Psychiatric Euthanasia and the Ontology of Mental Disorder

In the Netherlands and Belgium, it is lawful for voluntary euthanasia to be offered on the grounds of psychiatric suffering. A recent case that has sparked much debate is that of Aurelia Brouwers, who was helped to die in the Netherlands on account of her suffering from borderline personality disorder. It is sometimes claimed that whether or not a mentally ill person ’ s wish to die is valid hinges on whether or not that wish is a symptom of the person ’ s mental disorder. This article addresses the philosophical problems raised by this claim, with a speci ﬁ c focus on the diagnosis of borderline personality disorder. After considering descriptivist and causal conceptualizations of mental disorder, I argue that the current approach to borderline personality disorder in psychiatry precludes the possibility of dissociating the wish to die from the disorder. I then examine the implications of this analysis for the question of whether or not the request for voluntary euthanasia in the case of borderline personality disorder can be considered valid. Ultimately, I conclude that the inability to dissociate the wish to die from the disorder does not invalidate the wish in the case of borderline personality disorder.


Introduction
On Friday, 26 January 2018, Aurelia Brouwers took her own life at the age of 29 by ingesting poison supplied by a physician in a euthanasia clinic in the Netherlands.Voluntary euthanasia is permitted by the state in the Netherlands, and so Aurelia Brouwers' death was lawful.However, the case has attracted considerable media attention worldwide.This is because Aurelia Brouwers, unlike most people who are helped to die through voluntary euthanasia, did not have a diagnosis of a terminal medical condition.Rather, the reason for her being allowed to die was that she was suffering from a mental disorder, namely borderline personality disorder.
The case of Aurelia Brouwers is not the first case of voluntary psychiatric euthanasia.In 2016, 60 people in the Netherlands were helped to die through voluntary euthanasia on the grounds of psychiatric suffering. 1The number increased to 83 people in 2017.In Belgium, there were a total of 179 cases of euthanasia for psychiatric suffering or dementia between 2002 and 2013, and the prevalence has risen each year since 2008. 2 Nonetheless, the vast majority of cases of voluntary euthanasia are motivated by suffering associated with progressive medical conditions such as cancers, arterial diseases, and neurodegenerative disorders.
This article offers a philosophical examination of a question asked by the journalist Linda Pressly in a BBC News article about Aurelia Brouwers' case: 'But could a death wish have been a symptom of her psychiatric illness?' 3 This is a poignant question, because whether or not a person's wish to die is attributable to a mental disorder is often considered to have bearing on the normative question of whether or not that wish ought to be respected.Reginald Deschepper and colleagues suggest that it can be 'difficult to differentiate between a request based on a genuine and constant form of unbearable suffering and a request as a symptom of a severe depression'. 4Similarly, Paul Applebaum suggests that '[a]lthough laws permitting physician-assisted death specify that the person must be competent to make that decision, determining competence is difficult when the wish to die is part of the underlying syndrome itself.' 5 The suggestion implicit in the aforementioned question, then, is that a mentally ill person's wish to die may be deemed invalid if this wish can be attributed to the person's mental disorder.However, it is widely held that determining whether or not such a wish is a symptom of the disorder is not straightforward.The psychiatrist Kit Vanmechelen, who supports the provision of voluntary euthanasia for psychiatric suffering, when asked about whether Aurelia Brouwers' wish to die could have been a symptom of her mental disorder, is quoted as replying, 'I think you never can be 100% sure of that.' 6 Similarly, the psychiatrist Frank Koerselman, who opposes psychiatric euthanasia, is quoted as asking, '[h]ow could I knowhow could anybody knowthat her death wish was not a sign of her psychiatric disease?' 7 Hence, commentators on both sides of the debate about psychiatric euthanasia acknowledge that dissociating a person's wish to die from the person's mental disorder presents a serious challenge.
My aims herein are twofold.First, I will assess whether or not it is possible to dissociate the person's wish to die from the person's mental disorder in the case of borderline personality disorder.Second, I will consider what implications this has for whether or not the person's wish to die is deemed valid.As we shall see, the answers to both of these problems will depend on how the mental disorder is conceptualized.For the reason of scope, I shall not be addressing the broader debate about whether or not voluntary euthanasia is ethically defensible, although I appreciate that one's stance on this issue is likely to be relevant to whether or not one is willing to follow the argument of this article.Given that my discussion is centered on a case which occurred in a context where voluntary euthanasia is permissible, the argument of this article takes it for granted that there are some conditions under which requests for voluntary euthanasia can be considered valid.
The discussion will proceed as follows.In Section 2, I shall provide some preliminaries on the diagnosis of borderline personality disorder, the legal framework of voluntary euthanasia in the Netherlands, and why a wish's being a symptom of a disorder might be considered to invalidate the wish.In Section 3, I shall show that in order to establish whether or not it is possible to dissociate a wish to die from the person's mental disorder, we need to look at how mental disorder is conceptualized in psychiatry.Under a descriptivist ontology, which characterizes mental disorders as clusters of symptoms, the way in which the diagnosis of borderline personality disorder is defined precludes the possibility of dissociating the wish to die from the mental disorder.A causal ontology, which posits that mental disorders refer to the states that cause the symptoms, allows the theoretical possibility of dissociating the wish from the mental disorder, but I shall argue that the appropriate sort of causal ontology that allows this possibility is not available in the case of borderline personality disorder.The upshot is that the dominant approaches to defining borderline personality disorder in contemporary psychiatry do not allow us to dissociate the person's wish to die from the person's mental disorder.In Section 4, I shall examine the implications this has for the validity of the request for psychiatric euthanasia.Contrary to what is commonly assumed, I will argue that the inability to dissociate the person's wish to die from the person's borderline personality disorder does not justify deeming the wish to be invalid.This is due to the way in which the diagnosis of borderline personality disorder is considered to be inextricably connected to the person's character.

