Understanding autism: disorder, disability, and diversity

This dissertation examines autism through the conceptual lenses of disorder, disability, and diversity, while also considering these concepts through the lens of autism. Disorders, we tend to think, are harmful and best addressed through medical intervention, whereas diversity is something to be respected and celebrated. Disability, on the other hand, resists easy characterization, since it involves both impairment (making it like disorder) and a complex social identity (making it like diversity). Autism has been described alternatively as all three, which makes it a perfect test case for our thinking about what these things are (a matter of conceptual analysis and ontology) and how they relate to what we think ought to be done (a matter of ethics and politics). The structure of the dissertation is as follows: Chapter 2 provides some background for the project by distinguishing between biological, cognitive, and behavioral levels of analysis for autism, in addition to introducing the medical model of psychiatry and outlining several prominent views in medical ontology. Chapter 3 proposes a dilemma for "Autism Spectrum Disorder" as a diagnostic category: If understood behaviorally, then our diagnostic manuals identify autism only trivially, revealing nothing interesting about the world. If understood biologically, then heterogeneity threatens the category's clinical utility (usefulness in developing medical interventions) and projectability (usefulness in making inferences). In Chapter 4, I discuss the social model of disability, first noting that facts about the cause or valence of disability entail nothing about the proper response to it. I then consider several iv arguments that might be given to close this is-ought gap. In Chapter 5, I present the neurodiversity view, challenging the arguments given in support of its claim that autism is a mere-difference, while defending its political aims of respect, inclusion, and accommodation. Finally, in Chapter 6, I continue my critique of neurodiversity by providing a limited defense of the practice of selecting against autism, since it may be said to express something about the nonideal social world rather something negative or false about autism or autistic people.

UNDERSTANDING AUTISM: DISORDER, DISABILITY, AND DIVERSITY by Ryan Hummel Nelson A dissertation submitted to the faculty of The University of Utah in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Philosophy The University of Utah August 2019 Copyright © Ryan Hummel Nelson 2019 All Rights Reserved The University of Utah Graduate School STATEMENT OF DISSERTATION APPROVAL The dissertation of Ryan Hummel Nelson has been approved by the following supervisory committee members: Leslie Francis , Chair April 24, 2019 Margaret P. Battin , Member April 24, 2019 Brent Michael Kious , Member April 24, 2019 James Tabery , Member April 24, 2019 Natalia Washington , Member April 24, 2019 and by Matt Haber the Department of and by David B. Kieda, Dean of The Graduate School. Date Approved Date Approved Date Approved Date Approved Date Approved , Chair/Dean of Philosophy ABSTRACT This dissertation examines autism through the conceptual lenses of disorder, disability, and diversity, while also considering these concepts through the lens of autism. Disorders, we tend to think, are harmful and best addressed through medical intervention, whereas diversity is something to be respected and celebrated. Disability, on the other hand, resists easy characterization, since it involves both impairment (making it like disorder) and a complex social identity (making it like diversity). Autism has been described alternatively as all three, which makes it a perfect test case for our thinking about what these things are (a matter of conceptual analysis and ontology) and how they relate to what we think ought to be done (a matter of ethics and politics). The structure of the dissertation is as follows: Chapter 2 provides some background for the project by distinguishing between biological, cognitive, and behavioral levels of analysis for autism, in addition to introducing the medical model of psychiatry and outlining several prominent views in medical ontology. Chapter 3 proposes a dilemma for “Autism Spectrum Disorder” as a diagnostic category: If understood behaviorally, then our diagnostic manuals identify autism only trivially, revealing nothing interesting about the world. If understood biologically, then heterogeneity threatens the category's clinical utility (usefulness in developing medical interventions) and projectability (usefulness in making inferences). In Chapter 4, I discuss the social model of disability, first noting that facts about the cause or valence of disability entail nothing about the proper response to it. I then consider several arguments that might be given to close this is-ought gap. In Chapter 5, I present the neurodiversity view, challenging the arguments given in support of its claim that autism is a mere-difference, while defending its political aims of respect, inclusion, and accommodation. Finally, in Chapter 6, I continue my critique of neurodiversity by providing a limited defense of the practice of selecting against autism, since it may be said to express something about the nonideal social world rather something negative or false about autism or autistic people. iv For Fern, as you are. TABLE OF CONTENTS ABSTRACT....................................................................................................................... iii ACKNOWLEDGMENTS ............................................................................................... viii Chapters 1. INTRODUCTION .......................................................................................................... 1 2. SETTING THE STAGE: (WHAT) IS AUTISM? ........................................................ 11 2.1 Introduction ........................................................................................................ 11 2.2 Levels of Analysis for Autism ........................................................................... 13 2.3 ASD and the Medical Model of Psychiatry ....................................................... 23 2.4 ASD and Medical Ontology............................................................................... 28 2.5 So What? ............................................................................................................ 34 2.6 References .......................................................................................................... 39 3. A DILEMMA FOR “AUTISM SPECTRUM DISORDER” ........................................ 45 3.1 Abstract .............................................................................................................. 45 3.2 Classification and Ontology in the DSM ........................................................... 46 3.3 The Triviality Problem....................................................................................... 49 3.4 Open Questions .................................................................................................. 53 3.5 The Heterogeneity Problem ............................................................................... 58 3.6 Conclusion ......................................................................................................... 62 3.7 References .......................................................................................................... 63 4. THE IS AND OUGHT OF DISABILTY ..................................................................... 67 4.1 Abstract .............................................................................................................. 67 4.2 Introduction ........................................................................................................ 68 4.3 Models of Disability .......................................................................................... 69 4.4 The Is-Ought Gap .............................................................................................. 74 4.5 Bridging the Gap ................................................................................................ 78 4.6 Conclusion ......................................................................................................... 85 4.7 References .......................................................................................................... 86 5. A CRITIQUE OF NEURODIVERSITY ...................................................................... 89 5.1 Abstract .............................................................................................................. 89 5.2 The Neurodiversity Movement .......................................................................... 90 5.3 Arguments and Critique ..................................................................................... 95 5.4 Preserving the Political; or Having and Eating Cake....................................... 102 5.5 References ........................................................................................................ 106 6. NEURODIVERSITY IN AN IMPERFECT WORLD ............................................... 110 6.1 Abstract ............................................................................................................ 110 6.2 Introduction ...................................................................................................... 110 6.3 Neurodiversity and the Social Model of Disability ......................................... 114 6.4 Causing, Treating, and Preventing Autism ...................................................... 118 6.5 Perfect and Imperfect Worlds .......................................................................... 126 6.7 References ........................................................................................................ 132 7. CONCLUSION ........................................................................................................... 136 viii xi ACKNOWLEDGMENTS I am grateful to Margaret Battin, Jeff Botkin, Paul Carbone, Brent Kious, Kendra Lystad, Jim Tabery, and Natalia Washington for their contributions to this work. Thank you also to my colleagues at the University of Utah’s Department of Philosophy, the members of the University of Utah’s Bioethics Journal Club, and to my cofellows at the Tanner Humanities Center for their valuable feedback. I am especially indebted to Leslie Francis, whose endless effort, insight, and generosity made this project possible. Financial support for this project was provided by the University of Utah’s Department of Philosophy, the Steffensen-Cannon Scholarship, and the Tanner Humanities Center. viii CHAPTER 1 INTRODUCTION Discussions surrounding autism tend to be remarkably polarizing. Some characterize the condition, with striking vividness, as a monster that steals children in the night, a self-effacing mask one can never remove, a shape-shifting plague with obscure causes and unpredictable manifestations. For them, autism represents an epidemic to which the only acceptable response is an aggressive push for prevention, treatment, and cure. Others provide a dramatically contrastive conception of autism as an entirely neutral—or even valuable—human difference, akin to racial and ethnic diversity. In the provocative language of this view, the prevention of autism amounts to eugenics, and eradication to genocide. Such polarization can leave one feeling suspended between gaping dichotomies: Is autism a public health emergency to be avoided at all costs, or something to be celebrated (and perhaps cultivated)? Do efforts to eliminate it stem from a beneficent desire to improve health and quality of life, or an intolerant insistence on conformity? Are awareness organizations like Autism Speaks, in all their support for medical solutions to the “autism puzzle,” heroic or hateful? At the center of this controversy lies the distinction between disorder and difference—or, more precisely, the distinction between forms of human difference that amount to disorders and those that amount to mere-differences. It is tempting to think that 2 the distinction is based upon a straightforward assessment of harm: if a difference is harmful, it is a disorder; if not, it is a mere-difference. As quickly becomes apparent, however, this will not do, since there are a number of mere-differences which involve harm or disadvantage. Structural sexism and racism, for example, confer tangible disadvantages upon women and people of color. Yet being a woman or person of color is not to have a disorder, since their associated harms derive from (exogenous) injustice, rather than some kind of (endogenous) dysfunction. This suggests the amendment that disorders involve inherent biological harm, while mere-differences do not. But this proposal, too, faces difficulties. For one, the tangled strands of biological and social influences are not so easily pulled apart. Since an individual’s level of functioning is largely a matter of how well she is able to function within a given environment, and many of the environments in which one operates are fundamentally social, functioning is often subject to the arbitrary structure of the built world. While a condition like paraplegia may be severely limiting in an environment modeled after an Escher drawing, it may pose little to no limitation in a world designed to be perfectly accessible to wheelchair users. Another difficulty arises from the fact that the concept of disorder is often colored by value judgments. Although an objective statistical analysis of a population can tell us which differences are atypical, it arguably cannot tell us which are beneficial and desirable (like exceptional strength or intelligence), and which detrimental and to be avoided. This latter task may turn out to be a matter of values, in which case there is reason to doubt the notion that disorder is a purely empirical phenomenon that can be identified absent social context. Of course, certain differences—such as having cancer cells in one’s pancreas— 3 seem to be clear instances of disorder, while other differences—such as being female— seem to be clear instances of a mere-difference. For a condition like autism, though, things are not so simple, which makes it a nice test case for the difference-disorder debate. But why does this distinction matter in the first place? One crucial reason is that the distinction bears on our judgments about ethics and justice. We tend to think, for example, that disorders (like pancreatic cancer) ought to be addressed through medical intervention (such