The child's right to bodily integrity and autonomy: A conceptual analysis

Abstract

It is widely accepted that children enjoy some form of a right to bodily integrity. However, there is little agreement about the precise nature and scope of this right. This paper offers a conceptual analysis of the child's right to bodily integrity, in order to further elucidate the relationship between the child's right to bodily integrity and considerations of autonomy. Following a discussion of Leif Wenar's work on the structure and justification of rights, I first explain how the adult's right to bodily integrity can be distilled into separate elements that may plausibly be justified by different moral considerations. In particular, I claim that this analysis suggests that whilst the adult's right to bodily integrity is not wholly reducible to bodily autonomy, autonomy nonetheless remains entwined with our understanding of this right in a number of ways. On the basis of this discussion, I go on to outline three important complexities that arise when we consider the child's right to bodily integrity, before particularly focusing on the question of how third parties should determine whether or not to perform a physical interference upon a child who lacks decision-making capacity. Here, I raise some objections to Earp and Mazor's recent attempts to answer this question, before briefly defending an ‘autonomy-based interests’ account of permissible interference, an account that shares in what I take to be the spirit, if not the precise letter, of these earlier views.

The right to bodily integrity (RBI) affords significant protection against various kinds of intervention. It is amongst the most fundamental rights that individuals enjoy,^1,2 and it has been described as the core legal value in health law. ³ It is also widely accepted that children enjoy some form of a RBI, ⁴ a point that has been invoked in ethico-legal discussions of controversial medical interventions, ⁵ genital cutting^6,7 and corporal punishment. ¹

Nonetheless, there is little agreement about the precise nature and scope of the child's RBI. This is partly a reflection of the fact that there is also considerable divergence about the nature and scope of the adult's RBI. As one initial illustration of this, consider how different theorists have conceptualised the relationship between the adult's RBI and autonomy. It is often claimed that there is a close relationship between the two; Hill ⁸ identifies autonomy and bodily security as key aspects of the RBI, whilst others have emphasised that informed consent is ‘fundamental’ to the adult's RBI. ⁹ However, Herring and Wall have recently stressed that the RBI is not reducible to bodily autonomy, and that we should draw a clear distinction between bodily autonomy and bodily integrity.^2,10

Both views capture important truths; the reason that they are able to do so without contradiction is that rights are complex, molecular concepts incorporating a number of different elements. ¹¹ As I hope to establish in what follows, with greater conceptual clarity about this complex right in adults, we can begin to gain greater insight into some of the unique issues that arise in thinking about the child's RBI, and the role that considerations of autonomy should play in our understanding of this right.

A final introductory point of clarification; in this paper, I shall be concerned with the child's RBI conceived as a moral rather than legal right. Given the controversies surrounding the legal status of the RBI, it is, I believe, crucial to understand the justification of the moral RBI before offering an argument for the place of that right within the law. ^a

The structure of the adult's RBI

As Leif Wenar ¹¹ has argued, any statement of the form ‘a person has a right to X’ can be further distilled into more basic elements. The adult's RBI is a good illustration of this point. When we say that someone has a RBI, one of the fundamental things that we are saying (amongst others) is that third parties have a duty to not perform certain actions upon the right-bearer. We may say that the RBI thus incorporates a negative claim. ^b

Even this brief analysis raises important questions. First, what is the content of this negative claim, or what duties does the RBI imply? Second, what justifies this claim? These are complex questions that I can only answer briefly here. We can begin by saying that the RBI implies a duty on others to refrain from physically interfering with the right-bearer's body. The precise implications of this statement will depend on: (a) our understanding of what constitutes a physical interference (is mere touching sufficient, or must it involve penetration of the skin? ² Must the interference substantially deviate from what is in the person's interests to infringe the claim? ⁹ ) and (b) how we understand the boundaries of the body. ⁶ In any case though, we may note that the severity of a given physical interference (and the extent to which it is contrary to a person's interests) plausibly admits of degree ² ; this may depend on a number of factors, such as the invasiveness and reversibility of the interference. ¹⁴

