Self-Rule in Sick Selves

(1)

Exploring the Limits of Personal Autonomy in Contemporary

Philosophy and Psychiatric Practice

Floor Cuijpers Student number:10135839

Name supervisor: dhr. dr. T. R. V. (Thomas) Nys Name reader: dhr. dr. H.W.J.M. (Henri) Wijsbek

Master Thesis submitted in partial fulfilment of the requirements for the degree of Master of Philosophy

Date: 24-07-2017 Word count: 20.0049

(2)

(3)

This is a painting by Michaël Borremans: The Advantage, 2001.

(4)

Preface

Exactly one year ago, I decided to quit my studies in medicine and start a master in philosophy. While studying medicine, it was the field of psychiatry that mostly intrigued me. What are the presuppositions when we say that someone shows ‘dysfunctional behavior’, or ‘lacks the requisite ability to be a self-governing individual’? Moreover, what do such

statements reveal about our implicit intuitions considering ‘normal’ or ‘functional behavior’?

Such questions highly interest me, yet the medical curriculum does not leave much time to elaborate on these issues.

Fortunately, studying philosophy allowed me to engage with similar questions. During my ethics course, for instance, I was introduced to the work of Susan Wolf and Harry Frankfurt, who both critically reflect on some of our most basic assumption on what it means to be an autonomous person. This led me to presume that these philosophical theories might also aid in clarifying what disputes regarding the autonomy of psychiatric patients (or the implicit ideal of what it means to be a functioning person) are all about. I soon realized however, that characterizing this relationship between mental conditions and personal autonomy introduces many puzzles.

Here, I would like to thank Thomas Nys for our inspiring meetings, his support in guiding me trough this, at times complex, process of writing a thesis and the many interesting theories and philosophers he introduced me to.

(5)

Despite the fundamental value our present day society attaches to the notion of ‘personal autonomy’, the meaning of the term differs greatly depending on the context in which it is used (Dworkin, 1988). In health-care settings for instance, contemporary medical ethics dictates that patients ought to be treated as autonomous agents, i.e. they need to be allowed to make decisions about their treatment for themselves. By contrast, those who suffer from ‘diminished autonomy’ due to illnesses or disabilities are entitled to protection (Beauchamp & Childress, 2011). This reveals that autonomy is not only understood as an individual right to self-governance that others should respect; autonomy is also interpreted as a capacity, where only those who fulfill certain criteria are entitled to this right. Yet what exactly does it mean to say that a subject has or lacks the requisite capacity to be treated as an autonomous agent? And how can this be capacity be assessed?

These challenging questions are central to the theory and practice of psychiatric health-care. The intuition often expressed in psychiatric health settings is that sufferers from mental disorders make choices that they would have never made prior to the onset of their conditions. Put differently, “he or she is not ‘autonomous’ or ‘self-governing’, but is ‘governed’ by the illness’’. The justification for imposing treatment would then be “respect for the right for autonomy, expressed as the attempt to restore autonomy to someone who presently lacks it’’ (Matthews, p. 67, 2000). This argument is based on the premise that a patient’s action or choice should only be recognized as autonomous if it is a true expression of oneself. Thus, considering what justifies infringing on patients’ autonomy requires addressing questions such as “What is the nature of the self who is choosing?”.

Such questions are not only central to debates in psychiatric practice, on the contrary, they also play a crucial role in philosophical theories of personal autonomy. Interestingly, philosophers regularly invoke the example of the mentally ill, where the absence of autonomy is presumed, in order to explicate, conceptualize or demarcate different notions of autonomy (Wolf 1987, Christman, 1991).

Thus, the relationship between autonomy and mental illness concerns both philosophers and practitioners of clinical psychiatry. However, the approaches of these two fields differ. Whereas current debates in psychiatric practice revolve around the question if,

(7)

and how, psychiatric illnesses might impinge on the mental capacities required for autonomous agency; theorists of personal autonomy at times refer to mental illnesses to illustrate how we could better understand the notion of autonomy in the first place.

Research Questions

In this thesis, I will investigate the link between mental illnesses and the concept of autonomy as it appears in psychiatric practice, as well as how this link is described in contemporary philosophical theories on personal autonomy. Such an exploration might at first glance be understood as a project belonging to the field of ‘Philosophical Psychopathology’ (Murphey, 2017). Philosophers writing in this tradition attempt to incorporate empirical results from psychiatry in order to open up new ways of thinking about contemporary theories of autonomy (Graham & Stephens, 2007), or conversely reflect upon the philosophical literature on autonomy in order to better understand why illnesses such as compulsive disorders and addiction might infringe on patients’ autonomy (Prinz, 2007).

In this thesis however, I will take on a different approach. Rather than trying to illuminate one concept by turning towards the other discipline, I will follow the path proposed by Gerald Dworkin and “study how the term personal autonomy is connected with other notions, what role it plays in justifying normative claims, how the notion is supposed to ground ascriptions of value’’ (1988a, p. 11).

In doing so, I will first focus on practical issues regarding autonomy, by scrutinizing the debate around informed consent and competence assessments in psychiatric patients. What does this debate tell us about the underlying assumptions regarding autonomous agency guiding contemporary psychiatric practice in Western societies? I will then turn towards contemporary philosophical literature on personal autonomy. In this literature, mental illnesses are frequently introduced to exemplify impairments to autonomy, perhaps guided by the idea that we can understand the notion of autonomy better if we understand its (presumed) failure. What, however, are the underlying assumptions autonomy theorists make when reflecting upon impaired autonomy in those living with mental illnesses? And do these assumptions relate to those that guide psychiatric practice?

(8)

Outline

In the first part of this thesis I will reflect upon the difficulties that arise when physicians attempt to evaluate whether sufferers from mental illnesses are competent for autonomous decision-making. More specifically, I will critically examine the various ‘competence tests’ which have been introduced: ‘objective’ instruments which aim to assist psychiatrists in judging whether patients should be considered capable of clinical decision-making, and therefore, autonomous. There is an ongoing debate on what type of criteria should be included in these tests. A critical reflection on this discussion shows that what might initially appear to be a practical issue - a dispute over the effectiveness of these tests- soon reveals more fundamental disagreements within psychiatric practice on what it should mean to be an autonomous agent.

