Hostname: page-component-6898cfcb55-zsp5v Total loading time: 0 Render date: 2026-11-11T11:22:56.854Z Has data issue: false hasContentIssue false
  • English
  • Français

Autonomy, vulnerability and capacity: a philosophical appraisal of the Mental Capacity Act

Published online by Cambridge University Press:  19 February 2013

Catriona Mackenzie and
Wendy Rogers
Catriona Mackenzie*
Affiliation:
Department of Philosophy, Macquarie University, Sydney, NSW,Australia
Wendy Rogers
Affiliation:
Department of Philosophy and Australian School of Advanced Medicine, Macquarie University, Sydney, NSW, Australia
*
catriona.mackenzie@mq.edu.au
Rights & Permissions [Opens in a new window]

Abstract

The UK Mental Capacity Act provides an important legislative framework for protecting persons who are vulnerable, by virtue of partial or total lack of capacity, from abuse, coercion, exploitation, disrespect or unwarranted intrusions on their privacy and liberty. In this article we argue, however, that in order to effectively implement the Act's governing principles and promote the autonomy of persons whose capacities are impaired, health and social welfare professionals must go beyond a primarily cognitive approach to capacity assessment and refer to a range of more demanding autonomy conditions relating to authenticity, diachronic coherence and consistency, accountability to others, and affective attitudes towards oneself. We suggest that these conditions are in fact implicit in the guidelines, scenarios and practical advice contained in the Act's accompanying Code of Practice. To ground our analysis, we discuss two cases of end-of-life decision-making, and consider how these cases might have been assessed had the Mental Capacity Act been applied. We also address the question of whether the conditions for autonomy we identify raise the bar too high to be met by persons who are vulnerable because their capacity is partially compromised.

Information

Type
Articles
Information
International Journal of Law in Context , Volume 9 , Special Issue 1: Mental capacity and value neutrality , March 2013, pp. 37 - 52
Copyright
Copyright © Cambridge University Press 2013

Introduction

Recent capacity legislation, such as the UK Mental Capacity Act 2005 (MCA), aims to strike a balance between two competing principles in medical ethics: respect for (patient) autonomy, and the best interests principle. The MCA balances these competing principles by making a default presumption in favour of individual autonomy and then invoking the best interests principle in cases involving persons who do not have the capacities required to make decisions for themselves. This order of priority is explicitly reflected in the five governing principles of the Act:

Box 1

Key principles of the MCA described in the Mental Capacity Act 2005 Code of Practice (hereafter the Code: Department for Constitutional Affairs 2007, p. 20).

  1. 1. Every adult has the right to make his or her own decisions and must be assumed to have capacity to make them unless it is proved otherwise.

  2. 2. A person must be given all practicable help before anyone treats them as not being able to make their own decisions.

  3. 3. Just because an individual makes what might be seen as an unwise decision, they should not be treated as lacking capacity to make that decision.

  4. 4. Anything done or any decision made on behalf of a person who lacks capacity must be done in their best interests.

  5. 5. Anything done for or on behalf of a person who lacks capacity should be the least restrictive of their basic rights and freedoms.

Principle 1, the key principle, is that of respect for autonomy. Principle 4 is the best interests principle, while Principle 5 invokes human rights principles as constraints on how to interpret and implement the best interests principle, seeking to protect persons who lack the capacity to make their own decisions from abuse, coercion, exploitation, disrespect or unwarranted intrusions on their privacy and liberty. The MCA thus recognises that those who lack capacity or whose capacity is partially compromised are vulnerable to such harms, and to paternalism and stereotyping. Its aim is to promote the autonomy of vulnerable persons whenever possible and, when not, to provide protections under the best interests principle.

We regard Principles 1, 4 and 5 as providing important legislative rights protections both for persons who are capable of making self-determining decisions and for those who are not. But our discussion in this article will focus primarily on how the principle of respect for autonomy (Principles 1–3) is understood within the MCA, rather than the best interests principle (Principles 4 and 5). We think there are some tensions and problems with the way that autonomy is understood within the MCA and the accompanying Code. In particular, we will argue that while the MCA adopts a primarily cognitive, procedural approach to assessment of capacity, and hence to assessment of autonomy competence, in practice this approach to capacity assessment needs to be underpinned by a range of thicker, more demanding autonomy conditions.Footnote 1 We suggest that these conditions are in fact implicit in the guidelines contained in the Code concerning how health and welfare professionals should implement the Act. These conditions raise questions about the plausibility of the MCA's attempt to bracket the values of both patients and health professionals in capacity assessments.

In Part I, we outline the MCA's approach to capacity assessment. We then explain why we think any such primarily cognitive, procedural approach to autonomy must be underpinned by more demanding autonomy conditions and show how these conditions are implicit in the Code.Footnote 2 In Part II, we discuss two cases of end-of-life decision-making, analysing how these cases might have been assessed had the MCA been applied. We show why, in both cases, the application of capacity tests within the MCA would require reference to conditions for autonomy that are more demanding than those apparently specified by the MCA. We also address the question of whether these autonomy conditions raise the bar too high to be met by persons who are vulnerable because their capacity is partially compromised.

I Capacity and the conditions for autonomy

The MCA understands capacity in terms of the ability to make a specific decision at a particular time. Such decisions extend from simple day-to-day decisions – deciding what to wear, what to eat, whether to go on an organised outing – to complex and important life decisions, for example concerning refusal of treatment or bequeathing property. Principle 1 expresses the idea that respect for autonomy means that a person should be presumed to have the right and the authority to make such decisions of practical importance to her life, for her own reasons, whatever those reasons might be. Principle 3 expresses the anti-paternalist idea that others are obliged to respect our autonomous decisions, even if we do not always exercise our autonomy wisely.

The presumption of capacity means that capacity must be assessed as deficient before the protections of the MCA are triggered. The MCA (s. 2(1)) defines lack of capacity in the following way:

‘For the purposes of this Act, a person lacks capacity in relation to a matter if at the material time he is unable to make a decision for himself in relation to the matter because of an impairment of, or a disturbance in the functioning of, the mind or brain.’ (cited in the Code, p. 42)

The Act sets out a two-stage test for assessing capacity. Stage 1 asks whether the person has an impairment or disturbance in the functioning of their brain or mind, while Stage 2 enquires whether this impairment or disturbance prevents the person from making the specific decision in question. Therefore, to fall under the Act, there must be proof that the person in question does have some kind of disability in their mental functions, together with a further assessment that this impairment impedes them from making the specific decision in question. The MCA uses a four-part procedural test to assess whether or not a person can make the decision. The first three factors are cognitive and relate to capacities to understand and use information, whilst the fourth relates to communication (see Box 2). The test is thus procedural because it stipulates the standards of information processing that must be met in order to judge a person as having capacity, but brackets the content of the decision from the assessment process.

