#moral language

Introduction

It is widely agreed that moral judgment should play no role in the practice of medicine. The rationale is straightforward. Moral judgment, understood as the attribution of moral traits to an agent, can impair clinical judgment in ways that might harm the patient. Moral judgment’s capacity to impair clinical judgment is especially clear when the attributed traits are morally disvalued traits, such as self-indulgence or laziness.1Conditions associated with unhealthy “life-styles” are a case in point. A 350 pound man suffering from severe osteoarthritis wants something “really strong” for his pain, maybe some Percocet. The doctor recommends Celebrex instead and mentions in passing that the man’s “obesity” is a large part of the problem. The extra weight placed on his hips and knees is exacerbating the arthritis. However, the doctor doesn’t mention any of the more serious health risks associated with obesity, as she believes that people who are overweight are overweight because of the bad choices they knowingly make. She doesn’t buy into the “disease model” of obesity that so many of her colleagues do. The man, who would have been an excellent candidate for bariatric surgery, dies several years later as a result of complications brought on by his excessive weight.

In other cases, it might be the patient’s personality, rather than her medical condition, that elicits the negative moral judgment. Maybe she’s one of those patients who tends to show up, unannounced, at the doctor’s office, always with the vaguest of complaints, none of which ever seems to merit medical attention. The woman insists that something “really weird” is going on and it’s frightening her. The “palpitations” are back but this time she’s also feeling nauseous. She even has the hiccups!2 The doctor, inwardly rolling his eyes at the self-dramatization, asks that the woman remain in the waiting room until he finds time to “squeeze her in.” By the time she’s being seen in the emergency room, several hours later, the stroke has already caused considerable (and irreversible) damage.

In the case of psychiatric conditions, moral judgment’s capacity to impair clinical judgment can be especially troublesome. Such judgment might prevent the clinician from developing the empathy necessary for effective psychotherapeutic treatment. As Nancy Potter [2] explains, “When clinicians view patients’ primary character as morally objectionable, it’s difficult for clinicians to feel empathy and for patients to either receive or elicit it” ([2], 155). There is thus an important contrast between moral judgment in psychiatric cases and moral judgment in non-psychiatric cases. A physician who judges her morbidly obese patient as self-indulgent is not thereby prevented from recommending bariatric surgery or consultation with a nutritionist. A physician who judges his hypertensive patient as a “drama queen” is not thereby prevented from sending her straight to the emergency room on the off chance that she may have suffered a stroke. Yet a psychotherapist who judges her Borderline patient as a “pain in the ass” [2] is going to have considerable difficulty developing the empathy necessary for effective psychotherapeutic treatment.

Potter’s particular focus is on Borderline Personality Disorder and the manipulative behavior with which that condition is associated. However, as Potter points out, manipulative behavior is not a trait that appears in the DSM diagnostic criteria for Borderline Personality Disorder. It would thus be possible, at least in principle, to diagnose a patient with that disorder without ever conceptualizing them as “manipulative.” However, several of the traits appearing in the diagnostic criteria for Borderline Personality Disorder are expressed in overtly moral3 language. These include “impulsivity” and “inappropriate anger.” The same is true of the other “Cluster B” Personality Disorders: Antisocial, Histrionic, and Narcissistic. “Deceitfulness,” “self-dramatization,” and “sense of entitlement” are among the traits featured in the diagnostic criteria for these disorders. Indeed, given the current DSM criteria, diagnosis with a Cluster B Personality Disorder is not even possible without the attribution of morally disvalued traits to the diagnosed patient [4, 5].

This leads immediately to a problem: How is the clinician to avoid moral judgment in cases where a disorder’s diagnostic criteria are expressed in overtly moral language? 4 How, for instance, is the clinician to avoid judging her Antisocial patient morally if accurate diagnosis requires that she label him as “consistently irresponsible,” “impulsive,” and “deceptive”?

After considering and rejecting a couple of initially promising responses to this question, I suggest that the answer lies in disambiguating the notion of moral judgment so as to respect the distinction between having morally disvalued traits and having moral responsibility for those traits. Potentially harmful attitudes arguably result from neglecting this distinction: from assuming that morally disvalued traits are traits for which the agent invariably has moral responsibility.

In developing and defending this view, I make use of H.P. Grice’s [6] distinction between “conventional” and “conversational” implicature. I begin by drawing attention to the sorts of morally judgmental attitudes elicited in non-clinical settings involving the attribution of Cluster B traits. I have in mind ordinary everyday settings where someone might be characterized as “deceptive” or “impulsive,” or as having a “sense of entitlement” or a tendency to “self-dramatization.” I then suggest that the morally judgmental attitudes that tend to accompany such characterizations are usefully construed in terms of Grice’s notion of conversational (vs. conventional) implicature. The view, in essence, is that any morally judgmental attitudes that might accompany application of Cluster B diagnostic criteria are not entailed by the language of those criteria; they are instead implicated by the use of that language in particular conversational settings. I go on to point out that the conversational settings provided by clinical (vs. ordinary) contexts are not such as to implicate morally judgmental attitudes when morally laden diagnostic criteria are applied.

After setting out the proposed view, I compare and contrast it with Potter’s [2] view on how to reduce moral judgment (of the sort associated with moral responsibility) in clinical settings. I conclude with a few brief remarks on the irony of retaining overtly moral language in a manual for the diagnosis of ostensibly “medical” disorders.