Whereas the first section of this thesis focuses on practical contexts in which autonomy is at issue, in the second part of this project I will turn to a number of important works in the philosophical literature on personal autonomy. I observe that, despite the fundamental differences in these approaches to personal autonomy, many theorists seem to share an important underlying intuition: to achieve the status of autonomy, one’s acts or desires need to flow in some way from what can be considered one’s true self. In other word, an agent acts autonomously when ‘an authentic expression of his or her will’.

There are a number of difficulties with such requirements. In the third and final part of this thesis, I will consider the requirement of authenticity for autonomy. More specifically, I will explore how Robert Noggle (2005) problematizes the notion of authenticity as a condition for autonomy.

(9)

PART I: AUTONOMY IN MENTAL HEALTH CARE

“Why do you assume to have the right to decide for someone else? Don’t you agree it’s a terrifying right, one that rarely leads to good? You should be careful to. No one is entitled to it, not even doctors.’’ But doctors are entitled to the right – doctors above all’’, exclaimed Dontsova with deep conviction. By now she was really angry. Without that right there’d be no such thing as medicine!’’

Solzhenitsyn, Cancer Ward,Autonomy

and informed consent, Gerald Dworkin

(10)

1. Autonomy and Competence in Psychiatric Practice

§1.1 Informed Consent and Competence

Directing our life in accordance to a self-chosen plan, based on what we regard as valuable is of crucial importance in today’s liberal society. In health-care settings, this value is reflected in patients’ right to make their own choices considering their physical and mental health, a right safeguarded by the principle of ‘Informed Consent’. According to this principle, medical interventions can only be performed on patients who are firstly, fully informed about the type of treatment they might receive, and secondly, have given their permission for the treatment at stake (Beauchamp & Childress, 2011). As such, this principle should protect patients against potential manipulation, coercion or other forms of maltreatment in medical settings, in this way safeguarding their right for ‘self-governance’ (Dworkin, 1988c, p. 100). Moreover, it is also in accord with the liberal ethos, which states that no person ‘not even a doctor’, has the right to impose his or her (medical) values on those who do not share them (Taylor, 2010).

However, not every patient is entitled to this right of Informed Consent. Indeed, the liberal ideal of self-governance rests on the condition that the individual is capable of exercising self-governance; in other words, one needs to be considered ‘competent’ (Beauchamp & Childress, 2011). Within medical contexts, this notion of competence is often defined as the “capacity to make (health-) choices’’ (Beauchamp & Childress, p. 70).1

At times, physicians fear that their patients lack the capacities that are required to make the decision at stake. In order to safeguard the patient from any harm, he or she can be declared ‘incompetent’, in which case someone else decides for the patient. Such “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non- “non-

1_{Much debate exists around this formulation of competence. It has been argued, for instance, that competence is}

not a categorical principle- something that is fixed, but should rather be understood as a dynamical concept that

changes with every decision. Following this line of thought it might be better to speak of “ competence for some

task, competence to do something’ (Buchanan & Brock 1989, 84). This task might differ, and with that,

someone’s competence. In this regard, someone might be competent of deciding between pulling a tooth or not,

yet would be incompetent when this decision encompasses her heart surgery. The term ‘decisional relativity’

now emphasizes that the evaluation of competence needs to be directly tied to a particular decision, for a particular patient, at a specific place and time.

(11)

consensual treatments” (Matthews, 2000, p. 59) occur for instance in individuals who are in a coma and therefore physically unable to express their wishes, but also in patients who suffer from severe psychoses, treatment can sometimes start without firstly obtaining the patient's’ approval -even when this goes against his or her direct will.2

In the Dutch health care system the term ‘wilsbekwaamheid ’ is frequently used to refer to the notion of competence (Ruissen, Meynen & Widdershoven, 2011). This notion should, according to the ‘WBGO-wet’, be understood as “a patient who is rendered capable of fairly appreciating those issues which are at stake’’.3 This formulation can be interpreted in many ways though, an issue I will come back to. In the following sections, I will use the term competence as well as ‘wilsbekwaamheid’, where in both cases I refer to “the state in which a patient's decision-making capacities are sufficiently intact for their decisions to be honored” (Grisso & Appelbaum 1998, p.11).

§1.2 Competence in Psychiatry

Within psychiatric health care, the terms competence and self-governance play a special role. Whereas in somatic health-care, physical barriers such as a paralysis or a coma might prevent patients from expressing their will- thereby rendering them incapable of decision-making, in psychiatric patients something different is going on. Frequently, sufferers from mental illnesses are still capable of expressing a will, yet it is exactly the genuineness of this ‘will’ which is sometimes questioned by their treating psychiatrists. The intuition giving rise to these doubts is that in some cases, psychopathological changes interfere with the mental life of a patient to such an extent “that he or she is no longer self-governing, yet is instead “governed’ by the illness’’ (Matthews, p. 67, 2000). The will or choice expressed might then not be truly belonging to the patient, yet is merely the result of pathological disturbances.

2_{These forms of coerced treatment are only accepted when there is a risk for severe damage to the patient’s}

health or when there is a danger other people might be harmed. An additional requirement is that the treating physician has looked for alternative treatments. Rijksoverheid (2017). Informatiepunt Dwang en Zorg. WBGO-wet. Wilsbekwaamheid (Accessed May 2017)

3_{“[E]en meerderjarige patiënt die in staat kan worden geacht tot een redelijke waardering van de belangen ter}

zake’’. Rijksoverheid (2017). Informatiepunt Dwang in de Zorg, WBGO-wet. Wilsbekwaamheid (Accessed

(12)

Such deliberations take place, for instance, in cases of euthanasia requests. According to a new proposed law “[E]uthanasie bij Voltooid Leven’’, elderly people who consider their life as “accomplished’’ should be allowed to receive euthanasia, provided that their wish is not the result of either external pressures (by for instance medical staff or family members) nor the effect of mental disturbances which render them incompetent, such as a depression or psychosis (Rijksoverheid, 2016).4_{As such, prior to any intervention, psychiatrists need to}

distinguish between a ‘competent’ and ‘incompetent’ request, and examine whether the will expressed belongs to the patient, or in fact, derives from mental changes due to a psychiatric illness.