Box 2

Inability to make decisions (MCA, s. 3).

  1. (1) For the purposes of section 2, a person is unable to make a decision for himself if he is unable—

    1. a) to understand the information relevant to the decision,

    2. b) to retain that information,

    3. c) to use or weigh that information as part of the process of making the decision, or

    4. d) to communicate his decision (whether by talking, using sign language or any other means).

Importantly, both the MCA and the Code acknowledge that capacity, and lack of capacity, are not always clear-cut but a matter of domain and degree. Principles 2 and 3, in particular, aim to provide further direction regarding how the principle of respect for autonomy should be interpreted and implemented with respect to persons whose capacities may be impaired in some domains or to some degree; that is, persons who are neither clearly competent nor clearly lacking in capacity. These principles are explained and elaborated at length in the Code through numerous examples, which seek to show how a person may have capacity to make decisions in some domains, for example with regard to simple day-to-day decisions, yet be vulnerable to harm or exploitation when it comes to other decisions, such as those regarding financial investments. Principles 2 and 3 thus seek to strike a balance between, on the one hand, a strong anti-paternalist stance and, on the other, recognition that persons whose capacity is partially impaired are vulnerable to specific harms and may require specific forms of assistance and protection.

At this point it is worth exploring in a little more detail what vulnerability means and how the concept of vulnerability operates in the MCA. Vulnerability is usually defined as being at increased risk of harm and/or having diminished capacity to protect oneself from harm or to safeguard one's own interests.Footnote 3 Because vulnerability can be due to a range of different factors, it is useful to distinguish inherent and situational sources of vulnerability. Inherent vulnerability refers to vulnerability that is intrinsic to the person and arises from specific characteristics, such as disability, mental illness, mental incapacity or frail old age. Situational vulnerability refers to vulnerability that arises from the personal, social, political, economic or environmental situation of a person or social group.Footnote 4 Many sources of situational vulnerability are social and relational: what makes a person vulnerable is her social situation or interpersonal relationships, which may involve abuse, coercion, domination or exploitation. Although these two kinds of vulnerability are distinct, they often overlap because persons who are inherently vulnerable are frequently in social circumstances or interpersonal relationships that subject them to wrongful harms by others, thereby compounding their initial vulnerability.

Within the UK, the treatment and protection of vulnerable persons or groups has been a major concern of social policy.Footnote 5 Recent reports and legislation use various descriptors to identify the vulnerable. For example the No Secrets report (Department of Health, 2000) defines vulnerable adults as individuals in need of community care services due to a mental or other disability, age or illness; and who are unable to take care of themselves or protect themselves against serious harm or exploitation (cited on p. 246 of the Code). There is potential for considerable overlap between those falling under protections such as the Safeguarding Vulnerable Groups Act 2006 (SVGA) and the MCA, as the presence of a mental impairment that affects capacity (thereby triggering the MCA) will affect many of the people to whom the SVGA definition of ‘vulnerable adult’ applies.Footnote 6 The focus of the MCA, however, is protecting those who are vulnerable due to partial or total lack of capacity.

As well as protecting those who are vulnerable through partial or total lack of capacity, Section 10 of the Code identifies a further subgroup who are described as ‘particularly vulnerable’. In order to protect the interests of the particularly vulnerable, the MCA mandates the appointment of an Independent Mental Capacity Advocate (IMCA):

‘The purpose of the IMCA service is to help particularly vulnerable people who lack the capacity to make important decisions about serious medical treatment and changes of accommodation, and who have no family or friends that it would be appropriate to consult about those decisions.’ (Department for Constitutional Affairs, 2007, p. 178)

The extra layer of protection provided by IMCAs is triggered by the absence of support people (family or friends) who might act as advocates for the person who lacks capacity.

The Act thus identifies three distinct pathways for decision-making with persons who are vulnerable because they lack capacity or have partial capacity. The first pathway is to support autonomous decision-making wherever possible for those whose capacities are comprised. The second pathway is to use a best interests standard, which requires that any decision made on behalf of a person who lacks capacity is in the person's best interest and is in line as much as possible with what is known about his or her beliefs and values, according to the advice of the person's significant others. The third pathway is only taken in cases where a person lacks capacity and has no significant others. In such cases, the MCA deems the person particularly vulnerable, and provides for an independent advocate to protect their interests.

We think that much of the practical advice provided in the Code concerning how to promote the autonomy of vulnerable persons is laudable and valuable. However, we think that this advice exposes some tensions between the MCA's procedural and cognitive approach to capacity assessment and its aim of promoting the autonomy of persons who are vulnerable because their capacities are compromised. For the remainder of this part we will argue that this is because determining whether or not a person's decision is autonomous is more complicated than determining whether or not the decision meets the cognitive test set out in the Act. It also requires reference to a range of more substantive conditions relating to authenticity, diachronic coherence and consistency, accountability to others, and affective attitudes towards oneself.Footnote 7

The philosophical literature typically distinguishes different kinds of conditions necessary for autonomy: freedom conditions, competence conditions and authenticity conditions. These conditions reflect different dimensions of the concept of autonomy, in particular autonomy as self-determination and autonomy as self-governance. To be self-determining is to be able to determine one's own beliefs, values, goals and wants, and to make choices regarding matters of practical import to one's life free from undue interference. The obverse of self-determination is determination by other persons, or by external forces or constraints. Freedom conditions are thus especially relevant to autonomy as self-determination: coercion, manipulation and threats; physical intimidation and violence; social and political repression; and paternalistic interference by others or the state, are all threats to the exercise of self-determination.Footnote 8 Principle 1, and the strong anti-paternalist stance of the MCA, aim to protect this dimension of autonomy; that is, to protect persons who may be vulnerable by virtue of partial capacity from having their decisions overridden by health professionals, social and community workers or family members.