as surgery, chemotherapy, and radiation), whereas the disadvantages associated with mere-differences (like being female) ought to be addressed through social reform (namely, the abolition of structural sexism). We also tend to think that people with certain disorders are owed things others are not: those with learning disabilities ought to receive extra time on standardized tests; those with severe mental illness warrant special consideration in the criminal justice system; those with limited mobility ought to be able to access public spaces. Such judgments are thought to establish moral rights or entitlements, as well as obligations for parties ranging from clinicians and insurance providers to voters and policy makers. The issues at hand, then, are both descriptive and normative. The descriptive issue concerns whether autism is a disorder or a mere-difference; the normative issue concerns what kinds of rights and duties autism generates. Put more generally, the questions I consider in this dissertation are these: (1) How should we understand autism? and (2) How should we respond to it? In what follows, I confront both questions, with the aim of addressing ontological and conceptual questions about autism as they relate to relevant ethical and political questions. Ontology is concerned with what kinds of things there are and what those things 4 are like. In Chapter 1, I lay groundwork for the project by considering the ontology of autism. Getting clear on what kind of thing autism is, though, presents a serious challenge. One of the sources of the challenge is the fact that autism can be analyzed at the levels of biology, cognition, and behavior (by, say, examining the brains of autistic people, investigating how autistic people attribute mental states to others, and observing the ways autistic people communicate). Such complexity is characteristic of many psychiatric conditions, the nature of which has long been debated. A central question is this debate is whether psychiatric conditions—like autism and depression—are medical conditions like any other—like asthma and heart disease—or whether they constitute a different class of things that warrant a different kind of response. Such ontological complexity makes autism difficult to “locate.” Is it best understood as a brain condition with cognitive and behavioral manifestations? Or is it a cognitive or behavioral condition caused by underlying biological mechanisms? Further complicating matters is the fact that autism appears, by all accounts, to be broadly heterogeneous and variable, which is one reason clinicians refer to the condition as Autism Spectrum Disorder. And despite the fact that it is common to hear reference to “highfunctioning” and “low-functioning” autistics, there is no neat high-to-low spectrum, since one’s level of functioning can vary upon numerous dimensions of cognition and behavior. Sorting out answers to these questions is worthwhile for the simple reason that how we conceive of autism determines which people get diagnosed with autism; and diagnosis is significant for both its potential benefits and detriments. Among the potential benefits of an autism diagnosis is a grounding for claims to various services and accommodations, conceptual resources for self-understanding in autistic people, and an impetus for others to 5 approach autistic people with patience and empathy. Detrimentally, an autism diagnosis can generate stigma and exclusion, fatalism and pessimism, and labelling effects, which occur when one’s receiving a label actually makes it more likely that one will fulfill the expectations that accompany the label. These effects should inspire care in conceptualizing and diagnosing autism. With this in mind, I devote Chapter 2 to evaluating the way in which autism is actually diagnosed in contemporary medical practice. The diagnostic process for nonpsychiatric conditions most often goes like this: a patient presents with some symptoms; the clinician conducts a differential diagnosis and generates a diagnostic hypothesis; the patient undergoes a test to determine whether the hypothesis is correct and, if so, the patient is diagnosed and treated. Psychiatric conditions like autism, by contrast, are typically diagnosed on the basis of clinical presentation alone. Whether a patient in fact has autism is simply a matter of determining whether a patient’s behavior satisfies the behavioral diagnostic criteria in an accepted diagnostic manual. In the United States, the gold standard for psychiatric diagnosis is the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM), currently in its fifth edition, which refers to autism as “Autism Spectrum Disorder” (ASD). The DSM-5 provides two primary diagnostic criteria for ASD: (1) persistent deficits in social communication and social interaction across multiple contexts; and (2) restricted, repetitive patterns of behavior, interests, or activities. These criteria rely on clinically significant behavioral signs, rather than a particular biological pathology or etiology. Admittedly, there are good reasons for classifying autism in behavioral terms. Even so, I argue that this approach to classifying autism faces a dilemma: If all it takes to 6 have ASD is satisfying the DSM’s diagnostic criteria for ASD, then it seems that there is no possibility that the DSM gets ASD wrong, so long as the criteria are correctly applied. If, on the other hand, the diagnostic criteria for ASD serve as mere symptoms of autism, then the biological heterogeneity seen among autistic people threatens the usefulness of an ASD diagnosis in developing interventions and predicting the course of the condition. The upshot of this dilemma is that the category “ASD” is either trivial or unhelpful. Although autism is diagnosed and managed by medical professionals, it is important to remember that the medicalization of disability is controversial. Among the critics of medicalization are proponents of the social model of disability, who typically understand the disadvantages associated with disability to be (primarily or solely) the result of discrimination against individuals with impairments, rather than the result of impairment per se. Presumably, the debate over the source of disability’s disadvantage is meant to have some amount of normative force: If something like the social model of disability is correct, then perhaps we should favor social change, rather than medical intervention, in response to disability. But, of course, claims about what disability is entail nothing about what ought to be done. In Chapter 4, therefore, I consider several arguments that might be given to bridge disability’s is-ought gap, concluding that arguments invoking identity and diversity get closest to doing so. When it comes to autism, these kinds of arguments are not difficult to come by. Indeed, the neurodiversity view makes both a conceptual claim about what kind of thing autism is a political claim about how it ought to be addressed. Conceptually, the neurodiversity view weighs in on the difference-disorder debate by arguing that certain neurological differences currently classified as disorders are best understood as forms of 7 diversity. Politically, it holds that, rather than being medicalized and “treated,” neurodiversity ought to be respected in the way other human differences—such as differences in race and sexual orientation—are respected. I engage the neurodiversity view directly in Chapter 4 by challenging several arguments given in support of its conceptual claim, while defending its political aims of respect, inclusion, and accommodation. Despite neurodiversity’s compelling challenge to the status quo of medicalization, I conclude that it falls short of establishing that autism is not a disorder. Fortunately, given the discussion in Chapter 3, the fact that autism is a disorder does not entail anything about whether it ought to be addressed with medical intervention. My view is ultimately that social change and medical intervention need not be considered mutually exclusive, since it is perfectly consistent to hold that (i) autism is a disorder; (ii) we ought to be more inclusive and accommodating of people with autism; (iii) and prevention, treatment, and cure may not be the best response to all cases of autism. On this view, we can pursue neurodiversity’s political aims of social change and accommodation without relying on implausible conceptual claims about disorder. The advantage of this approach is that it supports autism activism while also recognizing the diversity of experience among autistic people and respecting the autonomy of those who opt for medical intervention. In Chapter 5, I take up the closely related issue of whether it is permissible to use medical intervention—such selective abortion and preimplantation genetic diagnosis—to prevent autism, supposing (hypothetically) that such intervention were available. Beyond the ethics of abortion and discarding human embryos, the practice of selecting against 8 disability has been criticized on the grounds that it falsely assumes that disability amounts to a bad-difference rather than a mere-difference (the epistemic objection); and it expresses something negative about, and thus cause harm to, existing people with disabilities (the expressivist objection). In considering these objections, I provide a limited defense of autism prevention by arguing that both rely upon an equivocation between ideal and nonideal circumstances. The claim that autism is a mere-difference is a counterfactual appeal to an ideal society: If the social world were more inclusive, being autistic would not be disadvantageous. A decision to select against autism, by contrast, is made in the nonideal circumstances in which we find ourselves—and these circumstances are indeed nonideal, as we learn from disability scholars who emphasize the pernicious influence of ableism. Of course, there may well be instances of autism prevention that express something negative about autistic people. Suppose, for example, that expectant parents chose to abort a fetus that screened positive for autism on the basis of the belief that autistic people are possessed by an evil spirit. Such a belief certainly appears to be both false and potentially harmful to existing autistic people. But the attitudes that undergird prevention efforts are not necessarily false or problematic and may reasonable in light of the limitations presented by nonideal circumstances. In such cases, prevention may be said to express something about the social world rather than something about autism per se. This being so, I conclude that the epistemic and expressivist objections fall short of justifying a blanket ban on the practice of autism prevention. At this point, it is worth asking what positive contribution the dissertation makes. After all, the narrative so far might seem rather negative. A cursory read could leave one with the impression that my view is simply that the DSM is wrong (Chapter 2), the social 9 model of disability is wrong (Chapter 3), and the neurodiversity view is wrong (Chapters 4 and 5). Beyond the critiques, however, this project has several things to offer. First, although autism is fascinating in its own right, I see it in many ways as a vehicle for discussing more general issues in the philosophy of medicine and psychiatry (e.g., the concept of disorder), the philosophy of disability (e.g., the social model of disability), and biomedical ethics (e.g., genetic screening). Given its complexity, contentiousness, and familiarity, autism provides a compelling entry point into these consequential discussions. Second, I hope for the project to convey a kind of pragmatism about the subject matter. When thinking about concepts like “autism” and “mental disorder,” it is important to ask what it is we want the concepts to do. Such theorizing can have very real implications for individuals and institutions, and should thus be regarded as an importantly moral exercise. This being so, I support a pragmatic approach to medicine, psychiatry, and disability—one that keeps in its sight the goal of actually making people’s lives better. The DSM, the social model of disability, and the neurodiversity movement are tools; our task is to determine how best to use them, and toward what end. Finally, I hope to highlight opportunities for progress. Despite the polarizing effect of autism discourse, there are areas of consensus. One thing everyone agrees upon, for instance, is that things could be made better for autistic people. My contention is that we should use all the tools at our disposal in pursuit of this aim. Focusing on medical intervention alone ignores the important role social factors play in the lived experience of disability, while focusing on activism alone limits choice and places a burden on the shoulders of individuals who may not have the luxury of waiting for social change. Over the course of my research, I have spent time with clinicians and researchers looking for 10 more effective ways to diagnose and treat autism, as well as autistic people and advocates fighting to make the world more accepting and accommodating of disability. Both have something to offer the other, and both have something to offer the rest of us. 11 CHAPTER 2 SETTING THE STAGE: (WHAT) IS AUTISM?1 2.1 Introduction Taylor’s parents are concerned about him. Though he is only 4 years old, they have noticed that he is not like their other kids: he avoids eye contact; he lines up his toys instead of playing with them; he has a difficult time interacting with the children in his preschool class. So, they decide to take him to see a developmental pediatrician. After observing him for some time, and asking his parents some questions, the pediatrician diagnoses Taylor with Autism Spectrum Disorder (ASD). This is a common-enough story. Recent prevalence data reveal that as many as one in 40 children in the United States has a diagnosis of ASD (Kogan et al., 2018). But what does it mean to say that someone has autism? Answering this question requires getting clear on what we mean by “autism.” In what follows, I attempt to do so by addressing several closely related issues. First, I discuss the different levels of analysis for autism by examining the condition through the lenses of behavior, cognition, and biology. Though the diagnostic category “ASD” is treated by clinicians as a description of behaviors, autism can be described also in terms of psychology and biology. Are the behaviors used to diagnose ASD mere symptoms of an underlying disorder? Or do the behaviors themselves 1 This chapter has been formatted for stand-alone publication. 