Borderline personality disorder and voluntary euthanasia
Before commencing the critical discussion, I present some background about Aurelia Brouwers' diagnosis and the legal framework of voluntary euthanasia in the Netherlands, where she was helped to die.Borderline personality disorder is a psychiatric diagnosis that is given when someone exhibits an enduring pattern of behavior characterized by emotional lability, impulsivity, volatile interpersonal relationships, unstable self-image, acts of self-harm, and recurrent suicidal ideation.This pattern is typically reported to have a chronic course, although many people with borderline personality disorder do report that the distressing symptoms lessen over passing years. 8However, it is estimated that between 8% to 10% of people with borderline personality die from suicide. 9As with most psychiatric illnesses, the etiology of borderline personality disorder is multifactorial and complex, but an adverse social environment in childhood is known to be a significant causal factor.Notably, parental neglect, emotional mistreatment, physical abuse, and sexual trauma have been shown to contribute causally to the formation of borderline personality disorder by affecting psychosocial development in childhood. 10he diagnosis of borderline personality disorder is made on the basis of a clinical assessment in order to determine the presence of the relevant symptoms.According to the fifth edition of the American Psychiatric Association's Diagnostic and Statistical Manual of Mental Disorders (DSM-5), which is the dominant diagnostic manual in psychiatry in the present day, the diagnosis of borderline personality disorder can be made if the person exhibits at least five out of the following nine symptom criteria: (1) Frantic efforts to avoid real or imagined abandonment [. ..]; (2) A pattern of unstable and intense interpersonal relationships characterized by alternating between extremes of idealization and devaluation; (3) Identity disturbance: markedly and persistently unstable self-image or sense of self; (4) Impulsivity in at least two areas that are potentially self-damaging [. ..]; (5) Recurrent suicidal behavior, gestures, or threats, or self-mutilating behavior; (6) Affective instability due to a marked reactivity of mood [. ..]; (7) Chronic feelings of emptiness; (8) Inappropriate, intense anger or difficulty controlling anger [. ..]; (9) Transient, stress-related paranoid ideation or severe dissociative symptoms. 11further requirement for the diagnosis is that the expression of this pattern of behavior must be 'stable across time and consistent across situations'. 12orderline personality disorder is a controversial diagnostic category which has received much critical attention in the philosophy of psychiatry.For example, Nancy Potter argues that the diagnostic practices and clinical discourses related to borderline personality disorder are influenced by problematic cultural norms and value judgments regarding gender. 13Building on this, Andrea Nicki suggests that assigning the label of borderline personality disorder to survivors of childhood trauma inappropriately medicalizes their experiences and fails to appreciate the genuine reasons underlying their problems. 14Indeed, there is significant stigma associated with the diagnosis of borderline personality disorder, with sufferers being viewed in negative terms by healthcare professionals and by the public. 15It is beyond the scope of this article to examine these criticisms in detail, but I mention them here to highlight the broader context within which the challenges regarding borderline personality disorder and voluntary euthanasia are situated.While I agree with some of these criticisms, the present article will largely be taking the current approach to borderline personality disorder in mainstream psychiatry for granted, as the aim is to unpack some of the philosophical problems that result from this mainstream approach.
In the Netherlands, voluntary euthanasia is regulated by the Termination of Life on Request and Assisted Suicide (Review Procedures) Act, which took effect in 2002.For euthanasia to be lawful under this legislation, the attending physician must: (a) be satisfied that the patient's request is voluntary and well-considered; (b) be satisfied that the patient's suffering is unbearable, with no prospect of improvement; (c) have informed the patient about his situation and his prognosis; (d) have come to the conclusion, together with the patient, that there is no reasonable alternative in the patient's situation; (e) have consulted at least one other, independent physician, who must see the patient and give a written opinion on whether the due care criteria set out in (a) to (d) have been fulfilled; (f) have exercised due medical care and attention in terminating the patient's life or assisting in his suicide. 16e legislation requires that the suffering must be unbearable with no prospect of improvement, but there is no requirement for this suffering to be due to a medical, as opposed to a psychiatric, condition.Nonetheless, the guidance states that it is 'usually necessary to consult an independent psychiatrist or another expert if the patient's suffering is caused by a psychiatric disorder', in order to 'obtain an assessment of whether the request is voluntary and well-considered and whether the suffering is without prospect of improvement'. 17lthough there is no requirement in the legislation that a valid wish to die must not be a symptom of a mental disorder, the judgment that the wish is attributable to a mental disorder might nonetheless be considered to the count against the claim that the request for euthanasia is 'voluntary and well-considered'.Accordingly, the validity of the wish to die may be considered to hinge on the possibility of dissociating the wish to die from the mental disorder.In order to understand why a wish's being attributable to a mental disorder might be considered to invalidate the wish, we need to look more closely at the relation between mental disorder and voluntary action.

