Accueil
https://fr.wikipedia.org/wiki/Harm_principle
https://plato.stanford.edu/entries/perfectionism-moral/#HarPri
The Harm Principle
Those who reject the principle of state neutrality entrust the state with the task of promoting the good. This can give rise to the worry that a perfectionist account of politics is insufficiently sensitive to the harm of coercion and to the value of liberty. If the state need not be neutral among rival understandings of the good, then is the door not open for the coercive imposition of state policies designed to promote the good? In fact, many versions of perfectionism drawn from the history of political thought have paid little heed to the value of individual liberty. It is an important matter, then, to what extent perfectionist politics can be reconciled with a proper regard for individual liberty.
To approach this issue, it will be helpful to consider the so called harm principle. The harm principle, as articulated by writers in the liberal political tradition from Mill to Feinberg, is often taken to be an essential safeguard for individual freedom in political society. At least at first pass, the harm principle substantially restricts the power of governments to promote the good. It holds that governments cannot coercively interfere with persons unless doing so is necessary to prevent them from causing clear and direct harm to others. The harm principle requires interpretation and can be understood in different ways. But for present purposes our question is whether the harm principle is best understood to be an anti-perfectionist principle—a principle that provides reasons for rejecting or limiting perfectionist politics.
The first thing to say is that not all perfectionist policies are coercive. Governments can and do promote the good noncoercively. A government may promote the good, for example, when it intelligently subsidizes art. This perfectionist policy need involve no coercive interference whatsoever.[14] So the harm principle, even if sound, would not bar all perfectionist policies. It would rule out only coercive governmental policies designed to favor some options and discourage others. The question then is whether this restriction is itself anti-perfectionist.
Here we need to proceed with caution. The extent to which perfectionism licenses the coercive promotion of the good depends, among other things, on the degree to which autonomy or liberty is itself recognized to be a perfectionist good. On some versions of perfectionism the harm principle would be an anti-perfectionist principle, while on others it would not be. Consider, in this regard, Mill’s own defense of the harm principle. For Mill “individuality” is an essential component of a good human life. Mill’s notion of individuality can be understood to be a conception of autonomy. A person is autonomous for Mill if he leads his life on his own terms and develops his capacities and faculties according to “his own mode of laying out his existence” (Mill 1859, 64). The important point for present purposes is not Mill’s particular characterization of autonomy, but rather the structure of his view. Autonomy is understood to be an essential aspect of a good human life, not a separate norm. And the value of autonomy explains, at least in part, why Mill recommends the harm principle.
The Millian defense of the harm principle sits well with the perfectionist focus on good human lives. (Brink 2013) Its availability nicely illustrates how perfectionist politics can be consistent with a strong rejection of state coercion. But it is natural to suspect that Mill overstates his case. Even granting that “individuality” is an aspect of a good human life, we should wonder why it takes priority over all other aspects. If a governmental policy, say a policy that criminalizes the sale and use of dangerous recreational drugs, would prevent many from ruining their lives while infringing the individuality of only a few, then, contrary to Mill, the government may do better in discharging its duty to promote good human lives by adopting the policy than by not adopting it.
This point can be pushed further. Autonomy, it can be argued, requires that one have access to an adequate range of valuable or worthwhile options (Raz 1986). This adequacy requirement does not imply that every time an option is closed off one’s autonomy will be set back. Moreover, what may be of value is not autonomous agency per se, but valuable autonomous agency. Joseph Raz explains: “Since our concern for autonomy is a concern to enable people to have a good life it furnishes us with reason to secure that autonomy which could be valuable. Providing, preserving or protecting bad options does not enable one to enjoy valuable autonomy” (Raz 1986, 412). If valuable autonomy, and not autonomy per se, is what has perfectionist value, then when governments eliminate, or make it more costly for persons to pursue, worthless options, then they may do no perfectionist harm and much perfectionist good.
On Raz’s view, the harm principle is superceded by an autonomy principle that captures the truth in it while avoiding its exaggerations. The autonomy principle holds that the state has negative duties to respect the autonomy of its citizens as well as positive duties to promote and sustain social conditions that contribute to its realization. The harm principle, to the extent that it is sound, is vindicated because it follows, given certain contingent facts, from the autonomy principle. This leads Raz to reformulate the harm principle as a principle “that regards the prevention of harm to anyone (himself included) as the only justifiable ground for coercive interference with a person” (Raz 1986, 412–13). So construed, the harm principle would permit the coercive enforcement of at least some self-regarding duties.
Both Mill and Raz accept versions of the harm principle. But they accept it not as a limit on perfectionist politics, but rather as a principle that guides the proper promotion of the good. Their political theories are examples of perfectionist liberalism and their discussions of the harm principle show how perfectionist politics can be supportive of individual liberty. It can be objected, however, that the defense of individual liberty provided by perfectionist liberalism is insufficient. Recall that on the perfectionist view discussed here autonomy is an aspect of a good human life. It is not a separate norm. A strong and uncompromising defense of the harm principle, it may be thought, must be based on a different understanding of the value of autonomy, one that holds that the autonomy of persons cannot be infringed even when doing so is, all things considered, in their best interests. Autonomy, on this view, is a sovereign right, not an ideal to be promoted (Feinberg 1989)."
