-dates the interference which itself constitutes, in principle, a prerequisite for formulating the claim and for bringing the action. It cannot therefore be confused with the plaintiff’s interest in the success of the claim, which only arises with the emergence of the dispute. Besides, the content of the interest covered by the Code is not necessarily the same as that of the interest the interference with which constitutes the harm. The success of the claim, for the plaintiff, is not invariably conflated with the restoration of the interest interfered with. Thus, the plaintiff who claims physiological suffering and complains of an infringement of the fundamental interest that is his well-being will not see that well-being restored by the mere fact that he wins his action in liability. Requiring that the interest interfered with be legitimate does not therefore amount to conflating the substantive conditions and the conditions for admissibility of the action in liability.
The requirement that the interest interfered with be legitimate has also been criticised on the ground that it is supposedly the means for an undue intrusion of morality into civil liability law. This criticism seems unwarranted to me. It is true that the idea of legally protected legitimate interest has been used in its time by the courts to impose a solution that now seems narrowly and outdatedly moralistic – but today’s social morality is not that of yesterday and the risk of anachronism is great when attempting to make moral judgements with hindsight even of a few decades. In absolute terms, however, it does not seem illegitimate that moral considerations should be present in the law of liability. Why should morality be excluded from this branch of law when it innervates so many others, from criminal law to human rights, via contract law? More generally, and even if it is important to firmly uphold the distinction between law and morality, it seems to me vain to claim to seal off these two normative orders from one another. Law is nourished by the morality of the society in which it is steeped, and so it should be.
Besides, the requirement of a legitimate interest does not necessarily relate to moral considerations. The whole problem is to determine what a legitimate interest is. Now, the notion of legitimacy proves to be particularly slippery. It seems impossible to define a priori what a legitimate interest is. Or more exactly, in the current state of the French system of civil liability, which there is no question of fundamentally calling into question, any interest invoked by the plaintiff may be presumed legitimate. It is not a question, therefore, for the plaintiff of proving that the interest he claims has been interfered with is legitimate, but as the case may be, for the defendant to prove, or possibly the court to find, that this interest is not legitimate and that there is therefore no cause for remedy, for example, because it is against public policy to do so or it would be an infringement of the respect of human dignity. This is, in any event, the position implicitly adopted in case law, especially when it refuses to award a remedy for the loss of an unlawful gain. In a case of the kind, the interest interfered with is illegitimate in the eyes of the law and it is therefore unable, by means of civil liability, to grant what amounts to an advantage of which it otherwise prohibits the enjoyment or obtaining. A clear distinction must be made in this respect between the illegitimate or illicit character of the interest interfered with and the illicit situation of the plaintiff in the action for remedy. In the event the interest interfered with is unlawful (say, if the plaintiff complains of the accidental destruction of counterfeit goods), the action in liability cannot succeed, because there is no harm in the eyes of the law. However, were the plaintiff in an unlawful situation at the time of the wrongdoing, but the interest interfered with is not unlawful, there would be a harm under liability law, for which the plaintiff could seek remedy, save that the plaintiff’s unlawful situation should be considered a wrong reducing in part or in full his right to redress (this is the hypothesis of a passenger without a ticket who is injured when a train derails: he was in an unlawful situation at the time of the accident, but the interest interfered with, his health, is entirely lawful).
The legitimacy of the interest interfered with, or more specifically the absence of unlawfulness, is therefore a condition for the existence of a harm within the meaning of the law of liability. The harm can then be validly defined as an infringement of a legally protected legitimate interest. However, this definition remains very broad. True, it excludes remedies for certain infringements, in rather particular instances, but it does not as such constitute an effective means of narrowing the scope of liability. If therefore we are to ring-fence liability based on considerations of the interests it aims to protect, it is not possible to hold to the principle that all harm must be remedied. It is necessary to introduce different treatments for harm, based on the recognition of a hierarchy of protected interests.
2. NOT ALL HARM CAN NECESSARILY BE REMEDIED
Although French law now centres civil liability on harm (préjudice) and no longer on seminal acts (‘faits générateurs’), it maintains a fairly cursory and undifferentiated conception of harm and leaves little scope, for the time being, to the idea that not all interests deserve to be equally protected. However, not all interests, even if legitimate and recognised by law, are to be ranked equally. It can be argued that the process of diversification of interests that might be recognised by liability law must lead to a typology of those interests (section A), which might ultimately lead to differentiated protection mechanisms (section B).
2.1. Diversity and Typology of Protected Interests
Article 1382 of the Civil Code makes no distinction, nor a fortiori establishes any rank order, among interests interfered with when it asserts that ‘Anyone’s act that causes tort to another places the one through whose fault the tort occurred under an obligation to make it good’. It certainly marks the intention of the drafters of the Civil Code not to exclude a priori remedy for infringement of certain interests on the grounds that they might be less important or less legitimate than others. In this, this article stands clearly apart from foreign legislation and especially from paragraph 823 (1) of the BGB, which on the contrary contains a list – true, not an exhaustive one – of the interests interference with which may give rise to remedy.
However, it is not possible to abide by the intentions of the drafters of the Civil Code and refuse as a matter of principle to establish any distinction or any rank order among protected interests by civil liability law on the ground that ubi lex non distinguit, non debemus distinguere. Apart from the point that the exegetic argument is scarcely admissible when it comes to articles 1382 and following of the Civil Code, given the liberties that case law has taken with these texts, it is obvious that the authors of the Code could not have had in mind either the extraordinary extension of tort law that was to come nor the multiplicity of new interests to which social changes were to give rise. It is true that the 1804 speech by Tarrible, often cited, in which this statesman discusses the provisions of the Civil Code about to be adopted, asserts that any tort (dommage) must be remedied. Yet, at the same time, this speech attests to the relatively narrow scope of interests that the jurists under the Consulate could have had in mind. Physical damage to bodies or property seem to have been the only torts for which it was conceived at the time that a remedy might arise. Indeed, how might jurists of the time, unless they were extraordinarily prescient, have been able to imagine the many interests to which the evolution of society and sensitivity has given legal life: interference with the environment, privacy, a normal sex life, the common pleasures of existence, philosophical or religious convictions, and so on? Moreover, the appearance of new interests is a process that is certainly not running dry. On the contrary, the complexification of our society and the development of new activities cannot but create new interests, whether pecuniary or non-pecuniary, which civil liability might be called upon to protect.
The sociological evolution of Western societies is also a source of new forms of harm. Litigation pertaining to birth and life is highly significant in this respect. Fifty years ago in France, trials with plaintiffs asking for remedy for the harm caused by their own disabled life, incestuous conception, or the birth of a disabled child or sister would have been unthinkable. No one would have seen in those tragedies harm that might be remedied by an action in liability. The change in thinking means that nowadays we feel to be harm calling for remedy sufferings that once would probably have been viewed as inevitable or at any rate as not falling within the ambit of the law. It seems to me that this shows there is in the concept of harm a psycho-sociological dimension that is probably not given enough attention. Unhappiness or misfortune are not in themselves sufficient, in the minds of those within the frame for justice, to constitute a harm, unless there is a feeling that there is some anomaly, a departure from the normal course of things which could have been averted.