Mental Disorder as an Invalidating Condition
As noted above, mental disorder is sometimes considered to undermine the voluntariness of an action or choice.A comprehensive treatment of the philosophical literature on voluntary action is beyond the scope of this article, but there are some features of voluntariness that are recognized in the medical ethics literature as being especially relevant to patient decision-making and mental disorder.Three of these features are described by Gerben Meynen. 18First, an action might be considered voluntary if the action is an expression of the person's character.This could mean two different things, which are self-determination and self-governance. 19An action is self-determined if the person's character is the source of the action. 20An action is self-governed if the person's character endorses the action. 21Second, an action might be considered voluntary if it occurs for an intelligible reason. 22This feature has been emphasized in recent compatibilist accounts of voluntary action. 23Third, an action might be considered voluntary if it is possible for the person to act otherwise. 24This is often associated with incompatibilism, but in recent years there have been compatibilist accounts that consider it to be an important feature of voluntary action. 25The second and third features can be thought of as being underpinned by the first, insofar as it is the person's character that engages in intelligible reasoning and deliberates about whether or not to act.
These three features are not uncontested, and philosophers disagree about which of them are relevant to the analysis of voluntary action.Nonetheless, in the medical ethical context, they are widely considered to be important in the clinical assessment of patient autonomy.There is increasing recognition in the medical ethics literature that a request's being an expression of the patient's character is an important consideration when judging whether or not the request is voluntary.This is often termed the authenticity of the patient's decision. 26As noted by Scott Kim, 'the more the patient's decision is at odds with who the patient is or is known to be, the stronger the case that the decision is not authentic'. 27The presence of an intelligible reason is also widely recognized as an important consideration when judging whether or not a patient's request is voluntary. 28For example, the MacArthur Competence Assessment Tool for Treatment, which is a standard tool for assessing the ability of a patient to make a voluntary decision, specifically includes the ability to give a reason for the decision as a criterion. 29The ability to choose otherwise is considered important in medical ethics, insofar as valid consent for an intervention requires the patient to understand the alternatives to the proposed intervention. 30This indicates that the clinician considers it important that the patient chooses the intervention over any alternative possibilities.And so, although they may not be necessary and sufficient conditions for voluntary behavior in general, these three features are considered to be appropriate minimal criteria for voluntariness in the medical ethical context of assessing the validity of a patient's decision.
Meynen suggests that disease can compromise the voluntariness of an action by undermining any of these three features. 31For example, a limb movement during an epileptic seizure may not be considered a voluntary action because the action is not an expression of the person's character, but an effect of the epileptic seizure.The action may also be considered involuntary because it is merely the causal outcome of the epileptic seizure and does not occur for an intelligible reason.Finally, the action may be considered involuntary because it is impelled by the epileptic seizure, such that the person could not act otherwise.Indeed, philosophers often couch the association between mental disorder and the loss of agency in terms of these three features.Galen Strawson suggests that responsibility can only be attributed to a person 'just so long as one is not caused to act by any of a certain set of constraints (kleptomaniac impulses, obsessional neuroses [. ..])', which implies that the character of a person suffering from kleptomaniac impulses or obsessional neuroses is not the genuine source of the act. 32Daniel Levy suggests that obsessive-compulsive disorder is associated with behavior that is 'mechanically dictated by stereotyped scripts', which implies that such behavior does not occur for an intelligible reasoning. 33According to Patricia Churchland, 'OCD [obsessive-compulsive disorder] patients often indicate that they wish to be rid of hand-washing or footstep counting behavior, but cannot stop', which implies that people with obsessive-compulsive disorder cannot act otherwise. 34Of course, most mental disorders do not undermine the ability to choose otherwise, but it is commonly supposed that there are at least some mental disorders that plausibly do, including severe cases of addiction and phobia. 35he above gives some indication of what underpins the worry that a wish to die may be deemed invalid if it can be attributed to a mental disorder.If the wish is attributable to the mental disorder, then the presumption is that the wish is not an expression of the person's character, but of the mental disorder.This would imply that the wish does not occur for an intelligible reason, because it is the product of the mental disorder rather than of the person's reasoning.In a particularly severe case, it might also be taken to imply that the person could not choose otherwise, inasmuch as a wish that is impelled by the mental disorder, if it is sufficiently strong, could override the person's deliberative control.The suggestion, then, is that a wish to die that is attributable to a mental disorder is not genuinely voluntary, which invalidates the request for voluntary euthanasia.
Although this line of thought has intuitive appeal, it relies on certain assumptions about the relations between the mental disorder, the wish to die, and the person's character.Notably, the suggestion that the wish can be an expression of the mental disorder rather than of the person's character assumes that the mental disorder is distinct from the person's character.It also assumes that the relation between the mental disorder and the wish is causal, such that the wish is something that is caused by the mental disorder.However, whether or not these assumptions obtain depends on the particular characteristics of the disorder in question, as well as on how the disorder is conceptualized.In the following section, I examine whether or not these assumptions obtain in the case of borderline personality disorder, with a view to establishing whether or not it is possible to dissociate the wish to die from the mental disorder for this diagnosis.