"Le principe de non-nuisance.
Ceux qui rejettent le principe de la neutralité de l'État lui confient la tâche de promouvoir le bien. Cela peut susciter l'inquiétude qu'un rapport perfectionniste à la politique ne soit pas suffisamment sensible au danger de la coercition et à la valeur de la liberté. Si l'État n'a pas besoin d'être neutre parmi les conceptions rivales du bien, alors la porte n'est-elle pas ouverte à l'imposition coercitive de politiques d'État destinées à promouvoir le bien ? En fait, de nombreuses versions du perfectionnisme tirées de l'histoire de la pensée politique n'ont guère tenu compte de la valeur de la liberté individuelle. Il est donc important de savoir dans quelle mesure la politique perfectionniste peut être conciliée avec un respect adéquat de la liberté individuelle.
Pour aborder cette question, il sera utile d'examiner le principe dit de non-nuisance. Le principe de non-nuisance, tel qu'il est formulé par les écrivains de la tradition politique libérale, de Mill à Feinberg, est souvent considéré comme une sauvegarde essentielle de la liberté individuelle dans la société politique. Au moins au premier abord, le principe de non-nuisance restreint considérablement le pouvoir des gouvernements de promouvoir le bien. Selon ce principe, les gouvernements ne peuvent pas intervenir de manière coercitive à l'encontre des personnes, sauf si cela est nécessaire pour les empêcher de causer un dommage clair et direct à autrui. Le principe de non-nuisance nécessite une interprétation et peut être compris de différentes manières. Mais pour les besoins de la présente étude, notre question est de savoir si ce principe est mieux compris comme un principe anti-perfectionniste -un principe qui fournit des raisons pour rejeter ou limiter la politique perfectionniste.
La première chose à dire est que toutes les politiques perfectionnistes ne sont pas coercitives. Les gouvernements peuvent promouvoir le bien de manière non coercitive, et ils le font. Un gouvernement peut promouvoir le bien, par exemple, lorsqu'il subventionne intelligemment l'art. Cette politique perfectionniste ne doit impliquer aucune ingérence coercitive quelle qu'elle soit. Ainsi, le principe de non-nuisance, même s'il est sain, n'empêcherait pas toutes les politiques perfectionnistes. Il n'écarterait que les politiques gouvernementales coercitives destinées à favoriser certaines options et à en décourager d'autres. La question est donc de savoir si cette restriction est elle-même anti-perfectionniste.
Nous devons ici procéder avec prudence. La mesure dans laquelle le perfectionnisme autorise la promotion coercitive du bien dépend, entre autres, de la mesure dans laquelle l'autonomie ou la liberté est elle-même reconnue comme un bien perfectionniste. Dans certaines versions du perfectionnisme, le principe de non-nuisance serait un principe anti-perfectionniste, tandis que dans d'autres, il ne le serait pas. Considérez, à cet égard, la propre défense de Mill du principe. Pour Mill, "l'individualité" est une composante essentielle d'une bonne vie humaine. La notion d'individualité de Mill peut être comprise comme une conception de l'autonomie. Pour Mill, une personne est autonome si elle mène sa vie selon ses propres principes et développe ses capacités et ses facultés selon "son propre mode d'organisation de l'existence" (Mill 1859, 64). Le point important pour les besoins de la présente étude n'est pas la caractérisation particulière de l'autonomie de Mill, mais plutôt la structure de son point de vue. L'autonomie est comprise comme un aspect essentiel d'une bonne vie humaine, et non comme une norme distincte. Et la valeur de l'autonomie explique, au moins en partie, pourquoi Mill recommande le principe de non-nuisance.
La défense par Mill du principe de non-nuisance s'accorde bien avec l'accent perfectionniste mis sur les bonnes vies humaines. (Brink 2013) Sa disponibilité illustre bien comment une politique perfectionniste peut être cohérente avec un fort rejet de la coercition étatique. Mais il est naturel de soupçonner que Mill exagère son cas. Même en admettant que l'"individualité" est un aspect d'une bonne vie humaine, on peut se demander pourquoi elle a la priorité sur tous les autres aspects. Si une politique gouvernementale, disons une politique qui criminalise la vente et l'utilisation de drogues récréatives dangereuses, empêchait beaucoup de gens de ruiner leur vie tout en ne portant atteinte à l'individualité que de quelques-uns, alors, contrairement à Mill, le gouvernement pourrait mieux s'acquitter de son devoir de promouvoir une bonne vie humaine en adoptant cette politique qu'en ne l'adoptant pas.