The Problem

Let’s begin by taking a closer look at the problem at hand, which can be articulated as follows:

How is it possible to avoid moral judgment when diagnosing a patient with a Cluster B Personality Disorder, given the overtly moral language in which the diagnostic criteria for those disorders are expressed?5

In order to appreciate the nature of the problem, it is important to look closely at the diagnostic criteria for the disorders in question: the Cluster B Personality Disorders. First, however, let’s consider briefly the more general notion of a “Personality Disorder.” According to the current edition of the DSM, the “essential feature” of a Personality Disorder,

[is an] enduring pattern of inner experience and behavior that deviates markedly from the expectations of the individual’s culture and is manifested in at least two of the following areas: cognition, affectivity, interpersonal functioning, and impulse control...This enduring pattern is inflexible and pervasive across a broad range of personal and social situations...and leads to clinically significant distress or impairment in social, occupational, and other important areas of functioning ([7], 686)...

The DSM-IV-TR divides Personality Disorders into three “clusters”: Cluster A (odd or eccentric), Cluster B (dramatic, emotional, or erratic), and Cluster C (anxious or fearful). Because it is only the Personality Disorders in Cluster B whose diagnostic criteria are expressed in overtly moral language, it is those disorders on which I will be focusing.

Let’s now have a quick look at some of the diagnostic criteria for each of the four Cluster B Personality Disorders: Antisocial, Borderline, Histrionic, and Narcissistic. The diagnostic criteria for Antisocial Personality Disorder include “deceitfulness” and “consistent irresponsibility.” Those for Borderline Personality Disorder include “impulsivity” and “inappropriate anger.” Those for Histrionic Personality Disorder include “inappropriate sexually seductive behavior” and “self-dramatization.” Those for Narcissistic Personality Disorder include “grandiose sense of self-importance” and “sense of entitlement.”

Perhaps not surprisingly, persons suffering from Cluster B Personality Disorders tend not to seek treatment for those conditions [8].6 While they might recognize that their lives are problematic in various ways, they are unlikely to connect this to their own thoughts, feelings, or behaviors. As they see it, their problems are caused by other people or by personal misfortune; perhaps they’ve simply been dealt a “bad hand.” In short, such individuals tend to lack “insight”: while they might recognize that there is a problem of sorts, they fail to recognize that the source of that problem is their own disordered personality. Because those suffering from Cluster B Personality Disorders tend not to seek treatment for those conditions, one might suspect that the problem of how to diagnose those disorders without the intrusion of moral judgment is more theoretical than practical. It’s a problem that would arise only if, contrary to fact, persons with Cluster B Personality Disorders tended to seek treatment for those conditions.

In fact, however, those with Cluster B Personality Disorders are well-known to clinicians. They sometimes seek help for other conditions, such as anxiety, depression, or substance abuse [9]. They are sometimes brought to the clinician’s attention by frustrated family members, who have no trouble recognizing that the problem is with the disordered personality of the individual in question. Those suffering from Antisocial Personality Disorder, in particular, are well-known to forensic psychiatrists [10]. This should hardly come as a surprise given the considerable overlap between behaviors that are antisocial and behaviors that are criminal, something which would explain the overrepresentation of Antisocials in the prison population [10]. There is a parallel here with patients suffering from schizophrenia and other mental disorders characterized by psychosis. Those suffering from schizophrenia do not generally seek treatment for their mental disorder, as they don’t generally recognize themselves as having a condition amenable to psychiatric treatment: they lack “insight” [11]. Nevertheless, they are well-known to clinicians. They are often brought to the attention of clinicians by concerned family members or friends and, because actions stemming from psychosis are sometimes criminal,7 patients suffering from schizophrenia are not unknown to forensic psychiatrists.

The problem of how to diagnose accurately without the intrusion of moral judgment is therefore a practical as well as a theoretical problem. Before putting forth my own proposal for dealing with this problem, I would like to consider two alternative proposals, both of which effectively concede that the moral language of the Cluster B diagnostic criteria is out of place in a manual for the diagnosis of psychiatric (or, more generally, medical) conditions.

The Eliminativist Proposal: Eliminate the Cluster B Personality Disorders from the DSM

Perhaps the most obvious, and certainly the simplest, strategy would be to eliminate the problem by eliminating the Cluster B Personality Disorders from the DSM. One might attempt to justify such a proposal on the grounds that the disorders in question are not in fact mental disorders but moral disorders. On behalf of such a view, one might propose, in the spirit of Thomas Szasz [12], the following argument: Given their current diagnostic criteria, virtually all of which involve deviations from moral norms, there is no reason to suppose that the Cluster B Personality Disorders are anything other than moral disorders, in which case they simply do not belong in a manual for the diagnosis of mental disorders.8 This is the very sort of argument recently proposed by bioethicist Louis Charland [4, 5] who, in contrast to Szasz, does believe that mental disorders exist. He simply questions the DSM assumption that the Cluster B Personality Disorders are mental rather than moral disorders.9

My response to any such argument is three-fold. First, the Cluster B Personality Disorders are arguably DSM “mental disorders” despite the overtly moral language of their diagnostic criteria. Second, there is empirical evidence suggesting that the Cluster B Personality Disorders, at least some of them, are associated with underlying neurological dysfunction [3, 14], thereby lending credence to the idea that they are genuine mental disorders, perhaps even neurological disorders.10 Third, the fact that some of the disorders in question occasionally respond well to standard forms of psychiatric treatment suggests, without proving, that they are indeed mental disorders. Let’s consider these three points in turn, beginning with the first.