In practice, this is not an easy task. Important to note is that a psychiatric diagnosis does not automatically renders a patient incompetent (WBGO-wet).5 Different mental diseases can influence a person’s behavior in various ways. Moreover, a similar mental illness might manifest itself very differently in one patient compared to another. As such, declaring a patient incompetent solely based on his or her psychiatric diagnosis would be stigmatizing as well as discriminating. To safeguard patients from such unfair treatments, there is consensus that “every individual should be regarded as competent, until proven otherwise’’.6 What however, does this ‘proven otherwise’ exactly mean? And how does one determine this?

These questions have shown to be very difficult. As stated before, the definition of ‘wilsbekwaamheid’ as a state in which one is “capable of fairly appreciating those issues that are at stake’’, can be interpreted in many (conflicting) ways.7 Is it, for instance, the physician or the patient who determines what these ‘issues at stake’ are? And what level of ‘appreciation’ is sufficient for a patient's’ choice to be respected? In short, although patients’ right to self-governance is premised on the condition that a patient is ‘competent’ for such exercise, how we should precisely understand what it means to be competent is unclear.

4_{Rijksoverheid (2016). Rapport Adviescommissie. ‘Voltooid leven’.(Accessed April 2017)}

5_{Rijksoverheid (2017). Informatiepunt Dwang in de Zorg, ‘Wilsbekwaamheid’. (Accessed May 2017)}

6_{Rijksoverheid (2017). Informatiepunt Dwang in de Zorg. ‘Wilsbekwaamheid’ (Accessed May 2017)}

(13)

This lack of consensus on the term (in)competence is remarkable however, when we consider the radical impact this assessment can have on a patient's’ life. For not only can an ‘incompetent’ statement lead to a denial of a euthanasia request (such as in the situation sketched above) at times this declaration might also result in coerced treatment or even an involuntary commitment to a psychiatric ward.8

Bearing in mind these drastic consequences, it would be troublesome if a patient’s competence assessment (and consequently, his or her right for autonomous decision making) would depend on the type of psychiatrist he or she happens to have. Ideally then, psychiatrists would adhere to a shared idea of what it means to be competent when they assess this capacity in their patients. Not only would such an agreement on this notion assist psychiatrists in judging their patients’ competence whenever they are in doubt, more importantly, an equal evaluation would be more fair towards the patient whose right for self-governance is at stake. Roughly said, it would be in line with what Beauchamp & Childress define as the principle of justice in health-care, “the obligations of fairness in the distribution of benefits and risks’’ (Beauchamp & Childress 2011, p. 64).

For these reasons, various competence ‘tests’ have been developed: tools established to help psychiatrists in evaluating their patients’ competence (Grisso & Applebaum, 1998). Ever since these tests came into being however, they have been criticized for different reasons (Charland, 1998) and the debate around these competence tests still continues (Meynen & Widdershoven, 2012). In general, we could state that this discussion revolves around two conflicting approaches to this idea of ‘measuring competence’, a procedural and a more substantive approach. In the next section I will provide an example of these two approaches, as well as an analysis of the various ways in which they have been criticized.

8_{These forms of coerced treatment are only accepted when there is a risk for severe damage to the patient’s}

health, or when there is a danger that other people might be harmed. An additional requirement is that treating physicians have looked for alternative treatment the patient does agree with. Rijksoverheid (2017). Informatiepunt Dwang en Zorg. WBGO-wet. Wilsbekwaamheid (Accessed May 2017)

(14)

§1.3 Competence Assessments: Two Approaches

1.3.1. A Procedural Approach: The MacArthur Competence Assessment Tool

The most widely used competence test, the MacArthur Competence Assessment Tool (developed by Grisso & Applebaum, 1998), aims to evaluate a patient’s competence by means of four criteria: his or her ability to communicate a choice, to understand the relevant information, to appreciate the medical consequences of the situation, and to reason about treatment choices, where this reasoning is usually said to include the ability to weigh risks and benefits

With these four criteria, this test focuses on the patient's ‘cognitive functions’ which are thought to be required for the process of reaching a certain decision, but leaves out the material (the values, beliefs and convictions) individuals use whilst deciding, and does not evaluate the actual outcome of the decision either. As such, this test is characterized as ‘content-neutral’ or ‘procedural’ (Hermann et al. 2016).

It is stated that this latter characteristic make this test well suited for liberal societies. The argument expressed here is that since assessments of competence take place against a background of a pluralistic culture, in which all people involved (whether they are nurses, psychiatrists or patients) hold on to different values, and since the imposition of certain values on other people is unwanted, “a test for competence must leave enough space for persons to choose which (irrational) values to adhere to, and consequently, which courses of actions to follow’’(Taylor, p. 64, 2010). Procedural tests, like the MacArthur-test, are considered to fit these demands, since they “refrain from judging the lifestyle, value system, viewpoint, or reasons underpinning a decision as more or less appropriate. As long as these procedural demands are met, people are allowed to make decisions on whatever grounds they choose- rational or irrational’’ (Hermann et al. 2016. p. 7).

1.3.2 Problems with the MacArthur Test: The Role of Values

(15)

actually use in reasoning when they weigh the risks and benefits of proposed treatment options’’ (2006, p. 283). In a similar vein, Buchanan and Brock have argued that making a decision necessitates “a conception of what is good” against which to weigh and evaluate alternative courses of action (1989, p. 24). From this follows that a method that excludes these components in evaluating competence, fails to capture what it aims to evaluate: a patient’s capacity to make a choice. Based on this argument, Charland concludes that the MacArthur test lacks “empirical validity”(2006, p. 284).

Moreover, at times it are precisely patients’ values and emotions, which cause psychiatrists to question their competence (Meynen & Widdershoven, 2012). In explaining their concerns, Meynen & Widdershoven refer to a large empirical study by Tan and colleagues (2007) in which the MacArthur tool was used to evaluate whether anorexia patients were competent enough to refuse treatment. According to the MacArthur test, the patient’s cognitive functions were fully intact (i.e. they were able to reason about treatment options and to explain and communicate their choice for refusing treatment in an adequate way), as such; all of these patients were considered competent.

However, this outcome shows exactly why there is a foundational problem with the MacArthur test, according to Meynen & Widdershoven (2012), for in fact, all of these patients should not have been judged as competent. After evaluating these patients according to the MacArthur tool, Tan et al. (2007) asked them why they wished to refuse treatment. As it turned out, most of them based their decisions on the value of ‘staying thin’. This value, however is a direct a result of the pathology anorexia nervosa, the illness they should be protected against, according to Meynen & Widdershoven (2012).