3. Can the Wish to Die Be Dissociated from the Mental Disorder?

A Descriptivist Ontology
In the case of a person diagnosed with borderline personality disorder expressing a wish to die, is the wish to die necessarily a symptom of the borderline personality disorder?In order to answer this question, we need to look at what the definition of borderline personality disorder is according to standard psychiatric practice.There has been a long-standing methodological debate in psychiatry about how best to define and classify mental disorders.Jennifer Radden notes that the dominant approaches have fallen broadly into two sorts, which she respectively calls 'descriptivist' and 'causal'. 36This way of dividing up the dominant approaches to conceptualizing mental disorders has also been used by other contemporary philosophers of psychiatry, including Dominic Murphy, Jonathan Tsou, and Elisabetta Lalumera. 37ccording to a descriptivist ontology, a psychiatric diagnosis refers to the cluster of observable symptoms exhibited by the person, without any allusion to the underlying causal structure.For example, in the case of a person who is diagnosed with borderline personality disorder on account of efforts to avoid abandonment, volatile relationships, unstable self-image, impulsiveness, and suicidal ideation, the diagnosis of borderline personality disorder under a descriptivist ontology simply refers to the conjunction of the person's efforts to avoid abandonment, volatile relationships, unstable self-image, impulsivity, and suicidal ideation.By contrast, according to a causal ontology, a psychiatric diagnosis refers to the underlying causal state that is believed to produce the cluster of observable symptoms.In the aforementioned example of the person diagnosed with borderline personality disorder, the diagnosis of borderline personality disorder under a causal ontology refers to the underlying state that is believed to produce the person's efforts to avoid abandonment, volatile relationships, unstable self-image, impulsivity, and suicidal ideation.Hence, a descriptivist ontology assumes a constitutive relation between the mental disorder and the cluster of symptoms, whereas a causal ontology assumes a causal relation between the mental disorder and the cluster of symptoms.
At this point, some clarification is required regarding the use of the term 'ontology'.Often in the philosophical literature, 'ontology' is used to refer to a theory of the fundamental constituents of the world.In the context of Radden's article, the 'ontology' of mental disorder pertains to what sorts of thing are picked out by psychiatric diagnoses.However, Radden's 'ontology' is not concerned with what are the fundamental constituents in the world that make up a mental disorder, but rather with what the conditions are that determine the meaning or reference of a mental disorder diagnosis.Hence, a descriptivist ontology of mental disorder, under Radden's account, does not claim that the symptoms of borderline personality disorder are fundamental parts that are irreducible to anything more basic, but rather that these symptoms are what determine the reference of the diagnosis of borderline personality disorder.Although these symptoms may have underlying causes, these causes are not included in the reference of the diagnosis.Radden clarifies this as follows: [. ..] ontological descriptivism is the view that categories such as depression so identified refer solely to those observable sign/symptom clusters and not to any underlying causal framework.This is not to assert that such signs and symptoms are uncaused, nor-although this will probably be true-that their causes are unknown, but merely that they are not part of the meaning, or reference, of depression. 38 order to keep my terminology consistent with that of Radden, I will be using the expressions 'descriptivist ontology' and 'causal ontology' as she uses them in her article.
As noted by Radden, Murphy, and Tsou, descriptivism remains the standard and most widely used approach to conceptualizing mental disorders in contemporary psychiatry. 39Notably, it is the approach that has been used by the editions of DSM since DSM-III: The approach taken in DSM-III is atheoretical with regard to etiology or pathophysiological process [. ...] This approach can be said to be 'descriptive' in that the definitions of the disorders consist of descriptions of the clinical features of the disorders.These features are described at the lowest order of inference necessary to describe the characteristic features of the disorder. 40rallel to the DSM system, the 10th edition of the World Health Organization's International Classification of Diseases (ICD-10) also uses a descriptivist approach that defines mental disorders in terms of clusters of symptoms. 41The purpose of this methodology is to establish a shared discourse whereby clinicians can reliably diagnose mental disorders on the grounds of observable clinical manifestations without being hindered by theoretical disagreements or uncertainties about the causes of these clinical manifestations.Accordingly, a person can be diagnosed with borderline personality disorder on account of efforts to avoid abandonment, volatile relationships, unstable self-image, and suicidal ideation, regardless of what had caused these symptoms.
An implication of this descriptivist ontology is that the presence of a symptom counts towards a diagnosis in virtue of its being one of the definitional criteria for the diagnosis and irrespectively of how or why that symptom had come about.As noted above, one of the DSM-5 diagnostic criteria for borderline personality disorder is the presence of suicidality, or '[r]ecurrent suicidal behavior, gestures, or threats'. 42And so, according to the standard descriptivist approach to mental disorder, the presence of suicidality counts towards the diagnosis of borderline personality disorder simply in virtue of its being one of the definitional criteria for the diagnosis.
To be clear, the claim here is not that the presence of suicidality is necessary or sufficient for a diagnosis of borderline personality disorder.A diagnosis of borderline personality disorder requires any five out of nine symptom criteria, of which suicidality is just one.While the presence of suicidality counts towards a diagnosis, it is neither necessary nor sufficient for the diagnosis, because the diagnosis could be made if other combinations of symptom criteria are met.Hence, a person might lack suicidality and still be appropriately diagnosed with borderline personality disorder.Rather, the claim here is that suicidality which occurs in the presence of a diagnosis of borderline personality disorder must be taken to be part of the borderline personality disorder if a descriptivist ontology is assumed.That is to say, under the standard descriptivist approach, if (a) A is diagnosed with borderline personality disorder and (b) A is exhibiting suicidality, then necessarily A's suicidality constitutes a symptom of A's borderline personality disorder.This is because the descriptivism of DSM-5 does not provide an option for excluding a person's suicidality from the cluster of symptoms that constitutes the person's borderline personality disorder.If we want to make a case that the person's suicidality should not be included as a symptom of the person's mental disorder, then we would need to appeal to further considerations that are not supplied by the descriptivist definition in DSM-5.