In assessing whether a disputed condition is a “mental disorder,” one must look beyond the condition’s diagnostic criteria. One must consider, in particular, whether satisfaction of those criteria yields a condition that simultaneously satisfies the criteria for being a mental disorder. According to the current DSM definition, a “mental disorder” is,

...a clinically significant behavioral or psychological syndrome or pattern that occurs in an individual and that is associated with present distress...or disability...or with a significantly increased risk of suffering death, pain, disability, or an important loss of freedom...Whatever its original cause, it must currently be considered a manifestation of a behavioral, psychological, or biological dysfunction in the individual ([7], xxxi)...

Although the details of this definition are subject to dispute, I am going to take it as providing sufficient conditions, however vague, for a disorder’s being a “mental” disorder. My rationale is simple: It represents the consensus, insofar as there is one, of the relevant experts: mainstream American psychiatrists. I am also going to assume, perhaps more controversially, that the Cluster B Personality Disorders, all four of them, do in fact meet the conditions specified in the current DSM characterization of “mental disorder.” My rationale for this latter assumption is two-fold. First, that the disorders in question (and the Personality Disorders more generally) are mental disorders presumably represents the consensus of the relevant body of experts: mainstream American psychiatry.11 That this is indeed the consensus is evident from the inclusion of the Cluster B Personality Disorders in the DSM, the American Psychiatric Association’s manual for the diagnosis of “mental disorders.”12 Second, expert consensus is, in this particular case, supported by common sense. Given the diagnostic criteria associated with the Cluster B Personality Disorders, it doesn’t require much imagination to see that individuals suffering from these disorders might be at increased risk for distress, disability, early death, or loss of freedom. The Antisocial ends up dead at age 25, murdered by a rival drug dealer. The Borderline dies of an overdose, her fifth suicide attempt in as many years. The Histrionic becomes clinically depressed after a series of disastrous affairs with married men who decided they wanted to stay married. The Narcissist, a once successful CEO, is sentenced to 50 years in prison after decades of “shady” business practices that turned out to be as illegal as they were immoral. The point here is a simple one: Chronic violations of moral (and sometimes legal) norms of the sort associated with the Cluster B Personality Disorders are likely to lead to the very sorts of conditions (distress, disability, early death, loss of freedom) specified in the DSM criteria for “mental disorder.”

Such a view is unlikely to move the Szaszian, who might respond with something to the effect that “what goes around comes around.” Difficult people, people who chronically and flagrantly violate moral norms, can hardly be expected to have easy lives, lives relatively unencumbered by distress or disability, or by the threat of premature death or incarceration. The fact that those diagnosed with Cluster B Personality Disorders might have “mental disorders” as defined by mainstream American psychiatry, suggests only that mainstream American psychiatry is either unwilling or unable to recognize so-called “mental disorders” for what they are: moral, as opposed to medical, problems.13

This response might have some force were it not for considerations having to do with the etiology of the Cluster B Personality Disorders. This brings us to a second reason for thinking that those disorders are genuine mental disorders: their association with detectable brain abnormalities (and thus with “biological dysfunction”).14 Imaging studies involving Antisocial patients reveal clear evidence of neurological dysfunction, as do imaging studies involving Borderline patients [14].15 As Szasz himself [12] reasonably suggests, evidence that a so-called “mental disorder” is associated with neurological dysfunction (or “brain lesions”) is evidence that the disorder is indicative of16 genuine illness, amenable in principle to medical treatment.

One might object that neurological differences are being identified (by neuroscientists) as “dysfunctions” only because they reinforce antisocial attitudes and behaviors. Thus, the question as to their putative status as genuine dysfunctions is effectively begged. However, one might reasonably counter that such differences rightly count as dysfunctions precisely because they reinforce the kinds of antisocial attitudes and behaviors that give rise to DSM “mental disorder” in the first place. We thus emerge with conditions (“mental disorders”) that, by definition, merit psychiatric treatment. This brings us to a third and final reason for retaining the Cluster B Personality Disorders in the DSM: their amenability to psychiatric treatment.

Although (DSM) Personality Disorders are notoriously difficult to treat, some Cluster B patients respond well to psychiatric treatment, psychotherapeutic as well as psychotropic [15]. For instance, some Borderline patients respond well to cognitive behavioral therapy, and to Dialectic Behavior Therapy in particular [16]. Additionally, the aggression associated with Antisocial Personality Disorder sometimes responds well to antidepressant medication [17], as does the anger characteristic of Borderline Personality Disorder [18]. The fact that some patients with Cluster B Personality Disorders respond well to psychotherapy and/or psychotropics suggests, although certainly doesn’t prove, that those conditions are in fact mental disorders and thus right at home in a manual for the diagnosis of such disorders. The fact that some Cluster B patients do not respond well to standard forms of psychiatric treatment, tells little against the idea that those patients are suffering from genuine mental disorders. Patients with schizophrenia or bipolar disorder don’t always respond well to available treatments, but this doesn’t support the conclusion that those conditions are not mental disorders—any more than the failure of some cancer patients to respond well to chemotherapy suggests that cancer is not a bodily disorder.