In short, Meynen & Widdershoven argue that the MacArthur test sets the threshold for competence too low, for even though patients cognitive reasoning processes might be intact, due to these value disturbances they should still be considered incompetent. These clinicians therefore propose a model of competence that not only focuses on cognitive capacities, but also incorporates an evaluation of patients’ values.

1.3.3 A Substantial Approach: Pathological versus Authentic Values?

Despite the fact the MacArthur-test is praised for its content-neutrality, not judging the desires or values that could motivate a person to make that decision, critics consider this

(16)

aspect a major weakness of the test, bringing in that it are exactly value disturbances which might render patients incompetent (Charland, 2007, Meynen & Widdershoven, 2012). These critics promote what has been defined as a more ‘substantive approach’ to competence (Hermann et al 2016), by proposing to include an evaluation of potentially pathological values as well. Yet, this account raises questions as well, for how should we understand this notion of a ‘pathological value’?

The answer to this question is not straightforward. As stated before, a psychiatric diagnosis does not automatically render someone incompetent. In a similar vein then, a value cannot be considered ‘pathological’ solely because it belongs to someone with a mental disorder. Rather, according to Meynen & Widdershoven (2012), sufferers from mental illnesses should be assessed as incompetent with regard decision-making when certain ‘pathological values’ start to inflict upon this patient’s ‘authentic self’. Shortly, they express the argument that in order to be deemed competent and hence, enjoy one’s right to autonomous decision-making, patients must base their decisions on ‘authentic’, rather than ‘pathological’ values. This however, raises another, potentially even more challenging question, for how do we differentiate between a person’s ‘authentic’ and ‘pathological’ values?

To illustrate this complexity, we might turn to a fictional case. Imagine a man who, after spending ten days in Nepal, decides to drastically change his diet. Whereas his family and friends would characterize him as a true ‘bon-vivant’, enjoying good company, his daily steak and glass of wine, he now only eats one plate of red beets a day (he only allows himself to eat purple food) and prefers to stay at home. Not only did the traveler drastically change his diet, while standing on the Mount Everest, he also decided to quit his job at a prestigious law office in order to meditate all day. It is these sudden changes which made his family question whether the man’s wish to quit his job is truly autonomous.

Indeed, apparently, the man acquired a novel set of values and goals that highly contrasts with his previous way of living. Yet, declaring this man incompetent to make health decisions would be odd. What is it about the traveler which, despite the intrusion of these novel (bizarre) values, leaves him to ‘act authentically’, while the anorexic patient is no longer considered authentic? Clearly, it cannot be the fact that the anorexic patient makes decisions that conflict with his or her own well-being, since in both cases, the newly acquired

(17)

values lead to serious health problems. However, the mere fact that the anorexia patient suffers from a mental condition is not enough either, as discussed before, in declaring a person incompetent and thus, not autonomous.

Some psychiatrists have elaborated on the link between pathological values and authenticity by stating that:

One implication of them [anorexic values] being pathological is that these values do not represent the true or authentic views of the person. In respecting the autonomy of the person it is her ‘authentic’ views that should be respected – that is the views that she would have (or did have) if she did not suffer from the mental disorder (Tan et al. 2006, p. 280).

In this regard, if we were to question ‘which values are authentic?’ The answer would be; those that are not inflicted upon by pathological values. However, if we then wonder ‘what types of values are pathological’? The answer would be: those that do not represent authentic views. Thus, based on these latter explanations, we cannot distinguish between a ‘novel bizarre value’ and a ‘pathological value’. What is actually stated here is that some values are pathological because they are not authentic, and authenticity is acting without the influence of pathological values. Hence, such a definition of authenticity is begging the question.

In short, more substantive approaches to testing competence, such as those by Meynen & Widdershoven, propose to include an evaluation of patients’ values as well. The argument is based on two propositions. Firstly, that choosing authentically is a necessary requirement for competence and thus for autonomy, and secondly, that there exist certain pathological values that hamper one’s authentic decision-making processes. The example above however, revealed that if we were to use this approach in evaluating competence, we also have to explain how and why certain new values are pathological and render choices inauthentic, whilst others do not, without referring to the patient’s diagnosis (this would be discriminating).

(18)

§1.4 Conclusion

In this chapter, we looked into the controversies regarding the assessment of patients’ competence, for even though this notion of competence forms a fundamental condition for patients’ right to autonomy, how we should understand or operationalize this term is unclear. This fuzziness around the notion of competence is reflected in the debate around the type of criteria that are relevant for such an assessment. On the one hand, it is argued that competence tests should focus only on patients’ cognitive skills, not judging the underlying values, in this way allowing for a variety of lives patients can lead (Grisso & Applebaum, 1998). Opponents of this apparent ‘neutral approach’ to competence bring in that not only does decision-making require a set of values, at times it are exactly value disturbances which compromise a patient's’ competence. Since content-neutral approaches such as the MacArthur-test do not take into account these (potential pathological) values, critics state that this test is not suitable for assessing competence in patients (Meynen & Widdershoven, 2012). In short, there seems to be a tension between constructing a test that is ‘thin enough’ to allow space for patients’ irrational decisions, yet ‘thick’ enough to safeguard those patients who should be entitled to protection.

Moreover, analyzing the clinical issue of competence assessments revealed that this debate is also guided by different assumptions about what it means, or what it should mean to be a competent agent. We have seen that the apparent content-neutrality of the Mac-Arthur test of competence differs highly from substantive approaches to competence, on which patients cannot endorse certain values if he or she is to be considered competent, and therefore autonomous. This latter account is guided by the assumption that only patients who make their decision based on ‘authentic values’ are competent, and hence, suitable for the right to autonomy. How to understand or define these types of values though, is left unanswered.

(19)

PART II: AUTONOMY CONCEPTIONS IN PHILOSOPHY

In the first part of this thesis, we explored practical contexts in which autonomy is at issue by analyzing the notion of autonomy within psychiatric practice. More specifically, we shed light on how various notions of ‘competence’; the condition for autonomous decision-making, take shape in different clinical tests developed to assist psychiatrists in evaluating this capacity in psychiatric patients.