To be self-governing is to be able to make choices and enact decisions that express, or are consistent with, one's values, beliefs and commitments. Whereas the threats to self-determination are typically external, the threats to self-governance are typically internal, and often involve volitional or cognitive failings.Footnote 9 Weakness of will and failures of self-control are common volitional failings that interfere with self-governance. Failures of understanding for whatever cause, or inability to use information to make decisions, can constitute cognitive threats to self-governance. Compulsions, phobias and neuroses, addictions, and delusions and other forms of psychosis are less common, but more severe, volitional and cognitive threats to self-governance. Such disorders and impairments can undermine a person's capacity to make and enact decisions in line with her values and commitments, and sometimes even to form a coherent set of values, beliefs and commitments.

The notion of self-governance implies both competence and authenticity conditions.Footnote 10 Competence conditions specify the range of competences a person must possess in order to be self-governing. There is extensive debate among theorists of autonomy concerning whether the competence conditions for autonomy should be specified minimally or more stringently. The MCA's test for capacity (see Box 2 above) specifies minimal cognitive competence conditions, in order to be as inclusive as possible in identifying those persons who should be presumed to be autonomous and to have the authority to make self-determining decisions. In doing so, the MCA effectively understands autonomy as equivalent to informed consent.Footnote 11 However, the Code also appeals to authenticity and other conditions that imply a more demanding conception of autonomy.

Authenticity conditions stipulate that, to count as autonomous, a person's decisions, values, beliefs and commitments must be her ‘own’ in some relevant sense; that is, she must identify herself with them and they must cohere with her ‘practical identity’, her sense of who she is and what matters to her.Footnote 12 Actions or decisions that a person feels were foisted on her, which do not cohere with her sense of herself, or from which she feels alienated, are not autonomous. But what are the processes involved in determining whether or not a person's decisions are authentic or coherent with her practical identity, and do genuinely express her values, beliefs and commitments? In answering this question, many theorists of autonomy appeal to a reflective endorsement test of some kind: a decision is authentically a person's own (and hence she is autonomous with respect to that decision) if, upon reflection in the light of the relevant information and in the light of her values, beliefs and commitments, she endorses that decision.Footnote 13

The MCA clearly appeals to authenticity conditions for autonomous decision-making. This is why Principle 3 and the accompanying Code stress that the important consideration is whether a person's decision is in line with her own values and beliefs, regardless of whether others may think that decision unwise. It thus imposes two requirements on health professionals and other care workers: to respect the values, beliefs and decisions – whatever their content – of those in their care, and to bracket their own values and beliefs in their interactions with patients or clients.

These requirements aim to prevent the use of capacity assessments to override decisions of clients or patients, when these decisions clash with the values of health professionals and other care workers. As such, they provide important safeguards against unwarranted paternalistic interference, two examples of which we discuss below, in Part II. But they also raise numerous complex questions, particularly in cases involving vulnerable persons, including: What are the processes involved in ascertaining whether a patient's values and beliefs are authentically her own, and that her decision is reflectively endorsed in the light of these values and beliefs and the available information? Are these values and beliefs always evident to the patient beforehand, or do they emerge during the process of decision-making? And to what extent can capacity be assessed independently of the content of the patient's decision or of the values and beliefs that inform it? These questions are clearly recognised within the Code as issues that need to be addressed by health professionals and others providing care to persons who are vulnerable by virtue of partial capacity. In providing advice about how best to respond to these issues, the Code appeals to further conditions for autonomy, which are in some tension with the cognitive approach to capacity assessment explicitly adopted in the MCA.

First, the Code appeals to conditions relating to diachronic consistency and coherence, which are in tension with the MCA's synchronic approach to capacity assessment. Synchronic approaches to autonomy analyse the necessary and sufficient conditions for a particular decision or action to be autonomous. However, synchronic conceptions of autonomy have been criticised as inadequate, we think rightly, because they fail to account for the temporally extended, or diachronic, nature of human agency.Footnote 14 Values, beliefs and commitments are inherently diachronic; that is, they are manifested in a person's patterns of action over time. It is therefore not possible to determine whether a decision is authentically a person's own – that is, consistent with her practical identity, and in line with her values, beliefs and commitments – by considering only the decision at the point of time at which it is made. Rather, the decision must be considered within the broader context of the person's history. Now on the one hand, the Code seems consistent with the MCA's synchronic approach to capacity and autonomy assessment because throughout the Code it is repeatedly stressed that ‘a person's capacity (or lack of capacity) refers specifically to their capacity to make a decision at the time it needs to be made’ (Code, p. 19).Footnote 15 On the other hand, the Code elsewhere adopts a diachronic approach to capacity and autonomy assessment, making it clear that to determine whether a decision is authentically a person's own, social welfare and health professionals and family members need to determine whether the decision is in line with the person's long-standing values, beliefs and commitments.Footnote 16 This is why the first and second decision-making pathways advocated by the MCA (supporting autonomous decision-making whenever possible; and when not, using a best interests standard) involve, among other things, seeking the advice of a person's family and friends about the person's long-standing beliefs and values.Footnote 17

Second, in requiring that all practicable help should be given to a person before they are treated as lacking capacity, Principle 2 and the accompanying advice in the Code concerning assistance with decision-making (see Chapter 3), adopt a dialogical approach to autonomous decision-making that is in tension with the MCA's primarily cognitive approach to capacity assessment. Dialogical approaches emphasise the relational and social dimensions of autonomous decision-making.Footnote 18 These social dimensions of autonomy involve several interconnected aspects. First, that we develop our values, beliefs and commitments, and make decisions in ongoing interaction with others. Authenticity conditions are often understood in quite individualistic terms, such that the involvement of others in the processes of value-formation and decision-making are seen as threats to autonomy. In contrast, dialogical approaches regard ongoing interaction with others as integral not only to the processes of value-formation and decision-making but also to the formation of our practical identities. Social and dialogical approaches recognise, of course, that some kinds of social interaction can thwart autonomous value-formation and decision-making.Footnote 19 But the corrective is not to think that we can only be autonomous if we are freed from social influence. It is rather to promote social relationships that foster autonomy. Principle 2 and the Code implicitly appeal to a dialogical approach to autonomy, in recognising that vulnerable persons or those with partial capacity will only be able to make autonomous decisions with extensive social scaffolding and support. A dialogical approach is also implicit in many of the examples provided in the Code as to how health professionals and other care workers can assist persons with a range of different capacity impairments to understand the information relevant to the decision, weigh that information in the decision-making process, and communicate their decision.Footnote 20