The problem thus remains: How is the clinician to avoid moral judgment when diagnosing a Cluster B patient, given the overtly moral language in which the diagnostic criteria are expressed? Perhaps the answer lies in replacing the morally laden language with morally neutral language. This brings us to what might be called the “linguistic proposal.”

The Linguistic Proposal: Replace the Morally Laden Language with Morally Neutral Language

The idea here is simple: replace the morally laden language of the Cluster B diagnostic criteria with morally neutral language. The challenge would be to ensure that those, and only those, diagnosable with the current morally laden criteria, would be diagnosable with the revised, morally neutral, criteria.

This proposal might sound perfectly reasonable, even obvious. Besides, what place does moral language have in a manual for the diagnosis of mental disorders—especially when the disorders in question appear in a manual published by a medical association: the American Psychiatric Association? Doesn’t everyone agree on at least this much: that morals have no place in medicine?17

There are, however, at least four problems with the linguistic proposal, all of them serious. One problem is that replacing the current morally laden diagnostic criteria with morally neutral diagnostic criteria is easier said than done. Perhaps one might try replacing “deceptive” with “tends to makes false statements.” Perhaps one might try replacing “inappropriate anger” with “uncommon anger.” But intuitively, the morally laden expressions are importantly different from their morally neutral counterparts. Indeed, they would appear to pick out different, if overlapping, groups of individuals; they would appear (in other words) to have different “extensions.”

The difficulty is that morally laden expressions are, in an important sense, “richer” than their morally neutral counterparts. They somehow speak to the agent’s state of mind, to her attitudes and motives. For this reason, they pick out a narrower class of individuals. For instance, it is possible to make false statements (even chronically) without being deceptive. The schizophrenic patient who insists that his neighbors are government agents is a case in point. He speaks falsely without intending to mislead those to whom he speaks. He is not deceptive. In contrast, when the Antisocial patient insists that it’s you and not your money that he wants, he wants you to believe what he knows to be false. He is decidedly deceptive. Similarly, it is possible to experience uncommon anger without experiencing inappropriate anger. The woman who is diagnosed with breast cancer and fired from her job on the same day she is served with divorce papers, experiences uncommon anger that is anything but inappropriate. In contrast, the Borderline patient who becomes enraged when she is not invited to her estranged stepfather’s fourth wedding, experiences uncommon anger that is perhaps also inappropriate.

A second problem with the linguistic proposal concerns whether its implementation would yield the desired results. Even if it were possible to replace morally laden language with morally neutral language, such replacement might have little effect on the intrusion of moral judgment in clinical settings. The clinician might subconsciously translate morally neutral concepts into morally laden ones. When she makes note of a “tendency to make false statements” she might think “deceptive.” When she makes note of “uncommon anger” she might think “inappropriate anger.” As just noted, these are importantly different concepts. The situation here has parallels in cases where evaluative terms are, or at least appear to be, roughly co-extensive with their non-evaluative counterparts. A physician might think “fat” when she informs her patient that he is “overweight.” A guidance counselor might think “stupid” when she informs a parent that, according to a recently administered standardized test, her child is “significantly below average in intelligence.” The point in both cases is the same: morally neutral language isn’t necessarily going to prevent the intrusion of (potentially harmful) moral judgment in clinical settings.

A third problem with the linguistic proposal concerns implications for treatment. Treatment would undoubtedly be less effective were the moral component of the patient’s symptoms not explicitly acknowledged by the clinician. Treating the Narcissist’s belief that he is “special” requires seeing that trait as more than just a belief in one’s own uniqueness. The Narcissist’s problem isn’t that he thinks he’s different from everyone else, it’s that he thinks he’s better than everyone else. Similarly, treating the Histrionic’s tendency to make “rash decisions” requires seeing those decisions as imprudent. The patient’s problem isn’t that she is spontaneous, it’s that she has a tendency to make important decisions with little or no regard for their likely consequences.

A fourth and final problem with the linguistic proposal is theoretical. Retaining moral language in a manual for the diagnosis of mental disorders reinforces the idea that mental disorders sometimes express themselves in morally disvalued traits. This is presumably just what mainstream American psychiatry believes. To the extent that this view is supported by clinical and other empirical evidence, there is no reason not to publicly acknowledge and embrace it by employing moral language (where appropriate) in the profession’s official diagnostic manual. Importantly, this does not mean that the Cluster B Personality Disorders are “moral” in the sense that they are the responsibility, and therefore the fault, of those who bear them. It does, however, mean that they are “moral” in the sense that they are effectively defined by observable deviations from conventional moral norms. We might mark this distinction by saying that, while the disorders in question are moral in “derivative” sense (involving deviations from conventional moral norms), they are not necessarily moral in a “fundamental” sense (reflecting autonomous moral agency).

At this point, it appears as though there are grounds for (i) retaining the Cluster B Personality Disorders in the DSM and (ii) retaining the distinctively moral language of their diagnostic criteria. The problem thus remains: How is the clinician to avoid moral judgment when the language of diagnosis is overtly moral? In an effort to answer this question, it might be wise to consider a more fundamental question, one that makes no explicit reference to language, namely:

How is the clinician to avoid moral judgment given that some patients (those with Cluster B Personality Disorders) are diagnosed largely on the basis of chronic violations of moral norms?