With respect to the justification of the claim against physical interference, it is helpful to begin by reflecting on the broader justification of rights. As Wenar ¹¹ illustrates, we can identify two broad approaches. According to the ‘interest theory’, the purpose of rights is to safeguard interests that are central to the right-holder's well-being. The alternative ‘will theory’ holds that rights safeguard a sphere of autonomy, acknowledging the individual's sovereignty over certain matters. Wenar ¹¹ himself proposes a ‘several functions’ theory, according to which rights need not have one unitary function. The several functions approach is compatible with the thought that different elements of the same right might be justified by different functions.

Both approaches can be invoked to answer the question ‘what justifies a person's claim against physical interference?’ One might argue that this claim protects a particularly strong interest, given the importance of the body to well-being. For instance, consider Herring and Wall's ² claim that the body is ‘… where we experience states of well-being, it is the way in which we flourish as humans, it is the medium through which we interact with others, and it is the way in which we execute our agency’. This approach suggests one way in which considerations of autonomy are relevant to understanding the RBI; our interest in exercising our autonomous agency plausibly contributes to the broader interest that the RBI protects. I shall return to this point below. Of course, one might also justify the RBI by appealing to ideals of self-ownership, and the idea that the individual should retain sovereignty over their body.

Crucially in this regard, the adult's RBI is often understood to incorporate further elements beyond the claim against physical interference per se. One such element is the individual's power to waive this claim. My having the RBI does not imply that my claim against physical interference must always be in force; instead, in many cases, I can choose to release another from their duty to not physically interfere with my body. ^c I can do so via the provision of valid consent; consent is a crucial mechanism by which a right-holder can authoritatively exercise the power to waive the claim enshrined in their RBI, absolving others of their duties to not interfere. ^d

The power of waiver incorporated into the adult's RBI is plausibly justified by considerations of autonomy; ¹⁵ the power safeguards the right-holder's sphere of decision-making authority over the duties of others. We should also note that whilst the interest (and ideals of self-ownership) that may ground the claim against physical interference plausibly apply to all individuals (no matter their age), not all people have the capacities that are necessary to make autonomous decisions about whether to waive a claim against physical interference. If the individual is not capable of autonomously deciding to exercise the power to waive a claim, then it is not clear that affording the individual this power as part of their RBI can be justified by considerations of autonomy.

To see this difference, a temporarily unconscious adult patient cannot make an autonomous choice about whether to waive a right against physical interference; but they plausibly still have an interest against physical interference that warrants weighty consideration. Accordingly, even if the RBI is a universal right (because the claim it enshrines is enjoyed by all human beings), that does not commit us to the view that all holders of this right should be afforded the power to waive the claim the RBI incorporates.

We can now begin to see how each of the approaches to the RBI briefly surveyed in the introduction captures important truths. The RBI is not wholly reducible to bodily autonomy; the claim against physical interference can plausibly be understood to be grounded in the strength of the interest it protects, as well as ideals of self-ownership; the claim is thus plausibly enjoyed by all individuals (including those who lack the capacity to make autonomous decisions). However, autonomy remains entwined with the RBI in a number of ways. Most importantly, consideration of autonomy provides the foundation for the power of waiver that is afforded to many holders of the RBI. In addition, our interest in bodily integrity might partly be constituted by our interest in being able to exercise our own autonomous agency.

Finally, considerations of autonomy also have implications for the normative force of the RBI. In the case of individuals with decision-making capacity, it is widely agreed that it would be impermissible to perform a non-consensual physical interference on that individual, even when doing so would plausibly be in their best interests. Consider, for example, the right of patients to refuse even life-saving medical treatment. Accordingly, when an individual has the power to waive their claim against physical interference, but chooses not to exercise it, the claim enshrined in their RBI is taken to trump other beneficence-based considerations in favour of the interference. As Vallentyne ¹⁶ notes, in such cases, ‘valid consent is necessary for permissible breach’.