Ce point peut être poussé plus loin. L'autonomie, peut-on dire, exige que l'on ait accès à un éventail adéquat d'options valables ou utiles (Raz 1986). Cette exigence d'adéquation n'implique pas que chaque fois qu'une option est fermée, l'autonomie de l'individu sera remise en cause. En outre, ce qui peut avoir de la valeur n'est pas une activité autonome en soi, mais une activité autonome valable. Joseph Raz explique : "Notre souci d'autonomie étant de permettre aux gens de mener une vie agréable, il nous donne des raisons d'assurer cette autonomie qui pourrait être valable. Offrir, préserver ou protéger de mauvaises options ne permet pas de jouir d'une autonomie valable" (Raz 1986, 412). Si une autonomie valable, et non l'autonomie en soi, est ce qui a une valeur perfectionniste, alors lorsque les gouvernements éliminent, ou rendent plus coûteuses pour les personnes, les options sans valeur, ils ne peuvent pas faire de mal perfectionniste et beaucoup de bien perfectionniste.
Selon Raz, le principe de non-nuisance est remplacé par un principe d'autonomie qui permet de saisir la vérité en elle tout en évitant ses exagérations. Selon le principe d'autonomie, l'État a le devoir négatif de respecter l'autonomie de ses citoyens, ainsi que le devoir positif de promouvoir et de maintenir les conditions sociales qui contribuent à sa réalisation. Le principe de non-nuisance, dans la mesure où il est sain, est justifié parce qu'il découle, compte tenu de certains faits contingents, du principe d'autonomie. Cela conduit Raz à reformuler le principe comme un principe "qui considère la prévention du préjudice à quiconque (y compris lui-même) comme le seul motif justifiable d'ingérence coercitive dans la vie d'une personne" (Raz 1986, 412-13). Ainsi interprété, le principe permettrait l'application coercitive d'au moins certains devoirs d'auto-observation.
Mill et Raz acceptent tous deux des versions du principe de non-nuisance. Mais ils l'acceptent non pas comme une limite à la politique perfectionniste, mais plutôt comme un principe qui guide la promotion adéquate du bien. Leurs théories politiques sont des exemples de libéralisme perfectionniste et leurs discussions sur le principe montrent comment une politique perfectionniste peut soutenir la liberté individuelle. On peut cependant objecter que la défense de la liberté individuelle assurée par le libéralisme perfectionniste est insuffisante. Rappelons que selon le point de vue perfectionniste discuté ici, l'autonomie est un aspect d'une bonne vie humaine. Elle n'est pas une norme distincte. On peut penser qu'une défense forte et sans compromis du principe doit être fondée sur une compréhension différente de la valeur de l'autonomie, qui considère que l'autonomie des personnes ne peut être violée même si, tout bien considéré, elle est dans leur intérêt. L'autonomie, selon ce point de vue, est un droit souverain, et non un idéal à promouvoir (Feinberg 1989)".
https://academic.oup.com/mind/article/125/500/1005/2277465
Mill distinguishes that ‘which concerns others’ from that ‘which merely concerns [the agent] himself’ (1859, p. 224). Moreover, he calls actions ‘self-regarding’ when they do not affect, or at least do not directly harm, others (1859, p. 281). Since Mill employs the notion of self-regarding action, which ought not to be interfered with, it is tempting to contrast this with other-regarding action, which is potentially liable to social interference. This traditional reconstruction of Mill’s views is problematic, both in its own terms and as an interpretation of Mill.
One important preliminary is that Mill’s harm principle actually concerns the reasons that may be given for interference, rather than what may be interfered with ( Ten 1980 , pp. 40–1; Skorupski 1989 , p. 343). Mill sometimes speaks as if restricting the reasons for interference amounts to restricting what could be interfered with, but this is only indirectly so. An action that does not itself harm others might none the less be interfered with in order to prevent harm to others ( Holtug 2002 , p. 360; Bird 2007 , pp. 181–6). For simplicity, I shall sometimes follow Mill in speaking of certain acts being within the protected sphere of liberty, though strictly it is only that certain reasons for intervention are excluded.
Commentators have generally focused on the distinction between self-regarding and other-regarding actions, with critics alleging either that such a distinction is impossible to draw or that no acts of any significance are self-regarding. I shall argue, first, that we should not seek to distinguish between actions that are self-regarding and actions that are other-regarding, but rather between actions that are self-regarding and actions that are not self-regarding. Then, more radically, I show that this distinction is of little significance. Mill could have defended all the conclusions he wanted to had he said that the only legitimate reason for interference is to prevent non-consensual harm. 1 Thus, we should focus on consent, rather than attempting to delineate a self-regarding sphere.