12 constitute the disorder? I then consider the minimal and strong interpretations of the medical model of psychiatry, the view that psychiatric conditions, like other medical conditions, are best investigated and intervened upon using the tools of the medical sciences. Where these interpretations diverge is over the level at which psychiatric conditions like autism are to be defined. In Section 2.3, I canvass the medical ontology literature, which asks what kinds of entities medical conditions are. Sorting out these issues may strike some as a philosophical endeavor of no practical consequence. As I argue in Section 2.4, however, the various ways of understanding autism have importantly different implications for controversies over diagnostic validity and autism’s status as a disorder, which in turn affect ethical and political judgments about what is owed to autistic individuals.2 The aim of this chapter, then, is to set the stage for the rest of the dissertation by laying out some of the general ideas that will be relevant to a philosophical investigation of autism and to make a case that such an investigation is worthwhile. 2 There is some controversy about how best to refer to persons diagnosed with ASD. Using “individuals with autism” conforms to the norm of “person-first language,” which is meant to avoid defining a person by her diagnosis. Yet some in the autism community prefer “autistic persons” or even “autistics,” on the grounds that this language captures the centrality of autism to their identity, while denying its status as a disorder. Throughout the dissertation, I will move somewhat arbitrarily between these different linguistic conventions. In doing so, I do not wish to make any philosophical commitments, unless otherwise noted. 13 2.2 Levels of Analysis for Autism What is autism? The answer to this question depends on whom you ask. Ask a psychiatrist or pediatrician, and he or she will say something about behavior, such as “autism is a social communication disorder characterized by restricted interests, repetitive activities, and deficits in interpersonal interaction.” Ask a psychologist or philosopher of mind, and you will likely hear something about cognition, like “autism is a cognitive disorder characterized by deficits in appropriately attributing mental states to others.” Ask a neurologist or neuroscientist, and the answer will involve biology—“autism is a neurodevelopmental disorder characterized by deficient mirror neuron activity in the inferior frontal gyrus,” for example. In these different answers, we see that autism can be described at the levels of behavior, cognition, and biology. This feature is not unique to autism—indeed, one commonly encounters in the philosophy of psychiatry a tripartite distinction between clinical, psychological, and material levels of explanation. But the multileveled nature of autism makes describing the condition no small task, particularly since there is controversy at each level. I will therefore begin with an overview of the prominent accounts of autistic behavior, cognition, and biology. 2.2.1 Behavior The earliest systematic descriptions of autistic behavior come from Leo Kanner and Hans Asperger. In a pioneering paper, Kanner (1943) presents eleven case studies of children he describes as having an “inability to relate themselves in the ordinary way to people and situations from the beginning of life” (p. 242). Though the children varied in 14 “the degree of their disturbance” and “the manifestation of specific features,” they shared “a number of essential common characteristics,” which, Kanner states, “form a unique syndrome, not heretofore reported” (1943, pp. 241-242). He called this syndrome early infantile autism. Independent, and nearly concurrent, observations from Asperger (1944/1991) document cases of autistic psychopathy, characterized by, inter alia, “severe and characteristic difficulties of social integration,” “peculiarities of eye gaze,” “a paucity of facial and gestural expression,” “originality of language formulations,” and “stereotypic activity.” Like Kanner, Asperger (1944/1991) notes both similarity and diversity among his subjects. “The autistic personality,” he states, “is highly distinctive despite wide individual differences” (p. 67). Current diagnoses of what is now called Autism Spectrum Disorder (ASD) are based upon behavioral criteria outlined in one of the two major diagnostic frameworks— the fifth edition of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (DSM-5) and the tenth edition of the World Health Organization’s International Classification of Diseases (ICD-10). The diagnostic criteria in the DSM-5 fall under two dimensions: (1) persistent deficits in social communication and social interaction across multiple contexts, and (2) restricted, repetitive patterns of behavior, interests, or activities (APA, 2013a, 299.00). These symptoms, which vary in severity, must be present in the early developmental period and cause clinically significant impairments that are not better explained by co-morbidities, such as intellectual disability. The DSM-5 reflects a significant shift away from the DSM-IV (APA, 2000) by replacing the previous diagnostic categories of autistic disorder, Asperger’s syndrome, childhood 15 disintegrative disorder, and pervasive developmental disorder not otherwise specified with ASD. This revision is supposed to represent “a new, more accurate, and medically and scientifically useful way of diagnosing individuals with autism-related disorders” (APA, 2013b, p. 1). The ICD-10, which was completed in 1992, is closer to the DSM-IV in its treatment of autism-related disorders. Included under the category “Pervasive Developmental Disorders” are childhood autism, atypical autism, Rett syndrome, Asperger syndrome, and pervasive developmental disorders. This group of disorders is characterized by “qualitative abnormalities in reciprocal social interactions and in patterns of communication, and by a restricted, stereotyped, repetitive repertoire of interests and activities” (World Health Organization, 1992, F84.0). Early drafts of the ICD-11, which was released in 2018 but will not come into effect until 2022, show a move towards autism as a spectrum disorder, following the DSM-5. This updated edition gives “Autism Spectrum Disorder” its own category, now categorized as a “Neurodevelopmental Disorder,” with subtypes specified by absence or lack of comorbidities. 2.2.2 Cognition A number of cognitive accounts of autism appear in the literature, four of which I will review here. The most widely discussed of these is the theory of mind thesis, which holds that “a core and possibly universal abnormality” (Baron-Cohen, 2000, p. 3) of people with autism is their failure to ascribe, or accurately ascribe, mental states to others. This thesis, defended prominently by Simon Baron-Cohen (1995), Uta Frith (1985), and Francesca Happé (1999), rests upon the idea that neurotypical individuals possess a theory 16 of mind allowing them to recognize others as having beliefs, preferences, emotions, and other intentional states.3 The view that autistic individuals possess a deficient theory of mind, or lack a theory of mind altogether,4 has become “widely known and accepted” (Frith, 2003, pp. 79-80). Proponents of this view cite empirical evidence from studies analyzing the performance of autistic individuals in pretend play (Baron-Cohen, 1987) and false belief tasks (Perner et al., 1989; Wimmer & Perner, 1983). Anna Stubblefield (2013) warns that, despite its prominence, the theory of mind thesis is both empirically questionable and morally problematic. She notes that some researchers (Begeer et al., 2010; Gernsbacher & Frymaire, 2005) have raised doubts about the methods of false-belief studies and the generalizability of their results. Furthermore, a number of individuals who identify as autistic have argued that the results of these studies are more a function of study design than a particular deficit in autistic participants (Donnellan et al., 2010). In addition to these empirical concerns, Stubblefield (2013) raises worries about the moral and practical implications of the theory of mind thesis: “The belief that autistics cannot empathize and that they are irredeemably strange,” she states, “reinforces attitudes that have led to past and ongoing harm to people labelled with autism” (p. 145). Stubblefield argues that these harms have resulted from, among other things, segregation in schools and communities, denial of basic human rights, and neglect of features arguably more central to autism, such as sensory and movement challenges. Even more concerning are allegations that autism may play a central role in acts of violence, 3 4 failure.” The term “theory of mind” was introduced in Premack & Woodruff (1978). This deficit is sometimes referred to as “mind blindness” or “mentalizing 17 with reports of a “significant statistical link between autism and mass murder” being made in both the popular press (McCoy, 2014, p. 21) and academic literature (Allely et al., 2017, p. 17). Some in the autism community have noticed the emergence of this narrative and are pleading with the media to “stop blaming autism for acts of mass violence” (Beresford, 2018, p. 5). This suggests that, even if the theory of mind thesis gets something right about autism, its significance must not be overstated. Difficulty attributing mental states to others should not, for example, be seen as some kind of irredeemable moral flaw that can justify the stigmatization and exclusion of people with autism. Keeping this in mind, let us now briefly consider three alternative (or perhaps complementary) cognitive accounts of autism. Frith and Happé (1994), early defenders of the theory of mind thesis, have more recently proposed the weak central coherence thesis (see Frith, 1989/2003; Frith & Happé, 1994; Happé, 1999; Happé & Frith, 2006). On this view, “autism is characterized by a specific imbalance in integration of information at different levels” (p. 121). Whereas neurotypical individuals easily synthesize a range of stimuli into a coherent whole, people with autism “have a processing bias for featural and local information and a relative failure to extract gist” (Happé & Frith, 2006, p. 6). This view is reminiscent of Kanner’s descriptions of his patients as having an “inability to experience wholes without full attention to the constituent parts” (Kanner, 1943, p. 246). Proponents argue that weak central coherence may account for many autistic behaviors—such as hypersensitivity and insistence on sameness—citing a large body of supporting empirical work. Happé and Frith (2006) are careful to note that weak central coherence should be seen as a bias rather than a deficit, since it can explain both the 18 difficulties and “islets of ability” (Kanner, 1943) associated with autism. For example, while individuals with autism tend to have difficulty with tasks involving face matching, their detail-oriented processing style may confer a number of advantages over neurotypical individuals, including higher rates of perfect pitch, superior rote memory, and reduced susceptibility to optical and auditory illusions (see Happé & Frith, 2006 for a review of this research). A third cognitive account of autism stems from the weak executive function thesis, according to which individuals with autism are impaired in at least some of their executive functions, such as planning, working memory, mental flexibility, and inhibition.5 Hill (2004) argues that this view is better equipped than the theory of mind and the weak central coherence theses to account for the repetitive behaviors and limited interests often seen in individuals with autism. Citing studies that measure the executive functioning of autistic participants,6 she concludes that “autistic individuals of all ability ranges are impaired in the executive function[s] of planning and… mental flexibility” (Hill, 2004, p. 31). Nichols and Stich likewise suspect that mental inflexibility is characteristic of autism, proposing that “autistic children have difficulty updating their beliefs on the basis of new information” (2003, p. 179), echoing Asperger’s observation that his subjects exhibited a “failure of instinctive situational adaptation” (1944/1991, p. 79). Finally, let us consider the extreme male brain thesis. This controversial view7 takes 5 (1991). 6 For an early discussion of executive dysfunction in autism, see Ozonoff et al. Using, for example, the Tower of Hanoi puzzle, which requires participants to move a stack of disks in ascending order of size from one rod to another. 7 For a well-developed challenge to the extreme male brain thesis, see Sample (2013). 19 male and female brains to differ in their respective abilities to empathize and systematize. Baron-Cohen (2002) cites empirical evidence purporting to show that systematizing is more developed than empathizing in the male brain, whereas the reverse is true in the female brain.8 Understood as an extreme version of the male brain, which he takes to be hyper-systematic and hypo-empathetic, this thesis is supposed to explain many of autism’s characteristic features as well as the fact that ASD is diagnosed at significantly higher levels in males. Asperger appears to have anticipated this view, as well, stating: “The autistic personality is an extreme variant of male intelligence. Even within the normal variation, we find typical sex differences in intelligence…In the autistic individual, the male pattern is exaggerated to the extreme” (1944/1991, p. 84). How are these various cognitive accounts of autism supposed relate to one another? Some have been optimistic about the prospects of a common cause or feature of autism that would link several of the accounts. Perhaps weak executive function can explain both mindblindness (Glüer & Pagan, 2003) and weak central coherence (Frith, 2003, p. 180). Or perhaps the core feature of autism is an “absent self,” resulting from a lack of balance between top-down control and bottom-up processing (Frith, 2003, p. 208). Many others, however, have supposed that any account attempting to stitch together autism’s features with a single thread will ultimately fall apart. Happé et al. (2006) note that social cognition accounts (e.g., the theory of mind thesis) have difficulty explaining rigid and repetitive behavior, and that domain-general processing accounts (e.g., the weak executive function thesis) struggle to explain social and communication impairments. They conclude that, 8 He is careful to note that, while “not all [males] have the male brain type, and not all [females] have the female brain type,” the male brain type is more common in males and the female brain type is more common in females (p. 249). 