In contrast, when individuals lack decision-making capacity, and their RBI is thus not understood to incorporate the relevant power of waiver, the strength of their claim against interference may not have this same trumping force over broader considerations of beneficence. To illustrate, suppose that doctors believe that an unconscious adult patient requires a life-saving medical procedure; the patient has not made an advance decision authorising it, nor have they identified a proxy decision-maker. Now, the fact that this patient has a claim against physical interference that he is incapable of waiving does not entail that it would be impermissible for doctors to perform this procedure. One could seek to morally justify the interference in one of the following ways.

First, one could try and establish that the patient would waive the claim if she hypothetically had the capacity to do so. This ‘substituted judgment’ model of hypothetical consent would be a way of attempting to honour the patient's autonomy, even if she is now unable to exercise it; it is a broadly autonomy-based justification of interference. ¹⁷ Alternatively, one could argue that even though the patient has not waived their claim against interference, the procedure would serve interests of the patient that outweigh their interest against physical interference. In other words, one would argue that the procedure is in the patient's best interests; this is a beneficence-based approach to justifying the interference. Here, there is no sense in which the patient's claim has been waived; rather, the interference is justified by the thought that the benefits of interference sufficiently outweigh the harms.

Both of these approaches face epistemic challenges; how can we know what a person lacking capacity would hypothetically have consented to if they did have the capacity to do so?^18–20 How can we know what is in a patient's best interests, given the many different reasonable conceptions of well-being, and the difficulties in ascertaining an individual's preferences?^21,22 To give a brief illustration of the latter issue, one might attempt to overcome the epistemic obstacles facing a best interests assessment by adopting a simplistic, medicalised conception of best interests, according to which a patient's best interests are simply ‘determined by the patient's clinical need’. ²³ On such an approach, a physical interference will be in a person's best interests if it adequately addresses a clinical need. However, this narrow view has been criticised due to its implicit paternalism; health is not the only thing that matters for well-being, and it is not always what matters most for some people. ¹⁵

Therefore, to adequately reflect the complexity of the concept of well-being, best interests assessments need to draw on a wide range of factors and evidence. In particular, it is widely agreed that assessments should include the patient's own views and preferences where possible; the individual's own views should carry some weight in determining what is in her best interests,^23,24 even if she lacks the capacity to waive her claims against interference. This reflects the thought that the individual herself has unique insight into the significance that different kinds of interests have in her own conception of well-being. It also acknowledges the anti-paternalist thought that individuals can differ significantly about what makes a life go well. ¹⁵

The child's RBI and an autonomy-based interests approach to its permissible infringement

In view of these clarificatory remarks about the RBI in adults, we can now identify three important complexities that arise when we consider the child's RBI.

First, it can be unclear whether a particular child's RBI should be understood to incorporate the power of waiver in the context of a specific decision. Some elder children and adolescents may plausibly possess the capacities that are necessary for making certain decisions about physical interference autonomously, at least to a degree suggesting that they represent borderline cases of decision-making capacity.^25,26 Second, in cases where the child herself is not afforded the power to waive their claim against interference, there is another set of questions about how much normative significance that claim should have, and whether parents may have power rights in relation to particular decisions involving their child's RBI. This might be justified in part on the basis that parents are in a privileged epistemic position to make a decision that is in their child's best interests; it might also potentially be justified on the basis that parents should have some zone of discretion and authority to direct the upbringing of their child. ⁵ Both of these are important issues that I must set aside here; I direct the interested reader to comprehensive treatments of these issues elsewhere.^5,27

Instead, I shall focus on a third issue that arises in discussing the child's RBI; when the child lacks capacity (and the associated power to waive their claim against physical interference), and assuming that parents do not have the relevant power to unilaterally decide the matter upon whatever basis they see fit, how should the relevant parties determine whether it is justifiable to perform a physical interference upon the child?