In any event, there is no reason why the trend towards the multiplication and diversification of harm should be reversed, quite the contrary. If people can today obtain redress for the harm caused by the birth of a disabled sister, why not tomorrow grant compensation to someone who complains of the birth of a perfectly ‘normal’ younger brother whose arrival disrupts their tranquillity and the harmony between the parents and reduces their prospects as to inheritance? And were the statute no longer to provide that no one can claim a harm merely because they have been born, what would one day prevent someone from suing their parents for having let them come into the world, so bringing them into what is for some a vale of tears? At the time of the Perruche affair, many deemed such a prospect unconscionable particularly on the ground that life in itself, independently of any disability, is not a remediable harm. But if the harm is nothing more than the infringement of a legitimate interest, why should not the very fact of being alive be a harm for some, a fortiori if the plaintiff has some disabling illness which the parents knowingly took the risk of passing on to their child? Even if such a prospect can be put aside, the emergence of other harm, which at present may still seem fanciful, cannot be ruled out. It would not be surprising if, sooner or later, someone might have the idea of suing their parents for bringing them up badly, thus depriving them of better prospects or causing them suffering for the rest of their lives. Now, in the current state of French law, if negligence or wrongdoing in the upbringing of children is proved, and given the very broad definition of harm that is usually taken, it is hard to see what legal argument might justify dismissing such a claim, short of reactivating perhaps article 371 of the Civil Code. It is not very difficult, besides, to imagine other forms of harm, however weird and wonderful, the existence of which could be asserted one day by a plaintiff and for which remedy could not be dismissed on the basis of the current conditions for implementing civil liability.
True, the Cour de cassation, under cover of applying the law, knows how to sweep aside claims for remedy on the basis of considerations of opportuneness, especially by playing on the poorly defined notion of causality. However, this way of avoiding remedy for harm the recognition of which is deemed socially inopportune, apart from it being somewhat hypocritical, has obvious limits, notably when the negligence or wrongdoing is incontrovertible and the relationship of cause and effect with the alleged harm is too direct to be denied. It seems to me therefore that if the movement of diversification of harm were to continue and take on worrying scope, there would be no other choice but to consent to question whether all harm was equally intended to be remedied. Now, if there is one criterion that can justify the remediable character or otherwise of a harm, it is primarily, it would seem, the nature of the underlying interest interfered with. Not all interests have the same status and it is possible to outline a typology of interests protected by civil liability, which might itself serve as a basis for a differentiated protection of those interests.
The first interest that civil liability protects is the physical integrity of the human being. The need to make good these infringements and their consequences (pecuniary and non-pecuniary) for the direct victim barely requires discussion. It is first these harms that the drafters of the Civil Code had in view and it was to remedy them that most of the systems of strict liability or compensation that exist in French law arose and developed. A glance abroad reveals that even legal systems which, like English or German law, have a restrictive conception of remediable harm admit by way of priority the remedy of injury to the human body and its consequences.
The second interest protected by liability is the material integrity of tangible property. Remedy for damage to such property and its pecuniary consequences is very widely accepted in French law as in foreign legal systems. Yet, probably not all property has the same status, because some is almost an extension of the person or an essential feature of their living conditions, whereas other property is not as important.
Another category of interests is that of intangible interests to which the social and legal order accord particular importance, which is reflected by the existence of a specific regime of remedy for infringements of them. These interests may vary with place and time. In the current state of French law, these intangible interests are especially reputation and privacy. The particular importance of these interests and the necessity to reconcile them with the exercise of other rights or freedoms, and particularly free speech, is manifest from the special regime to which their protection is submitted in part or in full.
The first three categories are quite clearly demarcated. In the case of physical integrity of people and property, the scope of infringements is bounded by the limits of the physical world. True, the consequences of such infringements may be variable and the law recognises new ones from time to time (think of the acceptance of sexual harm or the harm of contamination). Even so the material substrate of these harms largely limits their diversity. As for specially protected intangible interests, the list of them is limited by definition, even if it likely to change over time.
Then comes a fourth category, which brings together all the intangible interests that do not enjoy any special protection. These intangible interests long had a fairly limited variety of forms. These were essentially ties of affection and the economic interest derived from the pursuit of a gainful activity or the receiving of an income. Excepting the case law saga over compensation for the loss incurred by unmarried partners, French law has always tended to accept the broad compensation of infringement of such interests, in which it stands apart from most foreign systems of law. It has thus widely accepted compensation of indirect harm, whether pecuniary or non-pecuniary, for close relatives of the direct victim, and the compensation of what foreign legal systems usually name pure economic loss, that is, economic harm independent of any physical harm to a person or to property. That being so, and despite acceptance in principle of compensation for infringement of such intangible interests, case law sometimes displays a degree of reluctance, as indicated by its hostility to compensating economic harm incurred by the employer or joint contractor of the direct victim. Now, this reluctance could increase because it is this fourth category that the new forms of harm referred to above are coming to swell and will continue to swell. All share in common the fact that they result from the infringement of intangible interests. Because of the unlimited character of the universe of intangible things, there is no natural bound to the domain of intangible interests that might be interfered with and give rise to claims for remedy. As said, changes in society and claimants’ imaginations are capable of throwing up new interests without end. The concern to maintain liability within reasonable limits may then limit the remedy for harm of this type, as in most foreign law systems. Besides, there are reasons inherent in the nature of these harms that makes it less obvious that there is any necessity to remedy them.
In itself, the intangible character of the interests in question might justify less protection from the law. Admittedly, law has long recognised intangible phenomena and contemporary law is marked by the increased importance of intangible values. Even so it would not be shocking for liability law to give priority, if priority there must be, to the remedying of troubles occurring in the physical universe. For although they are not necessarily more serious, their material character lends them an objective quality and a density that law cannot ignore, which is not true of infringements of intangible interests. Such infringements, in many instances, also have the characteristic of being felt in a purely subjective manner by those who complain of them. That does not make the infringements any less real, but it does raise problems as to how to measure and evaluate them, which, although long familiar, continues nonetheless to be a difficulty. How can one measure the grief of someone whose relative has died? These problems become all the more acute when the harm invoked is properly metaphysical: who can claim to measure and evaluate the suffering related to existence, to disquiet or to jealousy, apart from the lawyer who must draft pleadings and the judge who has no choice but to do so?
To all of these difficulties, a further one can be added, when the harm referred to is that, because of the act for which remedy is claimed, future events did not occur as they should have, thereby causing a loss of income. In such an instance, the plaintiff invokes not a tangible change in the actual situation, as with harm to a tangible interest, but the hypothetical change to some future condition. Apart from the fact that one can then question the real and true character of the tort, it may appear debatable that the plaintiff should secure remedy because of the occurrence of an event that negatively affects his chances of gain, when the plaintiff shall owe nothing if an event occurs that on the contrary increases his chances of gain. Through the remedy for pure economic loss, the law provides litigants with a form of insurance against certain events that may adversely affect their economic activity whereas it does not make them pay any price for all the windfalls that the course of events may bring their way. There would be nothing shocking about this if pure economic loss were not merely virtual. Systematically remedying it is, in a way, to keep for the plaintiff the benefit of the chances of gain and to transfer to a third party the risk of loss inherent in any economic venture. Economically, the systematically remediable character of economic loss seems therefore open to criticism.