Next, a central claim of social or dialogical approaches is that as social agents we are subject to others' demands that we account for our decisions, explain the reasons for them, and take responsibility for those decisions. We refer to these requirements as ‘accountability conditions’ for autonomy.Footnote 21 We want to suggest that to be autonomous a person must be capable of both understanding and responding to this social demand to be accountable to others. Being accountable does not mean that those others to whom we are accountable must agree with our reasons and decisions. But accountability does require that we are responsive to others' requests for explanations of our decisions, that we can make our reasons comprehensible to others, and that we are prepared to take responsibility for our decisions and their consequences. It also requires, in turn, that we are able to hold others to account for their decisions and actions. The MCA does not explicitly recognise the importance of accountability conditions for autonomy. However, we would argue that unless accountability conditions are built into the interpretation and implementation of Principle 3 (just because an individual makes what might be seen as an unwise decision …), the principle of respect for autonomy simply reduces to a consumerist model of medical decision-making, in which the role of the health professional is to satisfy the preferences of the patient, no matter how misguided or ill-informed these may be. Accountability conditions also put considerable onus on health professionals, in turn, to explain and discuss treatment options and recommendations with patients, and to try to understand and clarify the patient's reasons.Footnote 22

We acknowledge that accountability conditions seem to raise the bar for autonomy quite high and may seem unrealistic for persons who are vulnerable by virtue of partial capacity. Explaining and giving reasons for one's decisions in the light of one's values, and asking others to account for their decisions, seems to require not only high levels of capacity and articulacy, but also self-confidence and trust in one's own judgment. In other words, it seems to require a certain kind of social standing – that others treat one as capable and worthy of respect, and that one regards oneself as capable and worthy of respect.Footnote 23 Yet persons who are vulnerable because they have impaired mental functioning often lack such social standing, because they are not treated by others, and may not regard themselves, as having the status of autonomous agents. In response, however, we would argue that this is precisely the problem that Principle 2 and the advice in the Code aim to address: requiring that people are treated as having capacity, whenever possible, and given all practicable help to make and communicate their decisions, is a way of restoring their status as autonomous agents. In other words, we are suggesting that, despite the MCA's explicitly procedural approach to capacity assessment, Principle 2 and the Code are actually underpinned by a dialogical approach which implicitly recognises the status dimensions of autonomy, and which aims to find practical mechanisms to enact accountability conditions in cases involving persons who are vulnerable by virtue of mental impairment.

To sum up our argument so far, in this part we have tried to show that a primarily cognitive approach to capacity assessment, such as that adopted in the MCA, must be underpinned by more demanding autonomy conditions, in particular conditions relating to authenticity, diachronic coherence and consistency, accountability to others, and affective attitudes towards oneself. In the following part, we try to make this somewhat abstract analysis more concrete by showing the relevance of these conditions in two cases of end-of-life decision-making.

II Vulnerability, capacity and end-of-life decision-making

Case 1: A challenge to capacity based upon conflicting values

In a landmark UK High Court decision in 2002 (prior to the introduction of the MCA), the presiding judge, Dame Elizabeth Butler-Sloss, decided in favour of an appellant, Ms B, who had brought a suit of unlawful trespass against an NHS hospital.Footnote 24 In what follows we show how this case highlights both the relevance and the limitations of the MCA's cognitive test of capacity. We also suggest that Ms B's testimony and Dame Butler-Sloss's judgment implicitly appeal to the more demanding autonomy conditions discussed above, hence demonstrating their importance in implementing Principles 1–3 of the MCA.

Ms B was a forty-three-year-old woman who had suffered a cervical spine cavernoma, which left her a ventilator-dependent quadriplegic. She underwent neurological surgery, which resulted in only a minimal improvement of her condition, enabling her to move her neck slightly. Disappointed by the failure of the surgery to result in any substantial improvement, she requested that the ventilator be switched off, in accordance with a Living Will she had executed eighteen months earlier upon suffering her first spinal haemorrhage.

To determine whether Ms B was competent to make the decision to have ventilation removed, she underwent three psychiatric assessments within a period of two weeks, conducted by three different psychiatrists. She was judged competent by each assessment. After the last assessment, preparations were made to switch off the ventilator. However, the day after the final assessment, the assessing psychiatrist amended her finding and judged Ms B not competent. One of the other psychiatrists then amended his decision, the ventilator was not switched off and Ms B was prescribed anti-depressants. Two months later she was reassessed by the first psychiatrist, who found her competent. Between that time and the time of the hearing, seven months later, Ms B refused to participate in a weaning programme, to see if her dependency on ventilation could gradually be reduced, and continued to request that the ventilator be removed. Despite the fact that the hospital and the treating doctors judged her to be competent during that whole period, and despite a subsequent assessment before the hearing affirming this finding, the case put by the hospital trust was that she was not competent.

After assessing the medical evidence, the psychiatric assessments and the evidence of the patient, Dame Butler-Sloss found that Ms B was not merely competent but that she was ‘a most impressive witness’. From the evidence outlined in the court transcripts, including her testimony, Ms B would clearly have met the MCA's cognitive test for capacity. Contra the hospital trust's claim, there was no reason to think that she had an impairment or disturbance in the functioning of her mind or brain, which would impair her capacity to make the specific decision to switch off the ventilator. In terms of the information test, her testimony shows that she was more than sufficiently rational and emotionally balanced to understand the nature of her medical condition and her prognosis. She had informed herself about the available treatment options, their potential side effects, foreseeable risks and probabilities of success. Thus, she understood the information relevant to the decision, and was able to retain and use it as part of the decision-making process, and to communicate her decision to the treating health professionals.

Given that Ms B met the conditions for capacity outlined by the MCA, there seems to be good reason to be hopeful that, had the Act been in force at the time, it would have provided her with protection against the refusal by the hospital to accede to her request for ventilation to be withdrawn. Nevertheless, this case shows how the two-stage capacity test and the information test might be used to circumvent Principle 3 of the MCA and to challenge a patient's capacity if a patient rejects the proposed treatment. For the case put by the hospital to substantiate the claim that Ms B lacked capacity was, first, that she was suffering from an impairment of mind or brain, evidence for which was that she had been prescribed anti-depressants (corresponding to Stage 1 of the test), and second, that because she had not experienced weaning, her decision to refuse weaning was not based on sufficient understanding of the available information (corresponding to Stage 2 of the test).

The presiding judge dismissed these arguments, suggesting that what was really at issue was a clash of values between the patient and some of the medical staff, who felt deeply that to accede to Ms B's request would conflict with the values embedded in their professional roles and self-conceptions, as involving responsibilities to save life rather than end it. Principle 3 of the MCA (that just because an individual's decision seems unwise [to their family, to their medical carers], that person should not be treated as lacking in capacity) echoes the judge's comments and speaks directly to cases analogous to that of Ms B.