In the previous section, I adverted to two ways of responding to the corresponding question in the case of adults lacking capacity; the substituted judgment and the best interests approach. Both approaches have analogues in discussions of the child's RBI. For instance, Earp ²⁸ suggests that it may be permissible to perform a physical interference on a child who is unable to consent if they would consent to the intervention if they were able. Alternatively, Vallentyne ¹⁶ endorses a best interests approach to permissible interference, suggesting that ‘agents are permitted to impose short-term bodily harms on children—as long as the net long-term effects are suitably non-harmful’. ^e

The application of these approaches in the case of children inherits the epistemic challenges faced by their application in adults, as discussed above. However, the presumed consent model faces a particularly profound challenge in this context. In the case of individuals who have never had capacity (such as young children), it is not clear how we can apply a model of hypothetical consent; doing so requires us to hypothesise the values of a capacitous person who has not yet existed. ²⁹ Earp ²⁸ recognises this difficulty and suggests in response that we should only presume consent for interventions that are almost universally regarded as promoting well-being (when such judgments are stable over time), and that the threshold for permissible intervention is likely to be at or near the threshold of medical necessity. However, whilst this response avoids the conceptual difficulty of hypothesising the values of a person who has not yet existed, it appears that this presumed consent approach is no longer serving as an analogue of the patient's autonomy; instead, it is an analogue for an objective conception of well-being grounded by universally valued goods and the goods of medicine.

Here, one might also raise the concern that an appeal to such an objective standard of well-being may also provide only limited guidance in light of deep disagreements about well-being, even in the context of healthcare. Moreover, we might note a further anti-paternalist concern, namely that a standard grounded by universally agreed values will have trouble accommodating the possibility that future children might develop quite different values to those of currently existing generations. Interestingly, Mazor offers one way in which one might incorporate the child's own future views into a best interests approach, in a manner that could potentially avoid this sort of concern. He suggests that whether a physical interference is in the child's best interests should be understood to depend upon the strength of their interest in ‘being able to decide for herself once she is autonomous what is to be done with her body’. ⁹ In turn, Mazor ⁹ understands the strength of this interest in bodily autonomy to depend on how likely she is to regret and reflectively reject her parent's decision in the future.

However, it is not clear that the absence of evidence of likely regret speaks strongly in favour of the claim that interference would be in a child's best interests. Such an assumption overlooks the fact that we may be considering an intervention on a child who is still in the process of developing their values. Crucially, the values that the child develops, and that will form the basis for her later retrospective views about a physical intervention may very well depend on whether the intervention in question was carried out in the first place. The problem then is not simply that we do not know what the child values at the time of intervention; rather, the issue is that whether or not the intervention is performed may lead the child to engage with quite different communities and develop quite different values and retrospective views of the intervention itself. ¹⁵ For this reason, I broadly agree with Thomson when she claims that the child's RBI is ‘grounded in the child's interest in having their bodily integrity respected irrespective of whether they would or would not retrospectively endorse’ ¹⁰ a given physical interference.

Despite these concerns, both Mazor and Earp capture important insights. Earp is right to seek to elucidate the permissibility of interference with a framework that places significant justificatory weight on considerations of autonomy. However, rather than justifying infringement by hypothesising what the child would consent to, I believe it is more promising to incorporate considerations of autonomy into a beneficence-based framework, as Mazor suggests. The general thought here is that the strength of the child's claim against bodily interference is grounded, inter alia, by their interest in becoming an autonomous person. On such an approach, the child's interest in autonomy is one of the crucial interests that should be accommodated by a best interests assessment. However, as I shall now explain, this autonomy-based interest goes beyond the particular interest that Mazor highlights, namely the child's interest in being able to decide for herself once she is autonomous on what is to be done with her body.