More generally, to admit without discussion the remediable character of all harm resulting from infringements of intangible interests under the fourth category, is to open the floodgates to an uninterrupted extension of the domain of liability, with all the risks, especially economic risks, which that may involve, and for a social benefit that is still largely to be proved. It therefore seems necessary to contemplate ways of limiting the extension of remediable harm.
2.2. For Differentiated Protection of Interests
The idea of differentiated protection in tort law of interests that are otherwise recognised as legitimate by the legal system may seem shocking. It appears to run counter to the French tradition in this domain by which any harm is remediable when the other conditions as to liability are met. As Geneviève Viney wrote, however, ‘civil liability is not made for fully remedying any and all imaginable torts, but for meeting social demand for compensation which is more or less pressing and more or less legitimate depending on the nature of the interest interfered with and the gravity of the harm. To deny this gradation of social expectations is neither fair nor realistic from an economic point of view’.
There are now, besides, several exceptions to the equally remediable character of all types of tort. In at least one decision, the Cour de cassation dismissed a claim for remedy not because there was no harm but because the plaintiff failed to invoke any remediable harm. Liability for nuisance also attests that certain characteristics are sometimes required for there to be remediable harm, in addition to the mere existence of harm. The statute of 5 July 1985 on road traffic accidents also provides a typical example of the rank ordering of protected interests. While it does not a priori exclude the remedy of any type of harm, it subjects to separate regimes the tort resulting from physical injury and that resulting from damage to property. The case of product liability must also be cited. The EC directive of 25 July 1985 provides for compensation for harm resulting from infringement of the physical integrity of people and property provided that the products were intended for private use. It therefore excludes remedy for harm arising from damage to property for professional use and a fortiori remedy for pure economic harm. The reason is that this special regime was conceived with a view to protecting certain specific interests (essentially the interests of people who were physically injured or whose property was damaged by dangerous products) and not to allow compensation for all harm that might be caused by a defective product. Such a rationale, however, is largely alien to French law, which explains why the French legislature ‘forgot’ to introduce the restrictions laid down by the directive on the scope of remediable harm when the directive was transposed.
These examples show that the principle of differentiated protection of interests is both present and poorly accepted in French law. The Avant-projet de réforme du droit des obligations extends this ambivalence, renewing the general principle of the remediable character of any harm while subjecting harm arising from harm to physical integrity to a specific regime in order to facilitate compensation. It seems to me, however, that we should go further and take an extra step, not settling for facilitating the remedy of certain harms but explicitly admitting that other harms cannot be remedied, at least not in all circumstances. Even so there is need to clearly delimit the exceptions to the current principle, which asserts the remediable character of all harms.
First of all, the remediable character of all harm arising from infringement of the physical integrity of people cannot be called into question, at least by case law, whatever the basis for the liability implemented. For specially protected intangible interests, the conditions for remedying infringements are set out by the specific rules about them, which settles the matter. That leaves infringements of the integrity of tangible property and other intangible interests. It is for these harms alone that, in my opinion, the challenge to the principle that any harm is remediable should be limited.
Framed in this way, this challenge may take on two forms. The most extreme is to decide that some harms are never remediable. This is how the legislator’s intervention further to the Perruche decision was interpreted. Already some 10 or so years earlier the Cour de cassation had relied on the absence of any remediable harm to refuse to compensate a mother complaining of having given birth to a (healthy) child after a failed abortion. It is certainly no chance event that, in both cases, the refusal to recognise the remediable character of the harm claimed has arisen in connection with litigation over unwanted births. ‘Remedying’ a birth that is experienced as a harm raises obvious moral and metaphysical questions which it is not certain that law is in a position to answer. I personally believe that it could validly be decided that the very fact that someone has come into this world and exists cannot constitute a remediable harm, either for that person or for others. The reason for this is not necessarily because there is no harm, but because the courts are not and should not be authorised to say whether a life is worth living or not, whether it is a good or an evil for the subject and for others. It is not a matter of expressly denying or asserting the existence of a harm, but of abiding by a form of agnosticism and recognising that the claim for redress of this harm goes beyond the ambit of issues on which liability law is authorised to adjudicate.
Some will judge that this position denies a right of remedy to people who feel real suffering and may be confronted with a tragic situation. Yet the purpose of civil liability is not to ease all suffering nor to remedy all tragedies. Believing that liability law, or any law, can solve all the world’s problems is an almost magical conception of things. In the case of suffering related to birth and to life, it is not a matter of asserting that there is no harm, but of saying humbly that there are claims that liability law cannot answer. It can be argued, true, that liability law also has a symbolic function and that it is important that the court can recognise, in some cases, that a birth or a life may be felt to be a harm. To that it will be answered that it is precisely so as not to blank out this symbolic function of law that it is proposed, in this instance, not to deny the existence of a harm but only to deny that it can be remedied through civil liability. The symbolic recognition of what is subjectively perceived as a harm may thus take place, or at any rate not be ruled out, without the court asserting that in the eyes of the law, it would have been better that some particular human had not come into the world – for the symbolic violence of such an assertion risks being stronger than denying the existence of the plaintiff’s harm.
There is another type of harm for which it might be contemplated declaring, not that it does not exist, but that it is not remediable: it is the non-pecuniary indirect loss claimed by certain relatives of the direct victim. It is a good thing that, for a long time now, and unlike, say, English law, French law makes good the suffering experienced by certain relatives of the immediate victim and notably that of the father or mother. Conversely, the remedy for the harm incurred by more distant relatives, such as uncles and aunts, is far more open to debate. As has already often been said, it is particularly tricky to evaluate and its social usefulness is more than uncertain. Under the circumstances, it would not be shocking, I feel, to decide that only non-pecuniary indirect loss of the closest relatives of the immediate victim can be remedied (father, mother, grandparents, children and possibly siblings).
The categorical negation of the remediable character of certain harms, which is the first way of limiting remediable harms, can only be exceptional, though. The second way, which may relate both to interference with tangible property and intangible interests could also have a broader field of application. It would consist in delimiting, in a differentiated way, the types of remediable harm depending on the act that triggered liability.
In the event of negligence or wrongdoing, when liability is based on article 1382 of the Civil Code, all harm caused directly by the seminal act must be considered remediable, save harm referred to previously which, by nature, should never be remediable. In this way, the general character of the rule laid down by article 1382 would be saved, along with its capacity to cover new situations and to allow new rights or protected interests to emerge. There is no question then of challenging this general – but not absolute! – principle of French law, that is unknown in many foreign legal systems, which is that any tort caused by wrongdoing or negligence may give rise to remedy. It is simply a matter of according its proper place to it. Article 1382 does not assert that any tort must be made good, whatever the basis of liability. Accordingly, it is quite possible, without falling foul of either the letter or even probably the spirit of this founding text, to consider that in the event of liability resting on some foundation other than wrongdoing or negligence, the harm arising from infringement of tangible property or intangible interests cannot be remedied except insofar as its compensation corresponds to the ratio legis of the liability regime invoked.