In her judgment, Dame Butler-Sloss appealed primarily to the principle of respect for autonomy, arguing that the right to self-determination with respect to one's life and one's bodily integrity trumps other considerations, including the best interests principle and the principle of sanctity of life. What is particularly interesting about the judgment, for the purposes of our argument in this article, is that in determining that Ms B had capacity and hence the right to refuse treatment, the judge drew extensively on Ms B's testimony, which implicitly appealed to a number of the more demanding conditions for autonomy discussed in Part I. In bringing a suit of unlawful trespass against the hospital, Ms B was of course appealing to a right to self-determination, arguing that the hospital's refusal to accede to her refusal of treatment was coercive. Her testimony is explicit on this: ‘I felt that my path was being blocked and I was being pressurized to accept this option [weaning]’.Footnote 25 But her testimony also aims to show that her decision is authentically self-governing because it has survived a reflective endorsement test, even though in some respects the decision seems to violate conditions of diachronic coherence and consistency because it conflicts with her lifelong faith as a Christian. She says: ‘I have questioned myself about this and it has challenged my integrity. It has been a very difficult process to rationalise what I am doing in the context of my faith but … I have come to believe that people die and become disabled and God does not always intervene. It has also been difficult for me to contemplate leaving the people I love behind.’Footnote 26 This and other statements in her testimony explaining her reasons for refusing the weaning programme also attest to the importance of accountability conditions for autonomy. It is evident that Ms B entered into extensive and ongoing discussions with her clinicians, and that she understood and accepted the social demand to be accountable to others for her decision to refuse treatment. Her frustration that the hospital and the assessing psychiatrists did not reciprocate this accountability requirement is evident in her testimony. By taking the hospital trust to court, Ms B was demanding the accountability that she felt was her due, thereby demonstrating that she had a strong sense of trust in her own judgment and in her social standing as a person who should be treated with respect. This is evident in the following statement from her testimony: ‘I felt that my rights were being eroded and that is not something I tolerate really; it is not within my character to go along with that.’Footnote 27

Ms B was clearly an educated, intelligent and articulate woman. But her obvious competence raises the question of whether the more demanding conditions for autonomy we have identified do indeed raise the bar too high. To address this question, in what follows we consider another case of refusal of treatment in an end-of-life case, this time involving a vulnerable patient with impaired capacity. We argue that what this case shows is how implementing the principles of the MCA in accordance with these more demanding conditions, in particular the dialogical condition, may actually provide such patients with greater protection from unwarranted paternalism.

Case 2: A challenge to capacity based on the presence of brain impairmentFootnote 28

Charlie Davenport was a fifty-nine-year-old man with severe chronic obstructive airways disease (COAD). He had an intellectual disability precluding employment, but was able to live alone in the community with regular visits from a disability support officer. He had no known close family or friends and his financial affairs were managed by a public trustee. He was admitted to hospital when his COAD became worse. No formal consent to treatment was sought at this stage, but Mr Davenport co-operated with treatment including an IV cannula and oxygen via a face mask. He continued to deteriorate, such that the next level of treatment required admission to the Intensive Care Unit (ICU). At this point, Mr Davenport was asked if he consented to be moved to the ICU, and he withheld consent. This prompted an assessment of his capacity, following which it was determined that he lacked capacity to make this decision to refuse potentially life-saving treatment. He was then transferred to ICU where he was treated for five days. After making a recovery, Mr Davenport returned to the ward. Again he co-operated with treatment, but after a few days he deteriorated and his doctors recommend readmission to ICU. Once more Mr Davenport refused, once more his capacity was assessed and found wanting, and he was again transferred against his will to the ICU.

The treating doctors used a cognitive test for capacity similar to that in the MCA. It was clear that Mr Davenport had a global mental deficit, but it was by no means clear that he lacked capacity to make medical decisions. The doctors who initially treated him presumed that he had capacity to consent to that treatment (although no formal consent was obtained); it was only when his treatment preferences conflicted with medical opinion that his capacity was challenged. It may be that his deteriorating medical condition, which triggered the ICU admission, also affected his capacity to make a decision; or perhaps, like Ms B, it was his refusal of medical advice that initiated the challenge to his capacity. It is difficult to maintain that he was capable of making one set of medical decisions (i.e. accepting ward treatment) but not another (i.e. rejecting ICU treatment).Footnote 29 We do not know which part of the information test Mr Davenport was deemed to have failed; certainly communication was not a barrier in this case. Possibly his treating doctors felt that he did not fully understand the information (that he might die if not transferred to the ICU) or that he did not weigh or use that information in making his decision. Either of these are plausible; however, by the time it came to Mr Davenport's second ICU admission, he had experienced the treatment, and arguably, his refusal was well-informed. One of the grounds for challenging Ms B's capacity was that she had not experienced the weaning programme, leading to the claim that her treatment refusal was not based on sufficient understanding of the available information. This condition did not apply to Mr Davenport's second refusal, as he knew just what ICU admission involved. Like Ms B, this example suggests that as long as there is concordance between the values of the patient and the values of the treating doctors in terms of the aims of therapy, capacity is often presumed, even in patients who might otherwise be thought to have borderline capacity.

Despite the apparent similarity between the capacity test used in this case and that in the MCA, we suggest that the MCA would have offered greater protections, as an adequate application of Principles 1–3 must rely upon more demanding autonomy conditions than suggested by its cognitive approach to capacity assessment.

First, under Principles 1 and 2 of the MCA, those treating Mr Davenport would have been duty bound to try to ascertain his values and beliefs in relation to treatment, and to provide all possible support to him in his decision-making. These steps should have occurred at the time of his admission, rather than at the time that he refused transfer to the ICU. In order to determine whether or not his decisions (first, to accept treatment on the ward, second, to refuse ICU admission) were authentically his own (Principle 1), those assessing him would have had to draw upon Mr Davenport's own testimony, his known history of decision-making in similar circumstances, and any evidence or information from those who knew him well. This step, assessing the authenticity of his decisions, would have had to occur before or in tandem with the capacity test, otherwise Principle 1 would be violated. Further, as argued above in relation to Ms B, this process is diachronic: despite the demand that capacity must be determined for this decision at this time, we cannot make sense of the assessment unless we place that decision within the context of what is known about the patient and his beliefs and values.

