The limits of autonomy





If autonomy were to be seen as an absolute right to decide for oneself the consequence would be the risk of chaos with the vulnerable left to the mercy of the strong. Furthermore, if conceived as a right then it is logically impossible for it to be absolute. If A decided to exercise his or her autonomy by locking B inside a room this would interfere with B’s autonomy and B could not be said to have a right to autonomy unless A is restrained from exercising his or her right in this way. Thus, ignoring the problem of ability, any right to autonomy cannot be the freedom to do what one wants. That trite example suggests that any right to autonomy must be limited by having regard to the autonomy of others. It might be suggested that A has the right to do what he or she wants providing it does


not prevent B from doing what he or she wants. This formulation, however, would be overly restrictive and would paralyse much of our life. It may be reasonable if resources were plentiful and we lived completely independent lives. But, in a world in which we must compete for scarce goods and interact with others, autonomy would lose much of its value if A cannot do something simply because it would prevent B from doing it. Imagine if A wants to build a house on a particular spot next to the sea with good access to fresh water and a supply of food. If no one else wants to build there then he or she is free to do so. But, if B also wants to build there then there is a problem. Neither A nor B may exercise his or her autonomy if it prevents the other from doing so, which means that A can only build there if B does not want to and vice versa. However, if both want to build there then neither can and the plot must remain unused. This means that they must each select other plots but the same problem may recur ad infinitum meaning that neither can build anywhere. This would be a ridiculous state of affairs and so a middle ground ‘capable of public justification’ must be found,91 which suggests that any limitation must conform with the ‘morality of duty’ rather than the more individualistic ideal ‘morality of aspiration’.92
The harm principle
Perhaps the least contentious ground for limiting autonomy is to prevent harm to others. Mill explicated the most famous version of this principle in his essay ‘On Liberty’. He argued that:
the only purpose for which power can be rightfully exercised over any members of a civilised community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a sufficient warrant. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because, in the opinion of others, to do so would be wise, or even right. These are good reasons for remonstrating with him, or reasoning with him, or persuading him, or entreating him, but not for compelling him, or visiting him with any evil in case he do otherwise.93
This limitation of course depends on what is meant by ‘harm’. Since limiting autonomy may itself be seen as harmful, the outcomes that might justify limiting autonomy should be more than temporary or trivial upsets.



 The limitation must also require that the harm be wrongfully inflicted. One possible way of defining harm is to characterise it as a setback to someone’s interests.95 The problem with this is that, as a justification for limiting autonomy, such a definition would be overly restrictive. Returning to my earlier example of choosing a plot of land to build one’s house on, allowing A to build on the most desirable plot would certainly be a setback to B’s interests if building on that plot was one of B’s autonomous goals. Thus, it would constitute a harm and we could justify limiting A’s autonomy to prevent harm to B. However, as I have shown, this would result in neither being able to build and both would be harmed. This explains why ‘harm’ as a justification for limiting autonomy must incorporate both damage (setback to interests) and the notion of a wrong.96 The notion of a wrong is something that must be defined independently of autonomy and a full exploration is beyond the scope of this book. Suffice it to suggest that someone is wronged if they have a justifiable claim obstructed and that our justifiable claims are defined by the rules determined by the society in which we live. One difficulty in determining the limits of autonomy is that most, if not all, decisions made by any individual will have an impact on other people. This is true even if the decision seems to be essentially selfregarding. For example, imagine that A is a vegan with strongly held views that humans should not use other animals solely for their own benefit. A has developed a condition that will leave him severely disabled unless he accepts a transplant of tissue taken from a pig. Such a decision would contravene A’s deeply held autonomous views about the relationship between humans and animals but his decision to refuse treatment, which will leave him severely disabled, will harm his wife. It will also affect others who have a relationship with A and will place an additional burden on the community’s resources. It is arguable that A ought to consider the effect of his decisions on others and he should take them into account when making a decision. If possible A ought to make decisions that will be good for all concerned. It would be especially good of A to make decisions that put others’ interests ahead of his own. However, if morality and the meaningfulness of individual existence is to survive state coercion then A should not be prevented from exercising his autonomy unless the affected other has a justified claim that would be obstructed by his choice.