How, for instance, is the clinician to avoid moral judgment given that she regards her Antisocial patient as “true to form”: as consistently irresponsible, impulsive, and deceptive?

Answering such questions requires that we get clear on the operative notion of moral judgment. It requires, in particular, that we distinguish between having morally disvalued (or valued) traits and having moral responsibility for those traits.18 Although these two attributes are generally presumed to accompany one another, they arguably come apart in cases involving Cluster B Personality Disorders. This suggests that we have two importantly distinct notions of moral judgment, notions which it is important to keep separate. This leads to what might be called the “separatist” proposal.

The Separatist Proposal: Separate Moral Traits from Moral Responsibility

In ordinary, non-clinical settings, attribution of morally disvalued traits is accompanied by a presumption of moral responsibility. When we characterize someone as “deceptive” or has having a “sense of entitlement” we presume that they are responsible for having those traits. It is for this reason that we tend to adopt a morally judgmental attitude, an attitude of disapprobation, toward those to whom we attribute such traits. However, according to mainstream American psychiatry, as embodied in the current edition of the DSM, the characteristic symptoms of some mental disorders include morally disvalued traits. This is true of the Cluster B Personality Disorders in particular. Yet if such traits are, quite literally, the symptoms of mental disorder, of psychopathology, it is unclear how we can justifiably hold the (untreated) patient morally responsible for them. More generally, where morally disvalued traits are symptomatic of underlying pathology, patients’ moral responsibility for those traits is arguably diminished, even if it is not eradicated. After all, to the extent that personal traits, whether attitudes or behaviors, result from (or are reinforced by) underlying pathology, they do not result from autonomous choice.19

Thus, in clinical settings, having morally disvalued traits might not carry the usual presumption of moral responsibility. This is a familiar point, at least among those who regularly diagnose and treat psychiatric and/or neurological disorders. Hostility and suspicion are sometimes symptomatic of schizophrenia, irritability of depression, aggression of Alzheimer’s disease, and lack of empathy of autism. Vocalizations that are decidedly antisocial (and often quite vulgar) are symptomatic of Tourette syndrome [19]. However, the clinician does not, or at least need not, hold such patients morally responsible for their antisocial traits, whose etiology is presumed to involve pathology. For similar reasons, the clinician does not, or at least need not, hold her Cluster B patients morally responsible for the antisocial traits that lead to their diagnoses.

It might reasonably be objected that the Cluster B Personality Disorders are different in kind from conditions like schizophrenia, depression, Alzheimer’s disease, autism, and Tourette syndrome. Indeed, the former have been identified as character disorders even by some of those who recognize their psychopathological status [3]. Yet no one claims this of schizophrenia, depression, Alzheimer’s disease, autism, or Tourette syndrome—at least not today, in the 21st century. It might be argued that the (untreated) Cluster B patient has some degree of moral responsibility for her antisocial traits, as she has some degree of autonomy with respect to those traits.20 While this may well be true, it does not affect the present point, which is simply that morally judgmental attitudes needn’t accompany the application of Cluster B diagnostic criteria. Whether such attitudes might be appropriate in certain clinical settings (as when a patient fails to adhere to an agreed upon treatment plan) is another question.

Still, one might reasonably wonder whether it’s really possible to characterize someone as “deceitful,” “self-dramatizing,” or “lacking in empathy” without simultaneously adopting an attitude of moral disapproval. Perhaps the language of diagnosis has more of an influence on the clinician’s attitudes toward her patients than she realizes. An appeal to Grice’s [6] distinction between “conventional” and “conversational” implicature should help defuse this worry.

We often imply things in saying what we say. In some cases, our “implicatures” (as Grice calls them) arise from the conventional meanings of the particular words that we use in saying what we say. A student says to you, “I still don’t understand Grice’s notion of conventional implicature, but I’ll keep trying.” The student doesn’t actually say that she’s been trying for some time to understand the notion in question. However, the word ‘still’, in virtue of its conventional meaning, implies as much. The student doesn’t actually say that her continued attempts to understand Grice’s notion of conventional implicature, in the face of repeated failures, are somehow unexpected. However, the word ‘but’, in virtue of its conventional meaning, implies as much.

In other cases, implicatures arise from assumed conformity to what Grice calls the Cooperative Principle, which (in essence) counsels us to make appropriate contributions to the conversations in which we participate. Suppose you ask your student whether she found Grice’s “Logic and Conversation” challenging. She responds by saying, “I thought Grice was supposed to be easy.” You assume that she is being cooperative in Grice’s sense, that (in other words) she has provided you with an appropriate response to your query. You are therefore able to infer that, in saying what she said, the student has implicated that she found the reading in question challenging. (For details on how such inferences transpire, see [6]).

Let’s now apply Grice’s views on implicature to the issue at hand. Consider four Personality Disordered individuals, all new colleagues of yours: Adam the Antisocial, Brianne the Borderline, Heather the Histrionic, and Nicholas the Narcissist. Imagine now that the following brief exchange occurs between you (A) and another colleague (B).

A: What do you think of our new colleagues, morally speaking?

B: Well, Adam is often deceptive. While Brianne is prone to inappropriate anger, Heather tends to be self-dramatizing. And Nicholas clearly has a sense of entitlement.