The relevance of the above to the debate about psychiatric euthanasia is that a descriptivist ontology of mental disorder precludes the possibility of dissociating a © 2020 The Authors.Journal of Applied Philosophy published by John Wiley & Sons Ltd on behalf of Society for Applied Philosophy person's wish to die from the person's mental disorder in the case of borderline personality disorder.Given that an enduring wish to die is an indication of suicidality, and suicidality is one of the symptom criteria for borderline personality disorder, the enduring wish to die must be classed as a symptom of the person's borderline personality disorder if a descriptivist ontology of mental disorder is assumed.Furthermore, given that the cause of the symptom does not matter for the purpose of making the diagnosis, the suicidality constitutes part of the illness irrespective of whether or how it is causally related to the person's other symptoms.

A Causal Ontology
Under the standard descriptivist approach that is widely used in contemporary psychiatry, an enduring wish to die that occurs in the context of a diagnosis of borderline personality disorder necessarily constitutes a symptom of the borderline personality disorder in virtue of its being one of the definitional criteria for the diagnosis.For it to be theoretically possible to dissociate the wish to die from the mental disorder, we require an ontology of mental disorder that suggests that the wish to die could at most only be contingently a symptom of the mental disorder.A causal ontology can accommodate this contingency.
According to a causal ontology, a diagnosis refers not to a cluster of observable symptoms, but to the underlying causal state that is believed to produce the cluster of symptoms. 43This is the kind of ontology that is often, though not always, assumed in modern medicine. 44For example, chickenpox is associated with a rash, fever, and malaise, but the diagnosis of chickenpox does not refer to the conjunction of rash, fever, and malaise.Rather, chickenpox refers to a varicella zoster infection, which is the underlying causal state that produces the rash, fever, and malaise.
In psychiatry, there have been strong calls to move away from the descriptivism of DSM-5 and toward a causal ontology of mental disorder. 45The hope is that this could bring psychiatry in line with the rest of medicine.With respect to borderline personality disorder, there have been attempts to characterize the disorder causally at different levels, including in terms of its psychodynamic, cognitive, and neurobiological features. 46However, given the multifactorial complexity and causal heterogeneity associated with the diagnosis, it is unlikely that the condition can be characterized at any single level of analysis.Rather, a causal ontology of borderline personality disorder is likely to require a pluralistic characterization that includes factors at social, psychological, and biological levels, as well as a concession that many of the factors that enable the relevant pattern of behavior are relational processes in the interpersonal enivornment rather than factors internal to the individual. 47his proposed shift from standard descriptivism to a causal ontology marks a change in how we conceptualize the relation between symptom and disease.At least since David Hume's analysis of causation in An Enquiry Concerning Human Understanding, it has generally been accepted, at least in philosophy, that causal connections are contingent. 48And so, an implication of a causal ontology is that the symptoms of a disease are only contingently connected to the disease.For example, a varicella zoster infection might be the cause of someone's rash, fever, and malaise, but it is possible to have a varicella zoster infection without having a rash, fever, and malaise, or to have a rash, fever, and malaise without having a varicella zoster infection.This contingent connection makes it theoretically possible for a given symptom and the disease to both be present, but nonetheless be causally unconnected.For instance, a person could have a varicella zoster infection and a rash, but the rash could be caused by something other than the varicella zoster infection, such as an allergic drug reaction.In this case, the rash is not a symptom of the varicella zoster infection, even though it occurs in the presence of the varicella zoster infection.
Likewise, assuming a causal ontology of mental disorder opens up the theoretical possibility of dissociating a person's wish to die from the person's mental disorder.In contrast to a descriptivist ontology, which suggests that suicidality in the presence of a diagnosis of borderline personality disorder must be classed as a symptom of the illness due to its constituting part of the definitional criteria for the disorder, a causal ontology suggests that it is theoretically possible for suicidality to occur in the presence of a mental disorder without its necessarily being attributable to the mental disorder.Just as a rash in a case of varicella zoster infection could have been caused by something other than the varicella zoster infection, it could be speculated that the wish to die in a case of a mental disorder could have been caused by something other than the mental disorder.For example, it might be suggested that the wish to die arose through a process of intelligible reasoning that can be attributed to an aspect of the person's character that is unaffected by the mental disorder.