20 while the various cognitive theories may well explain particular characteristics of autism, no one theory is able to account for the full range of autism’s features. 2.2.3 Biology Those who endorse behavioral or cognitive accounts of autism do not deny that factors such as genetics and neurobiology contribute to autism in important ways. Even so, autism is not typically defined in terms of its underlying biology (Sample, 2013, p. 77), though unifying biological theories of autism have been proposed. Kelleher and Bear, for example, modestly suggest that aberrant synaptic protein synthesis “represent one possible pathway leading to autistic phenotypes, including cognitive impairment and savant abilities” (2008, p. 401). More recently, Markram and Markram (2010) have advanced the view that autism is a product of hyper-reactivity and hyper-plasticity of local neural microcircuits, rather than a deficiency in brain functioning. Several factors explain why autism, along with most other psychiatric conditions, are defined by behavioral and cognitive features rather than biology. First is the biological heterogeneity seen among individuals diagnosed with ASD. Although autism is considered to be “among the most heritable of all mental disorders” (Lichtenstein et al., 2010), the heterogeneity of genetic findings in autism presents a “significant challenge to the concept of autism as one ‘disease’ or even ‘diseases’” (Lord & Jones, 2012, p. 491). In terms of neurobiology, it has been argued that each of autism’s impairments involve independent neural substrates (Happé et al., 2006). Second is the inchoate state of autism science. Unlike in single-gene conditions, such as Huntington’s disease, or conditions involving a unique pathology, such as Ewing’s sarcoma, no unifying biological feature has been 21 identified in autism. It is possible that such a feature will be discovered in the future, but many are pessimistic. Happé et al. argue that “it is time to give up on the search for a monolithic cause or explanation” for autism (2006, p. 1219). A third reason for focusing on behavior and cognition is that these are the features we actually care about: if there were no clinical manifestations of autism, the underlying biology would arguably be of no interest. Though a single biological feature of autism has proven elusive, researchers have proposed biological explanations for each of autism’s features. Social deficits have been linked to a dysfunctional mirror neuron system in the interior frontal gyrus (Dapretto et al., 2006) and disruptions in a brain network comprising the medial frontal cortex, temporoparietal junction, superior temporal sulcus, and temporal poles (Happé et al., 2006). Others have proposed that the diminished social motivation seen in autism point to abnormalities in the orbitofrontal–striatum–amygdala circuit and oxytocin regulation (Chevallier et al., 2012). Impaired communication has been associated with atypical cortical-cortical connectivity (Anagnostou & Taylor, 2011) and “a lack of activation of superior temporal sulcus voice-selective regions in response to vocal sounds” (Happé et al., 2006, p. 1219). Rigid and repetitive behavior has been linked to increased volume in the caudate (Happé et al., 2006, p. 1219). Despite this fragmentation, there may be good reason to be open to a biological definition of autism. For one thing, given that wide variation is seen also at the behavioral and cognitive levels of autism, heterogeneity is not a unique challenge to biological accounts. For another, there is biologic variation even among individuals with paradigmatic examples of unified biological conditions, such Huntington’s and Down syndrome. The 22 joints of biology are not as clearly carved as those of, say, chemistry. As such, one cannot simply dismiss the notion that autism is a biological type or kind on grounds of heterogeneity, since heterogeneity is ubiquitous in biology. Instead, one must give a principled account of how much heterogeneity is allowed within a biological type and show that autism exceeds that threshold. Moreover, while pessimism about the prospects of defining autism as a unified biological disorder seems justified, it might be that the autism spectrum comprises several biological subtypes, and that the labels “autism” and “ASD” are just placeholders for conditions whose types have not yet been identified. Indeed, some conditions previously considered forms of autism, such as Rett syndrome and Fragile X, “have been reclassified as separate disorders principally because their specific biological basis and etiology has been identified” (Sample, 2013, p. 77). As the science of autism becomes increasingly precise, it is possible that the spectrum will continue to splinter, as crude psychiatric categories like “feeble-minded” have done in the past.9 Perhaps the splintering will stop at more or less naturally defined kinds, or perhaps it will continue to the level of the individual. Regardless of where one stands on the prospects of a biological account of autism, one must acknowledge that there has been a neuroscientific shift in autism research, and psychiatry more generally. Gone are the days of blaming “Refrigerator Mothers” for autism,10 as psychogenic explanations have been replaced by pathogenic ones. “Autism 9 Of course, this process will be imperfect. Just as we may discover some divisions to be made on the autism spectrum, we may discover that previously made divisions were inappropriate, as was the judgment of the APA when it combined previously distinguished conditions under the “ASD” umbrella. 10 Kanner introduced the “Refrigerator Mother” theory of autism, but later abandoned it. Its true champion was Bruno Bettelheim (1967). 23 research,” Verhoeff notes, “is guided by the depiction of autism as a scientific and physical object that can be discovered and identified with systematic biomedical and neuroscientific investigation” (2013, p. 1). This approach reflects the medical model that took hold of psychiatry in the late 20th century, which I explore in the next section. 2.3 ASD and the Medical Model of Psychiatry The increased attention paid to the biological underpinnings of autism stems from a broader shift towards the medical model of psychiatry, which holds that mental illnesses are best investigated and intervened upon using the methods and tools of modern biomedicine (Black, 2005). Notwithstanding the influential work of Thomas Szasz (1974), who challenged the very idea of mental illness, this view has come to dominate contemporary psychiatry. There are two broad ways of understanding of the medical model, which Murphy (2009) calls the minimal interpretation, which prioritizes the behavioral and cognitive levels of psychiatric disorder, and the strong interpretation, which prioritizes the biological level. In this section, I look at these interpretations and ask how they fit with the various ways of understanding autism. 2.3.1 The Minimal Interpretation Proponents of the minimal interpretation of the medical model (see Guze, 1992; McHugh & Slavney, 1998) think of “mental disorders as collections of symptoms that occur together and unfold in characteristic ways,” while making “no commitments about the underlying causes of mental illness” (Murphy, 2009, p. 103). This view falls under the medical model because it acknowledges that psychiatric conditions are associated with 24 biological markers and depend upon biological processes; it is a minimal interpretation because it does not define or classify psychiatric conditions in terms of these biological markers or processes. Rather than focusing on etiology, the minimal model is concerned primarily with symptomatology and prognosis. This way of thinking about mental illness is commonly attributed to Kraepelin, who employed something like the minimal interpretation of the medical model in his process of differential diagnosis. Siding with the “somatists” of his day, Kraepelin embraced an empirical view of psychiatry, according to which mental disorders are “the psychic manifestations of more or less subtle changes in the brain” and “illnesses to be treated by a doctor” (1899a, p. 2). At the same time, he acknowledged the complexity of psychopathology and urged caution in proclaiming particular causes of mental disorders (1899a, p. 3). Given the limits of psychiatric knowledge, Kraepelin argued, “only the overall picture of a medical case from the beginning to the end of its development can provide justification for its being linked with other observations of the same kind” (1899b, p. 3). This has led some to label the minimal interpretation a neo-Kraepelinian view.11 Murphy, however, notes that the label can be misleading, since Kraepelin saw a syndromebased approach to mental disorders as an unfortunate product of psychiatry’s immaturity. He was hopeful that, as the field progressed, etiology and symptomatology would converge and that “pathological anatomy promises to provide the safest foundation for the classification” of mental illness (1899b, p. 2). Whether or not it is “Kraepelinian,” the DSM is commonly interpreted as endorsing 11 This label was introduced in Klerman (1978). 25 a minimal interpretation of the medical model. In the DSM-5, mental disorders are defined as “syndromes comprising clinically significant disturbances of cognition, emotion or behaviour that reflect underlying dysfunctions” (American Psychiatric Association, 2013, p. 20). Here we see a gesture towards etiology without it supplanting course and symptoms as a condition’s classifying features. The DSM’s definition appears to aim at diagnostic reliability, or “agreement across measurements and observers,” as opposed to diagnostic validity, which is concerned with “what is really there” (Murphy, 2015). This approach reflects the clinical goals of prediction and control, which a pessimist might say is the best we can hope for in a domain as causally and ontologically messy as psychiatry. Interestingly, the DSM-5’s diagnostic criteria for ASD are not entirely without reference to etiology. It states that symptoms “must be present in the early developmental period” and “not better explained by intellectual disability (intellectual developmental disorder) or global developmental delay” (American Psychiatric Association, 2013a, 299.00).12 Early drafts of the ICD-11 similarly exclude patients with diagnoses of developmental language disorder, schizophrenia, and other primary psychotic disorders. While ASD is often characterized as a mere description of behaviors, it is in fact a description of behaviors that do not share an etiology with certain other psychiatric conditions. This is not to deny that ASD is characterized in the tradition of the minimal interpretation of the medical model; rather, it is to note that even the minimal interpretation makes certain appeals to cause. The minimal interpretation is appealing in that it makes psychiatric classification 12 The DSM-4 included a narrower provision: “The disturbance is not better accounted for by Rett’s Disorder or Childhood Disintegrative Disorder” (American Psychiatric Association, 2000, 299.00). 26 possible in the absence of definitive biological understanding of psychiatric disorders. One problem, however, is that it is not clear how the minimal interpretation would allow for diagnostic false positives and negatives, except for those resulting from clinician error in applying the criteria, an issue I address in Chapter 2. This is relevant insofar as we think it is intelligible to ask whether someone diagnosed with ASD really has autism or whether the documented prevalence of autism is accurate. The possibility of false positives and negatives allows us to make sense of statements like “My son is diagnosed with ASD, but I think he’s just quirky.” How can we evaluate such statements? On the minimal interpretation of the medical model, we can quibble about diagnostic criteria and the clinical application of those criteria, but once those debates are settled there is nothing more to be said. The search for some kind of objective measure with which to validate psychiatric diagnoses has provided motivation for the strong interpretation, to which I now turn. 2.3.2 The Strong Interpretation On the strong interpretation of the medical model, psychiatric conditions are “not just sets of co-occurring symptoms, but destructive processes taking place in biological systems” (Murphy, 2009, p. 104). This interpretation reflects a broader trend in medicine, which has been moving away from syndrome-based views of disease since the 19th century (see Carter, 2003). Proponents argue that psychiatry, as a branch of medicine, should aim beyond mere prediction and control to gain an understanding of the underlying causal structure of mental illness (see Andreasen, 2001; Murphy, 2006).13 Hempel (1965) appears 13 Or, as I discuss in Section 3.5, psychiatry should seek to understand the underlying causal structure of mental illness so as to aid in prediction and control. 27 to have anticipated the strong interpretation, predicting that advances in psychiatry would lead to classification based upon the objective features of mental illness rather than its clinical course. As noted, both the DSM and ICD are generally seen as following a minimal interpretation of the medical model. Concerns that the resulting diagnostic categories “fail to align with findings emerging from clinical neuroscience and genetics” (Insel et al., 2010, p. 748) has prompted the National Institute of Mental Health to introduce the Research Domain Criteria (RDoC), a new framework for psychiatric research and classification. RDoC understands mental illnesses as disorders of brain circuits, which are (at least potentially) detectable by the tools of neuroscience, seeking to stabilize psychiatric diagnoses on a foundation of biology. The strong interpretation is in many ways aspirational, which can be seen as both a detriment and an attribute of the view. As we have seen, the strong interpretation may simply not yet be operational for a number of psychiatric categories, ASD among them. Furthermore, even if the science advanced to a point at which we could identify a number of subtypes that fall under the ASD umbrella, there may be practical and social benefits to grouping individuals who present similar signs and symptoms, just as there are benefits of acknowledging various social groups that are not distinguished by some biological feature. On the other hand, the strong interpretation provides both a clear basis for false positive and negative diagnoses and the possibility of certain medical benefits. The most obvious potential benefit is that seeking a causal understanding of psychiatric conditions will help with the development of more effective interventions. A further possibility is that focusing on biological markers will provide the opportunity for intervention before signs 28 and symptoms become clinically significant, as physicians do when they treat hypertension before myocardial infarction (see Insel, 2009). How this will apply to developmental conditions like autism, which are present from early childhood, is unclear.14 In any case, recent excitement about the promise of biomarkers and “bioprediction” (see Baum, 2016) suggests that the future of psychiatry will involve prevention, rather than mere management, of signs and symptoms. 