If everything we do may affect others, then the self-regarding sphere will be non-existent. Sympathetic interpreters have sought to resist this conclusion, typically by restricting what counts as harming others. 2
John C. Rees (1960 , pp. 174–80) suggested that we should focus on actions that affect the interests of others. Almost anything I do may causally affect you, but not everything that causally affects you affects your interests. Thus, there is plausibly a range of individual actions that are not other-regarding. This sphere is expanded still further by the observation that not all effects that I might have on your interests license interference; perhaps only harms, i.e. negative effects, permit intervention. (I return to this matter in § 11.)
Further, whether an individual’s interests are affected depends on our account of people’s interests, which Mill derives from his utilitarianism (1859, p. 224). A correct account of interests may show that there are some things we have no interest in avoiding, even if we wish to. Jeremy Waldron (1987) has argued that, in Mill’s view, we have no interest in avoiding distress or offence that comes from moral confrontation. My expressing a view contrary to your deeply held ethical or religious convictions may cause you distress but this does not negatively affect your interests, because you have no interest in avoiding challenges to your convictions. If correct, this further expands the sphere of protected liberty, by showing that not all actions taken to be harmful really are.
Note that ‘other-regarding’ cannot be understood analogously to ‘self-regarding’, at least not if we wish the two categories to be mutually exclusive and exhaustive (as presumably they should be; cf. Brink 2013 , p. 140).
It is natural to assume that other-regarding actions are all those actions that affect other people’s interests and self-regarding actions are all those actions that affect the agent’s own interests. But then some trivial actions, like yawning, may be neither self- nor other-regarding, if they do not affect anyone’s interests. More problematically, these two categories are not mutually exclusive. Many actions will be both self-regarding and other-regarding, because they affect the interests of the agent and of others.
This distinction may do the necessary work if we say that only purely self-regarding actions (i.e. actions that are self-regarding but not other-regarding) should be immune from interference. On this account, the state can interfere with other-regarding actions, whether or not those actions are also self-regarding. That is, the state can interfere with both purely other-regarding actions (actions that affect others but not the agent) and with actions that are both self- and other-regarding, but not with those that are purely self-regarding. This, I think, gets matters substantively right, but it requires us to specify that the protected sphere is not all self-regarding action, but only what is purely self-regarding. However, this is not how Mill uses the phrase ‘self-regarding’ since he says that an act ceases to be self-regarding when it directly harms others (1859, p. 281).
For Mill, ‘self-regarding conduct’ must mean something like conduct that affects only the agent’s interests. However, ‘other-regarding’ cannot similarly mean action that affects only other people’s interests, or else actions that affect both the agent and others will again be neither self- nor other-regarding. Other-regarding actions––to fit the sense required––must be any that affect other people, whether or not they also affect the agent. Thus, ‘other-regarding’ cannot be understood analogously to ‘self-regarding’.
These points hold even on certain refined versions of the distinction. If, for example, we define ‘self-regarding’ action as what directly affects the agent, and ‘other-regarding’ action as what directly affects others, then again an action might be both (if it directly affects both the agent and others) or neither (if it does not directly affect anyone). The categories might be made exclusive and exhaustive if, for instance, we were to define ‘self-regarding’ as what primarily affects the agent, and ‘other-regarding’ as what primarily affects others. This secures a distinction of the sort required, since any action will be either self-regarding or other-regarding, and not both, but it does not suit Mill’s purposes. It is not plausible that society can only intervene with an individual’s action when those actions have more of an effect on others than on the agent herself; surely society may prevent an action that does significant harm to others, even if it has a greater effect (benefit or harm) on the agent.
To do the work required, ‘self-regarding’ must refer to what affects only the agent, but ‘other-regarding’ must refer to whatever affects others, so the two terms cannot be understood analogously. These problems stem from introducing the term ‘other-regarding’ to refer to whatever actions are not self-regarding. Much confusion would have been saved if commentators had never invented the phrase––which is not Mill’s own––and instead spoken of actions being either ‘self-regarding’ or ‘non-self-regarding’. 5 A self-regarding action is one that (directly) affects only the agent herself. Non-self-regarding actions are simply those that are not self-regarding, because they (directly) affect others, regardless of any effects they have on the agent.
This is our first, rather modest, conclusion. It is unhelpful to introduce the notion of ‘other-regarding’ action. In saying this, I do not suggest that those who introduced the phrase ‘other-regarding’ are substantively confused about Mill’s position; ‘other-regarding’ is generally used to mean what is more accurately called non-self-regarding. My point is merely that using the term in this technical sense invites confusion. It would be clearer––and more in keeping with Mill’s text––to avoid the potentially misleading phrase ‘other-regarding’ altogether. This goes some way to resolving problems with the notion of self-regarding action; in principle, a clear line can be drawn between the self-regarding and the non-self-regarding. Whether there is any action of consequence on the self-regarding side remains an open question.