Above, I noted that autonomy plays a crucial role in the adult's RBI. Most obviously, considerations of autonomy provide the moral foundation for how adults may waive their claim against physical interference via the provision of valid consent. However, I also noted that the claim against physical interference itself may be grounded, inter alia, by the individual's interest in exercising their own autonomous agency. In a similar vein, even when a child's RBI does not incorporate the power of waiver, considerations of autonomy can still play a significant role in grounding the child's interest against physical interference. Of course, safeguarding an interest in autonomy may require quite different things in adults and children. Respecting an adult's autonomy plausibly requires respecting their occurrent choices about what kind of life to lead. However, when we are considering a child who has not yet developed a robust conception of what would make their life go well, considerations of autonomy instead give us strong grounds for placing a particular emphasis on safeguarding the space and freedom that they need to do this. ¹⁵ In the absence of good evidence that the particular child in front of us is (and will continue to be) reflectively committed to a particular good, our decision-making about protecting their autonomy should be significantly guided by a concern with safeguarding their ability to develop their own judgments about what matters to them, and to act on those judgments in the future.

This broad interest has also been said to ground a separate right; namely, a child's right-in-trust to an open future. ³⁰ On the approach I am outlining here, there are some important overlaps between the RBI and this right, since some physical interferences can close important future options; one obvious example would be performing a sterilising physical procedure upon a child that would prevent them from procreating when they reach adulthood. ³⁰ Nonetheless, the two rights are not entirely co-extensive on the approach that I am suggesting. First, there are clearly many ways in which one might violate a child's right to an open future without physical interference; for example, one might do so by preventing a child from accessing social or educational opportunities. Second, and more importantly for my purposes, there may plausibly be ways of acting in accordance with the right to an open future that do not actually promote the development of the child's autonomy. As critics of the right to an open future have pointed out, it is impossible to maintain an entirely open future for one's child, and doing so may overburden the child with too much choice, or only enable them to engage with their myriad available options in a shallow and superficial way.^31–34 In order to avoid these complications, I shall simply parse my arguments in terms of the fundamental interest that I take to be of central importance here; namely, the child's interest in developing their autonomy. The suggestion that I am raising here is that certain physical interferences may serve to undermine this interest in autonomy by taking away crucial options that are related to the child's developing sense of self in a particularly intimate way.

Four points of clarification are immediately in order. First, this interest in autonomy is not necessary for grounding a claim against physical interference. In some cases, a child may plausibly have a (perhaps weaker) claim against a physical intervention even if that intervention would not significantly undermine the development of her autonomy. For example, a child plausibly has an interest against having their ears pierced, even if this would not pose a particular threat to the development of her autonomy. I shall say more about what kinds of intervention might plausibly threaten the development of autonomy towards the end of the paper; the initial point I am making here is that this is a particularly significant interest to attend to in thinking about whether an infringement of a child's RBI is in their best interests.

Second, in a similar vein to the right to an open future, there are many other ways in which the development of a child's autonomy can be stymied without performing a physical intervention. Nonetheless, this does not undermine the narrower point that I am making here, namely that if a physical intervention does threaten to significantly undermine the child's developing autonomy, then that is sufficient to ground a particularly strong claim against the interference in question. I shall also suggest below that physical interventions can represent a particularly salient kind of threat to the development of the individual's future autonomy, due to the crucial role that our embodiment plays in our autonomous agency. ³⁵

Third, it is important to be clear that in appealing to the significance of safeguarding the child's interest in developing autonomy in our thinking about the interest that the RBI serves to protect, I am not suggesting that the child's RBI should be understood as simply seeking to protect the child's life from going poorly. ^f Rather, it should be understood as seeking to protect, inter alia, the particular contribution that autonomy can make to how well one's life goes. Of course, autonomy can sometimes lead people to make decisions that make them worse off. Nonetheless, whilst there is scope to debate the extent of autonomy's contribution to well-being, there is, I suggest, a great deal of truth in Mill's suggestion that there is a fundamental prudential value in an individual's laying out their own mode of existence ‘… not because it is best in itself, but because it is his own mode’. ³⁶ It is the interest in becoming a person who can lay out their own mode of existence that the child's RBI may be understood as seeking, at least in part, to protect. This can sensibly be described as an autonomy-based interest.