With a practical background in place, I will now examine different perspectives on autonomy as they appear in contemporary philosophical theories. What stands out in this discussion is that despite their different ideas about what autonomy is or what it ought to be, many theorists refer to a similar group of examples to demarcate this notion: in sufferers from mental illnesses, it is stated, autonomy is often undermined. Especially addictions and compulsions form popular examples in philosophical literature to illustrate how certain influences on our behavior prevent us from governing ourselves. How can these similar examples serve the same purpose in different- and at times even competing- theories of autonomy? And what are the underlying assumptions guiding such descriptions?

With these questions in mind, I will shed light on several contemporary approaches to autonomy, adhering to the distinction introduced by Catriona Mackenzie between procedural and more substantive autonomy accounts (2000, p.12). In general terms, those who follow procedural models believe that an agent's’ autonomy depends on their critical reflection of their values as “his or her own”, instead of feeling “indifferent or external towards these desires” (p. 13). Mackenzie further divides between structural and historical versions of procedural theories, which according to her, have different opinions on what kind of critical reflection processes are necessary for autonomy. On the other hand, more substantive approaches to autonomy claim that “internal reflection on one’s values and desires is not enough to secure autonomy”, and propose to add more constraints on the type of values an autonomous agent is guided by (Mackenzie, 2000, p.14).

This part of the thesis is divided into two chapters. In the first chapter, I will engage with procedural theories of autonomy, focusing first on one of the most well-known structural accounts, introduced by Harry Frankfurt and grounded in the notion of higher-order desires

(20)

(1988). Secondly, I will turn to more historical versions of these procedural theories, brought forward by Gerald Dworkin (1988) and John Christman (1999). Subsequently in chapter two of this part, I shall outline the critiques of these procedural theories of autonomy, which leads me to discuss more substantive approaches, focusing on one by Susan Wolf.

1. Procedural Accounts of Autonomous Agency

§1.1 Structural Versions: Frankfurt’s Higher-Order Desires

One of the most influential approaches to autonomy, introduced by Harry Frankfurt, is based on the notion of “second-order desires” (1988a, p.11). In his famous essay Freedom of the Will and the Concept of a Person, Frankfurt makes a distinction between two types of human desires. First-order desires constitute the wish to perform certain types of action, such as watching reality TV all day long. Second-order desire constitutes a type of attitude towards this first order desire, for instance, the desire to no longer be guided by the desire to watch reality TV all afternoon (instead to be productive and write a thesis proposal). On Frankfurt’s account, to have freedom of will (or to be an autonomous agent9) we need not only the ability to “translate our desires into action” (1988a, p.20), but also to “identify” with these desires (1988a, p.18). In this sense, identifying means distinguishing between desires that an agent considers “one’s own” from those desires that one is either indifferent to or regards as “external” to oneself (Frankfurt, 1988b, p.170).

To illustrate this account, Frankfurt asks us to consider two types of agents who, according to him, lack autonomy. Interestingly, he refers to a psychiatric condition in order to explain his point. The first type of agent who does not act autonomously is someone who does not have any second-order desires at all. Frankfurt calls such agents “wantons’’ instead of persons (p. 16), listing small children and animals as well as individuals who are addicted to drugs and constantly desire to take this drug, without having any type of attitude towards this desire (1988a, p. 17). As Frankfurt sees it, even though these individuals might have

9_{Frankfurt's view is not exactly a theory of autonomy, yet rather an account of freedom of the will.}

Nevertheless, “the account has been absorbed into the literature on autonomy as a model of that notion.” (Christman, John, "Autonomy in Moral and Political Philosophy", The Stanford Encyclopedia of

(21)

“freedom of action”, they lack the capacity to identify with this desire (the ability to accept or disregard the desire at stake as ‘one’s own’) and hence, do not have freedom of will (Frankfurt, 1988a, p. 16).

Merely having second-order desires is in Frankfurt’s view not enough for autonomy either; an autonomous agent also needs to satisfy a second condition. Again, Frankfurt asks us to consider a drug addict who does possess a second-order attitude towards his drug cravings-he wishes to no longer desire heroin-yet is unable to act upon this second order desire and continues shooting heroin. Even though this “unwilling addict” (p.17) has a clear aversive attitude towards his daily drug cravings (he wants them to go away), the fact that he is unable to act upon this second-order desire (and stop using heroin) still prevents him from having freedom of the will (1988a, p. 17).

Both examples reveal that in Frankfurt’s account, someone acts autonomously when there is match between this person’s second-order desire (a desire as to which first-order desire should move one to action) and the first-order desire, which does in fact move this agent to action (Frankfurt, 1988a, p. 18). Otherwise stated, an autonomous agent is someone who is not just guided by the desires one ‘happens to have’, but rather acts on the basis of those desires one identifies with, those one regards one’s own.

As such, Frankfurt’s account of identification opens up new ways of thinking about the practical issues regarding autonomy in mental health practices, discussed in the previous chapter. Within such contexts, imposing treatment (which might violate a patient’s right to self-governance) is sometimes justified by the argument that sufferers of mental disorders make decisions that they would not have made prior to the disease. Consequently, “he or she is not ‘autonomous’ or ‘self-governing’, but is ‘governed’ by the illness’’ (Matthews, 2000, p. 67).

Frankfurt’s account of identification sheds a different light on this debate. In fact, we have seen that Frankfurt handles the own attitudes of agents towards their desires as the ultimate determinants with regard to the assessment of whether or not they are autonomous, where acting in accordance with the desires one identifies with constitutes autonomous action (1988a, p.18) From this perspective, even patients who suffer from severe addictions might still be considered autonomous, provided that they identify with the desire (they should have

(22)

accepted this desire to use the substance at stake as ‘their own’). To illustrate this better, Frankfurt characterizes such a “willing addict” in the following way.

(H)e is altogether delighted with his condition. He is a willing addict, who would not have things any other way. If the grip of his addiction should somehow weaken, he would do whatever he could to reinstate it; if his desire for the drug should begin to fade, he would take steps to renew its intensity. (Frankfurt 1988a, p. 24)

In short, the willing addict both wants to want the drug, and is also moved by this desire. In contrast to what this quote might suggest, Frankfurt is not claiming that the willing addict enjoys control over his (first-order) desire to take the drugs. Just like the unwilling addict, he argues, the willing addict is “physiologically addicted” and “his will is out of his control”. What differentiates these two agents though, is that the willing addict identifies with his drug cravings, in other words, “by his second-order desire that his desire for the drug should be effective”, he, “made this will his own”. Frankfurt states that “given that it is therefore not only because of his addiction that his desire for the drug is effective, he may be morally responsible for taking the drug” (1988a, p. 25).