Most of the strict liability regimes in French law were devised and developed for the more or less openly expressed purpose of remedying the consequences of interference with the physical integrity of people. It is for example obvious that, if case law accepted to discover a general principle of liability for the acts of things in article 1384 (1) of the Civil Code, it was first to remedy the torts arising from bodily accidents caused by industrial machinery and subsequently by motor vehicles. When the courts decided that the non-negligent act of a child, as a direct cause of tort, sufficed for the parents to be held liable, it was not primarily so as to compensate the plaintiff’s pure economic loss but so that his physical injury could be compensated. Consequently, it would seem perfectly acceptable to me that, in the case of application of a strict liability regime, the courts should refuse to compensate some interference with tangible property or intangible interests and especially pure economic harm, on the grounds that making good this type of harm does not come within the ambit of the said regime. It seems to be less a question, then, of defining a priori a restrictive list of interests protected by each strict liability regime than of excluding compensation for certain types of harm when such compensation appears alien to the reason for being of the strict liability regime implemented.
The foregoing lines have no purpose other than to contribute to reflection on the framing of civil liability in French law by thinking about harm and its remediable character. This is merely an outline and I am well aware that the avenues of enquiry proposed may appear iconoclastic or even heretical. The mere suggestion that one should, or perhaps one day might have to limit the scope of remediable harms may seem incongruous in the context of French law. Such an idea might appear to run counter to the concern Professor Viney has always shown for victims. It seems to me, however, that taking to heart the fate of the victims does not necessarily mean remedying all types of harm that a plaintiff may claim for. In any event, and even if she might find the ideas expressed here very different from her own, I am sure that the dedicatee of this book will not hold me too strictly to account for that. She has never criticised her students for thinking differently from her. That in my view is the hallmark of a true maître
My thanks to Guillaume Canselier for his careful and critical reading of the paper, his valuable suggestions and for pointing out the excellent article by Tony Weir, referred to below. I also thank Christine Farenc for her thought-provoking comments.
 In English law, these checks result in particular from the limits on the existence and scope of duties of care in the tort of negligence and from the requirements in characterising causation. For a synthetic and committed presentation of these mechanisms, see T. Weir, Tort Law (Oxford, Oxford University Press, 2002). In German law, it is the restrictive list of interests protected by para. 823, (1) of the BGB and the requirements as to causation that contain tort law. See, for example, F. Ferrand, Droit allemand (Paris, Dalloz, 1997) no 360 ff.
 In French law, three cumulative conditions are normally required for liability to be recognised: harm (dommage/préjudice), a ‘seminal act’ (fait générateur), i.e. an act or fact that caused damage and justifies the imposition of a liability, and a causal connection (causalité) between the two.
 See in this respect the disappearance of the requirement of unlawfulness and even of the abnormal character of the wrongful act for holding parents liable for the acts of their child (Cass. 2e civ., 10 mai 2001, Levert, Bull. civ. II, n° 96, D. 2001, p. 2851, rapp. Guerder, note Tournafond, D. 2002, somm. P. 1315, obs. Mazeaud, JCP éd. G. 2001.II.10613, note Mouly, JCP éd. G. 2002.I.124, n° 20, obs. Viney, Defrénois 2001, p. 1275, obs. Savaux, RTD civ. 2001.601, obs. Jourdain) and the assimilation of any contractual failing to tort (Cass. ass. plén., 6 octobre 2006, Bull. civ. ass. plén., n° 9, D. 2006, p. 2825, note G. Viney, JCP éd. G. 2006.II.10181, avis Gariazzo et note M. Billiau, Resp. civ. et assur. 2006, études 17, L. Bloch, RTD civ. 2007.123, obs. P. Jourdain, on which see J.-S. Borghetti, ‘Breach of contract and liability to third parties in French law: how to break deadlock’, ZEuP 2010, n° 2, p. 279).
 This extension of liability has been marked by the regular emergence in recent decades of new counts for non-pecuniary harm such as aesthetic damage, loss of amenity, sexual harm or the specific harm of contamination, see especially on this G. Viney, P. Jourdain and S. Carval, Les conditions de la responsabilité, 4th edn (Paris, L.G.D.J., 2013) no 265.
 One thinks especially here of the frequent albeit implicit resort to the theory of equivalent conditions for assessing the existence of causation, especially in the event of a chain of causes. See, for example, in the event of contamination related to a transfusion further to a road traffic accident, the case law on remedies sought by blood transfusion centres against the drivers who caused the accidents (and see recently Cass. 2e civ., 25 janvier 2007, Bull. civ. I, n° 20, D. 2007, p. 443, obs. I. Gallmeister, JCP éd. G. 2007 act. 71, obs. C. Bloch, ibid. II.10035, note Ch. Radé, ibid. I.185, n° 5, obs. Ph. Stoffel-Munck, RDC 2007.723, obs. J.-S. Borghetti, Resp. civ. et assur., 2007, comm. 116, note H. Groutel, and the many case law references cited by Ch. Radé, previous note).
 Plaintiffs whom French jurists, quite symptomatically, tend to characterise systematically as victims, regardless of the outcome of the claim in tort. See the relevant remark by S. Whittaker, ‘La responsabilité pour fait personnel dans l’Avant-projet de réforme du droit de la responsabilité : donner voix aux silences du Code civil’, RDC 2007.89, p. 99.
 See on these issues J.-S. Borghetti, ‘The Culture of Tort Law in France’, JETL 2012, no 3, p. 158.
 See G. Viney, Introduction à la responsabilité, 3rd edn, (Paris, L.G.D.J., 2008) no 65, asserting that ‘one must be aware that the amounts that can be devoted at any given time in a given country to compensating victims cannot exceed a certain limit beyond which the system may too seriously impede economic dynamism’.
 See especially L. Engel, ‘Vers une nouvelle approche de la responsabilité. Le droit français face à la dérive américaine’, Esprit, June 1993, p. 5; S. Schiller, ‘Hypothèse de l’américanisation du droit de la responsabilité’, Arch. philo. dr., 2001, p. 177.
 Cass. ass. plén., 17 novembre 2000, Bull. civ. ass. plén., n° 9, D. 2001, p. 332, notes D. Mazeaud and P. Jourdain, JCP éd. G. 2001.II.10438, concl. Sainte-Rose and note F. Chabas; a good many of the countless commentaries prompted by the decision are identified by Viney and Jourdain, Les conditions de la responsabilité, n° 269-6, note 41.
 Article 1 I(1) of statute no. 2002-303 of 4 March 2002 provides that: ‘No one may make a claim for harm on the basis of their birth alone’. This statute is now codified in article L 114-5 of the Code of Social Action and Families. The passing of this provision has not, however, neutralised all the negative consequences the Perruche decision had on the practice of fœtal ultrasound scanning in France: see D. Moyse and N. Diederich, L’impact de l’« arrêt Perruche » sur les échographistes et les gynécologues obstétriciens, rapport à la Mission de recherche Droit et Justice, January 2005.