It might be argued that A’s wife does have quite a strong claim – arising out of their relationship – requiring that A accepts the necessary treatment to prevent the disability. However, for state coercion to be justified it would need to be clearly established that such a claim would be created when entering into such a relationship. In other words, A would need to consent in advance and this would mean that unless he could predict the need to compromise his principles his consent would not be normatively effective in relation to the pig organ transplant. If his views are deeply held it is unlikely that he would enter such an agreement voluntarily. Being able to define our own relationships, and that includes the obligations that arise from them, is an important part of what it means to be a person. State interference is only warranted where it is not possible for one of the parties to autonomously negotiate the nature of the relationship. As indicated earlier, choices that are essentially self-regarding may burden the community. If the cost to the community is too high then it may be justifiable to limit individual autonomy. This may be explained by Feinberg’s garrison model: in a situation when the community’s very existence is under threat then the selfish decision of a single person may tip the balance. Under those circumstances the community may be justified in seeking to preserve itself by overriding an autonomous choice.97 This justification, however, is simply the harm principle applied to the community rather than to individual others. It depends on seeing the destruction of the community as harm and on the existence of a minimal obligation towards the community that would be breached by failing to assist it in such times of need. While it is theoretically possible that this model may be relevant in the context of individual medical treatment it does not pertain in present-day Britain nor is it foreseeable in the near future. However, it may also be justifiable under the harm principle to limit an individual’s autonomy to prevent lesser levels of community harm. Certain interests may be considered as public interests if they are sufficiently widely held. These interests may be raised as justifiable limits on autonomy either when the specific interests of a sufficiently large number of individuals are harmed or when a ‘common’ interest is harmed.98 Public-health threats, such as the risk of transmitting an infectious disease, may justify coercion. Similarly, threats to the environment may be coercively prevented even though it may not be possible to identify an individual directly harmed. The protection of these community interests is simply an extension of the harm principle to cover ,

those harms where the victims are the indeterminate members of the community.
Other limiting principles
Apart from the harm principle, there are four principles that might justify limiting an individual’s autonomy: prudence offence to others self-harm and morality.99 Although allowing that actions offending good manners or decency may be prohibited if done in public,100 all of these possible constraints may be rejected as interfering with the general development of people’s ability to decide their own lives and to actually decide for themselves how their life should go, which is what gives life its personal value.101 These judgements justify argument, persuasion and remonstration but not coercion. Certainly if, when it matters, our autonomy is restricted to making decisions that others see as wise it would be fatally undermined as a concept. Apart from the impact such a strategy would have on our ability to identify with any of our choices when we know we could not have chosen otherwise, it also threatens the principle that all moral persons are equal. Dominant views of rationality would hold sway and subjugate the ‘incomprehensible’ minority. It might be argued that if autonomy requires rationality then an irrational action is not autonomous and does not, therefore, need to be protected.102 For example, the US President’s Commission suggested that ‘A second limitation on self-determination arises where a person’s decisionmaking is so defective or mistaken that the decision fails to promote the person’s own values or goals.’103 This constraint protects the overtly autonomous act but not the other acts of an autonomous person. One problem with this approach is that acts may be autonomous but not obviously so, for example, where people autonomously choose to act in a way that does not appear to coincide with their goals or interests. A related problem is the incomprehensibility that arises because observers are unable to understand the person’s goals. This results from a difference in the conception of the ‘good life’ and what goals a person ought to aim for. It is also likely, as Mill claimed, that – subconscious influences notwithstanding – we know ourselves better than others do. While I may not always be able to explain my actions, my own knowledge of my interests and goals and my own understanding of what life means for me will almost certainly be better than some other person’s appreciation of those crucial decision-making factors. Thus, an external judge of rationality may easily be mistaken about the logic of my decision. It may further be argued that we learn best by being allowed to make mistakes.



If I am only allowed to make rational decisions I will not be allowed to make mistakes and my ability to reason will improve less quickly. This argument can only apply to those mistakes that will not cause permanent and significant harm to my ability to be autonomous. It would be self-defeating to argue that I should be allowed to make mistakes so that I can learn better how to make decisions if, as a result, I am unable to exercise that rationality. Related to this argument is perhaps the stronger point that, since no one is infallible and I am the person who will have to live with the consequences, I should be the one who makes the decision. If, because you profess an expertise, I choose to rely on your judgement I am responsible for the decision and should accept the consequences with equanimity. Certainly, I have no justifiable reason to blame you for my misfortune. However, if you force a choice on me against my will and I am harmed as a consequence that harm will be amplified precisely because you chose to act against my will. Just as stifling free speech suggests infallibility,104 so too does preventing free choice. Such an unwarranted profession of infallibility makes a bad outcome seem all the worse especially as it is I rather than you who will have to live with the consequences. Apart from when the harm principle is invoked, limiting autonomy on the basis of morality is problematic. The reality of moral relativity and the existence of a pluralistic society make it difficult to base rules on particular moral principles. Again the problem of a dominant view is raised and allowing such moral judgements to justify coercive law hints at an arrogant intolerance. As Hayry noted, ‘there are an infinite variety of interpretations of what is moral, and to a person defending any one of them, its opponents will always appear more or less irrational’.