2.4 ASD and Medical Ontology Both the minimal and strong interpretations of the medical model of psychiatry tell us that psychiatric conditions are medical conditions not different in kind from any other. But what kinds of things are medical conditions? This is a question of medical ontology. In the most general terms, one can be either a realist or an antirealist about medical conditions, which are commonly referred to as “disorders” or “diseases.” Realism in medical ontology is the view that diseases are naturally occurring entities to be discovered, while antirealism holds that diseases are constructed and imposed for theoretical or practical purposes. In this section, I canvass in more detail the major positions in medical ontology and how they relate to autism and the two interpretations of the medical model of psychiatry. 14 Perhaps the greatest promise for early intervention upon developmental conditions lies in maternal-fetal medicine. 29 2.4.1 Realism Realism in psychiatry is the view that mental illnesses are discrete, naturallyoccurring entities which validate or invalidate psychiatric diagnoses. Though realists may often hold a strong interpretation of the medical model of psychiatry, the entity in question need not be biological (Broome, 2006). Simon (2011) distinguishes between realism about disease tokens and disease types. Realists about disease tokens think that there is some disease entity that unites a given patient’s symptoms. Realists about disease types go further, claiming that disease entities can be sorted into types or kinds which are themselves real. Though realists about tokens may often be realists about types, these two views are conceptually distinct. One might endorse a mixed realism, according to which tokens are real but types are not. A token realist, then, would interpret a statement like “Taylor has autism” as meaning that there is some thing, call it autism, underlying a cluster of Taylor’s cognitive and behavioral features. A type realist would interpret a statement like “Taylor and Fatima have autism” as meaning that the very same thing, autism, underlies Taylor’s symptoms and Fatima’s symptoms. A mixed realist would acknowledge that both Taylor and Fatima have a medical condition called autism, but would maintain that “autism” here is a homograph or pragmatic construct and thus should not be taken to denote two tokens of the same type. This view may be appealing when it comes to autism (and other spectrum disorders) since we might want to say that any given person diagnosed with ASD has a real disorder, but that the kind “autistic” is just a heuristic or category constructed for pragmatic reasons. We have seen that one can be a realist about disease tokens or disease types. Next, 30 we should ask what kinds of entities these views posit. A number of options present themselves (see Simon, 2011), the most prominent three of which I will review here. First is the view that disease, qua ontological entity, is a cluster of signs and symptoms. L. J. Rather (1959) provides an early defense of this view, which he attributes to Sydenham’s famous work concerning the clinical manifestations of gout. For a signs-and-symptoms realist, “autism” would refer to some constellation of clinically observable behaviors. While biological and cognitive features may cause or contribute to these behaviors, the behaviors themselves (on this view) are the thing that is autism. The focus of signs and symptoms will bring to mind the minimal interpretation of the medical model of psychiatry (discussed in Section 2.3.1), which defines mental illness in terms of its clinical presentation. But those who endorse the minimal interpretation need not be signs-and-symptoms realists, since it might be that a constellation of signs and symptoms do not constitute a real entity. Stubblefield makes this clear in the context of ASD when she states “it is important to understand that there is no such ‘thing’ as ‘autism’—because to diagnose someone as autistic is simply to observe that he or she displays a certain number of a list of characteristics” (2013, p. 148). A second variety of realism sees diseases as disordered physical states of the body. Rudolph Virchow understood somatic disease this way, stating “the disease entity is an altered body part, or … an altered cell or aggregate of cells, whether tissue or organ” (1895/1958, p. 192). This view, applied to psychiatry, understands disease entities in terms of biology, which makes it compatible with the strong interpretation of the medical model (discussed in Section 2.3.2). With regard to autism, the disease-as-state view would hold 31 that it should be understood at its most basic level,15 with observable cognitive and behavioral characteristics considered mere symptoms of some underlying disordered bodily state. Third is a realist position advanced by Caroline Whitbeck (1977), which Simon (2011, p. 71) calls the “disease-as-process” view. For Whitbeck, a disease is a diachronic process, distinguishable from other abnormal conditions like injuries, which can be characterized synchronically. Diseases are unique in that they follow a specific course over time. It is therefore not enough to consider the signs and symptoms or underlying bodily state of a disease; rather, disease must be understood as a particular sequence of signs, symptoms, and states. This view may be appealing when talking about autism, since it takes the “developmental” piece of neurodevelopmental disorder seriously. Though controversial, the influence of realism about autism can be found in both historical and contemporary writings. Both Kanner and Asperger appear to have been realists. Kanner (1965) states, “we can state unreservedly that…[autism] does represent a ‘definitely distinguishable disease.’ This disease, specific—that is, unique, unduplicated— in its manifestations, can be explored per se” (p. 418). Asperger (1944/1991), for his part, writes transparently that “the autistic personality type is a natural entity” (p. 67). Verhoeff notes that, despite the fact that autism has come to be understood as falling on a spectrum, “the idea of autism as a distinct nosological entity with a particular essential core deficit remains pervasive” (2012, p. 19). Critics of realist views about autism will point to the heterogeneity of signs, symptoms, and biomarkers found among individuals diagnosed with ASD as evidence that 15 Which is why the disease-as-state view may be called a reductionist view. 32 autism is not a distinct entity. As mentioned, however, heterogeneity is pervasive in biology and therefore (one could argue) not a unique problem for autism qua entity. If we are looking for biological entities as homogenous and clearly defined as paradigmatic examples of natural kinds—gold, for example—we are likely to fail. Even distinct disorders like Down syndrome vary rather widely in phenotype (e.g., intellectual ability) and genotype (e.g., mosaic and complete trisomy 21). Unless one is prepared to be an antirealist about all biological entities, then, making a case for antirealism about autism requires more than pointing out heterogeneity among autistics. One must argue also that there is some threshold of heterogeneity that biological entities must not cross, and that autism does cross. Having noted the debate surrounding autism and heterogeneity, let us now look briefly at antirealist views in medical ontology. 2.4.2 Antirealism Simon (2011) takes antirealist positions about disease to be one form or another of constructivism, which is the view that “disease is created by the inquiring intellect, is carved out of a block of marble by the chisel strokes of the sculptor” (King, 1954, p. 200). Broome (2006) helpfully distinguishes between pragmatist and eliminativist views about psychiatric classification. Pragmatism holds that psychiatric categories are empirical generalizations that have practical benefits for diagnosis, prognosis, and treatment. Unlike realism, which often takes on an essentialist slant, pragmatism “sees a classification as fulfilling certain concerns and purposes, that may well be in conflict, and considers that disorders themselves do not necessarily have to have a discrete essence or be a natural 33 kind” (Broome, 2006, p. 304).16 Eliminativism, alternatively, holds that current psychiatric categories are likely to disappear as scientific psychiatry advances to becomes better equipped to identify the particulars of disease tokens, a possibility touched upon above. In Broome’s words, eliminativism …suggests that the entities of psychiatry can and should be reduced to their biological underpinnings, and that it is unlikely that such entities will survive the reduction. This position may eliminate our “folk” psychopathology and classification…ontological primacy will be given to neurological explanations, and the entities that make up such explanations, instead. (2006, p. 304) Despite the neuroscientific turn in autism research seen in recent decades, it is common to see autism described in antirealist terms. Biklin (2005), for instance, notes that “autism is a concept that is developed and applied, not discovered” (p. 12). Broderick (2011) stresses the importance of rhetoric in our understanding of autism: “Because autism is fundamentally discursive, it cannot exist outside of, or prior to, rhetoric.” Verhoeff, too, seems to be an antirealist about autism: [T]he idea of an essential core or a natural autism entity that is waiting to be identified is misguided, and the corresponding dominant neuroscientific approach to autism obscures an array of social, cultural and psychological issues important in understanding the phenomenon we call “autism” (2012, p. 21). There seems to be a tension, therefore, in our thinking about autism: On the one hand, we have moved beyond primitive psychogenic explanations of autism as a result of embracing a scientific view of psychiatry that understands psychiatric disorders as brain disorders. On the other, we are reluctant to think of autism as a naturally occurring entity free from social baggage. 16 Returning to the type/token distinction discussed in the previous section, it should be noted that one could be a realist about disease tokens and a pragmatist about disease types, since pragmatism is concerned with the way psychiatric conditions are grouped. 34 2.5 So What? Thus far, I have discussed three levels of analysis for autism, two interpretations of the medical model of psychiatry, and a number of “isms” related to medical ontology. While the philosophically-minded reader might find these topics interesting in and of themselves, others by now are likely asking themselves a critical question: So what? Throughout this chapter, I have hinted at how the philosophical tools being introduced may be put to use. In this section, I will be explicit about why these theoretical debates matter. This will serve both to review what has been covered in this chapter and to lay out issues to be addressed in subsequent chapters. 2.5.1 Diagnostic Validity Imagine undergoing a medical test for something like leukemia. Clearly, it will matter greatly to you that the test is accurate. A false positive will result in unnecessary treatment and emotional distress; a false negative will result in neglecting the disease and a potentially worse prognosis. It is important, then, that the test for leukemia be diagnostically valid—that one tests positive if and only if one really has the disease. We might say the same thing about tests for autism and other psychiatric conditions. When someone claims that ASD is over-diagnosed, or that a child is quirky, not autistic, or that there is an epidemic of undiagnosed autism, she is questioning the validity of our diagnostic tools. But the truth of a very practically relevant claim like “ASD is overdiagnosed” depends in large part on the positions we take in the seemingly esoteric philosophical debates over the medical model of psychiatry and medical ontology. Suppose, for example, one adopts a minimal interpretation of the medical model, 35 which (recall) defines psychiatric conditions in terms of their signs and symptoms. On this view, diagnostic validity is threatened only by misapplication of agreed upon diagnostic criteria. There is no worry that a given cluster of signs and symptoms do not correspond with a disorder, since the disorder is defined in terms of these signs and symptoms. In other words, diagnostic validity follows trivially from proper application of diagnostic criteria. Of greater concern for the minimal interpretation is this proper application, or diagnostic reliability, which requires that clinicians arrive at the same diagnosis given the same clinical presentation.17 If, alternatively, one adopts a strong interpretation of the medical model, or takes a more robustly realist position on disease ontology, then it is possible for the signs and symptoms meant to track a disorder to come apart from the disorder, leaving room for falsely positive or negative diagnoses. As we know from somatic medicine, a given symptom or cluster of symptoms may indicate a number of disorders, or perhaps no disorder at all, as is the case with (say) night sweats and coughing. On the other side of the same coin, a single biological dysfunction can present itself in different ways. Capgras syndrome and Cotard syndrome, for example, are thought to share a neural etiology, while having distinct (as opposed to merely heterogeneous) clinical and phenomenological presentations (Young, Leafhead, & Szulecka, 1994). In Capgras syndrome, one experiences the delusion that a close friend or relative has been replaced by an impostor. In Cotard syndrome, one experiences the delusion that one is dead. For a proponent of the 17 The same might be said of antirealists of various stripes, insofar as they deny that there is anything there to validate a diagnosis. The antirealist may, however, still acknowledge the pragmatic benefits of developing diagnostic categories that are applied consistently, i.e., categories that are reliable. 