 See especially Cass. ass. plén., 9 mai 1984, Fullenwarth, Bull. ass. plén., n° 4, JCP éd. G. 1984.II.20255, note Dejean de La Bâtie, D. 1984, jur. p. 525, concl. Cabannes, note F. Chabas, Defrénois 1984, p. 557, note Legeais, RTD civ. 1984.508, obs. J. Huet.
 See A. Guégan-Lécuyer, Dommages de masse et responsabilité civile, préf. P. Jourdain, avant-propos G. Viney (Paris, L.G.D.J., 2006).
 See especially Viney, Jourdain and Carval, Les conditions de la responsabilité, no 269-5 and the references cited.
 See especially L. Cadiet, ‘Les métamorphoses du préjudice’, in Les métamorphoses de la responsabilité (Paris, P.U.F., 1998) p. 37.
 See especially R. Rodière, note sous Cass. 1re civ. 21 octobre 1952, JCP éd. G. 1953.II.7592; F.-P. Bénoit, ‘Essai sur les conditions de la responsabilité en droit public et privé (problèmes de causalité et d’imputabilité)’, JCP éd. G. 1957.I.1351; L. Cadiet, Le préjudice d’agrément, PhD thesis Poitiers, 1983, esp. no 323 ff.; S. Rouxel, Recherches sur la distinction du dommage et du préjudice, PhD thesis Grenoble, 1994; I. Poirot-Mazères, ‘La notion de préjudice en droit adminstratif français’, RDP 1997, p. 519; Y. Lambert-Faivre, Droit du dommage corporel, 4th edn (Paris, Dalloz, 2000) no 86; C. Cormier, Le préjudice en droit administratif français, préf. D. Truchet (Paris, L.G.D.J., 2002) and the many references cited, esp. p. 50 ff.; Guégan-Lécuyer, Dommages de masse et responsabilité civile., no 60-62; Ph. le Tourneau, Droit de la responsabilité et des contrats, 10th edn (Paris, Dalloz, 2014) no 1302 ff.
 For a discussion of this hypothesis, see especially Cadiet, Le préjudice d’agrément, no 326.
 See, for example, X. Pradel, Le préjudice dans le droit civil de la responsabilité, avant-propos J.-L. Baudouin and P. Deslauriers, préf., P. Jourdain (Paris, L.G.D.J., 2004); Viney, Jourdain and Carval, Les conditions de la responsabilité, no 246-1. Cf. the more nuanced position of S. Prochy-Simon, V° Dommage, in D. Alland and S. Rials (eds), Dictionnaire de la culture juridique (Paris, Lamy-P.U.F., 2003).
 It should be specified that the analysis of the process by which harm is caused and the determination of the conditions for bringing an action in liability are two separate issues. Refining the analysis of the process which leads from the occurrence of a seminal act to the finding of a reparable loss or damage does not necessarily mean imposing new conditions for the action for redress to meet with success. Thus, differentiating between tort and harm does not imply that the plaintiff is required to prove both. It allows us to realise that, in the event of a challenge, the proof demanded of the plaintiff bears, as the case may be, on the tort or on the harm. It must further be pointed out that the analysis of the process may be further refined beyond the fourfold distinction of seminal act, causation, tort, harm; see, for example, the analyses proposed by Bénoit, ‘Essai sur les conditions de la responsabilité en droit public et privé’ and A.-S. Dupont, Le dommage écologique (Zurich, Schulthess, 2005), no 16. The problem is obviously to strike a balance between the precision of the analysis and its manageable and operative character.
 See, for example, Dupont, Le dommage écologique, no 18.
 See, for example, on Italian law, E. Guerinoni, ‘La nozione general di danno’, in P. Cendon (ed.), I Danni riscarcibili nella responsibilità civile. I - Il danno generale (Turin, UTET, 2005), p. 53, no 15; on English and German law see T. Weir, ‘La notion de dommage en responsabilité civile’, in P. Legrand Jr (ed.), Common Law d’un siècle l’autre (Montreal, Les éditions Yvon Blais, Inc., 1992), p. 1 and pp. 11 ff.
 See especially articles 3, 5, 28 and 29 of the statute of 5 July 1985. Article 1386-2 of the Civil Code, inspired by the directive of 25 July 1985 on liability for defective products provides for reparation of ‘injury arising from harm to the person’ and of injury ‘resulting from harm to property other than the defective product itself’. See also article 2226 of the Civil Code, as modified by the statute of 17 June 2008, on the time limits applicable to claims for compensation for harm resulting from bodily injuries.
 The distinction can also be found in the two academic projects for reforming the law of obligations, the Avant-projet de réforme du droit des obligations (Projet Catala) and the Projet de réforme de l’Académie des sciences morales et politiques (Projet Terré). While it does not formally endorse the distinction between the two concepts, the Projet Catala implicitly gives two separate meanings to the terms ‘dommage’ (tort) and ‘préjudice’ (harm) and distinguishes between harm arising from interference with physical integrity (articles 1379 ff) and harm arising from interference with property (articles 1380 ff). The Projet Terré is much more explicit and clearly distinguishes between ‘dommage’ and ‘préjudice’: see Ph. Remy and J.-S. Borghetti, ‘Présentation du projet de réforme de la responsabilité délictuelle’, in F. Terré (ed.), Pour une réforme du droit de la responsabilité civile (Paris, Dalloz, 2011), p. 62, no. 16.
 See Cass. 1re civ., 28 octobre 2003, Defrénois 2004, 37894, p. 383, note R. Libchaber, JCP éd. G. 2004.II.10006, note G. Lardeux, JDI 2004, p. 499, obs. G. Légier, cited by C. Bloch, La cessation de l’illicite, PhD thesis Aix-Marseille 2006, no 120.
 See especially Cormier, Le préjudice en droit administratif français.
 See in doctrine, at any rate, R. Merle and A. Vitu, Traité de droit criminel, 6th edn (Paris, Cujas, 1984) no 488.
 The binding character of the wrongdoing or negligence of the direct victim on the indirect victim, in particular, can be quite well accounted for if it is considered that the wrongdoing or negligence is a cause of the tort incurred by the direct victim, from which arises the harm for this victim and the harm for the indirect victim. The indirect victim complaining of harm that is a consequence of the tort incurred by the direct victim, it is logical that the remedy of this harm should be affected by the causes that lead to the tort; see especially C. Lapoyade Deschamps, ‘Quelle(s) réparation(s) ?’, Resp. civ. et assur. special issue, June 2001, p. 62, no 23; Rouxel, Recherches sur la distinction du dommage et du préjudice, p. 276. This thesis also contains a detailed analysis of most of the practical consequences that might be drawn from the distinction between tort and harm.
 No claim is made, however, that this distinction is a panacea. Besides the point, often denounced, that it introduces further complexity into the subject of liability, there are hypotheses in which it lacks consistency. While it seems most illuminating when the tort is physical damage to a bodily entity (person or thing) or interference with a clearly defined right, it seems equally, in the event of interference with a state of affairs or an activity, that it can be far more difficult to clearly distinguish between this interference and the harm that may result from it; see O. Berg, La protection des intérêts incorporels en droit de la réparation des dommages, préf. G. Viney, avant-propos R. Frank (Brussels and Paris, Bruylant-L.G.D.J., 2006) no 56.