-Ben Saunders, "Reformulating Mill’s Harm Principle", Mind, Volume 125, Issue 500, October 2016, Pages 1005–1032: https://academic.oup.com/mind/article/125/500/1005/2277465
https://plato.stanford.edu/entries/perfectionism-moral/#HarPri
The Harm Principle
Those who reject the principle of state neutrality entrust the state with the task of promoting the good. This can give rise to the worry that a perfectionist account of politics is insufficiently sensitive to the harm of coercion and to the value of liberty. If the state need not be neutral among rival understandings of the good, then is the door not open for the coercive imposition of state policies designed to promote the good? In fact, many versions of perfectionism drawn from the history of political thought have paid little heed to the value of individual liberty. It is an important matter, then, to what extent perfectionist politics can be reconciled with a proper regard for individual liberty.
To approach this issue, it will be helpful to consider the so called harm principle. The harm principle, as articulated by writers in the liberal political tradition from Mill to Feinberg, is often taken to be an essential safeguard for individual freedom in political society. At least at first pass, the harm principle substantially restricts the power of governments to promote the good. It holds that governments cannot coercively interfere with persons unless doing so is necessary to prevent them from causing clear and direct harm to others. The harm principle requires interpretation and can be understood in different ways. But for present purposes our question is whether the harm principle is best understood to be an anti-perfectionist principle—a principle that provides reasons for rejecting or limiting perfectionist politics.
The first thing to say is that not all perfectionist policies are coercive. Governments can and do promote the good noncoercively. A government may promote the good, for example, when it intelligently subsidizes art. This perfectionist policy need involve no coercive interference whatsoever.[14] So the harm principle, even if sound, would not bar all perfectionist policies. It would rule out only coercive governmental policies designed to favor some options and discourage others. The question then is whether this restriction is itself anti-perfectionist.
Here we need to proceed with caution. The extent to which perfectionism licenses the coercive promotion of the good depends, among other things, on the degree to which autonomy or liberty is itself recognized to be a perfectionist good. On some versions of perfectionism the harm principle would be an anti-perfectionist principle, while on others it would not be. Consider, in this regard, Mill’s own defense of the harm principle. For Mill “individuality” is an essential component of a good human life. Mill’s notion of individuality can be understood to be a conception of autonomy. A person is autonomous for Mill if he leads his life on his own terms and develops his capacities and faculties according to “his own mode of laying out his existence” (Mill 1859, 64). The important point for present purposes is not Mill’s particular characterization of autonomy, but rather the structure of his view. Autonomy is understood to be an essential aspect of a good human life, not a separate norm. And the value of autonomy explains, at least in part, why Mill recommends the harm principle.
The Millian defense of the harm principle sits well with the perfectionist focus on good human lives. (Brink 2013) Its availability nicely illustrates how perfectionist politics can be consistent with a strong rejection of state coercion. But it is natural to suspect that Mill overstates his case. Even granting that “individuality” is an aspect of a good human life, we should wonder why it takes priority over all other aspects. If a governmental policy, say a policy that criminalizes the sale and use of dangerous recreational drugs, would prevent many from ruining their lives while infringing the individuality of only a few, then, contrary to Mill, the government may do better in discharging its duty to promote good human lives by adopting the policy than by not adopting it.
This point can be pushed further. Autonomy, it can be argued, requires that one have access to an adequate range of valuable or worthwhile options (Raz 1986). This adequacy requirement does not imply that every time an option is closed off one’s autonomy will be set back. Moreover, what may be of value is not autonomous agency per se, but valuable autonomous agency. Joseph Raz explains: “Since our concern for autonomy is a concern to enable people to have a good life it furnishes us with reason to secure that autonomy which could be valuable. Providing, preserving or protecting bad options does not enable one to enjoy valuable autonomy” (Raz 1986, 412). If valuable autonomy, and not autonomy per se, is what has perfectionist value, then when governments eliminate, or make it more costly for persons to pursue, worthless options, then they may do no perfectionist harm and much perfectionist good.
On Raz’s view, the harm principle is superceded by an autonomy principle that captures the truth in it while avoiding its exaggerations. The autonomy principle holds that the state has negative duties to respect the autonomy of its citizens as well as positive duties to promote and sustain social conditions that contribute to its realization. The harm principle, to the extent that it is sound, is vindicated because it follows, given certain contingent facts, from the autonomy principle. This leads Raz to reformulate the harm principle as a principle “that regards the prevention of harm to anyone (himself included) as the only justifiable ground for coercive interference with a person” (Raz 1986, 412–13). So construed, the harm principle would permit the coercive enforcement of at least some self-regarding duties.
Both Mill and Raz accept versions of the harm principle. But they accept it not as a limit on perfectionist politics, but rather as a principle that guides the proper promotion of the good. Their political theories are examples of perfectionist liberalism and their discussions of the harm principle show how perfectionist politics can be supportive of individual liberty. It can be objected, however, that the defense of individual liberty provided by perfectionist liberalism is insufficient. Recall that on the perfectionist view discussed here autonomy is an aspect of a good human life. It is not a separate norm. A strong and uncompromising defense of the harm principle, it may be thought, must be based on a different understanding of the value of autonomy, one that holds that the autonomy of persons cannot be infringed even when doing so is, all things considered, in their best interests. Autonomy, on this view, is a sovereign right, not an ideal to be promoted (Feinberg 1989)."