Finally, not all physical interferences undermine this interest; in fact, some physical interventions may serve to protect the development of the child's autonomy rather than undermining it. For example, medical interventions can be necessary for ensuring that a child will retain or develop certain abilities in the future. An important feature of some such cases is that although performing the physical intervention now rather than at some later point (perhaps when the person has the capacity to make their own decision) may take away one important choice for the individual, this diminishment of choice may enhance the development of their autonomy all things considered. To illustrate, considerations of autonomy can still speak in favour of performing an irreversible intervention in a child if that would safely and effectively mitigate a debilitating condition that would otherwise serve to significantly limit the child's (and future adult's) abilities. This is an important point to acknowledge, as it suggests one way in which autonomy-based considerations may override the sort of concern that Mazor highlights, namely the child's concern to decide for herself once she is autonomous whether a particular physical intervention should occur.

In a similar vein, the approach I am sketching here can align with Earp's claim that we may permissibly perform medically necessary interventions in children. However, there are some important differences in our justification for this claim, which it will be helpful to explicate. Recall that Earp suggests that we may presume consent for physical interventions that are at or near the threshold of medically necessity; I noted above some problems with the theoretical apparatus of presumed consent in this context, as well as some broadly anti-paternalist concerns about that line of justification. On the alternative approach that I am suggesting here, many medically necessary interventions may be justified because they will also likely be necessary for protecting the child's autonomy-based interests.

Theoretically, the starkest contrast between my view and Earp's will arise in cases in which a medically necessary intervention might plausibly be understood to infringe upon the child's autonomy-based interests. I use the term ‘theoretically’ because such cases will likely be rare on suitably nuanced understandings of the concepts of ‘autonomy’, ‘health’ and ‘medical necessity’. To briefly illustrate, if medically necessary interventions aim to restore health, and if one's health can plausibly be understood as a key factor in determining the scope of: (i) one's future options and (ii) the extent to which one will be able to act effectively on one's choices, then the scope for these sorts of conflicting cases significantly diminishes. Again, this is true even if the decision to perform an intervention in childhood does take away one important choice from the individual (namely, the choice about whether to have undergone the intervention in question); restricting some choices can be justified by autonomy-based considerations, all things considered.

A robust defence of the claim that such conflicting cases will only rarely arise would take us into deeper waters about the definition of health and medical necessity than I can enter into here. Accordingly, let us assume that some such conflicting cases can arise. I take it that Earp's position on such cases is that establishing the medical necessity of a paediatric procedure is typically sufficient for establishing that we can presume consent for it and a fortiori its permissibility on the approach he defends. The view I am outlining here differs; medical necessity does not serve as a trump that can alone establish the permissibility of an intervention. Rather, the health-based interests that medically necessary interventions serve to protect would have to be weighed against the autonomy-based interests that they may set back in conflicting cases. This means that the approach I am suggesting may complicate rather than simplify matters. However, I take that to be a virtue of the theory; indeed, I believe we should be wary of affording such dialectic power to a concept that is as contested as ‘medical necessity’, especially given the anti-paternalistic thought that health, especially on narrower biostatistical understandings of that concept, ³⁷ is not always the most important value for some people.

Like Earp's model then, an autonomy-based interest approach ultimately appeals to an objective component of well-being in justifying permissible physical interventions, namely autonomy. Crucially though, by emphasising the importance of the child's ability to develop their autonomy and conception of the good, the approach seeks to safeguard the sort of individuality that anti-paternalists advocate. Unlike the presumed consent model, autonomy here does not justify a physical interference because it grounds the authority of a hypothetical token of consent. Rather, the interest in autonomy is understood to be one that could potentially outweigh the interest that is protected by the child's claim against certain forms of physical interference.

Moreover, on the approach suggested here, the importance of autonomy's contribution to well-being need not be taken to depend on the future endorsement or rejection of different interventions, in the manner that Mazor suggests. Nor are the autonomy-related concerns restricted to the frustration of the future adult's retrospective preferences about particular physical interventions. Instead, the deeper concern on this autonomy-based approach is that if certain choices and freedoms are taken away from the child by virtue of a physical interference, then this may circumscribe their ability to develop their own broader autonomous agency. ³⁰ The interest in developing the capacity for autonomy is then, I suggest, an interest that is fundamental to the child's well-being, and one that should significantly inform our understanding about what is in their interests. Accordingly, whether a physical interference is justifiable in a child who cannot consent depends not only on whether the medical benefits of interference outweigh contemporaneous harms to the child (such as the physical pain it causes); we must also consider the extent to which the intervention will safeguard or circumscribe the future development of the child's autonomy.