 The most convincing solution to the problem of ‘bodily harm’ and resort to third-party payers seems to be that which proposes to distinguish clearly between bodily injury and the resulting pecuniary and non-pecuniary loss or damage; this approach is recommended notably by Loïc Cadet in his PhD thesis and is now largely taken up in doctrine. It has been endorsed by the Nomenclature Dintilhac, a classification of counts of harm (‘chefs de préjudice’) potentially resulting from a bodily injury, which is theoretically non-binding and yet widely applied by the courts; see Ph. Brun, Responsabilité civile extracontractuelle, 3rd edn, (Paris, Litec, 2014) no 621 ff.
 A hypothesis must be pointed out here, which is to be best of my knowledge the only one in which the distinction could have had the effect of radically impeding the success of an action in liability. It is the particularly sensitive case in which a disabled child seeks redress for the harm caused by his or her congenital disability. It has been argued that this disability cannot be a tort since it does not constitute an impairment of a pre-existing condition. Now, in the absence of any tort, there cannot strictly be any harm (see especially M. Deguergue, ‘Les préjudices liés à la naissance’, Resp. civ. et assur. special issue, May 1998, p. 14). This reasoning appears logically flawless. However, it is voided by the legislative provision passed in 2002 which provides that no one can make an argument for harm from the mere fact of having been born (see note 9 above). Moreover, regardless of this text, having regard to the tragic situation, it may seem extremely stringent to refuse any compensation to a child born with a disability on the basis of a narrowly technical argument (see Viney, Jourdain and Carval, Les conditions de la responsabilité, no 249-6). It seems to me therefore that if some justification must be found for refusing to compensate the tort claimed by someone born with a disability, it is on the basis of more general considerations relating to the concept of remediable harm, which shall be considered a little later on.
 This combination is used here insofar as case law does not in principle distinguish between the two terms and holds them to be synonyms.
 This hesitation as to the way harm is apprehended is not, it seems, peculiar to France; see various approaches suggested in Quebec to determine non-pecuniary harm, P. Deslauriers ‘Les pertes non pécuniairies : compte-rendu/constat/critiques’, Revue juridique Thémis, 2005, p. 371, esp. p. 376.
 Cass. 2e civ., 22 février 1995, Bull. civ. n° 61 (2 arrêts), D. 1996, jur. p. 69, note Y. Chartier et somm. p. 233, obs. D. Mazeaud, RTD civ. 1995.629, obs. P. Jourdain, stating that ‘the vegetative condition of a human being not excluding any count for compensation, his or her harm must be made good in all its constituent parts’.
 See especially Cass. 1re civ., 5 novembre 1996, Bull. civ. I, n° 378, JCP éd. G. 1997II.22805, note J. Ravanas, ibid. I.4025, n° 1, obs. G. Viney, RTD civ. 1997.632, obs. J. Hauser, asserting that ‘under article 9 of the Civil Code, the mere finding of an infringement of privacy opens up a right to redress’. (This case law may be likened to the decisions in which the Cour de cassation asserts that ‘an employee’s observance of an unlawful non-competition clause necessarily causes him or her harm the scope of which it is for the courts to determine’; on this solution, which falls within the realm of contractual liability, see especially A. Pinna, ‘La mesure du préjudice consécutif à l’exécution d’une clause de non-concurrence illicite’, JCP éd. G. 2006.I.192.) It could, admittedly, be argued that the violation of a right in itself constitutes harm. However, it is an abstract, not to say virtual harm. In any event, this explanation is hardly satisfactory, because it does not say why the solution applies to certain very specific rights only. Thus, case law has never asserted, as far as I know, that the violation of a right of ownership necessarily entailed harm for the holder of that right.
 Distinguishing between tort and harm might also clarify the question of ecological tort/harm. In the event of interference with the natural environment, there is a tort but not necessarily any harm, in the sense that there is not always a person or entity recognised by law who can sue for the actual negative consequences this tort has caused them. Directive 2004/35/EC of 21 April 2004 on environmental liability with regard to the prevention and remedying of environmental damage adopts an interesting approach because it aims primarily to prevent and repair ‘in kind’ damage (defined as a measurable negative modification of a natural resource or a measurable deterioration in a service related to natural resources, which may arise directly or indirectly). Although it does not take up the conceptual distinction between tort and harm, the directive has a structure that is compatible with this distinction and, above all, it avoids the problem of determining the persons or entities who can avail themselves of ecological harm and concentrates instead, ahead of that, on the issue of tort (dommage, there damage).
 To avoid this outcome, it is not a question of forcing the plaintiff to prove both the tort and the harm. However, one might contemplate having a simple inference by which the existence of a tort means it is presumed there is a harm corresponding to the usual harm in similar instances, it being for the defendant to overturn this inference by proving that in the case in point, the plaintiff has incurred a lesser harm, if any. In the hypothesis by which there is an infringement of privacy, however, the Cour de cassation applies in the current state of affairs a form of irrebuttable presumption of the existence of harm.
 On the principle of full compensation, see especially G. Viney and P. Jourdain, Les effets de la responsabilité, 3rd edn (Paris, L.G.D.J., 2010) no 57-60.
 Besides the reservations formulated by Viney and Jourdain, Les effets de la responsabilité, no 58-1, see for example the remarks by C. Grare, Recherches sur la cohérence de la responsabilité délictuelle, préf. Y. Lequette (Paris, Dalloz, 2005) no 271 ff; A. Tunc, La responsabilité civile, 2nd edn (Paris, Economica, 1989) p. 176.
 See le Tourneau, Droit de la responsabilité et des contrats, no 1309.
 Cass. 2e civ., 28 octobre 1954, Bull. civ. II, n° 328, JCP éd. G. 1955.II.8765, note R. Savatier.
 See especially P. Jourdain, ‘Le préjudice et la jurisprudence’, Resp. civ. et assur. special issue, June 2001, p. 45 no 12; Viney and Jourdain, Les conditions de la responsabilité, no 249-6.
 See, for example, for Germany D. Medicus, Schuldrecht I. Allgemeiner Teil, 16th edn, (Munich, Verlag C. H. Beck, 2005) no 595; for Italy Guerinoni, ‘La nozione general di danno’, no 15-1; for Switzerland, Dupont, Le dommage écologique, no 423 ff.
 At least in tort law, because in contract, article 1149 of the Civil Code could be tied in with this theory.
 See, for example, Ph. Brun, Responsabilité civile extracontractuelle no 185; J. Carbonnier, Droit civil, Volume 2 (Paris, P.U.F., 2004) no 1121; Viney, Jourdain and Carval, Les conditions de la responsabilité, no 248. This definition is sometimes criticised because of its imprecision; see, for example, L. Cadiet, Le préjudice d’agrément, esp. no 283.
 See, for example, Ph. Brun, ‘Personnes et préjudice’, Rev. gén. de droit, 2003, p. 187 and p. 199. As T. Weir writes, ‘La notion de dommage en responsabilité civile’, préc., p. 35, ‘it is for the legal system to decide whether a given effect [of the proximate cause] is good or bad’.