At initial glance, a causal ontology of mental disorder might seem to provide a framework that allows us to determine whether or not a mentally ill person's wish to die is attributable to the person's mental disorder, which could have normative significance for the debate about psychiatric euthanasia.However, I argue that an appropriate causal ontology that allows this possibility is implausible in the case of borderline personality disorder.This is because the characterization of borderline personality disorder as a persistent and pervasive condition calls into doubt whether it is possible to portion off an aspect of the person's character that is unaffected by the condition.We saw earlier that the DSM-5 diagnosis of borderline personality disorder requires the pattern of behavior to be 'stable across time and consistent across situations'. 49Also, the ICD-10 defines a personality disorder as 'a severe disturbance in the characterological constitution and behavioral tendencies of the individual'. 50The depiction of borderline personality disorder as a disorder of a person's characterological constitution that is consistent across situations suggests that it is inextricably connected to the person's character.The implication is that in the case of someone diagnosed with borderline personality disorder who is seeking voluntary euthanasia, the person's wish to die is almost inevitably influenced by the mental disorder, because the mental disorder pervades the person's character.As Carl Elliott suggests, 'if a personality can be compared to a machine, that machine is not one that is separate from and external to a person's identity [. ..] to change a personality is to change the person herself'. 51f course, this is not to say that everything a person with borderline personality disorder does is a symptom of his or her mental disorder.Rather, it is to say that there are certain kinds of thought and behavior which cannot straightforwardly be disentangled from the mental disorder, including suicidal ideation and self-harm.This is because these kinds of thought and behavior reflect the particular cognitive and affective dispositions that are supposed to underpin the clinical manifestations of the disorder.These dispositions include the tendencies for dichotomous thinking, catastrophizing, negative self-evaluation, and internalization of failure. 52Insofar as these dispositions can induce and deepen distressing affective states, they make suicidal ideation and self-harm more likely.Accordingly, they can partly account for why suicidal ideation and self-harm are more prevalent in people with borderline personality disorder than in people without borderline personality disorder.However, people with borderline personality disorder also have many beliefs, interests, and desires that do not reflect the aforementioned dispositions.Given that these are no more prevalent in people with borderline personality disorder than in people without borderline personality disorder, they can more straightforwardly be disentangled from the mental disorder.
The characterization of borderline personality disorder as a persistent and pervasive condition contrasts it with more circumscribed conditions which can be more easily demarcated from the core characters of their sufferers.For example, the distinct episodes of mania and depression associated with bipolar disorder are considered to involve changes in the person's character for limited periods of time.During periods of remission, the person's core character is considered to remain intact.In the case of a phobia, the person's thought and behavior are influenced in a certain context involving a specific stimulus, but the person's core character is considered to be intact in other contexts. 53In such a case, the mental disorder can be characterized as being separable from the core character of its sufferer, thus making it theoretically possible to distinguish a desire that is causally influenced by the mental disorder from a desire that is not causally influenced by the mental disorder.However, this is not possible in the case of borderline personality disorder, which is supposed to pervade the person's characterological constitution.As we shall see, this has implications for the question of whether or not a wish is invalidated by its being attributable to a mental disorder.
So far, I have shown that neither of the main approaches to conceptualizing mental disorder in contemporary psychiatry allows the possibility of dissociating the wish to die from the mental disorder in the case of borderline personality disorder.Under a descriptive ontology, the wish is classed as part of the borderline personality disorder in virtue of its being one of the definitional criteria for the diagnosis.While a causal ontology might open up the theoretical possibility of dissociating a person's wish to die from the person's mental disorder, an appropriate causal ontology that allows this possibility is unavailable for borderline personality disorder.This is due to its being inextricably connected to the person's character.And so, under the current approaches to conceptualizing borderline personality disorder, a wish to die that occurs in the context of a diagnosis of borderline personality disorder prima facie constitutes a symptom of the borderline personality disorder.In the following section, I consider the normative significance this has for whether or not a wish's being a symptom of a mental disorder invalidates the wish in the context of psychiatric euthanasia.

The Descriptivist Reading
Depending on whether a descriptive ontology or a causal ontology of psychiatric disorder is assumed, the claim that something is attributable to a mental disorder can mean different things.Under a descriptive ontology, 'S is a symptom of D' translates to 'S is a constitutive part of D'.Under a causal ontology, 'S is a symptom of D' translates to 'S is a causal outcome of D'.In what is to follow, I consider whether or not either of these readings supplies a justification for the normative judgment that an enduring wish to die that occurs in the context of borderline personality disorder is invalid.
Let us begin with the descriptivist reading.Recall that under a descriptivist ontology, the wish to die is a symptom of borderline personality disorder in virtue of its being one of the diagnostic criteria that make up the mental disorder.Importantly, such a descriptivist ontology makes no allusion to the causal structure underlying the observable syndrome.Hence, it does not matter how or why the wish to die came about for it to be classed as a symptom of the mental disorder.
The implication of the above is that a diagnosis of borderline personality disorder does not, if a descriptivist ontology is assumed, provide an explanation of the person's wish to die.The diagnosis serves as a descriptor for the presence of the wish to die in conjunction with other symptoms, but it does not tell us what had caused the wish or the reason underlying the wish.As Radden notes about descriptivism in the case of major depressive disorder, '[i]n spite of the commonplace and seemingly irresistible tendency to see explanatory advantage in the assertion that the symptoms of depression are caused by depression, if we accept descriptivism, there is none.' 54 This suggests that a wish's being a symptom of borderline personality disorder does not supply any substantive justification for considering the wish to be invalid according to the descriptivist reading.Recall from the earlier discussion about the relation between mental disorder and voluntary action that a mental disorder may be considered to undermine the voluntariness of a wish if it indicates that the wish is not an expression of the person's character, that the wish does not occur for an intelligible reason, or that the person could not have chosen otherwise. 55However, if a descriptivist ontology is assumed, then the diagnosis of borderline personality disorder does not indicate any of these.Because such a descriptivist diagnosis makes no allusion to the causal structure underlying the cluster of symptoms, it does not tell us anything about the authenticity of the wish, the intelligibility of the reasoning behind the wish, or whether or not the person could wish otherwise.The wish's being a symptom of borderline personality disorder is entirely compatible with its being voluntary according to the three aforementioned features.