It should be clear that Frankfurt’s notion breaks with the idea that mental illnesses automatically diminish a patient’s autonomy, and hence brings in a new perspective to the debate around autonomy in mental illnesses. Indeed, following Frankfurt’s approach, even though mental illnesses such as addiction might govern patients’ will (or first-order desire), the potential presence of second-order desires might turn this addictive desire into these patients’ ‘own will’, in which case they still act out of their free will, and their autonomy is not undermined.

Though this model opens up new ways of thinking about autonomy and addiction (leaving enough space for the idea that autonomy should be related to making up one’s own mind), Frankfurt’s account of the willing addict also led to resistance. Mark Leon for instance, argues that this willing addict does not use the drugs because he is drawn to it and

(23)

“identifies” with this desire, he uses drugs because he is “irresistibly addicted’’ (Leon, 2001, p. 437). More specifically, he states “given that the willing addict’s desire for the drug is an irresistible desire, he would still continue to use the heroin ‘even if he no longer desires to do so’”. Consequently, Leon claims that just like Frankfurt’s unwilling addict, this agent is “merely a passive by-stander to the operation of the addiction’’ and ’the willing addicts does not have free will” (p. 438, 2001).

James Taylor (2005) claims that the idea that the addict does not have freedom of will because he would continue using heroin regardless is based on a questionable assumption. Namely, “ for a person to act out of one’s free will, he or she needs to be able to act other than he or she did’’. Frankfurt has convincingly argued against exactly this latter assumption that dominated “moral intuitions”, a notion he famously called the “principle of alternate possibilities” (Frankfurt, 1988b, p.2).

To understand the consequences of Frankfurt’s willing addict argument, we will have to take a small detour and investigate why Frankfurt states that the principle of alternate possibilities is false. In what follows, I shall reflect upon the way he argued against this principle, subsequently returning to our debate by reflecting upon Taylor’s ideas of what this means for the willing addict.

Briefly, according to Frankfurt, the principle of alternate possibilities does not hold. He states, “a person may well be morally responsible (or have freedom of will) for what he has done even though he could not have done otherwise” (Frankfurt, 1988b, p.2). Frankfurt‘s arguments are based on the following thought experiment.

Suppose someone -Black, let us say- wants Jones to perform a certain action. Black is prepared to go to considerable lengths to get his way, but he prefers to avoid showing his hand unnecessarily. So he waits until Jones is about to make up his mind what to do, and he does nothing unless it is clear to him (Black is an excellent judge of such things) that Jones is going to decide to do something other than what he wants him to do. If it does become clear that Jones is going to decide to do something else, Black takes effective steps to ensure that Jones decides to do, and that he does do, what he wants him to do. Whatever Jones's initial preferences and inclinations, then, Black will have his way. However Black never has to show his hand

(24)

because Jones, for reasons of his own, decides to perform and does perform the very action Black wants him to perform. (Frankfurt,1988b, p.7).

In short, Frankfurt claims that because Jones performs the action for ‘reasons of his own’, even though he could not have done otherwise, neither free action nor free will requires that a person needs to be able to do anything other than what he did. How does this notion relate to freedom of will in the willing addict?

According to Taylor, we can say something similar of the willing addict. Even though there was no other way for the addict than to take the heroin due to irresistible desire, similar to Jones “the willing addict's act of taking the drug flowed from his own motivational set and he was moved to perform it by a desire that he volitionally endorsed” (p. 241, 2005). Following Frankfurt’s line of reasoning, Taylor concludes that “the fact that there was no other option for the addict than to take the drug to which he is addicted does not undermine the claim that he acted both freely and of his own free will’’(2005, p. 241).

In short, according to Taylor, the argument that the willing addict had no other option than to take the drug does not undermine Frankfurt’s statement that the willing addict acted of his own free will when he took the drug that he was addicted to. That is not to say, however, that Taylor fully agrees with Frankfurt’s account of freedom of will, grounded in the notion of identification. On the contrary, Taylor is skeptical regarding approaches to autonomy that adhere to a “subjective analysis of what it is for a person to identify with her desires” (p. 253, 2015). For how do these theories account for the possibility that a person’s attitude towards his or her first-order desire might have resulted from manipulation? (Taylor, 2005) Or, to refer to the example of the willing addict, what ensures us that this addict’s second-order desires were not also a result of the addiction?

It seems that on Frankfurt’s account, to answer this question we would need another desire (a third-order desire) to determine whether the patient truly identifies with this second-order desire. Yet this solution does not seem to solve the problem. In fact, as we will see in the following section, by raising this question, Taylor touches upon what is often seen as the most problematic part of Frankfurt’s account of identification.

(25)

§1.2 Criticizing Frankfurt’s Account: The Regress-Problem

In the previous section, we observed how in Frankfurt’s analysis identification with one’s desires is sufficient for autonomy. Moreover, following his account, even those individuals who appear to behave out of their own control (i.e. due to a severe addiction) might still be rendered autonomous provided that they have a second-order desire that is in line with their first-order desire and behavior. Such examples illustrate that there is some sort of “hierarchical structure” (Mackenzie, 2000, p.14) in Frankfurt’s approach to autonomy: second-order desires belong more truly to the agent compared to first-order desires. In other words, “second-order desires reveal what one ‘really’ wants, that they are the side of one’s “true self” (Anderson, 2008, p. 10).

This argument is met with skepticism. Critics of such hierarchical approaches, such as Gary Watson, raise that the mere fact that human beings have different types of desires does not yet provide hierarchical orders to them (1975, p. 218). Indeed, what is it about this second-order desire that makes it a more ‘true expression of the self’, than the first-order desire? Moreover, as Taylor argued, these second-order desires might have themselves resulted from manipulation or indoctrination (Taylor, 2005).

To refer to the willing addict, what prevents us from claiming that the addict’s identification with the desire to use drugs was in fact a result of the addiction, rather than at the side of the patient’s ‘true self? It seems that in Frankfurt’s terms, in order to determine whether this was the case and make sure that the patient truly identifies with this drug craving, there is need of another (third-order) desire.