 There is no cause to contemplate here the three conditions usually posited by case law for the harm to be remedied, namely its direct, certain and personal character, since these features characterise the harm without defining it intrinsically.
 This may be why it was not expressly taken up in the Avant-projet de réforme du droit des obligations, which did, though, adopt a very similar definition of harm, since its article 1343 says ‘an ascertained harm consisting in the interference with a lawful, pecuniary or non-pecuniary, individual or collective interest is capable of remedy’. In substance, for the Avant-projet, harm is therefore defined as interference with a lawful interest.
 On the case-law saga of compensation of common-law wives, see J. Dupichot, Des préjudices réfléchis nés de l’atteinte à la vie ou à l’intégrité corporelle (Paris, LGDJ, 1969).
 See especially Brun, ‘Personnes et préjudice’, p. 187 and p. 199; B. Starck, H. Roland and L. Boyer, Obligations. 1. Responsabilité délictuelle, 5th edn (Paris, Litec, 1996), no 100; Weir, ‘La notion de dommage en responsabilité civile’, p. 10.
 See especially F. Ost, Droit et intérêt, Vol. 2 – Entre droit et non-droit : l’intérêt, (Brussels, Publication des Facultés universitaires Saint-Louis, 1990), p. 51-59. Cf. Starck, Roland and Boyer, Obligations. 1. Responsabilité délictuelle, no 101.
 See especially L. Cadiet, Le préjudice d’agrément, esp. no 340.
 A distinction which implies not confusing moral responsibility and legal liability.
 This difficulty in grasping, in law, what is legitimate occurs for all ideas referring to legitimacy; see, for example, the notion of legitimate expectation, J.-S. Borghetti, La responsabilité du fait des produits, préf. G. Viney (Paris, L.G.D.J., 2004) no 439 ff.
 See especially Cass. 2e civ., 24 janvier 2002, D. 2002, p. 2559, note D. Mazeaud, JCP éd. G. 2002.II.10118, note C. Boillot, ibid. 2003.I.152, n° 22, obs. G. Viney, Defrénois 2002, p. 786, obs. R. Libchaber, RTD civ. 2002.306, obs. P. Jourdain, Dr. et patr. avril 2002, p. 92, obs. F. Chabas, Resp. civ. et assur. 2002, chron. 11, note S. Hocquet-Berg and Cass. 2e civ., 22 février 2007, Bull. civ. II, n° 49, JCP éd. G. 2007.I.185, n° 1, obs. Ph. Stoffel-Munck, Resp. civ. et assur. 2007, comm. 146, obs. S. Hocquet-Berg, asserting that ‘a victim cannot obtain remedy for the loss of remuneration unless that remuneration was lawful’.
 See le Tourneau, Droit de la responsabilité et des contrats, no 1392; obs. Stoffel-Munck, Resp. civ. et assur. 2007.
 See Brun, ‘Personnes et préjudice’, p. 199.
 See the celebrated decision Cass. crim., 20 février 1863, S. 1863, 1, 321, rapport Nouguier, asserting that ‘article 1382, in ordering in absolute terms the remedy for anyone’s act that causes harm to another, does not limit in any way the nature of the harmful deed, nor the nature of the harm felt’.
 ‘[Article 1382 of the Civil Code] encompasses in its breadth of scope all kinds of torts and subjects them to uniform remedy which takes as its measure the value of the harm suffered. From homicide to the slightest injury, from the burning of a building to the collapse of frail property, everything is subjected to the same statute; everything is declared capable of evaluation which will compensate the injured party for whatever torts they have incurred’ (Discours du corps législatif, cited notably by G. Viney, obs. sous Cass. 2e civ., 27 mai 1999, JCP éd. G. 2000.I.197, n° 6).
 Do we not speak today of ‘harm’ related to the disappointments of certain participants in the game Second Life!?
 See M. Deguergue, ‘Les préjudices liés à la naissance’, p. 15.
 See especially Cass. ass. plén., 17 novembre 2000, préc. note 8.
 See especially TGI Lille, 6 mai 1996, D. 1997, jur. p. 543, note X. Labbée, RTD civ. 1999.64, obs. J. Hauser; Cass. crim., 4 février 1998, JCP éd. G. 1999.II.10178, note I. Moine-Dupuis; CA Grenoble, 29 June 2005, Resp. civ. et assur. 2006, comm. 48, note Ch. Radé.
 See especially Cass. 1re civ., 16 juillet 1991, Bull. civ. I, n° 248, JCP éd. G. 1992.II.21947, note A. Dorsner-Dolivet and the decisions cited by Viney, Jourdain and Carval, Les conditions de la responsabilité, no 249-5 (birth of a disabled child); TGI Reims, 19 juillet 2005, JAC n° 59, décembre 2005, note I. Corpart (harm caused by the birth of a disabled sister).
 The same is probably true today in most non-Western countries. Regardless of considerations related to the different forms that civil liability may take on in legal systems very different from our own, can one imagine a family in Africa or the Middle East going to court over the birth of a child, even if disabled?
 But see O. Berg, La protection des intérêts incorporels en droit de la réparation des dommages, no 60-61.
 TGI Reims, 19 juillet 2005, JAC n° 59, décembre 2005, note I. Corpart.
 See, for example, M. Deguergue, ‘Les préjudices liés à la naissance’, p. 19.
 It seems a trial of the sort has been held in the United States (see A. Tunc, La responsabilité civile, no 1, note 4. This remarkable book contains other interesting examples of harms that are a priori incongruous but for which US plaintiffs have sought remedy; see esp. no 1, notes 2 to 9 and p. 171, note 3). It should be pointed out in passing that the very fact of suing one’s parents for bringing one up badly would seem to be strong evidence of the wrongdoing or negligence complained of.
 I confess to being surprised that amorous relationships, which are sometimes the occasion for evident wrongdoing and intense suffering, do not give rise to more tort cases, apart from divorce suits or suits for abusively ending engagements. From a strictly technical standpoint, it is hard to see why a broken heart should not constitute a harm to be remedied. It is probably that we are still reluctant to bring affairs of the heart within the legal realm, which we should certainly be glad about; imagine if Hermione had sued Pyrrhus for tort instead of having him killed, then instead of writing Andromaque Racine would have had to settle for a new version of Les Plaideurs (The Litigants)…
 See, for example, Cass. 2e civ., 24 février 2005, Bull. civ. II, n° 53 and the observations of G. Viney, JCP éd. G., 2005.I.149, n° 1.
 The question of making good the consequences of these infringements for third parties (indirect harm) will be considered later.
 Cf. the speech by Tarrible (note 56 above).
 See, for example, Weir, ‘La notion de dommage en responsabilité civile’, p. 44. The particular importance and the priority character of remedy for infringement of physical integrity are also underscored by the existence of a resolution (75) 7 of 14 March 1975 of the Council of Europe ‘On Compensation for Physical Injury or Death’ (reproduced in Lambert-Faivre, Droit du dommage corporel, p. 68).