And so, under a descriptivist ontology, the claim that a wish to die is a symptom of a mental disorder has no bearing on the question of whether or not it is valid.In the case of a person with a diagnosis of borderline personality disorder who expresses a wish to die, the diagnosis serves as a mere descriptor for the presence of the wish to die in conjunction with other symptoms.As a mere descriptor with no causal explanatory content, it does not provide any justificatory support for deeming the wish to be invalid.It appears, then, that such justificatory support depends on extradescriptivist considerations that are not supplied by the descriptivist reading.

The Causal Reading
Let us now turn to the causal reading.Under a causal ontology, a wish is attributable to a mental disorder in virtue of its being a causal outcome of some state that is supposed to constitute the mental disorder.However, mental disorder may cause a behavior in different ways, which are not always adequately distinguished.For example, in forensic psychiatry, it is suggested that a behavior can be excused if it is caused by a mental disorder, but what is meant by 'caused by a mental disorder' is often not qualified. 56Hence, further analysis is needed to clarify which kinds of causation by mental disorder undermine voluntariness and which kinds do not.
Radden distinguishes two different ways in which behavior can be the causal outcome of a disease: (10) 'Direct Effect' cases where the disease or condition directly results in the particular motive or desire (e.g. a hostile impulse) occasioning the [. ..] action.(11) 'Indirect Effect' cases where the disease or condition is responsible for a pervasive personality change, and as a result of that change, which rearranges and affects attitudes, beliefs, emotions, and sometimes even memories, the [. ..] action is taken. 57though Radden's distinction is in the context of a discussion about criminal behavior, I suggest that can be applied to behavior more broadly, including suicidal behavior.This is in no way meant to suggest that suicidal behavior is morally equivalent to criminal behavior, but it is merely to acknowledge that Radden's distinction has more general applicability beyond its original intended use.Direct-effect cases include cases of delirium, intoxication, and head injury.Radden argues that in such a direct-effect case, the sufferer is deemed not to be responsible for his or her action because the effect of the disease amounts to ignorance or compulsion.Indirect cases include chronic and progressive conditions, where the psychological effects of prolonged suffering and the physiological effects of disease progression result in enduring changes in the values, beliefs, and attitudes of the sufferers.Such indirect-effect cases may not involve ignorance or compulsion.Radden argues that in those cases where ignorance or compulsion is absent, we would not be inclined to excuse the sufferers from being responsible for their actions.We can couch this in terms of the features of voluntary action discussed earlier.In a direct-effect case, the action is caused by the disease independently of the person's character.Therefore, the action is not voluntary because the person's character does not produce or endorse the action.Also, the fact that the action is not an expression of the person's character but of the disease implies that it cannot be the result of intelligible reasoning and that the person could not have acted otherwise.Hence, as Matt King and Joshua May note, '[a] mental disorder plausibly excuses, then, if its symptoms yield an action that entirely bypasses one's agency, such that an outcome is the result of no action at all or an entirely unintentional one.' 58 By contrast, in an indirect-effect case, the disease influences the person's character, which in turn produces the action.Given that the person's character still produces and endorses the action in this kind of case, the action cannot be deemed involuntary on the basis of its relation to the person's character.Rather, it can only be deemed involuntary if it lacks an intelligible reason or if the person could not act otherwise, neither of which are inevitable in an indirect-effect case.
Let us apply the above to the discussion about voluntary euthanasia.I suggest that a wish to die can be deemed invalid in the first kind of case, where the disease directly produces the suicidal desire.This is because the wish is not an expression of the person's character, which in turn implies that the wish does not occur for an intelligible reason.In the second kind of case, where the disease indirectly results in the wish to die by influencing the person's character, I suggest that we would not be inclined to consider the person's wish to be invalid if it can be shown that the wish occurs for an intelligible reason and that the person could have chosen otherwise.Of course, it is possible that such a wish might lack an intelligible reason or that the person is unable to choose otherwise.However, the lack of an intelligible reason and the inability to choose otherwise do not inevitably follow from the wish's being an indirect effect of the disease.Rather, they need to be independently demonstrated.
Indeed, it is often the case that a wish to die that is an indirect effect of a disease is genuinely voluntary.Recall that the vast majority of requests for voluntary euthanasia are motivated by suffering associated with progressive medical conditions.In the case of a person with a progressive medical condition, the wish to die is an indirect effect of the disease, insofar as suffering from the disease is what influenced the person's character to acquire the values, beliefs, and attitudes that occasion the wish to die.Had the person not had the disease, he or she would not have requested to die.However, in many cases, we would agree that these wishes are underpinned by intelligible reasons and that the sufferers are able to consider alternative choices.Accordingly, we would not be inclined to deem these wishes to be invalid, unless one is morally opposed to voluntary euthanasia for other reasons.This suggests that a wish's being an indirect effect of a disease is insufficient to justify judging the wish to be invalid.
Sometimes, the wish to die may be a direct effect of a mental disorder.For example, in a case of a person currently suffering from acute schizophrenia, suicidality may be the direct product of passivity symptoms and disorganized thought.In such a case, the wish to die may be deemed invalid, because the passivity symptoms suggest that the wish is not an expression of the person's character and that the person could not have chosen otherwise, while the disorganization of thought suggests that the wish lacks an intelligible reason.Likewise, a wish to die in a person with bipolar disorder who is currently suffering from a severe depressive episode may be deemed invalid, because the wish is thought to result from the depressive state rather than from the person's character.And so, direct-effect cases require the disease states to be distinct from the core characters of the sufferers.The enduring wish to die that is associated with borderline personality disorder, however, is closer to being an indirect effect of the mental disorder, insofar as the mental disorder involves the person's character and the wish is a result of the person's character.As noted earlier, borderline personality disorder is supposed to be a condition that pervades the person's characterological constitution.This suggests that the borderline personality disorder cannot be straightforwardly demarcated from the person's character.