B’s response can be broken down into four distinct assertions: one about Adam, one about Brianne, one about Heather, and one about Nicholas. In each case, there is an implicature to the effect that the subject is, to some degree, deserving of disapprobation. Clearly, B doesn’t think much of his new colleagues, “morally speaking.” He holds them in low regard. The question is: Is the implicature here conventional or is itconversational? In other words, does the implicature arise on account of the conventional meanings of the words and phrases uttered: ‘deceptive’, ‘inappropriate anger‘, ‘self-dramatizing‘, ‘sense of entitlement’? Or, does it arise on account of assumed conformity to Grice’s Cooperative Principle? In all four cases, the implicature is arguably conversational rather than conventional.

To see this, consider Grice’s best known test for determining whether an implicature is conventional or conversational: his cancelability test. An implicature is said to be “cancelable” just in case it can be denied without contradiction or incoherence. If it is cancelable, then it is conversational rather than conventional [6]. Thus, when asked whether she found “Logic and Conversation” challenging, your student might have replied, “I thought Grice was supposed to be easy—and I guess I was right,” thereby establishing the conversational nature of the implicature that accompanied her original utterance, an implicature to the effect that she found Grice’s paper challenging.

Given this test, the implicatures associated with your colleague’s response to your query emerge as conversational rather than conventional. For those implicatures can all be canceled. To see this, just consider how your colleague might have elaborated his response:

Adam is often deceptive but that’s not to suggest that he is deserving of disapprobation. He’s rightly come to believe that it’s a “dog eat dog world” and he doesn’t want to be the dog that gets eaten. “Look out for number 1” as they say! Brianne is prone to inappropriate anger but neither is she deserving of moral condemnation. She was sexually abused as a young child and her misplaced anger is what they call “anger coming out sideways.” Her problem is not so much moral as it is epistemic: once she understands who is responsible for her anger (her abuser), she will know just where she should be directing it. Heather tends to be self-dramatizing but that’s not to suggest that she is to blame for such behavior. She has an underactive amygdala and without the excitement that being the center of attention brings, boredom would likely spiral into depression and possibly even suicide. Who can blame her for the self-dramatization, it’s basically a survival mechanism. Nicholas has a sense of entitlement but that’s not to suggest that he falls short morally. On the contrary, he’s read his Nietzsche and is brave enough to admit what the rest of us are too cowardly to admit: that true morality demands arrogance. More power to him!

What these considerations suggest is that to judge that a person violates conventional moral norms, and to represent those violations in overtly moral language, is not to thereby adopt a morally judgmental attitude toward that person. It is not to thereby express disapprobation. This in turn suggests that it is possible, at least in principle, for clinicians to recognize and make note of their patients morally deviant behavior, without thereby adopting a morally judgmental attitude. In that case, application of the Cluster B diagnostic criteria does not inevitably lead to a morally judgmental attitude on the part of the diagnosing clinician.

The point can be made vivid by contrasting the rather tame language of the Cluster B diagnostic criteria with the more colorful language used by ordinary folk in describing those who chronically and flagrantly violate moral norms. In the latter case, implicatures involving the expression of disapprobation are not so easily canceled. The Antisocial might be described as a “user,” the Borderline as a “rageaholic,” the Histrionic as a “drama queen,” and the Narcissist as being “full of himself.” Here, it is not only implied that moral norms have been violated but also that the offender is a suitable object of disapprobation. The latter implication is not so easily canceled. There is a clear tension in saying of a “user” or a “drama queen” that they are not to be faulted for their morally disvalued traits. Such tension suggests that the speaker’s particular choice of words reflects a morally judgmental attitude. The ordinary person, in contrast to the clinician, is going to use words and phrases that carry disapprobation on their sleeves. This is only to be expected given the folk understanding of the Cluster B Personality Disorders: that they are actually character disorders, rather than mental (or psychiatric) disorders.21

It might be objected that, while the clinician’s application of Cluster B diagnostic criteria might notconventionally implicate that the patient is deserving of disapprobation, it does conversationally implicate something to that effect. It’s not the language itself that generates the implicature but rather the use of the language in a particular conversational setting.

The problem with any such objection is that it fails to take into consideration the nature of the “conversational”22 setting provided by the mental health clinic. The clinician, in applying Cluster B diagnostic criteria to her patient, is not responding to the question: What do I think of this patient, morally speaking? She is not, in other words, “speaking” in a conversational context where what is at issue is the patient’s moral character. She is instead responding to the question: What do I think of this patient, clinically speaking? She is “speaking” in a conversational context where what is at issue is the patient’s mental health, not his moral character. In such a context, application of the Cluster B diagnostic criteria would implicate, conversationally, that the patient in question is likely suffering from a mental disorder, one that happens to manifest itself in morally disvalued attitudes and behaviors. Application of such criteria would not conversationally implicate that the patient “suffers from” a moral disorder, for his moral character is not what is at issue.23

Of course, none of this guarantees that the clinician will not end up adopting a morally judgmental, and potentially harmful, attitude toward her Cluster B patient. It means only that the application of morally laden diagnostic criteria does not imply, either conventionally or conversationally, the adoption of such an attitude.

Comparison with Potter

Before concluding, I would like to compare and contrast the proposed view with Potter’s view regarding moral judgment and the Borderline patient. Potter’s view is that the manipulative behavior associated with Borderline Personality Disorder is sometimes the expression of a “deep need for relationship” ([2], 152). The clinician’s recognition of this fact might, and in fact should, prevent her from adopting a morally judgmental attitude toward her patient. Such recognition would then increase the likelihood of successful, empathically based, psychotherapeutic treatment.