However, it is exactly this point where Frankfurt’s theory of identification falls short. Gary Watson, for instance, states that “since second-order volitions are themselves simply desires, to add them to the context of conflict is just to increase the number of contenders” (Watson, p. 218, 1975). In a similar vein, Robert Noggle claimed that such a solution would merely “move the problem back a step’’ (2005, p.90), for what ensures us that this third-order desire is more authoritative of a person’s will and grants it the autonomy status? In summary then, it seems that Frankfurt’s approach runs into difficulties when it comes to determining which desire in the potential infinite chain can be can be regarded as truly autonomous, a problem famously articulated by Gary Watson as the “regress problem’’ (Watson, 1975, p.

(26)

217).

To account for this regress problem, different strategies have been introduced. Frankfurt himself also acknowledged the deficiencies in his account of higher-order desires. In his later essay ‘Identification and Wholeheartedness’, he states that “the mere fact that one desire occupies a higher level than another in the hierarchy seems plainly insufficient to endow it with greater authority or with any constitutive legitimacy” (Frankfurt, 1988c, p. 166). To end this possible infinite regress, Frankfurt introduces a second criterion; for someone’s desires to be autonomous, this person should also be “satisfied” with them (1988, p. 166). How we should understand this notion of satisfaction becomes clear when Frankfurt’s writes: “to be satisfied with something does not require that a person have any particular belief about it, nor any particular feeling or attitude or intention”. Rather, he writes elsewhere, we might see it as a state of the will which is “untroubled by doubts or desires to change”, and compares it to “being or becoming relaxed” (Frankfurt, 1999, p.105). Applying this revised account including the condition of ‘satisfaction’ to the willing addict leads to the following: if the drug addict forms second-order desires about her addiction (she wants to want the drugs), and is at that current moment not troubled by conflicting desires nor any desire to change, then there should be no reason to deny that this addict is an autonomous agent, as per Frankfurt.

It is questionable however, whether Frankfurt’s account of satisfaction really ends the regress problem. For how do we know when we are in a state of satisfaction? If “satisfaction” consists of an absence of any particular feelings or beliefs towards one’s state (as Frankfurt claims), then how does this fit with Frankfurt’s own theory of second-order desires, in which an autonomous agent not only acts on the basis of his or her will, but also forms (active) attitudes towards this will? Bas Peeters expressed this criticism by asking “if the satisfied agent has no active interest in bringing about a change, why then should the agent still be considered a genuine person? After all, if he simply does not care about his psychic condition, then being satisfied seems to come very close to the indifference of the wanton.” (2015, p. 23). In short, it is questionable whether Frankfurt’s revised account of satisfaction really solves the regress problem, an issue I will return to later.

At this point, we have seen that Frankfurt’s original, as well as his revised account of identification as satisfaction are both neutral with respect to evaluating the content of an

(27)

agent’s desires as well as to the origin of these desires. It is precisely this latter aspect of Frankfurt’s theory, which has led to much resistance. John Christman states that this view “accepts that 'a person's desires can be determined to be autonomous or not by taking a time-slice of the person and asking what her attitude would be about the desires she has at the time” (Christman 1991, p. 9). Yet, he concedes, “how the agent evaluates the desire in itself may have little to do with the process of desire-formation” (p. 9). Rather, exactly during this process of acquiring desires and beliefs is where Christman suspects autonomy might be hampered.

In a similar vein, Gerald Dworkin, argues that “second-order reflections cannot be the whole story of autonomy. For those reflections, the choice of the person one wants to become may be influenced by other persons or circumstances in such a fashion that we do not view those evaluations as being the person’s own” (Dworkin, 1988, p. 18).

Thus, for both Dworkin and Christman, the fact that Frankfurt’s account does not include the way an agent came to the desires at stake creates a pitfall to his approach. More specifically, their objection against this ‘time-slice’ approach is grounded in the concern that reflection on one’s desires might have been influenced by certain forms of manipulation or brainwashing, in which case the idea that these desires are the most authoritative basis of one’s ‘own will’ becomes questionable. To account for such scenarios, both critics developed a notion of autonomy that also incorporates the processes and circumstances leading up to an agent’s desires at stake. In the following section, I will shed light on these so called “historical approaches” to autonomy (Mackenzie, 2000, p. 11).

§1.3 Historical Versions: John Christman & Gerald Dworkin

In the previous section, we observed how Frankfurt’s notion of autonomy has been criticized for introducing a ‘time-slice-model’. Such criticism is guided by the belief that certain types of processes lead to autonomous desires and beliefs whilst other kinds do not, something that such a ‘time-slice approach’ cannot check for (Christman, 1999). To secure that agents’ autonomous desires “arose in the right way” (Dworkin, 1988), theories that also account for the circumstances under which an agent is reflecting have been developed (Christman 2009,

(28)

Dworkin, 1988). In these conceptions, autonomy is not only determined by an agent’s desires at a given moment in time, but also evaluates the processes and circumstances leading up to these desires.

To understand the main differences between Frankfurt’s account and these so called “historical approaches” (Mackenzie, 2000, p. 11) it might be insightful to analyze how both theories would deal with the case-study introduced in the first chapter of this thesis. To briefly recall, in the previous chapter we discussed the case of a man who, after returning from a ten-day holiday in Nepal, decided to drastically change his life. He began to meditate for more than six hours a day, changed his diet (he now only eats purple food) and is now considering quitting his job as a lawyer to meditate even more. In combination, it seems that during these ten days, the man acquired new desires and beliefs that highly contrast with his previous manner of living, which made his family question whether the man’s wish to quit his is job is truly autonomous.

Applying Frankfurt’s account of identification and satisfaction, whether or not we should perceive of the man’s decisions as truly autonomous can be determined by the traveler’s point of view alone. In short, if the man both wants to desire to eat purple food and meditate all day, and if he is untroubled by doubts or desires to change, then according to Frankfurt we have no reason to think he is not autonomous.

John Christman and Gerald Dworkin argue that these conditions would not be enough to guarantee that the traveler acts autonomously. They are concerned that such an account fails to check whether the process of desire-formation was free of influences by other people or circumstances, which would hamper one’s autonomy. What if during his trip in Nepal, the traveler joined a cult that indoctrinated him to adopt the desires of his cult leader (meditation, eating purple food) through a series of highly spiritual sessions?

Therefore, according to Christman, the way in which the traveler came to these novel desires and beliefs would be most troubling, rather than how the man evaluates these desires at the given moment in time. Following this intuition, Christman proposes that “the key element of autonomy is the agent's acceptance or rejection of the process of desire formation or the factors that give rise to this desire formation, rather than the agent's identification with the desires” (1991, p.2), and also “whether any factors are present during these evaluations which effectively undercut a person's ability to make these judgments about her past” (1991,

(29)

p.10).