 French law is more reluctant, though, to accept non-pecuniary loss (préjudice moral) arising from the loss of a thing, even if it has accepted it in the case of loss of an animal; see esp. Cass. 1re civ., 16 janvier 1962, D. 1962, jur. p. 199, note R. Rodière, JCP 1962.II.12557, note P. Esmein, RTD civ. 1962, p. 316, obs. A. Tunc.
 See especially the provisions on defamation in the 1881 statute on the freedom of the press which, in circumscribing freedom of speech, delimits by the same token the protection granted to reputation; article 9 of the Civil Code on privacy. In many foreign legal systems, and especially in English law, the particular importance accorded to a specific intangible interest is generally reflected by the existence of a specific tort to remedy any infringement of that interest.
 It is fairly significant that freedom (of movement), which features in the list of interests expressly protected by paragraph 823(1) of the BGB and by a specific tort in English law (false imprisonment) should not be the subject of particular attention in France in the area of civil liability.
 One might be tempted to construe the existence of these specific rules as the sign that the ordinary law of liability does not readily lend itself, whatever one may think, to the protection of purely intangible interests.
 See Viney, Jourdain and Carval, Les conditions de la responsabilité, n° 312 and the references cited. The refusal to make good this type of economic loss is generally justified by invoking the absence of any causal connection, but given the flexibility with which causation is often appraised in French law, this refusal to admit the existence of a causal connection in these instances is motivated less by technical reasons than it is the reflection of the courts’ fear of giving unbounded scope to liability.
 For criticism of the assimilation of lost expectation of gain to harm, see also Weir, ‘La notion de dommage en responsabilité civile’, p. 47.
 But see the celebrated argument by B. Starck, Essai d’une théorie générale de la responsabilité civile considérée en sa double fonction de garantie et de peine privée (Paris, Rodstein, 1947), which proposes separate treatment for bodily and material damage one the one hand and non-pecuniary damage and economic loss on the other.
 Obs. sous Cass. 2e civ., 27 mai 1999, loc. cit. See also A. Tunc, La responsabilité civile, p. 176-179; Viney, Introduction à la responsabilité, no 65.
 Among which can be cited the (largely occult) applications of the limitation on the scope of actions in tort. This principle is that in the event of violation of a rule, only those types of harm the rule was designed to prevent should be remediable. Officially, the limitation on actions in tort has no place in French law, but it is sometimes implemented by the courts and tribunals in the event of breach of a rule whose purpose is clearly delimited; see the examples cited by Viney, Jourdain and Carval, Les conditions de la responsabilité, no 441.
 Cass. 1re civ., 25 juin 1991, D. 1991, jur. p. 567, note Ph le Tourneau, JCP éd. G. 1992.II.21784, note J.-F. Barbiéri, RTD civ. 1991.791, obs. P. Jourdain, asserting ‘the existence of the child she conceived cannot alone constitute a legally remediable harm for the mother, even if the birth occurred after an unsuccessful operation to terminate the pregnancy’. See also recently Cass. 2e civ., 12 juillet 2007, n° 06-16.869, ruling in a case in which a man sought to enforce a guarantee for amounts incumbent on him for the maintenance of his child against the mother of the child, arguing that she had wrongfully accepted to have sexual intercourse with him without taking measures to avoid or counter the risk of conception; the supreme court ruled that the appeal court had not violated the law by deciding that, ‘having freely and fully consented to have unprotected sexual intercourse with Mrs Y … at their first time of meeting, Mr X…, a sexually experienced man, on whom it was incumbent, as much as on his partner, to take measures to avoid procreation, failed to establish either the negligence or wrongdoing of the mother of the child for having engaged in such intercourse or for having subsequently acted to secure affiliation and contribution for the maintenance of the child, nor the existence of any remediable direct or indirect harm’ (emphasis added).
 Article 9 of the Directive.
 Borghetti, La responsabilité du fait des produits, n° 303-307.
 This creates a mismatch between French law (article 1386-2 of the Civil Code) and the directive. The ECJ ruled that there was no violation of EU law on this count, however: ECJ,. 4 June 2009, Moteurs Leroy Somer v. Dalkia France, C-285/08.
 Article 1351.
 The Projet Terré explicitly takes this course: F. Terré (ed.), Pour une réforme du droit de la responsabilité civile.
 It is important to clarify that this principle does not have, it seems to me, any constitutional value in that it is not expressly set out in any text of this standing and that it cannot be considered as being implicitly enshrined in positive law, given the many statutory and case law exceptions to it, some of which have been referred to above. See Viney and Jourdain, Les effets de la responsabilité, no 59-2 ff.
 Out of concern to frame the remedying of harm to intangible interests, O. Berg proposes accepting such remedy only when the victim can show a ‘legitimate expectation’ to obtain redress (O. Berg, La protection des intérêts incorporels en droit de la réparation des dommages, n° 316 ff; see also G. Viney, obs. sous Cass. 2e civ., 27 mai 199, loc. cit.). While I share the concern to limit compensation to intangible interests, I am more sceptical about the operative character of the criterion proposed.
 See note 9 above. The law now provides that ‘No one may make a claim for harm on the basis of their birth alone’. However, this can be construed in at least two ways: either as denying any harm relating to birth, or as simply excluding remedy for the harm without necessarily denying its existence. The first corresponds more closely to the letter of the law, but the second corresponds to what to me seems a preferable theoretical approach.
 Cf. the more radical position expressed by Weir, ‘La notion de dommage en responsabilité civile’, esp. p. 17, considering that life in itself is neither good nor bad and that, that being so, neither birth, nor the loss of expectation of life, nor the loss of awareness of being alive (coma) are harms.
 Many commentators have come out in favour of excluding or at least circumscribing remedy for indirect non-pecuniary loss; see, for example, le Tourneau, Droit de la responsabilité et des contrats, n° 1562; Viney, Jourdain and Carval, Les conditions de la responsabilité, n° 267.
 It should be made clear that the refusal to consider certain harms as remediable in civil liability law does not amount to calling into question the protection that the law can and should afford to the interests the interference with which has constituted those harms. In particular, in a society as wealthy as ours, it is unconscionable that law should not provide for mechanisms for helping the disabled and their close relatives.
 It would be best, though, if the wrongdoing and the causation were actually established, which is far from being always the case.
 This proposition is obviously not unrelated to the theory of limitation on the scope for bringing actions in tort (see note 80 above). It seems best to me to apply this theory openly when the action in tort is based on the violation of a clearly finalised textual rule.
 One might contemplate applying this restrictive conception of remediable harm to the hypothesis of the tortious liability of a contract debtor to third parties for breach of contract (see Cass. ass. plén., 6 octobre 2006, note 2 above). Although case law now seems to liken any contractual failing to tort, this seems abusive (cf. Ph. Stoffel-Munck, ‘La relativité de la faute contractuelle’, RDC 2007.587). It seems rather that we have in this a new and independent proximate cause of liability and so in some sense a special civil tort, which would justify not remedying certain harm on this basis – provided that the ratio legis of the new regime were determined, which is not without difficulties of its own. On the issue of liability in tort to third parties for breach of contract in French law, see more generally Borghetti, ‘Breach of contract and liability to third parties in French law: how to break deadlock’.