Second, Principle 2 requires providing support; such support recognises that we often do not make decisions in isolation, but come to discover our own views through exploring and testing these with others (i.e. the dialogical model). In Mr Davenport's case, lacking close family or friends and being faced with a significant medical decision, he may have been deemed particularly vulnerable and appointed an IMCA to work with or speak for him during the capacity assessment. The problem with this is that it may be a catch-22: an IMCA is appointed only to those who lack capacity, and therefore cannot support such persons to make autonomous decisions; and if the person can make an autonomous decision, then they are not seen to need the services of an IMCA. However, this may not be a practical problem, as IMCAs are charged with determining to the extent possible the person's past and present wishes, and using these to guide decisions (Department for Constitutional Affairs, 2007, p. 186). Thus, whether the advocate deems that the person is capable of making an autonomous decision, or whether they deem the person incapable yet are guided by an investigation of that person's wishes, this clearly dialogical approach extends the reach of the affected person's autonomy to the greatest extent possible. In order to come to a robust conclusion about the authenticity of Mr Davenport's decision, the advocate would have had to talk this decision through with him and interpret it in the light of other information available about Mr Davenport's views and beliefs. This dialogical process is particularly important when the person in question has an intellectual disability that might hinder their consideration of all of the relevant information. A decision that has been talked through, that has been revisited over time, and that has remained diachronically constant can be taken as more indicative of the person's values than one declared with no evidence of detailed consideration or consistency over time.

Third, Principle 3 precludes challenging capacity based solely on the outcome of the decision. The outcome of the decision, judged unwise or not, should not have been taken into account, so long as the other criteria were met; that is, abiding by Principle 3 would have avoided the apparent paternalism and double standards of the treating team.

The provision of support for people with borderline capacity thus underlines the importance of the authenticity, diachronic coherence and accountability conditions for autonomy. Such support enables individuals to think through and explain their reasons, and thereby to own them in a more robust way than a decision taken in isolation. The accountability condition requires taking responsibility for decisions by demonstrating that one's reasons for holding a particular view relate to one's values. This is a high standard to meet, especially for people with impaired mental function, who may have difficulty articulating the connections between their values and their reasons. Nonetheless, as examples from the Code indicate (see, for example, Department for Constitutional Affairs, 2007, p. 60), this is the aim of providing support, since it is not until people have explained themselves to some extent that we can distinguish between capable but unwise decisions, and identical decisions based upon misunderstanding of relevant information.

This social dimension of autonomy, being able to discuss and test one's provisional decisions before owning them, is relevant to the treatment in the MCA of the particularly vulnerable, as one of the criteria for being identified as such is the absence of significant others who can speak with and for the individual in question. This provision applies only to those who lack capacity, rather than those with partial capacity, but, as discussed above, it is possible that the provision of an IMCA will in some cases help a person to execute their own decision. At the very least, even for those who have permanently lost capacity, IMCAs, in trying to determine the person's previous and current wishes, are extending the reach of that person's autonomy to the greatest extent possible.

In summary, it is unlikely that we can protect Mr Davenport's autonomy without reference to the conditions we have specified, since without providing support, or suspending dismissal of unwise decisions, we will have an impoverished understanding of his capacity. When a person has a mental disability, we need to invest more effort into understanding their values and decisions; we need to investigate aspects of the accountability, diachronic consistency and authenticity conditions for autonomy, using whatever resources will help to achieve this. Even so, there will be difficult and borderline cases, but unless we embrace a more demanding conception of autonomy, people like Mr Davenport are likely to be found lacking in capacity more often than is warranted.

Conclusion

The MCA provides an important legislative framework for protecting persons who are vulnerable, by virtue of partial or total lack of capacity, from abuse, coercion, exploitation, disrespect or unwarranted intrusions on their privacy and liberty. We have argued, however, that in order to effectively implement its governing principles and promote the autonomy of persons whose capacities are impaired, health and social welfare professionals must go beyond a primarily cognitive approach to capacity assessment and refer to a range of more demanding autonomy conditions relating to authenticity, diachronic coherence and consistency, accountability to others, and affective attitudes towards oneself. We have suggested that these conditions are in fact implicit in the guidelines, scenarios and practical advice contained in the Code. In response to the concern that these conditions may seem to raise the bar for autonomous decision-making too high, we wish to emphasise that autonomy is a matter of domain and degree. As our discussion of the case of Mr Davenport shows, considerable effort and support may be required to assist patients with lower levels of capacity to make autonomous decisions and to ascertain that these decisions do indeed cohere with the person's values and practical identity. But just because a person may need extensive assistance to make autonomous decisions, this does not mean that less demanding standards for autonomy assessment should apply in their case. Doing so opens the door to unwarranted paternalism and disrespect, which is precisely the problem the MCA is seeking to address. We acknowledge, however, that applying the more demanding autonomy conditions we have outlined in this article, and that we have suggested are implicit in the Code, may be difficult in crisis situations and does put an extra burden on welfare and health professionals. Proper implementation of the MCA therefore requires these professionals to be properly supported and adequately resourced.Footnote 30

Footnotes

1 In referring to the MCA's capacity test as ‘cognitive’ and ‘procedural’, we mean first that it focuses on a person's cognitive capacity to understand information and use that information in making a decision, and second, that capacity assessment is based solely on whether or not the person satisfies the cognitive test. In other words, the content of the person's decision, and the preferences, beliefs and values that inform it, are expressly bracketed from the capacity assessment.

2 As we discuss in more detail in ftn 7 below, procedural approaches to autonomy can be more or less demanding in their specification of the conditions necessary for autonomy. The MCA's primarily cognitive test for capacity is on the less demanding end of the spectrum. Procedural theories on the more demanding end of the spectrum (e.g. Christman, Reference Christman2009) include some of the conditions we discuss below, in particular authenticity and diachronic coherence and consistency.

3 Goodin (Reference Goodin1985), for example, defines vulnerability as susceptibility to harm or threats to one's interests. Although everyone is potentially vulnerable to such threats, what makes some persons or groups especially vulnerable is that they lack or have diminished capacity to protect themselves. Vulnerable persons then are those with reduced capacity, power or control to protect their interests relative to other persons.

4 For a related distinction, see Dunn, Clare and Holland (Reference Dunn, Clare and Holland2008). In Rogers, Mackenzie and Dodds (Reference Rogers, Mackenzie and Dodds2012), we make a tripartite distinction between inherent, situational and pathogenic vulnerability. Vulnerabilities are pathogenic, in our view, when they arise either from conditions that are unjust in various ways, or when social policy interventions that aim to ameliorate inherent or situational vulnerability have the paradoxical effect of increasing vulnerability. A key feature of pathogenic vulnerability is the way that it exacerbates the sense of powerlessness engendered by vulnerability.