What sorts of physical interventions will raise particular problems for the approach I am suggesting here? Space here allows only for a brief sketch, and there is much more to say about how the interest in future autonomy should be weighed against other components of well-being – however, the following observations are apposite.

We can begin with the prosaic but important point that there will plausibly be a higher bar to pass for autonomy-based justifications of irreversible physical interventions than for reversible interventions. ¹⁴ Although I have suggested that preserving the individual's later choice about whether to undergo a particular physical intervention is not the only relevant consideration for autonomy in this context, that is not to say that it is irrelevant. Crucially, reversible interventions preserve the possibility for the agent to exercise an important choice (about whether to maintain certain effects of the intervention) at a later point in a way that irreversible interventions do not. In a similar vein, if an intervention can be delayed until the child develops the ability to make an autonomous decision about the intervention without infringing upon their other interests, then the autonomy-based interest approach will support delaying the intervention until that time.

These are important autonomy-based considerations, but they may not constitute decisive strikes against the performance of an intervention. In some cases, reversible interventions may not be available, and there may be sufficient costs to delaying an intervention to outweigh the benefits of preserving that choice; for instance, the intervention may be more effective if it is performed earlier in the child's development, or there may be risks of the patient's condition deteriorating if a procedure is delayed. Indeed, such considerations have been militated in favour of performing even highly invasive procedures in paediatric patients, including deep brain stimulation in serious treatment refractory conditions.^38,39 Again though, such procedures can admit of an autonomy-based justification if they serve to safeguard the child's ability to develop their autonomous agency.

Beyond this broad observation of the defeasible reasons to preserve the individual's choice about whether to undergo an intervention, it is possible to identify three dimensions that we should attend to in thinking about how to safeguard a child's autonomy-based interests. First, we should attend to the capacities that will be necessary for the individual to make autonomous decisions in the future; these may include both cognitive capacities (such as the ability to understand information about what their options are like) and broader emotional ⁴⁰ and relational capacities^41,42 that may also be understood to play an important role in an agent's ability to make autonomous decisions. Notably, one capacity that relational theorists of autonomy have particularly stressed is the capacity for self-worth;^43,44 I shall return to this point below.

Second, we should also attend to the abilities that the individual will need to act on the basis of their autonomous choices in the future. Here, we face considerable epistemic challenges, since we do not yet know what the child may later value and choose to pursue in their adult life. However, we can at least be guided in our judgments on this point by a consideration of the sorts of freedoms, abilities and capacities that are necessary for the pursuit of a wide range of life plans, or what Rawls called ‘primary goods’. ⁴⁵ To be clear though, the rationale for intervening in accordance with these goods is not that we are presuming what the child will come to value or that such goods are universally valued per se. Rather, the thought is that, in the absence of knowing what the child will come to value, the best we can do is give them tools that will be useful to them whatever kind of life they may end up wanting to lead.

Finally, we must also consider what kind of options a child must have in order to sufficiently develop their autonomous agency. Autonomous agents form and sustain their preferences in the light of their available options; options foreclosed in childhood may thus not feature in the individual's later reflections on what sort of things to pursue in life. ¹⁵ For this reason, the development of autonomy plausibly requires a certain sphere of freedom, even if this need not amount to the sort of ‘fully open future’ that concerns critics of the right to an open future. ⁴⁶ Determining just how open a child's future must be to safeguard their capacity for autonomy is a hugely complex question that I cannot hope to answer here, but I shall conclude with some reflections on why this set of issues is deeply implicated with our understanding of the child's RBI.