if a particular view of morality is enforced this implies that the followers of a different morality are somehow less equal. If living by a particular code is not harmful to others then the value of such a code can only be coercively denied by a society that sees those individuals as less worthy of respect. Since moral equality is one of the assumptions behind my argument, harmless moral (or ‘immoral’) beliefs/actions cannot justify limiting autonomy. It is arguable that offensive actions fall within the purview of the law. It may be legitimate, as Mill argued (see p. 33), to restrict offensive activities by banning them from public display. In general, however, it would be a greater invasion of liberty and autonomy to coercively prevent an action seen by some as offensive. What counts as offensive is, to a large extent, simply a matter of opinion. Categorising something as offensive is an appeal to individual sensibilities and is not subject to reason but to feelings and, as such, it is less objective than the more legitimate harm principle. To infringe one person’s autonomy because another finds the idea of it offensive is to treat the actor as less equal because it subjugates his or her autonomy to the irrational emotions of the other; it becomes a conflict between two different wants.




Even though it may be rational to have an interest in not being offended, the content of that interest is populated by feelings. If the offensive actions solely further the actor’s interest in being offensive, then there is a conflict of equal interests. Other autonomous interests, however, will usually be weightier than the interest in not being offended and setbacks to those other interests may be more permanent than the relatively temporary impact of offensiveness. The interest in not being offended may be infringed by offensive behaviour but the offence will not persist beyond the incident. The setback caused to others by banning the ‘offensive’ activity may, however, have a much more permanent and total impact. In the context of medical care this argument is particularly apt. The types of treatment that may be seen as offensive include operations like abortion and sex change.106 The impact of preventing an abortion because such operations offend a section of society would be huge, both for the individual woman and for society. History demonstrates the misery and harm caused to women through having to deal with unwanted pregnancy when abortion is unlawful.


I will consider the issue of paternalism and the restriction of autonomy to prevent self-harm in more detail in Chapter 2 following a discussion of the principle of beneficence. For now I will confine the debate to an examination of whether it is legitimate to restrict present autonomy in order to protect the autonomy of the future self. One argument that would support restricting present autonomy in order to protect future autonomy relies on a utilitarian position that if autonomy is a good thing then it should be maximised and what really matters is not how much autonomy I can exercise now but how much I am able to exercise over my whole life. This may mean that present autonomy should not be respected if it would undermine my future autonomy.107 Parfit’s reductionist argument, based on the relative unimportance of personal identity, that ‘we ought not to do to our future selves what it would be wrong to do to other people’, forms the basis for a second argument.108 If the harm principle justifies restricting autonomy to protect others from harm then, if my psychological connectedness to my future self is no greater than it is to current third parties, just as my autonomy may be justifiably restricted to prevent harm to others so it may be constrained to protect my future self. As I noted earlier, for autonomy to be constrained the harm principle requires the act to be wrongful. What this implies is that for harm to future autonomy to count as a justifiable limit I must owe myself a duty not to cause such a setback to my autonomy.



This is problematic not least because if I owe myself the duty there is no good reason why I cannot waive it and that undermines any such duty. This consequence may be avoided if I am held to be the subject of the duty but the duty is owed to the community (or state). However, for such a duty to be justifiable, the harm caused by infringing my present autonomy must be less than the harm that would otherwise be caused to my future autonomy. Furthermore, any claim that I owe a duty to protect my future autonomy must survive the following arguments. The starting point is that many choices restrict my future options. To use Parfit’s railway analogy, if I choose the right – rather than the middle or left – track I lose all of the opportunities associated with the middle and left track. In this way, every time I exercise my autonomy I restrict my future autonomy. Avoiding making choices cannot solve this problem because it defeats the point of being autonomous .














Post a comment

0 Comments