5 See, for example, the reports No Secrets (Department of Health, 2000) and In Safe Hands (National Assembly for Wales, 2000), both of which are concerned with the care of vulnerable adults, who are defined as people aged eighteen and over who:

  • • need community care services due to a mental disability, other disability, age or illness, and

  • • may be unable to take care of themselves or protect themselves against serious harm or exploitation (cited on p. 246 of the Code).

6 The Safeguarding Vulnerable Groups Act 2006 defines vulnerable adults at s. 59 as follows:

  1. (1) A person is a vulnerable adult if he has attained the age of 18 and—

    1. (a) he is in residential accommodation,

    2. (b) he is in sheltered housing,

    3. (c) he receives domiciliary care,

    4. (d) he receives any form of health care,

    5. (e) he is detained in lawful custody,

    6. (f) he is by virtue of an order of a court under supervision by a person exercising functions for the purposes of Part 1 of the Criminal Justice and Court Services Act 2000 (c. 43),

    7. (g) he receives a welfare service of a prescribed description,

    8. (h) he receives any service or participates in any activity provided specifically for persons who fall within subsection (9),

    9. (i) payments are made to him (or to another on his behalf) in pursuance of arrangements under section 57 of the Health and Social Care Act 2001 (c. 15), or

    10. (j) he requires assistance in the conduct of his own affairs.

      Dunn et al. (Reference Dunn, Clare and Holland2008) are critical of over-reliance on these kinds of external objective descriptors, which in their view equate vulnerability with risk and fail to give due weight to whether or not persons defined as vulnerable according to these descriptors in fact subjectively experience themselves as being vulnerable. They urge the need for a more holistic approach to vulnerability assessment, incorporating both objective ‘risk’ factors and the person's subjective, experiential perspective on their situation. Their concern is that recent public policy and legislative developments err too much on the side of ‘protection’ and intervention and not enough on empowerment. We are sympathetic to this concern as well as to the more holistic approach to vulnerability assessment they advocate.

7 As we acknowledge in ftn 2 above, procedural theories on the more demanding end of the spectrum incorporate both authenticity and diachronic coherence conditions, while preserving content-neutrality. The two additional conditions we stipulate – relating to accountability and self-affective attitudes – are more substantive in specifying that autonomy is not just a capacity concept but also a status concept, requiring that one stands in certain kinds of social relations to others, and holds certain kinds of affective attitudes towards oneself. For a more detailed overview of the distinction between procedural and substantive conditions for autonomy, see Mackenzie and Stoljar (Reference Mackenzie, Stoljar, Mackenzie and Stoljar2000b).

8 For more detailed discussion, see Dworkin (Reference Dworkin1988).

9 These two dimensions of autonomy are of course interconnected in complex ways. Constraints on a person's freedom to exercise self-determination can also impair the development and exercise of capacities for self-governance. Conversely, if a person is unable to govern herself she is more likely to be subject to the will of others.

10 For further discussion of the distinction between competence conditions and authenticity conditions, see Christman (Reference Christman, Christman and Anderson2005; Reference Christman2009).

11 In medical contexts, respect for patient autonomy is often taken to be equivalent to informed consent. However, the relationship between autonomy and informed consent is complex and contested. Some philosophers and bioethicists, such as O'Neill (Reference O'Neill2002), argue that the two concepts are quite distinct and that informed consent should be understood in terms of non-coercion and non-deception rather than autonomy. We agree with Beauchamp and Childress (Reference Beauchamp and Childress2008) that the importance of informed consent arises from the principle of respect for autonomy but, for reasons outlined in the text below, we think that autonomy is more demanding than informed consent. Further complexities concerning the relationship between informed consent and autonomy are raised by the differences between express, tacit, implied and presumed consent, and between generic and specific consent. These complexities are discussed by both O'Neill and by Beauchamp and Childress but are beyond the scope of our concerns in this article.

12 Korsgaard characterises ‘practical identity’ as a normative self-conception, ‘a description under which you value yourself, a description under which you find your life worth living and your actions worth undertaking’ (Reference Korsgaard1996, p. 101). For further discussion of the importance of practical identity to authenticity, see Christman (Reference Christman2009).

13 There is extensive discussion and debate in the philosophical literature on autonomy concerning how best to understand the notion of reflective endorsement. For an overview of some of the issues raised by this debate, see Mackenzie (Reference Mackenzie2008).

14 This criticism is developed most extensively in Bratman (Reference Bratman2007) and Christman (Reference Christman2009). Christman proposes a historical constraint on authenticity. Thus he defines authenticity as ‘non-alienation upon (historically sensitive, adequate) self-reflection, given one's diachronic practical identity and one's position in the world’ (2009, p. 155).

15 This statement is repeated in the Code on pp. 40 and 64. See also the scenario on p. 21 regarding Mrs Arnold's capacity to make financial decisions. As outlined in the scenario, her son should assume that his mother has capacity to manage her affairs. He should also consider each of her financial decisions as she makes them, giving her any help and support she may need to make these decisions herself.

16 See the example and discussion on p. 25 of the Code, which suggests that in relation to prima facie ‘unwise’ decisions there might be need for further investigation, taking into account the person's past decisions and choices. Note also that in discussing how to apply the best interests principle to persons who lack capacity, the Code insists on the importance of making decisions that are authentic for the person, based upon their prior beliefs and values (see, e.g., pp. 80 and 83).

17 We acknowledge that in frontline capacity assessments, particularly in crisis situations, it may be difficult to establish whether a person's actions or decisions are diachronically coherent and consistent. Thanks to Andy Hill for pressing us on this point.

18 For an overview of relational approaches to autonomy, see Mackenzie and Stoljar (Reference Mackenzie and Stoljar2000a). The best-known exponent of the dialogical approach is Charles Taylor; see, for example, Taylor (Reference Taylor1989; Reference Taylor, Hiley, Bohman and Shusterman1991). For an important recent discussion elaborating the connections between relational and dialogical approaches, see Westlund (Reference Westlund2009).

19 For further discussion, see Mackenzie and Stoljar (Reference Mackenzie, Stoljar, Mackenzie and Stoljar2000b).

20 See, for example, the scenarios on pp. 34 and 38 of the Code.

21 For further discussions of the connections between autonomy and accountability or answerability to others, see Benson (Reference Benson, Christman and Anderson2005a; Reference Benson and Taylor2005b) and Westlund (Reference Westlund2009).