As Lewis and Holm ³⁵ have convincingly argued, embodiment plays an indispensable role in our autonomous agency. Drawing on a phenomenological approach, these authors suggest that our individual access to our own values, desires and motivations is mediated through our intentional bodily engagement with the world. We also forge our and express our unique practical identities through these bodily engagements. For these reasons, they claim that ‘having a practical identity and thereby the capacity for bodily intentionality is a necessary component of one's capacity for autonomy’, ³⁵ and that experiences of bodily disunity, or alienation from one's body can lead to crises in our personal identities and disrupt our capacity to reflect on what we have reasons to do.

Even this brief summary of Lewis and Holm's rich discussion is sufficient to highlight the importance of considerations of autonomy to the child's RBI. Physical interventions in children can occur in crucial periods in which the child is beginning to forge the practical identity that will form the basis for their later autonomy. Furthermore, interventions into certain parts of the body can plausibly threaten to evince forms of bodily disunity that might lie at the core of substantial threads of that burgeoning identity. Of course, some interventions, such as piercing a child's ears, may be unlikely to engender this problematic form of bodily alienation. However, others can affect parts of the body that may be deeply entwined with individual's burgeoning sense of self and identity, such as, for instance, the genitals, or the brain.

This is not to say that interventions into these parts of the body must thereby constitute impermissible violations of the child's RBI. It is rather to say that we have reasons to be particularly mindful of interventions into parts of the body that may be construed as threatening to either impose a particular kind of practical identity upon the individual, or to undermine an identity that they are seeking to develop. The point here is that such interventions can represent a particularly salient kind of influence over the child's development of autonomy for the reasons Lewis and Holm highlight. Given the significance of a child's interest in developing autonomy, and the role of the body in their developing autonomy, such physical interventions will require particularly robust forms of justification if they are to be permissible. Such justifications will need to clearly establish how the intervention is necessary to serve other constituents of an overall best interests assessment.

Indeed, in such circumstances, other factors might plausibly become necessary for deeming the intervention to be in the child's overall best interests, even if the intervention may be construed as undermining the development of their autonomy. For instance, the permissibility of some significant interventions might plausibly require, where possible and appropriate, the assent of the child undergoing the intervention, or at least the absence of dissent, even if they lack formal decision-making capacity. The rationale for this is that both the performance of some interventions and their persisting effects can plausibly have an expressive function, conveying to the recipient (however mistakenly) that they lack moral worth in an important sense; their bodily integrity and the development of their autonomy have, after all, both been deemed to be fungible, albeit for their own broader interests. This is significant in the context of the present discussion in so far as we agree with relational theorists that self-worth is a crucial component of autonomous agency. Including a child in the decision-making process for such procedures through a process of assent can plausibly mitigate concerns about the negative impacts of the procedure on self-worth; as Sibley et al. ⁴⁷ put the point in their analysis of the importance of assent, ‘a child will come to understand he is of value by being treated as if he is of value’. This insight is, I suggest, of particular importance in cases of physical interferences that may appropriately be construed as undermining the development of the child's autonomy, even if they are justified by the claim that they promote the individual's bests interests all things considered.

Conclusion

The child's RBI raises a number of important complexities, and there are many that I have not attempted to address here. These include whether and when the child's RBI could include the power of waiver, and how the child's RBI might relate to power rights enjoyed by their parents. I have instead been interested in the narrower question of how considerations of autonomy should be incorporated into a plausible understanding of when infringements of the child's RBI might be justified. Rather than incorporating these considerations via a hypothetical consent model for permissible interference, or a best interests approach that prioritises the preservation of choice about whether to undergo particular kinds of interference, I have outlined a broader approach that affords particular salience to the moral reasons to safeguard the broader development of the child's autonomy, and suggested why this set of issues is deeply implicated with the child's RBI.

My goals here have therefore been modest; I have not attempted to explain how we should weigh these autonomy-based interests against other factors that should be incorporated into an assessment of the child's best interests, or to make concrete judgements about particular examples of contentious physical interventions. However, I hope to have somewhat clarified the nature of how autonomy-based considerations should feature in our thinking about the complex issues attending the child's RBI in particular cases.