"Le principe de non-nuisance.
Ceux qui rejettent le principe de la neutralité de l'État lui confient la tâche de promouvoir le bien. Cela peut susciter l'inquiétude qu'un rapport perfectionniste à la politique ne soit pas suffisamment sensible au danger de la coercition et à la valeur de la liberté. Si l'État n'a pas besoin d'être neutre parmi les conceptions rivales du bien, alors la porte n'est-elle pas ouverte à l'imposition coercitive de politiques d'État destinées à promouvoir le bien ? En fait, de nombreuses versions du perfectionnisme tirées de l'histoire de la pensée politique n'ont guère tenu compte de la valeur de la liberté individuelle. Il est donc important de savoir dans quelle mesure la politique perfectionniste peut être conciliée avec un respect adéquat de la liberté individuelle.
Pour aborder cette question, il sera utile d'examiner le principe dit de non-nuisance. Le principe de non-nuisance, tel qu'il est formulé par les écrivains de la tradition politique libérale, de Mill à Feinberg, est souvent considéré comme une sauvegarde essentielle de la liberté individuelle dans la société politique. Au moins au premier abord, le principe de non-nuisance restreint considérablement le pouvoir des gouvernements de promouvoir le bien. Selon ce principe, les gouvernements ne peuvent pas intervenir de manière coercitive à l'encontre des personnes, sauf si cela est nécessaire pour les empêcher de causer un dommage clair et direct à autrui. Le principe de non-nuisance nécessite une interprétation et peut être compris de différentes manières. Mais pour les besoins de la présente étude, notre question est de savoir si ce principe est mieux compris comme un principe anti-perfectionniste -un principe qui fournit des raisons pour rejeter ou limiter la politique perfectionniste.
La première chose à dire est que toutes les politiques perfectionnistes ne sont pas coercitives. Les gouvernements peuvent promouvoir le bien de manière non coercitive, et ils le font. Un gouvernement peut promouvoir le bien, par exemple, lorsqu'il subventionne intelligemment l'art. Cette politique perfectionniste ne doit impliquer aucune ingérence coercitive quelle qu'elle soit. Ainsi, le principe de non-nuisance, même s'il est sain, n'empêcherait pas toutes les politiques perfectionnistes. Il n'écarterait que les politiques gouvernementales coercitives destinées à favoriser certaines options et à en décourager d'autres. La question est donc de savoir si cette restriction est elle-même anti-perfectionniste.
Nous devons ici procéder avec prudence. La mesure dans laquelle le perfectionnisme autorise la promotion coercitive du bien dépend, entre autres, de la mesure dans laquelle l'autonomie ou la liberté est elle-même reconnue comme un bien perfectionniste. Dans certaines versions du perfectionnisme, le principe de non-nuisance serait un principe anti-perfectionniste, tandis que dans d'autres, il ne le serait pas. Considérez, à cet égard, la propre défense de Mill du principe. Pour Mill, "l'individualité" est une composante essentielle d'une bonne vie humaine. La notion d'individualité de Mill peut être comprise comme une conception de l'autonomie. Pour Mill, une personne est autonome si elle mène sa vie selon ses propres principes et développe ses capacités et ses facultés selon "son propre mode d'organisation de l'existence" (Mill 1859, 64). Le point important pour les besoins de la présente étude n'est pas la caractérisation particulière de l'autonomie de Mill, mais plutôt la structure de son point de vue. L'autonomie est comprise comme un aspect essentiel d'une bonne vie humaine, et non comme une norme distincte. Et la valeur de l'autonomie explique, au moins en partie, pourquoi Mill recommande le principe de non-nuisance.
La défense par Mill du principe de non-nuisance s'accorde bien avec l'accent perfectionniste mis sur les bonnes vies humaines. (Brink 2013) Sa disponibilité illustre bien comment une politique perfectionniste peut être cohérente avec un fort rejet de la coercition étatique. Mais il est naturel de soupçonner que Mill exagère son cas. Même en admettant que l'"individualité" est un aspect d'une bonne vie humaine, on peut se demander pourquoi elle a la priorité sur tous les autres aspects. Si une politique gouvernementale, disons une politique qui criminalise la vente et l'utilisation de drogues récréatives dangereuses, empêchait beaucoup de gens de ruiner leur vie tout en ne portant atteinte à l'individualité que de quelques-uns, alors, contrairement à Mill, le gouvernement pourrait mieux s'acquitter de son devoir de promouvoir une bonne vie humaine en adoptant cette politique qu'en ne l'adoptant pas.