36 strong interpretation, it does not follow from the fact that two individuals share a clinical presentation that they share a disorder; and it does not follow from the fact that two individuals present differently that they do not share a disorder. This possibility has led some to worry that psychiatry’s focus on diagnostic reliability rather than diagnostic validity has resulted in lumping together groups of people who, despite similar clinical presentations, may be different in important ways. Horwitz and Wakefield for example, argue that DSM’s criteria for major depressive disorder (2007) and generalized anxiety disorder (2012) can be applied both to individuals with pathological conditions and those experiencing normal human emotions—one must appeal to causes in order to determine who is truly disordered and in which way. This is how we can say that a person is merely sad, not depressed (see Zachar, 2014). We see, then, that the philosophical positions we take influence, and perhaps determine, how we navigate practical scientific concerns about identifying and diagnosing autism. In the next section, I outline several ethical and political implications of the philosophical positions we take on psychiatry and disease. 2.5.2 Disorder and Diversity An active controversy in the autism world—which I address in greater detail in Chapters 4 and 5—concerns whether autism is in fact a disorder, as the label “Autism Spectrum Disorder” suggests. In addition to facing potential challenges from the so-called anti-psychiatrists,18 autism’s status as a disorder has been called into question by 18 Recall that the antipsychiatry movement takes issue with the medical model of psychiatry across the board. See Szasz (1974). 37 proponents of the more recent neurodiversity movement (see Baron-Cohen, 2015; Doan and Fenton, 2013; Perry and Herrera, 2014; Singer, 1999). The neurodiversity view regards autism and certain other atypical forms of neurological development as “normal human difference that should be tolerated and respected in the same way as other human differences” (Jaarsma & Welin, 2012, p. 24). How are we to evaluate the claim that autism is not a disorder? I view it like completing a child’s shape sorting task, in which one must fit a block of a certain shape into the corresponding opening on a box. First, we must select a block by making certain ontological commitments about autism: Do we want to understand autism minimally, as a cluster of behaviors which may or may not share a common etiology; or strongly, as a cluster of biological or cognitive features which may or may not manifest in certain behaviors? Do we want to be realists about autism state or process? Do we want to think of particular cases of autism as tokens of a type? Second, we must select an opening on the box by adopting a concept of disorder. There are a number of options at this stage: We could be naturalists and hold that disorder consists in some kind of biological dysfunction (see Boorse, 1977); or normativists and hold that disorders are disvalued conditions (see Cooper, 2002); or adopt a hybrid view and treat disorders as harmful dysfunctions, where “harm” is defined normatively and “dysfunction” is defined naturally (see Wakefield, 1992).19 Third, we need to see whether the block fits into the opening by determining whether our ontological view of autism corresponds with the concept of disorder we have 19 For a compelling probabilistic twist on a hybrid view of mental disorder, see Baum (2016). 38 adopted. If, for instance, we understand autism as a mere cluster of behaviors and disorder as requiring some kind of biological dysfunction, then it seems that the block will not fit. But if we prefer a concept of disorder that requires harm or increased probability of harm and determine that autism is harmful in the relevant sense, then it seems that the block will fit. The results of this process of determining fit will have implications for both the ethics and politics of autism, which I explore in subsequent chapters. In the ethical realm, which conditions are considered to be disorders influence judgments about moral responsibility (people with certain disorders may have reduced responsibility, while people with mere differences do not), autonomy and authenticity (perhaps disorders threaten, while differences constitute, autonomy and authenticity), and the obligations of clinicians (physicians ought to treat patients with disorders, but have no obligation, and perhaps an obligation against, treating patients with mere differences). It might be argued that these implications stem from problematically binary ways of thinking about disorder and difference, treatment and enhancement, and so on. I am sympathetic to this worry, but, even so, we must acknowledge that such categories do, as a matter of fact, influence moral judgments and institutional policies. In the political realm, the way we understand autism and disorder will have ramifications for everything from the legal system to health insurance regulation. One of the central claims of the neurodiversity movement is that persons with autism should be accommodated rather than medicalized. Which views of autism and disorder will support this claim? Which will undermine it? My hope is that the work done in what follows will help to address these important questions. 39 2.6 References Allely, C. S., Wilson, P., Minnis, H., Thompson, L., Yaksic, E., & Gillberg, C. (2017). Violence is rare in autism: When it does occur, is it sometimes extreme? The Journal of Psychology, 151(1), 49-68. American Psychiatric Association (APA). (2000). Diagnostic and statistical manual of mental disorders: DSM-4. Washington, D.C: American Psychiatric Association. American Psychiatric Association. (2013a). Diagnostic and statistical manual of mental disorders: DSM-5. Washington, D.C: American Psychiatric Association. American Psychiatric Association. (2013b). DSM-5 Autism Spectrum Disorder fact sheet. Retrieved from http://www.dsm5.org/Documents/Autism%20Spectrum%20Disorder%20Fact%20 Sheet.pdf. Anagnostou, E. & Taylor, M. J. (2011). Review of neuroimaging in Autism Spectrum Disorders: What have we learned and where do we go from here? Molecular Autism, 2(4). Andreasen, N.C. (2001). Brave new brain. New York, NY: Oxford University Press. Asperger, H. (1944). ‘Autistic psychopathy’ in childhood. In U. Frith (Ed.). Autism and Asperger Syndrome (1991, pp. 37-92). Cambride, UK: Cambridge University Press. Baron-Cohen, S. (1987). Autism and symbolic play. British Journal of Developmental Psychology 5, 139-148. Baron-Cohen, S. (1995). Mindblindness: An essay on autism and theory of mind. Cambridge, MA: MIT Press. Baron-Cohen, S. (2000). Theory of mind and autism: A fifteen year review. In S. BaronCohen, H. Tager-Flusberg & D. J. Cohen (Eds.). Understanding other minds: Perspectives from developmental cognitive neuroscience (2nd ed., pp. 3-20). New York: Oxford University Press. Baron-Cohen, S. (2002). The extreme male brain theory of autism. Trends in Cognitive Sciences, 6(6), 248-254. Baron-Cohen, S. (2015). Neuroethics of neurodiversity. In J. Clausen & N. Levy (Eds.). Handbook of neuroethics (pp. 1757-1763). Netherlands: Springer. Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic child have a “theory of mind?” Cognition 21(1), 37-46. 40 Baum, M. L. (2016). The neuroethics of biomarkers: What the development of bioprediction means for moral responsibility, justice, and the nature of mental disorder. New York, NY: Oxford University Press. Begeer, S., Malle, B., Nieuwland, M., & Keysar, B. (2010). Using theory of mind to represent and take part in social interactions: Comparing individuals with highfunctioning autism and typically developing controls. European Journal of Developmental Psychology, 7(1), 104-122. Beresford, M. C. (2018). Dear media: Stop blaming autism for acts of mass violence. The Mighty. Retrieved from https://themighty.com/2018/09/stop-blaming-autismmass-violence/. Bettelheim, B. (1967). The empty fortress: Infantile autism and the birth of the self. New York: Free Press. Biklen, D. (2005). Autism and the myth of the person alone. New York, NY: New York University Press. Black, K. (2005). Psychiatry and the medical model. In E. Rubin & C. Zorumski (Eds.). Adult psychiatry (2nd ed., pp. 3-15). Malden, MA: Blackwell. Boorse, C. (1977). Health as a theoretical concept. Philosophy of Science, 44(4), 542-573. Brokerick, A. A. (2011). Autism as rhetoric: Exploring watershed rhetorical moments in applied behavior analysis discourse. Disability Studies Quarterly, 31(3). Broome, M. R. (2006). Taxonomy and ontology in psychiatry: A survey of recent literature. Philosophy, Psychiatry & Psychology, 13(4), 303-319. Carter. K. C. (2003). The rise of causal concepts of disease. Aldershot, UK: Ashgate. Chevallier, C., Kohls, G., Troiani, V., Brodkin, E. S., & Schultz, R. T. (2012). The social motivation theory of autism. Trends in Cognitive Sciences, 16(4), 231-239. Cooper, R. (2002). Disease. Studies in History and Philosophy of Biological and Biomedical Sciences, 33, 263-282. Dapretto, M., Davies, M. S., Pfeifer, J. H., Scott, A. A., Sigman, M., Bookheimer, S. Y., & Iacoboni, M. (2005). Understanding emotions in others: Mirror neuron dysfunction in children with Autism Spectrum Disorders. Nature Neuroscience, 9(1), 28-30. Doan, M., & Fenton, A. (2013). Towards reconceiving certain “autism-related” behavioral atypicalities as functional. In J. L. Anderson & S. Cushing (Eds.). The philosophy of Autism (pp. 47-72). Lanham, MD: Rowman & Littlefield. 41 Donnellan, A., Hill, D., & Leary, M. (2010). Rethinking autism: Implications of sensory and movement differences. Disability Studies Quarterly, 30(1). Frith, U. (2003). Autism: Explaining the enigma (2nd ed.). Oxford, UK: Blackwell. Frith, U., & Happé F. (1994). Autism: Beyond ‘theory of mind.’ Cognition, 50, 115-132. Gernsbacher, M. A., & Frymiare, J. (2005). Does the autistic brain lack core modules? Journal of Developmental and Learning Disorders, 9(3), 3-16. Glüer, K., & Pagin, P. (2003). Meaning theory and autistic speakers. Mind & Language, 18(1), 23-51. Guze, S. B. (1992). Why psychiatry is a branch of medicine. New York, NY: Oxford University Press. Happé, F. (1999). Autism: Cognitive deficit or cognitive style? Trends in Cognitive Sciences, 3(6), 216-222. Happé, F., & Frith, U. (2006). The weak coherence account: Detail-focused cognitive style in autism spectrum disorders. Journal of Autism and Developmental Disorders, 36(1), 5-25. Happé, F., Ronald, R., & Plomin, R. (2006). Time to give up on a single explanation for autism. Nature Neuroscience, 9(10), 1218-1220. Hempel, C. (1965). Fundamentals of taxonomy. In C. Hempel (Ed.). Aspects of scientific explanation (pp. 137-154). New York, NY: The Free Press. Hill, E. L. (2004). Executive dysfunction in autism. Trends in Cognitive Sciences, 8(1), 26-32. Horwitz, A. V., & Wakefield, J. C. (2007). The loss of sadness: How psychiatry transformed normal sorrow into depressive disorder. New York, NY: Oxford University Press. Horwitz, A. V., & Wakefield, J. C. (2012). All we have to fear: Psychiatry’s transformation of natural anxieties into mental disorders. New York, NY: Oxford University Press. Insel, T. R. (2009). Translating scientific opportunity into public health impact: A strategic plan for research on mental illness. Archives of General Psychiatry, 66(2), 128-133. Insel, T. R., Cuthbert, B., Garvey, M., Heinssen, R., Kozalk, M., Pine, D. S., … Wang, P. (2010). RDoC: Toward a new classification framework for research on 42 mental disorders. American Journal of Psychiatry, 167, 748–751. Jaarsma, P., & Welin, S. (2012). Autism as a natural human variation: Reflections on the claims of the neurodiversity movement. Health Care Analysis, 20(1), 20-30. Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child, 2, 266272. Kanner, L. (1965). Infantile autism and the schizophrenias. Behavioral Science, 10(4), 412420. Kelleher, R. J., & Bear, M. F. (2008). The autistic neuron: Troubled translation? Cell, 135(3), 401-406. King, L. S. (1954). What is disease? Philosophy of Science, 21, 193-203. Klerman, G. L. (1978). The evolution of scientific nosology. In J.C. Shewshow (Ed.). Schizophrenia: Science and practice (pp. 99-121). Cambridge, MA: Harvard University Press. Kogan, M.D., Vladutiu, C.J., Schieve, L.A., Ghandour, R.M., Blumberg, S.J., Zablotsky, … Lu, M.C. (2018). The prevalence of parent-reported Autism Spectrum Disorder among US children. Pediatrics, 142(6), p.e20174161. Kraepelin, E. (1899a). Psychiatry: A textbook for students and physicians vol. 1 general psychiatry (6th ed.). Canton, MA: Science History Publications 1990. Kraepelin, E. (1899b). Psychiatry: A textbook for students and physicians vol. 2 clinical psychiatry (6th ed.). Canton, MA: Science History Publications 1990. Lichtenstein, P., Carlstrom, E., Rastam, M., Gillberg, C., & Anckarsater, H. (2010). The genetics of Autism Spectrum Disorders and related neuropsychiatric disorders in childhood. American Journal of Psychiatry, 167(11), 1357-1363. Lord, C., & Jones, R.M. (2012). Annual research review: Re-thinking the classification of Autism Spectrum Disorders. Journal of Child Psychology and Psychiatry and Allied Disciplines, 53(5), 490-509. Markram, K., & Markram, H. (2010). The intense world theory: A unifying theory of the neurobiology of autism. Frontiers in Human Neuroscience, 4(224). McCoy, T. (2014, May 21). Study: ‘Significant’ statistical link between autism and mass murder. Washington Post. McHugh, P., & Slavney, T. (1998). The perspectives of psychiatry. Baltimore, MD: Johns Hopkins University Press. 43 Murphy, D. (2006). Psychiatry in the scientific image. Cambridge, MA: MIT Press. Murphy, D. (2009). Psychiatry and the concept of disease as pathology. In M. Broome & L. Bortolotti (Eds.). Psychiatry as cognitive neuroscience: Philosophical perspectives (pp. 103-117). New York, NY: Oxford University Press. Murphy, D. (2015). Philosophy of psychiatry. Stanford Encyclopedia of Philosophy. Retrieved from https://plato.stanford.edu/entries/psychiatry/. Nichols, S., & Stich, S. (2003). How to read your own mind: A cognitive theory of self-consciousness. In Q. Smith & A. Jokic (Eds.). Consciousness: New philosophical perspectives (pp. 157-200). New York, NY: Oxford University Press. Ozonoff, S., Pennington, B. F., & Rogers, S. J. (1991). Executive function deficits in high-functioning autistic children: Relationship to theory of mind. Journal of Child Psychology and Psychiatry, 32, 1107-1122. Perner, J., Frith, U., Leslie, A. M., & Leekham, S. R. (1989). Exploration of the autistic child’s theory of mind: Knowledge, belief and communication. Child Development, 60, 688-700. Perry, A., & Herrera, C. (Eds.). (2014). Ethics and neurodiversity. Cambridge, UK: Cambridge Scholars Publishing. Premack, D., & Woodruff, G. (1978). Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1(4), 515-526. Rather, L. J. (1959). Towards a philosophical study of the idea of disease. In Brooks and Cranefield (Eds.). The historical development of physiological thought (pp. 353373). New York, NY: Hafner. Sample, R. (2013). Autism and the extreme male brain. In J. L. Anderson & S. Cushing (Eds.). The philosophy of autism (pp. 73-101). Lanham, MD: Rowman & Littlefield. Simon, J. R. (2011). Medical ontology. In F. Gifford (Ed.). Philosophy of medicine (pp. 65-114). Oxford, UK: North Holland. Singer, J. (1999). Why can’t you be normal for once in your life? In M. Corker & S. French (Eds.). Disability and discourse (pp. 59-67). London, UK: Open University Press. Stubblefield, A. (2013). Knowing other minds: Ethics and autism. In J. L. Anderson & S. Cushing (Eds.). The philosophy of autism (pp. 143-166). Lanham, MD: Rowman & Littlefield. 44 Szasz, T. (1974). The myth of mental Illness: Foundations of a theory of personal conduct. New York, NY: Harper and Row. Verhoeff, B. (2012). What is this thing called autism? A critical analysis of the tenacious search for autism’s essence. Biosocieties, 7(4), 410-432. Verhoeff, B. (2013). The autism puzzle: Challenging a mechanistic model on conceptual and historical grounds. Philosophy, Ethics, and Humanities in Medicine, 8(17). Virchow, R. (1895/1958). One hundred years of general pathology. In L. J. Rather (Trans.). Disease, Life, and Man: Selected Essays (pp. 170-215). Stanford, CA: Stanford University Press. Wakefield, J. (1992). Disorder as harmful dysfunction: A conceptual critique of DSMIII-R’s definition of mental disorder. Psychological Review, 99(2), 232-247. Whitbeck, C. (1977). Causation in medicine: The disease entity model. Philosophy of Science, 44, 619–637. Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: Representations and constraining function of wrong beliefs in young children’s understanding of deception. Cognition, 13, 103-128. World Health Organization. (1992). ICD-10 classification of mental and behavioural disorders: Clinical descriptions and diagnostic guidelines. Geneva, Switzerland: World Health Organization. Young, A.W., Leafhead, K.M., & Szulecka, T.K. (1994). The capgras and cotard delusions. Psychopathology, 27, 226–231. Zachar, P. (2014). Philosophical pathology: A metaphysics of psychopathology. Cambridge, MA: MIT Press. CHAPTER 3 A DILEMMA FOR “AUTISM SPECTRUM DISORDER”20 3.1 Abstract The DSM-5’s diagnostic criteria for Autism Spectrum Disorder are based upon clinical manifestations rather than etiology or mechanism. This purely descriptive approach to classification is consistent with two ways of understanding autism: (1) The diagnostic criteria provided in the DSM are constitutive of autism, and thus to have autism is simply to satisfy the criteria; (2) The diagnostic criteria provided in the DSM are taken to be symptomatic of autism, in which case autism is independent of the diagnostic criteria. In this paper, I propose a dilemma for ASD as a diagnostic category: If (1), then the DSM identifies autism only trivially, revealing nothing interesting about the world. If (2), then the biological heterogeneity seen among those diagnosed with ASD threatens the category's clinical utility (usefulness in developing medical interventions) and projectability (usefulness in making inferences). I conclude, therefore, that ASD as a diagnostic category is either trivial or clinically unhelpful. 20 This chapter has been formatted for stand-alone publication. 46 3.2 Classification and Ontology in the DSM Autism, like so much in psychiatry, is not well understood. Despite promising research into its genetic and neurobiological bases, autism remains defined by its clinical manifestations—namely, behaviors—rather than its cause or mechanism. Psychiatry’s gold-standard diagnostic tool, the fifth edition of the American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (APA, 2013a), provides criteria for “Autism Spectrum Disorder” (ASD) along two such dimensions: (1) Persistent deficits in social communication and social interaction across multiple contexts and (2) Restricted, repetitive patterns of behavior, interests, or activities.21 This characterization of autism represents a purely descriptive method of classification, which sorts disorders according to their signs, symptoms, and course, as opposed to a theoretical method, which sorts disorders according to their underlying etiology and pathology. Although explicit about its approach to nosology, the DSM-5’s definition of “mental disorder” leaves its ontological commitments somewhat unclear. It states the following: A mental disorder is a syndrome characterized by clinically significant disturbance in an individual's cognition, emotion regulation, or behavior that reflects a dysfunction in the psychological, biological, or developmental processes underlying mental functioning…usually associated with significant distress in social, occupational, or other important activities. (APA, 2013a, p. 20) Here we see reference both to dysfunction and the clinical manifestations of 21 Dimensions (1) and (2) are each rated on a three-level severity scale; both must “be present in the early developmental period” and “cause clinically significant impairment,” and not “better explained by intellectual disability.” 47 dysfunction, making room for differing interpretations of role of diagnostic criteria. One interpretation holds that the diagnostic criteria for a disorder are constitutive of the disorder. On this view, to have autism is simply to satisfy the diagnostic criteria given in the DSM. Among its proponents is Stubblefield, who states “it is important to understand that there is no such ‘thing’ as ‘autism’—because to diagnose someone as autistic is simply to observe that he or she displays a certain number of a list of characteristics” (2013, p. 148; see also Bilken, 2005; Broderic, 2011; Verhoeff, 2012). This way of thinking reflects what Murphy (2009) calls the minimal interpretation of the medical model of psychiatry, whose proponents think of “mental disorders as collections of symptoms that occur together and unfold in characteristic ways,” while making “no commitments about the underlying causes of mental illness” (p. 103). An alternative interpretation holds that the diagnostic criteria for a disorder are merely symptomatic of the disorder. On this view, the DSM is seen as a tool to assist in identifying mental disorders, with autism being understood as a condition the diagnostic criteria for ASD are meant to identify. Verhoeff notes that, despite the clinical practice of treating ASD as a mere description of behaviors, “the idea of autism as a distinct nosological entity…remains pervasive” (2012, p. 19). This idea follows the strong interpretation of the medical model of psychiatry, according to which mental disorders are “not just sets of co-occurring symptoms, but destructive processes taking place in biological systems” (Murphy, 2009, p. 104). With these interpretations in mind, I will pose a dilemma for ASD as a diagnostic category. Adopting the view that diagnostic criteria constitute autism leads to the triviality problem: Defining autism in terms of DSM criteria makes it trivially true that the criteria 48 identify all and only cases of autism, since (on this view) to have autism is simply to satisfy the DSM criteria. Such triviality rules out the possibility of falsely positive or negative diagnoses,22 leaves us unable evaluate claims about over- and underdiagnosis, and ultimately elevates the DSM criteria beyond reproach. The alternative view—that the diagnostic criteria for ASD are mere symptoms of autism—avoids the triviality problem by driving a wedge of contingency between the DSM and what it purports to identify. But here lies the second horn of the dilemma, which I will call the heterogeneity problem: If the diagnostic criteria for ASD are meant to pick out some feature of the world, then it appears—given what we know about people with autism—that “ASD” is a biologically heterogeneous category. And, as I argue in Section 3.5, such heterogeneity threatens the scientific and practical usefulness of “ASD.” This is a concern about the projectability (usefulness in making inferences about individuals diagnosed with the condition) and clinical utility (usefulness in developing medical interventions to control, treat, and prevent the condition) of ASD as a diagnostic category. The unhappy conclusion of this dilemma is that ASD as a diagnostic category is either trivial or clinically unhelpful.23 22 As long as the criteria are correctly applied in the clinical setting. Clinical error is certainly a potential concern, but not one I will address here. 23 I do not want to claim that the analytic/empirical disjunct is conceptually exhaustive; I am claiming merely that the two most prominent interpretations of the role of the diagnostic criteria for ASD face serious problems. It may turn out that the best way to understand the DSM is as providing something other than analytic definitions or empirical hypotheses, such as (say) models or conventions; but I must leave this issue aside. 49 3.3 The Triviality Problem The APA has embraced a purely descriptive approach to classification and diagnosis since the publication of the DSM-III (APA, 1980). Whereas earlier editions of the DSM reflected the period’s psychoanalytic bent by classifying mental disorders etiologically (distinguishing organic and functional causes) and theoretically (distinguishing psychoses and neuroses), more recent editions have sought to bridge theoretical divides among clinicians and to avoid premature claims about the biology underlying mental illness (Tsou, 2011). This more cautious methodology is often seen as prioritizing interrater reliability over diagnostic validity. Interrater reliability is a quantitative measure of the probability that “two [professional] observers will reach the same diagnosis…for a given patent examined at approximately the same time” (Hyman 2010, 158). Diagnostic validity, on the other hand, “represents a correct attribution of disorder” (Horwitz & Wakefield, 2007, p. 8) and picks out “real natural phenomena” (Tsou, 2016, p. 416). Put simply: reliability seeks “agreement,” while validity seeks “what is really there” (Murphy, 2015). Admittedly, there are reasons for adopting the DSM’s descriptive approach— reasons that go beyond the institutional significance of reliability. For one thing, there is no definitive laboratory test (such as karyotyping for Down syndrome) or imaging technique (such as CT scans and MRIs for vascular dementia) with which to make the diagnosis of ASD. Clinicians instead use screening and diagnostic tools24 to determine whether patients meet DSM criteria. Also relevant is the fact that it is autism’s behavioral 24 Most commonly the Autism Diagnostic Observational Schedule, Second Edition (Lord et al., 2012) and the Autism Diagnostic Interview-Revised (Rutter et al., 2003). 50 features, rather its underlying biology per se, which are of primary clinical interest and concern, since these features are what often makes it so difficult for individuals with autism to get along in the world.25 Despite the advantages of diagnosing and classifying according to clinically manifested behaviors, the idea that such behaviors constitute autism raises a potential problem: it suggests that “ASD” is a purely stipulated category, which risks trivializing it and undermining its status as an empirical hypothesis about a mental disorder. To see the issue more clearly, consider two routine cases of diagnosis: • Case 1: A patient presents with a productive cough and dyspnea. After conducting a thorough evaluation, the physician comes to believe that the patient has emphysema. Before making a diagnosis, though, she wants to be sure. So, the physician orders imaging to determine whether the patient’s symptoms are caused by enlarged air spaces in the respiratory bronchioles. The results are positive, confirming that patient does indeed have emphysema. • Case 2: A patient presents with social difficulties, limited interests, and an unusual style of communication. After conducting a thorough evaluation, the physician determines that the patient satisfies the diagnostic criteria for ASD, thus concluding that the patient has autism. Compare the clinical reasoning that occurs in these cases. In Case 1, an assessment of signs and symptoms indicates a diagnostic hypothesis which is then confirmed by an objective test of some measure independent of signs and symptoms. The hypothesis, 25 I wish to remain neutral here on the issue of whether this difficulty is a product of autism per se or a neurotypical or ableist arrangement of society. See Chapters 4 and 5 for a discussion of this issue. 