22 See, for example, the scenario on p. 25 of the Code, as well as the accompanying text. See also Chapter 3.

23 There has been considerable discussion in the philosophical literature concerning the importance of self-affective attitudes such as self-respect, self-trust and self-esteem for the development and exercise of autonomy. For further discussion, see Anderson and Honneth (Reference Anderson, Honneth, Christman and Anderson2005), Benson (Reference Benson1994), Govier (Reference Govier1993), McLeod (Reference McLeod2002), and Mackenzie (Reference Mackenzie2008).

24 Re Ms. B. vs. an NHS Hospital Trust [2002] EWHC 429 (Fam). See Mackenzie (Reference Mackenzie2008) for a related discussion of this case in connection with theories of autonomy. Parts of the discussion here are adapted from this article.

25 Re Ms. B. vs. an NHS Hospital Trust [2002] EWHC 429 (Fam), at [12] per Dame Butler-Sloss.

26 Re Ms. B. vs. an NHS Hospital Trust [2002] EWHC 429 (Fam), at [13] per Dame Butler-Sloss.

27 Re Ms. B. vs. an NHS Hospital Trust [2002] EWHC 429 (Fam), at [12] per Dame Butler-Sloss.

28 This is a fictionalised account of a situation brought to the attention of one of the authors (WR) by a person who was involved in the care of the patient in question.

29 We recognise that both best practice and the MCA require that assessments of capacity should be specific to this decision, at this time, therefore it would have been appropriate to assess capacity for both decisions.

30 Thanks to Paul Biegler, Andy Hill and an anonymous reviewer for this journal for helpful critical feedback on an earlier version of this article.

References

Anderson, Joel and Honneth, Axel (2005) ‘Autonomy, Vulnerability, Recognition, and Justice’, in Christman, John and Anderson, Joel (eds), Autonomy and the Challenges to Liberalism. Cambridge: Cambridge University Press, 127–49.Google Scholar
Beauchamp, Tom and Childress, James F. (2008) Principles of Biomedical Ethics, 6th edn.New York: Oxford University Press.Google Scholar
Benson, Paul (1994) ‘Free Agency and Self-Worth’, Journal of Philosophy 91: 650–68.CrossRefGoogle Scholar
Benson, Paul (2005a) ‘Taking Ownership: Authority and Voice in Autonomous Agency’, in Christman, John and Anderson, Joel (eds), Autonomy and the Challenges to Liberalism. Cambridge: Cambridge University Press, 101–126.Google Scholar
Benson, Paul (2005b) ‘Feminist Intuitions and the Normative Substance of Autonomy’, in Taylor, James Stacey (ed.), Personal Autonomy: New Essays on Personal Autonomy and its Role in Contemporary Moral Philosophy. Cambridge: Cambridge University Press, 124–42.Google Scholar
Bratman, Michael (2007) Structures of Agency. Oxford: Oxford University Press.CrossRefGoogle Scholar
Christman, John (2005) ‘Autonomy, Self-knowledge and Liberal Legitimacy’, in Christman, John and Anderson, Joel (eds), Autonomy and the Challenges to Liberalism. Cambridge: Cambridge University Press, 330–57.CrossRefGoogle Scholar
Christman, John (2009) The Politics of Persons. Cambridge: Cambridge University Press.Google Scholar
Department for Constitutional Affairs (2007) Mental Capacity Act 2005 Code of Practice. Norwich: Stationery Office.Google Scholar
Department of Health (2000) No Secrets: Guidance on Developing and Implementing Multi-agency Policies and Procedures to Protect Vulnerable Adults from Abuse. London: Home Office.Google Scholar
Dunn, Michael C., Clare, Isabel C. H. and Holland, Anthony J. (2008) ‘To Empower or to Protect? Constructing the ‘Vulnerable Adult’ in English Law and Public Policy’, Legal Studies 28(2): 234–53.Google Scholar
Dworkin, Gerald (1988) The Theory and Practice of Autonomy. Cambridge: Cambridge University Press.Google Scholar
Goodin, Robert (1985) Protecting the Vulnerable. Chicago: University of Chicago Press.Google Scholar
Govier, Trudy (1993) ‘Self-Trust, Autonomy and Self-Esteem’, Hypatia 8: 99–120.Google Scholar
Korsgaard, Christine (1996) The Sources of Normativity. Cambridge: Cambridge University Press.Google Scholar
Mackenzie, Catriona (2008) ‘Relational Autonomy, Normative Authority and Perfectionism’, Journal of Social Philosophy 39(4): 512–33.Google Scholar
Mackenzie, Catriona and Stoljar, Natalie (2000a) Relational Autonomy: Feminist Perspectives on Autonomy, Agency and the Social Self. New York: Oxford University Press.CrossRefGoogle Scholar
Mackenzie, Catriona and Stoljar, Natalie (2000b) ‘Introduction: Autonomy Refigured’, In Mackenzie, Catriona and Stoljar, Natalie (eds), Relational Autonomy: Feminist Perspectives on Autonomy, Agency and the Social Self. New York: Oxford University Press, 3–31.Google Scholar
McLeod, Carolyn (2002) Self-trust and Reproductive Autonomy. Cambridge, MA: MIT Press.Google Scholar
National Assembly for Wales (2000) In Safe Hands: Implementing Adult Protection Procedures in Wales. Guidance Issued by the National Assembly for Wales under s 7 of the Local Authority Social Services Act 1970. Cardiff: National Assembly for Wales.Google Scholar
O'Neill, Onora (2002) Autonomy and Trust in Bioethics. Cambridge: Cambridge University Press.Google Scholar
Rogers, Wendy, Mackenzie, Catriona and Dodds, Susan (2012) ‘Why Bioethics Needs a Concept of Vulnerability’, International Journal of Feminist Approaches to Bioethics 5(2): 11–38.CrossRefGoogle Scholar
Taylor, Charles (1989) Sources of the Self. The Making of Modern Identity. Cambridge: Cambridge University Press.Google Scholar
Taylor, Charles (1991) ‘The Dialogical Self’, in Hiley, David R., Bohman, James and Shusterman, Richard (eds), The Interpretive Turn: Philosophy, Science, Culture. Ithaca, NY: Cornell University Press, 304–314.Google Scholar
Westlund, Andrea (2009) ‘Rethinking Relational Autonomy’, Hypatia 24(4): 26–49.CrossRefGoogle Scholar