Ce point peut être poussé plus loin. L'autonomie, peut-on dire, exige que l'on ait accès à un éventail adéquat d'options valables ou utiles (Raz 1986). Cette exigence d'adéquation n'implique pas que chaque fois qu'une option est fermée, l'autonomie de l'individu sera remise en cause. En outre, ce qui peut avoir de la valeur n'est pas une activité autonome en soi, mais une activité autonome valable. Joseph Raz explique : "Notre souci d'autonomie étant de permettre aux gens de mener une vie agréable, il nous donne des raisons d'assurer cette autonomie qui pourrait être valable. Offrir, préserver ou protéger de mauvaises options ne permet pas de jouir d'une autonomie valable" (Raz 1986, 412). Si une autonomie valable, et non l'autonomie en soi, est ce qui a une valeur perfectionniste, alors lorsque les gouvernements éliminent, ou rendent plus coûteuses pour les personnes, les options sans valeur, ils ne peuvent pas faire de mal perfectionniste et beaucoup de bien perfectionniste.
Selon Raz, le principe de non-nuisance est remplacé par un principe d'autonomie qui permet de saisir la vérité en elle tout en évitant ses exagérations. Selon le principe d'autonomie, l'État a le devoir négatif de respecter l'autonomie de ses citoyens, ainsi que le devoir positif de promouvoir et de maintenir les conditions sociales qui contribuent à sa réalisation. Le principe de non-nuisance, dans la mesure où il est sain, est justifié parce qu'il découle, compte tenu de certains faits contingents, du principe d'autonomie. Cela conduit Raz à reformuler le principe comme un principe "qui considère la prévention du préjudice à quiconque (y compris lui-même) comme le seul motif justifiable d'ingérence coercitive dans la vie d'une personne" (Raz 1986, 412-13). Ainsi interprété, le principe permettrait l'application coercitive d'au moins certains devoirs d'auto-observation.
Mill et Raz acceptent tous deux des versions du principe de non-nuisance. Mais ils l'acceptent non pas comme une limite à la politique perfectionniste, mais plutôt comme un principe qui guide la promotion adéquate du bien. Leurs théories politiques sont des exemples de libéralisme perfectionniste et leurs discussions sur le principe montrent comment une politique perfectionniste peut soutenir la liberté individuelle. On peut cependant objecter que la défense de la liberté individuelle assurée par le libéralisme perfectionniste est insuffisante. Rappelons que selon le point de vue perfectionniste discuté ici, l'autonomie est un aspect d'une bonne vie humaine. Elle n'est pas une norme distincte. On peut penser qu'une défense forte et sans compromis du principe doit être fondée sur une compréhension différente de la valeur de l'autonomie, qui considère que l'autonomie des personnes ne peut être violée même si, tout bien considéré, elle est dans leur intérêt. L'autonomie, selon ce point de vue, est un droit souverain, et non un idéal à promouvoir (Feinberg 1989)".
https://academic.oup.com/mind/article/125/500/1005/2277465
Mill distinguishes that ‘which concerns others’ from that ‘which merely concerns [the agent] himself’ (1859, p. 224). Moreover, he calls actions ‘self-regarding’ when they do not affect, or at least do not directly harm, others (1859, p. 281). Since Mill employs the notion of self-regarding action, which ought not to be interfered with, it is tempting to contrast this with other-regarding action, which is potentially liable to social interference. This traditional reconstruction of Mill’s views is problematic, both in its own terms and as an interpretation of Mill.
One important preliminary is that Mill’s harm principle actually concerns the reasons that may be given for interference, rather than what may be interfered with ( Ten 1980 , pp. 40–1; Skorupski 1989 , p. 343). Mill sometimes speaks as if restricting the reasons for interference amounts to restricting what could be interfered with, but this is only indirectly so. An action that does not itself harm others might none the less be interfered with in order to prevent harm to others ( Holtug 2002 , p. 360; Bird 2007 , pp. 181–6). For simplicity, I shall sometimes follow Mill in speaking of certain acts being within the protected sphere of liberty, though strictly it is only that certain reasons for intervention are excluded.
Commentators have generally focused on the distinction between self-regarding and other-regarding actions, with critics alleging either that such a distinction is impossible to draw or that no acts of any significance are self-regarding. I shall argue, first, that we should not seek to distinguish between actions that are self-regarding and actions that are other-regarding, but rather between actions that are self-regarding and actions that are not self-regarding. Then, more radically, I show that this distinction is of little significance. Mill could have defended all the conclusions he wanted to had he said that the only legitimate reason for interference is to prevent non-consensual harm. 1 Thus, we should focus on consent, rather than attempting to delineate a self-regarding sphere.