The supposed inseparability of borderline personality disorder from the person's character suggests that a wish to die that is a symptom of borderline personality disorder cannot be deemed to be involuntary on the basis of its relation to the person's character.Given that borderline personality disorder is supposed to be a property of the person's character, the claim that a wish to die is a symptom of borderline personality disorder implies that the wish to die is an expression of the person's character.Similar claims have also been made regarding moral responsibility and personality disorder.Barbara Wootton suggests that to excuse a person from responsibility because he or she has a personality disorder is simply to excuse the person because he or she has a badly behaving character. 59King and May argue that 'an agent is more responsible the more reflective of their moral selves their action is or the better integrated its motivations' and that some disorders 'may be more difficult to dissociate from one's real self'. 60The implication here is that an action cannot be excused merely on the basis of its being the result of a personality disorder, because the inability to dissociate the personality disorder from the person's character suggests that the action is reflective of the person's motivation.Therefore, while a wish's being a symptom of a mental disorder may be sufficient to undermine the validity of the wish in a direct-effect case, it is insufficient in an indirect-effect case such as borderline personality disorder, where the wish's being a symptom of the mental disorder amounts to its being an expression of the person's character.
Of course, we might be able to deem a wish to die that occurs in the context of borderline personality disorder invalid on other grounds.For example, in the case where a person with borderline personality disorder exhibits impulsive suicidal behavior during an acute dissociative episode, the wish to die may be appropriately judged to be involuntary because it does not seem to occur for an intelligible reason and the person could not have acted otherwise.However, what is important to note here is that the lack of an intelligible reason and the inability to act otherwise do not follow from the mere fact that the wish to die is an effect of borderline personality disorder.Indeed, many of the choices and actions of people diagnosed with borderline personality disorder are associated with intelligible reasoning and control, as Hanna Pickard notes: The behaviors that are constitutive of PD [personality disorder] are not mere bodily movements.They are kinds of action: the kinds of behavior over which we have choice and control.On the whole, PD service users possess both relevant capacities with respect to these behaviors [. ...] The evidence for this claim is relatively straightforward.service users routinely do choose to behave otherwise and alter entrenched patterns of behavior, when they have incentive, motivation, and genuinely want to do so. 61d so, a wish's being an effect of borderline personality disorder does not by itself indicate that the wish is invalid.Rather, to invalidate the wish, we would need to demonstrate independently that the person lacked an intelligible reason and could not have chosen otherwise.Once these have been established, the claim that the wish is attributable to borderline personality disorder carries no further justificatory weight.
The implication for psychiatric euthanasia in the case of borderline personality disorder is that the assessment of a person's request to die should, as recommended by the Termination of Life on Request and Assisted Suicide (Review Procedures) Act, focus on establishing whether the person's request is 'voluntary and well-considered', whether there is any prospect of improvement in the person's condition, and whether there is any reasonable alternative that could ameliorate the person's suffering. 62This is in line with what is recommended by the psychiatrists Werdie van Staden and Christa Kr€ uger regarding patient decision-making in general when they note that 'the clinical assessment of a particular patient's capacity to give informed consent in a case of mental disorder is better informed by the consideration of conditions necessary to give informed consent than by making inferences from the general features implied by a specific diagnosis'. 63Such a recommendation is especially relevant in an indirect-effect case such as borderline personality disorder, where the wish's being a symptom of the mental disorder does not preclude its being an expression of the person's character.

Conclusion
'Aurelia Brouwers argued she was competent to make the decision.But could a death wish have been a symptom of her psychiatric illness?' 64 In this article, I have argued that our current approaches to conceptualizing mental disorder do not allow a wish to die to be dissociated from the mental disorder in the case of borderline personality disorder.Under a descriptivist ontology, the wish to die is classed as a symptom of the borderline personality disorder by definition.Under a causal ontology, the persistence and pervasiveness of borderline personality disorder preclude the possibility of portioning off a wish to die that is not causally influenced by the mental disorder.
Nonetheless, I have argued that in the case of borderline personality disorder, the inability to dissociate a wish to die from the mental disorder does not have the normative implication it is sometimes assumed to have.Under a descriptivist ontology, the diagnosis of borderline personality disorder serves as a descriptor for the wish to die in conjunction with other symptoms, but it does not supply any information that could justify deeming the wish to be invalid.Under a causal ontology, a mental disorder can undermine the validity of the wish in a direct-effect case where it can be argued that the wish is not an expression of person's character.However, in an indirect-effect case such as borderline personality disorder, where the mental disorder is conceptualized as being inextricably connected to the person's characteriological constitution, the wish's being a symptom of the mental disorder amounts to its being the expression of the person's character.Accordingly, it fails to provide an independent justification for judging the wish to be invalid.The implication is that the assessment of a request for psychiatric euthanasia in the case of borderline personality disorder should focus on determining whether the request is voluntary, whether it is reasonable, and whether there is any reasonable alternative that could ameliorate the suffering, rather than on trying to determine whether or not the wish is a symptom of the mental disorder.

©
2020 The Authors.Journal of Applied Philosophy published by John Wiley & Sons Ltd on behalf of Society for Applied Philosophy Htut Maung, Department of Philosophy, School of Social Sciences, University of