The idea that understanding the Borderline patient’s motives might encourage empathy is an intuitive one. It can be brought out by appeal to a stereotypical situation with which we are all familiar. Ashley and Justin are college freshman. Ashley cares deeply about Justin but tells him she is unsure about just how much she cares; she believes that, by being elusive in this way, she might win him over. She’s simply playing “hard to get” and she knows, at some level, that she is doing just that. Justin doesn’t much care for Ashley but tells her he does anyway; he believes that, by telling her he cares deeply about her, he will greatly improve his chances of having sex with her. For Ashley has made quite clear to him that she is not “into” casual sex. Both Ashley and Justin are deceptive but there seems to be a sense in which Ashley’s deception is less deserving of disapprobation than Justin’s. Why is that? It clearly has to do with the differing motives behind the deception. Ashley lies because she wants a “relationship” with Justin; Justin lies because he wants to have sex with Ashley. Perhaps the manipulative behavior of the Borderline patient parallels Ashley’s rather than Justin’s deceptive behavior. Seen in this light, the Borderline patient’s manipulative behavior seems less blameworthy, and so less likely to elicit a morally judgmental attitude on the part of the clinician.

However, this way of looking at the morally questionable behavior of the Borderline patient doesn’t obviously extend to the other Cluster B Personality Disorders. Perhaps the morally questionable behaviors associated with these other disorders are closer, morally speaking, to Justin’s deception. The Antisocial might manipulate because he wants sex, money, or drugs; the Histrionic might manipulate because she wants to be the focus of attention; the Narcissist might manipulate because he wants your adulation.24

None of this undercuts Potter’s strategy for defusing morally judgmental attitudes in the case of the Borderline patient. It suggests only that a different sort of strategy might be required in order to deal with the broader problem of morally judgmental attitudes in the case of the Cluster B Personality Disorders more generally. I have suggested that, with respect to all four Cluster B Personality Disorders, there can be moral language (in the form of morally laden diagnostic criteria) without disapprobation. There is arguably diminished moral responsibility with respect to all four disorders, not so much because the motives for the questionable behaviors might be mitigating (as in the case of the Borderline patient), but because those motives are arguably shaped by mental (and ultimately neurological) dysfunction.25 Thus, it’s not necessarily (or only) the patient’s motives that are mitigating, it’s the pathological nature of their origins.

A Telling Irony: The Importance of Moral Language in Psychiatric Diagnosis

In concluding, I would like to make note of a telling irony that underlies the proposed view that the overtly moral language of the Cluster B Personality Disorders is not out of place in the DSM: an ostensibly “medical” manual for the diagnosis of mental disorders.

That morally judgmental attitudes have no place in clinical settings is surely a reasonable view. After all, such attitudes have the potential to impair clinical judgment in ways that might harm the patient. One might initially take this as implying that moral language should never appear in a psychiatric (or other medical) manual. However, I have suggested that the moral language of the Cluster B diagnostic criteria is perfectly in order. It’s not simply that moral language is necessary for accurate diagnosis and successful treatment of the disorders in question. Ironically, the very presence of such language mitigates against the intrusion of morally judgmental attitudes in clinical settings. For such language amounts to an explicit acknowledgement of the idea, embraced by contemporary American psychiatry, that medical conditions can manifest themselves in morally disvalued traits. This does not mean that “bad” people have diseased brains; that would amount to the medicalization of morals. It means rather that mental dysfunction can give rise to morally disvalued traits. Rather than implicitly endorsing morally judgmental attitudes toward those with mental disorders, retention of moral language in the DSM reminds the clinician that persons who display morally disvalued traits such as dishonesty, chronic irresponsibility, and impulsivity might not be suitable objects of disapprobation. For in at least some such cases, the patient’s “problem” might be more medical than moral.

Footnotes

However, patients can also be harmed when the attributed traits are morally valued traits. Consider the physician whose patient, perhaps someone whom he knows socially, shows signs and symptoms associated with HIV infection. The physician never tests for the presence of the virus because the patient is not “that kind” of woman, the kind of woman who might engage in “risky” behaviors. Within a year, the woman develops full-blown AIDS. It wasn’t the falsity of the physician’s judgment that harmed the patient, it was the fact that he made the judgment at all. For the woman might have gotten the virus from her husband.

All of the foregoing may be symptomatic of a stroke and are more common in women than in men [1].

My use of the term “moral” is to be understood broadly, so as to encompass traits generally regarded as “character flaws.” See Pickard [3] in connection with the idea that the Custer B Personality Disorders involve such flaws.

For recent work on the more general topic of values in psychiatry, see the papers in volume 15 (1) (2009) of Philosophy, Psychiatry, and Psychology.

This initial formulation of the problem does not disambiguate the notion of moral judgment, as the solutions considered first trade on the ambiguity inherent in that notion. In the above formulation, “moral judgment” is to be understood as defined above: as the attribution to an agent of moral traits.

The same is true of those suffering from Cluster A Personality Disorders. In contrast, those suffering from Cluster C Personality Disorders often seek treatment for those conditions [8].

Which is not to say that such actions are actions for which the patient is criminally responsible.