Dworkin would agree with a Christman reading that checking the traveler’s current state would not be enough to ascribe him autonomy. Yet, whereas Christman emphasizes that this process of reflecting on one’s desires should not be interfered by “Illegitimate External Influences” (1987, p. 291), Dworkin develops a slightly different notion. More precisely, Dworkin attempts to ensure that this critical reflection is not influenced by autonomy undermining circumstances, by requiring that this process occurs under the conditions of “procedural independence”. In his words, “if a person’s reflections have not been manipulated, coerced, and so forth and if the person does have the requisite identification [i.e. if the person has formed second-order desires] then they are, on my view, autonomous” (Dworkin 1988a, p. 20).

Following these historical approaches to autonomy, Frankfurt’s conditions of identification and satisfaction do not suffice to consider the traveler autonomous. To evaluate whether the traveler’s desires are autonomous, we would also need to make sure the way he came to these desires was not manipulated or constrained in any way. Concretely, we would need to know what happened during his time in Nepal.

There are a number of attractive features to these historical conceptions of autonomy. Firstly, in contrast to structural versions, these accounts support our intuition that the way in which a person came to the beliefs that he or she holds is also an important feature of autonomous agency. According to Joel Anderson (2008) there is another advantage of historical approaches over structural accounts. Interestingly, Anderson’s criticism of Frankfurt’s approach brings us back to the issue with which we began this exploration: autonomy ascription in sufferers from mental illnesses.

According to Anderson, in everyday life autonomy is something that is always “ascribed and contested’’ to others (2008, p. 9). He illustrates this by writing that “in the practical contexts in which autonomy is at issue – for example, a patient’s request for assisted suicide, the relevance for autonomy-ascriptions is undeniable” (2008, p. 9). For these reasons, Anderson claims that “it is a desideratum of any theory of autonomy that its account of what autonomy entails provide at least the basis for an account of the pragmatics of how autonomy is to be ascribed and contested’’ (2008, p.9).

(30)

everyday and practical uses of the term. However, if we follow Anderson’s rationale that an account of autonomy should assist in autonomy ascriptions, Frankfurt’s account provides little support. This approach treats agents’ own evaluations of their desires as the sole and ultimate determinants of autonomy. As such, Frankfurt’s account of identification seems to better suited for ‘self-ascription’ of autonomy, rather than ascribing this quality to others.10

Moreover, Anderson suggests that even if we were to evaluate whether our desires are autonomous from a first-person perspective, Frankfurt’s model leaves us without resolve (2008, p. 13). We have seen that for Frankfurt, the condition for autonomy, being satisfied, is understood as a first-person, reflexive state of the will in which someone is untroubled by doubts or desires to change. In this regard, the question whether one’s desires are truly autonomous is quite paradoxical. Indeed, as Anderson’s nicely states it “from the perspective of the deliberating agent, the question of one’s autonomy is supposed to answer itself by not being asked” (2008, p. 13).

Historical versions, on the other hand, are according to Anderson much better suited to deal with the “dynamics of autonomy-ascription” (Anderson, 2008, p. 15). He states that “particularly for those working in contexts of applied ethics, there is much that is attractive about the apparently straightforwardly public criterion introduced by Dworkin: if a person’s reflections have not been manipulated, coerced, and so forth and if the person does have the requisite identification, then they are, on my view, autonomous” (Anderson, 2008 p.15).

In short, according to Anderson, autonomy is something that is (in everyday life) mainly contested and ascribed to others. Based on this observation, Anderson holds that a theory of autonomy should at least include a basis of how to assign such a quality in someone else. In comparison to Frankfurt’s account, historical versions are in Anderson’s view better

10_{It should be noted that both philosophers are on a different mission. Whereas Frankfurt spells out certain}

conditions for an agent to have freedom of will at a certain moment in time (1988a), not intending to analyze any ethical or practical matters, Anderson emphasizes that a theory of autonomy should do just that. In specific, he wants to focus on the “practical use of such an autonomy theory” and argues that in daily life autonomy is

something that is always ‘’ascribed and contested’’(2008, p. 9). It is of course up for debate whether a theory of

(31)

in meeting this demands, since the way he sees it, such theories spell out more concrete and practical conditions required for autonomy. But do they really?

§1.4 Criticizing Historical Accounts: The Ad-Hoc Problem & Oppression

In the previous section we have seen that theorists who follow historical approaches emphasize that autonomous desires should not only be properly assigned to someone’s current state and reflections, yet that these also originated in the ‘right way’. Consequently, the challenge for these theorists becomes to explain what this ‘right way’ entails. More specifically, they need to distinguish certain circumstances or processes that safeguard or encourage, someone’s self-governing, from those that infringe on this capacity.

We have seen that Dworkin specifies this by emphasizing that autonomous reflection should occur under the conditions of “procedural independence” (1988, p. 18). Yet, how do we know that these conditions of procedural independence are met? According to Dworkin, “spelling out the conditions of procedural independence involves distinguishing those ways of influencing people’s reflective and critical faculties which subvert them from those which promote and improve them” (p. 18). This however, is only part of the answer, for it still requires a specification of what ‘critical faculties’ are necessary for autonomy.

Other procedural theorists sometimes explain a lack of autonomy by referring to instances of psychological conditioning, manipulation or mental disorders and then argue that such influences “interfere’’ with the capacity of self-governance (Watson, 1975, p. 205). Yet this is also only a partial explanation, for it leaves the question why these influences hamper autonomy, unanswered. Sarah Buss expresses this criticism stating that “insofar as accounts of autonomy stipulate that certain influences on an agent’s intention-forming process “interfere with”’ or “pervert”, this process, these accounts are incomplete, for they leave it mysterious why certain, and not others, are a threat to self-government” (2008).

Moreover, we could even argue that such explanations resemble the ad-hoc argument we analyzed in chapter one, brought forward within psychiatric practice. To recall this, in the context of mental health-care, it is sometimes suggested that in order for patients to be rendered competent (and therefore autonomous), he or she needs to act authentically, i.e. a

Self-Rule in Sick Selves - Exploring the Limits of Autonomy in Contemporary Philosophy and Psychiatric Practice