If everything we do may affect others, then the self-regarding sphere will be non-existent. Sympathetic interpreters have sought to resist this conclusion, typically by restricting what counts as harming others. 2
John C. Rees (1960 , pp. 174–80) suggested that we should focus on actions that affect the interests of others. Almost anything I do may causally affect you, but not everything that causally affects you affects your interests. Thus, there is plausibly a range of individual actions that are not other-regarding. This sphere is expanded still further by the observation that not all effects that I might have on your interests license interference; perhaps only harms, i.e. negative effects, permit intervention. (I return to this matter in § 11.)
Further, whether an individual’s interests are affected depends on our account of people’s interests, which Mill derives from his utilitarianism (1859, p. 224). A correct account of interests may show that there are some things we have no interest in avoiding, even if we wish to. Jeremy Waldron (1987) has argued that, in Mill’s view, we have no interest in avoiding distress or offence that comes from moral confrontation. My expressing a view contrary to your deeply held ethical or religious convictions may cause you distress but this does not negatively affect your interests, because you have no interest in avoiding challenges to your convictions. If correct, this further expands the sphere of protected liberty, by showing that not all actions taken to be harmful really are.
Note that ‘other-regarding’ cannot be understood analogously to ‘self-regarding’, at least not if we wish the two categories to be mutually exclusive and exhaustive (as presumably they should be; cf. Brink 2013 , p. 140).
It is natural to assume that other-regarding actions are all those actions that affect other people’s interests and self-regarding actions are all those actions that affect the agent’s own interests. But then some trivial actions, like yawning, may be neither self- nor other-regarding, if they do not affect anyone’s interests. More problematically, these two categories are not mutually exclusive. Many actions will be both self-regarding and other-regarding, because they affect the interests of the agent and of others.
This distinction may do the necessary work if we say that only purely self-regarding actions (i.e. actions that are self-regarding but not other-regarding) should be immune from interference. On this account, the state can interfere with other-regarding actions, whether or not those actions are also self-regarding. That is, the state can interfere with both purely other-regarding actions (actions that affect others but not the agent) and with actions that are both self- and other-regarding, but not with those that are purely self-regarding. This, I think, gets matters substantively right, but it requires us to specify that the protected sphere is not all self-regarding action, but only what is purely self-regarding. However, this is not how Mill uses the phrase ‘self-regarding’ since he says that an act ceases to be self-regarding when it directly harms others (1859, p. 281).
For Mill, ‘self-regarding conduct’ must mean something like conduct that affects only the agent’s interests. However, ‘other-regarding’ cannot similarly mean action that affects only other people’s interests, or else actions that affect both the agent and others will again be neither self- nor other-regarding. Other-regarding actions––to fit the sense required––must be any that affect other people, whether or not they also affect the agent. Thus, ‘other-regarding’ cannot be understood analogously to ‘self-regarding’.
These points hold even on certain refined versions of the distinction. If, for example, we define ‘self-regarding’ action as what directly affects the agent, and ‘other-regarding’ action as what directly affects others, then again an action might be both (if it directly affects both the agent and others) or neither (if it does not directly affect anyone). The categories might be made exclusive and exhaustive if, for instance, we were to define ‘self-regarding’ as what primarily affects the agent, and ‘other-regarding’ as what primarily affects others. This secures a distinction of the sort required, since any action will be either self-regarding or other-regarding, and not both, but it does not suit Mill’s purposes. It is not plausible that society can only intervene with an individual’s action when those actions have more of an effect on others than on the agent herself; surely society may prevent an action that does significant harm to others, even if it has a greater effect (benefit or harm) on the agent.
To do the work required, ‘self-regarding’ must refer to what affects only the agent, but ‘other-regarding’ must refer to whatever affects others, so the two terms cannot be understood analogously. These problems stem from introducing the term ‘other-regarding’ to refer to whatever actions are not self-regarding. Much confusion would have been saved if commentators had never invented the phrase––which is not Mill’s own––and instead spoken of actions being either ‘self-regarding’ or ‘non-self-regarding’. 5 A self-regarding action is one that (directly) affects only the agent herself. Non-self-regarding actions are simply those that are not self-regarding, because they (directly) affect others, regardless of any effects they have on the agent.
This is our first, rather modest, conclusion. It is unhelpful to introduce the notion of ‘other-regarding’ action. In saying this, I do not suggest that those who introduced the phrase ‘other-regarding’ are substantively confused about Mill’s position; ‘other-regarding’ is generally used to mean what is more accurately called non-self-regarding. My point is merely that using the term in this technical sense invites confusion. It would be clearer––and more in keeping with Mill’s text––to avoid the potentially misleading phrase ‘other-regarding’ altogether. This goes some way to resolving problems with the notion of self-regarding action; in principle, a clear line can be drawn between the self-regarding and the non-self-regarding. Whether there is any action of consequence on the self-regarding side remains an open question.
-Ben Saunders, "Reformulating Mill’s Harm Principle", Mind, Volume 125, Issue 500, October 2016, Pages 1005–1032: https://academic.oup.com/mind/article/125/500/1005/2277465