But see Pickard [3], who points out that not all Cluster B diagnostic traits are morally disvalued traits.

For further criticisms of Charland’s position, see Pickard [3] and Reimer [13].

More cautiously, neurological disorders might explain some of the characteristic features of so-called “mental disorders.” See Pickard’s [3] “twin earth” thought-experiment involving schizophrenia for details.

While British psychiatrists apparently see things differently, this may be partly because they fail to appreciate the similarities between the Personality Disorders and paradigm mental disorders like schizophrenia [9].

This is not to say that expert consensus is correct or that it might not change.

Because Szasz does not countenance “mental disorders,” he sees the relevant contrast as between moral and medical disorders.

The Cluster B Personality Disorders are also associated with “psychological” as well as “behavioral” dysfunction, thereby satisfying the DSM criteria for “mental disorder” three times over [13].

The other two Cluster B Personality Disorders, Histrionic and Narcissistic, have not been studied as closely by neuroscientists. However, the fact that their diagnostic criteria overlap considerably with the diagnostic criteria for Antisocial Personality Disorder, suggests that they are correlated with similar neurological dysfunction.

“Indicative of” but perhaps not “identical to,” for reasons brought out in Pickard’s [3] “twin earth” thought-experiment.

For compelling arguments in favor of the view that morals play an important role in psychiatric treatment, especially in connection with Cluster B Personality Disorders, see Pearce and Pickard [15] and Pickard [3].

This would mirror the distinction (drawn above) between “derivative” and “fundamental” senses of “moral.”

A difficult question, not addressed here, is the extent to which the psychopathology presumed to underlie Cluster B traits, impugns autonomy. Here, my focus is on defending the idea that attribution of such traits, in clinical settings, needn’t be accompanied by a morally judgmental attitude on the part of the diagnosing clinician.

Interestingly, Tourette patients often claim to have control over their antisocial vocalizations. However, it has been argued the patients’ moral responsibility is nevertheless diminished, as the urges that lead to such vocalizations are not expressions of the patients’ desires [19].

For arguments to the effect that they might be both, see Pickard [3].

In such settings, the diagnosing clinician is (in effect) both speaker and hearer.

This might explain why clinicians do not seem to mind morally laden diagnostic criteria: they take themselves to be diagnosing mental disorders, not (at least not primarily) moral disorders.

Relatedly, Borderline Personality Disorder is the only Cluster B Personality Disorder whose diagnostic criteria are not strongly suggestive of selfishness and/or disregard for others.

Importantly, this does not necessarily imply that there is diminished legal/criminal responsibility.

References

Larkin, M. 2002. Women’s “non-traditional” stroke symptoms may delay emergency treatment. Lancet360(9343): 1398.CrossRef

Potter, N. 2006. What is manipulative behavior anyway? Journal of Personality Disorders 20(2): 139–156.CrossRef

Pickard, H. 2009. Mental illness is indeed a myth. In Psychiatry as cognitive neuroscience: philosophical perspectives, ed. L. Bortolotti and M. Broome, 83–101. Oxford: Oxford University Press.

Charland, L. 2004. Moral treatment and the personality disorders. In The philosophy of psychiatry: a companion, ed. J. Radden. Oxford: Oxford University Press.

Charland, L. 2006. The moral nature of the cluster-b personality disorders. Journal of Personality Disorders20(2): 119–128.CrossRef

Grice, H.P. 1975. Logic and conversation. In Syntax and semantics 3, ed. P. Cole and J.L. Morgan, 41–58. New York: Academic.

American Psychiatric Association. 2000. Diagnostic and statistical manual of mental disorders. Fourth edition (text revision). American Psychiatric Publishing, Inc.

Tyrer, P., et al. 2003. Treatment rejecting and treatment seeking personality disorders: Type R and type S.Journal of Personality Disorders 17(3): 26–268.CrossRef

Kendell, R.E. 2002. The distinction between personality disorder and mental illness. The British Journal of Psychiatry 180: 110–115.CrossRef

Moran, P. 1999. The epidemiology of antisocial personality disorder. Social Psychiatry and Psychiatric Epidemiology 34(5): 231–242.CrossRef

Reimer, M. 2010a. Treatment adherence in the absence of insight: a puzzle and a proposed solution. Forthcoming in Philosophy, Psychiatry, and Psychology.

Szasz, T. 1960. The myth of mental illness. The American Psychologist 15: 113–118.CrossRef

Reimer, M. 2010b. Moral disorder in the DSM-IV: The Cluster B Personality Disorders. (unpublished typescript).

Vollm, B., et al. 2004. Neurobiological substrates of antisocial and borderline personality disorder: Preliminary results of an fMRI study. Criminal Behaviour and Mental Health 14(1): 39–54.CrossRef

Pearce, S., and H. Pickard. 2009. The moral content of psychiatric treatment. The British Journal of Psychiatry 195: 281–282.CrossRef

Robins, C., and A. Chapman. 2004. Dialectical behavior therapy: current status, recent developments, and future directions. Journal of Personality Disorders 18(1): 73–89.CrossRef

Farah, M. 2002. Emerging ethical issues in neuroscience. Nature Neuroscience 5(11): 1123–1129.CrossRef

Salzman, C., et al. 1995. Effect of fluoxetine on anger in symptomatic volunteers with borderline personality disorder. Journal of Clinical Pharmacology 15: 23–29.