51 however, could have been mistaken. If the patient’s symptoms were the result of (say) widespread narrowing of peripheral airways, then asthma, not emphysema, would have been the correct diagnosis. This falsifiability reflects a substantive validation process at work. In Case 2, however, the assessment of signs and symptoms both indicates and confirms the diagnosis of ASD. Assuming the physician’s assessment was accurate, there is no possibility of falsely positive or negative diagnosis, since, according to the view that autism is a mere description of behaviors, having autism just is exhibiting the behavioral diagnostic criteria for ASD. Herein lies what I am calling the triviality problem: If autism is simply a description of behaviors given in the DSM, then the DSM criteria identify autism trivially and without fail. This suggests a necessary connection between diagnostic criteria and the conditions they are meant to identify. In other words, it is treated as analytically true that anyone who satisfies the diagnostic criteria for ASD has autism, and anyone who fails to satisfy the diagnostic criteria does not. Presumably, though, the specificity and sensitivity of diagnostic criteria is not something that can be ascertained a priori; rather, diagnostic manuals like the DSM should be seen as making empirical claims to be confirmed a posteriori. The DSM-5 Taskforce itself has addressed this point, stating Our immediate task is to set up a framework for an evolution of our diagnostic system that can advance our clinical practice and facilitate ongoing testing of the diagnostic criteria that are intended to be scientific hypotheses, rather than inerrant Biblical scripture. (Regier et al., 2009, pp. 648-649) Even so, many in the literature have noted that the tendency to view the DSM as sacred writ persists among clinicians. For example, Hyman (2007/2010) worries that diagnostic criteria in the DSM have become reified, “despite the fact that they appear to be 52 poor mirrors of nature” (p. 731). Kendler (2016) agrees, noting that ongoing diagnostic reification betrays the category mistake of “taking an index of a thing for the thing itself” (p. 771). Once an index becomes the thing indexed, the empirical risks being treated as analytic. At first glance, the worry about there being a necessary connection between autism and the DSM criteria for ASD may appear to be a nonissue. After all, somatic medicine draws similar connections between diagnostic criteria and disorder.26 Take myocardial infarction (MI), for instance. If a clinician conducts an accurate evaluation of a patient and identifies damaging blockage of blood flow in the heart due to coronary artery disease, then the diagnosis of MI cannot be mistaken. It is analytically true, and unproblematically so, that anyone who has experienced damaging blockage of cardiac blood flow due to coronary artery disease has had an MI. Why not say the same kind of thing about autism? The first thing to note is that, in somatic medicine, symptomatology is commonly distinguished from pathology and etiology. This distinction is important because a given symptom or cluster of symptoms may indicate a number of disorders (or no disorder at all), while a given disorder can manifest in a variety of ways. Returning to our example, one can experience chest pain, dizziness, and shortness of breath without having had an MI; and one can have an MI without experiencing these particular symptoms. A given cluster of symptoms, in other words, is neither necessary nor sufficient for the diagnosis of MI. One concern, then, is that the diagnostic process for ASD, which begins and ends with clinical manifestations, conflates symptomatology with pathology/etiology and thereby 26 I use “disorder” here in the broadest sense, to include everything from disease to illness to injury to impairment, which are sometimes distinguished in the philosophy of medicine literature. See Chapter 1 for a discussion of the concept of disorder. 53 risks producing falsely positive and negative diagnoses. Of course, this concern assumes that there is (or ought to be) some subbehavioral definition of autism, which is precisely the issue in question. In the next section, I will attempt to motivate the idea that ASD diagnoses are perniciously trivial in a way that diagnoses in somatic medicine tend not to be. 3.4 Open Questions Imagine two clinicians in a heated disagreement about the status of a patient. Although they agree that the patient has experienced damaging blockage of cardiac blood flow due to coronary artery disease, one maintains that the patient has had an MI while the other denies it. It seems that the latter clinician in this scenario is simply incompetent, and failing to grasp the meaning of “myocardial infarction.” Now imagine that the clinicians are disagreeing over whether the patient has autism. Although they agree that the patient satisfies the DSM criteria for ASD, one maintains that the patient has autism while the other denies it. It seems less clear in this case that the disagreement is the result of incompetence or miscommunication. Perhaps the latter clinician is skeptical of the validity of the DSM criteria, or perhaps he or she thinks there is there is a better explanation for the patient’s clinical presentation. These possibilities suggest that there are open questions involved with ASD diagnosis that are not involved in the diagnosis of conditions like myocardial infarction. I will now pose three such questions meant to challenge the necessary connection between autism and the DSM criteria for ASD. The first open question concerns the diagnostic criteria themselves: Do the DSM 54 criteria for ASD identify all and only individuals with autism? Or, alternatively, would some other set of criteria better identify individuals with autism? These questions make sense to ask only if one assumes that there is some independent target at which the DSM criteria aim; and I take it that the questions are indeed sensible, unlike similar questions involving analytic truths. That there is an open question here should be unsurprising, given the highly contingent and mutable nature of the DSM (See Carey, 2012 and Maenner et al., 2014 for a discussion of changes in prevalence from DSM-IV to DSM-5). Not only could the DSM criteria be different than they in fact are, there could be alternative sets of criteria which diverge from those laid out in the DSM. Consider the following (fictional) case: One of the APA’s rival psychiatric associations, the Psychiatric Association of America (PAA), releases a diagnostic manual of its own. This manual contains slightly more conservative diagnostic criteria for ASD, in response to the concern of its members that the DSM groups vastly heterogeneous individuals under the same diagnostic label. The result is that there are some people who have autism according to the APA criteria, but not according to the PAA criteria. Does a person who has autism according to the APA criteria but not according to the PAA criteria really have autism? Answering this question seems to require more than getting clear on the meaning of the word “autism.” Rather, it requires giving some principled reason for preferring one set of criteria to the other. The question, in other words, is open and indicative of substantive disagreement, disagreement that should lead us to reject the idea of a necessary connection between diagnostic criteria for a condition and the condition itself. A second open question concerns the possibility of false negatives in ASD diagnosis. While asking whether a patient who experienced damaging blockage of cardiac blood flow really had an MI betrays ignorance about the meaning of “myocardial 55 infarction,” asking whether a patient who fails to satisfy the DSM criteria for ASD really does not have autism appears perfectly reasonable. Indeed, such questions seem to be at the heart of controversies about the underdiagnosis of autism among those who have been screened (see, for example, Carey, 2012; Gnaulati, 2013). Consider another case, which illustrates the fact that open questions remain even after settling upon a given set of diagnostic criteria: A patient diagnosed with ASD works with his friends and family to devise a set of rules and scripts that enable him to minimize his autistic behaviors to the point of clinical insignificance. Following the rules and scripts is a constant challenge for the patient, but he succeeds in doing so, resulting in a more typical-appearing style of social interaction and a slightly broader set of interests.27 Does this individual have autism? If having autism is just exhibiting the behaviors outlined in the DSM, then no. But one could argue that the very fact that he requires this set of rules, and that they are difficult for him to follow, suggests that he has autism. Just as those who maintain normal blood sugar levels with diet and an insulin pump still have diabetes, it seems possible for someone who fails to satisfy the DSM’s diagnostic criteria for ASD to have autism. Whether or not this is actually the case, the mere possibility suggests that the purported necessary connection between the criteria and the condition itself is not necessary after all. Interestingly, the DSM-5 Taskforce considers something like the above case, noting that autistic behaviors “may be masked by learned strategies later in life” (APA 2013a). If autism is just a description of behaviors, however, then it is not immediately clear what is being masked. In any case, the very possibility of masking strategies reveals an open question about the connection between behavior and diagnosis. 27 For a nonfiction account that resembles this case, see Finch (2012). 56 The third open question I will pose concerns the possibility of false positives in autism diagnosis. Whereas the second case raises the question of whether someone who had autism still does, this case raises the question of whether someone who does not have autism ever did. Parents of a 2-year-old boy bring the child to the psychiatrist after growing concerned about his development. They have noticed that he avoids eye contact, lines up his toys rather than playing with them in the typical way, and has been slow to develop speech. After conducting an evaluation, the psychiatrist determines that the boy satisfies the criteria for ASD, informs the parents that their child has autism, and recommends that he begin therapy. The parents never get around to scheduling the therapy, though. And they begin to notice that, as the child gets older, he begins to outgrow the tendencies which originally concerned them. By the time he reaches grade school, he behaves just as a typical child of his age would be expected to behave (see Shulman et al., 2019 and Steinberg, 2012). One way of explaining the boy’s development is that he had autism and outgrew it; another is that he never really had autism, despite satisfying the DSM criteria for ASD for several years as a young child (for related cases, see Blumberg et al., 2016; Nugent, 2012); Seltzer et al., 2003. Which is the better explanation? This, it seems, is an open question. If there were a necessary connection between autism and the diagnostic criteria for ASD, however, the former explanation would be the only possibility. The open questions posed in this section are meant to show that autism diagnoses are trivial in a way that diagnoses of conditions like myocardial infarction are not. One could argue, of course, that, even if the DSM fails to meet certain empirical standards, its approach to classifying and diagnosing autism can be justified in virtue of its social utility or as a “practical kind” (see Kendler et al., 2011). We might think that “ASD” is a useful category in that it provides a sense of shared identity among those with the diagnosis, access to various services and accommodations to those who can benefit from them, or as an explanation/justification for atypical behavior. 57 Practical considerations certainly do, and ought to, bear on psychiatric nosology and practice. Social utility, however, is multifaceted, and its varied goals must be weighed against one another. While one classification scheme may further some end, a different scheme may be more effective in furthering another end. For example, if a shared identity were of primary concern, then a greater number of categories would seem called for, given the breadth of categories like “ASD.” Indeed, some self-described “Aspies” resented the co-opting of Asperger Disorder (included in the DSM-IV) under the broader autism spectrum in DSM-5 (Giles, 2014). If, alternatively, access to care were the sole consideration, then a single institutionally recognized diagnostic category would be sufficient. Why attempt a controversial taxonomy of mental disorders if all we are after is something to justify ABA therapy to the insurance company or an IEP to the school district? Why take on the thorny task of distinguishing individuals with mental illness from those with mere problems in living when some number of the latter may benefit from psychiatric intervention? That the DSM attempts such distinctions suggests that it aspires to more than socially useful stipulations. In fact, the APA states explicitly that, in addition to being “medically and scientifically useful,” “ASD” is supposed to represent a “more accurate” approach to classifying and diagnosing autism-related disorders (APA, 2013b). There are thus two reasons to reject the proposal that trivial diagnostic categories can be justified on the grounds of social utility. First, there is the issue of underdetermination, since a number of possible classification schemes may contribute to social utility. Second, the APA itself claims to be doing more than putting forward a stipulated nosology of mental disorders. It seems, then, that the triviality problem remains 58 a problem for the view that the DSM criteria for ASD constitute autism. The obvious alternative is to view the DSM as a tool for identifying autism in the world, with the diagnostic criteria for ASD being understood as symptoms of the disorder. However, this response faces a problem of its own. 3.5 The Heterogeneity Problem In this section, I present the second horn of the dilemma, the heterogeneity problem: If autism is more than a mere description of behaviors, then the DSM criteria for ASD should be seen as symptoms of an underlying disorder. However, evidence suggests that the biology underlying autistic behaviors is broadly heterogeneous, which threatens the scientific and practical usefulness of ASD as a diagnostic category.28 It is neither novel nor controversial, of course, to point out the heterogeneity of “ASD.” In fact, “heterogeneity” is often built into descriptions of autism—it is commonly referred to in terms such as “a heterogeneous syndrome” (Abrahams & Geschwind, 2008) or “a heterogeneous set of neurodevelopmental disorders” (Froehlich & Fung, 2012). Such descriptions stem from a large body of empirical work revealing a complex web of potential causes, correlations, and contributing factors underlying autism. Jaarsma and Welin (2012) note that, “in the scientific community, autism is not regarded as a single disease but as a syndrome with multiple nongenetic and genetic causes” (p. 22; see also Lord & Jones, 2012). Geschwind and Levitt (2007), for example, state that “autism has many etiologies and should be considered not as a single disorder 28 I use “biological” here in the broadest possible sense, to include a number of levels of analysis, from genetics to neurochemistry to anatomy to epigenetics. 59 but, rather, as ‘the autisms’” (p. 103). Some have concluded pessimistically from these findings that “it is time to give up on the search for a monolithic cause or explanation” for autism (Happé et al., 2006, p. 1219). Others are holding out hope for a common pathway for autism (see, e.g., Geschwind & State, 2015) or, more modestly, “pathophysiologically and etiologically defined subtypes of ASD” (Geschwind, 2009, p. 367). Even the optimists, though, accept that “the causes of this complex human condition are manifold” (Geschwind & State, 2015, p. 1109). The problem at hand is not heterogeneity as such; as noted in Chapter 2, heterogeneity is ubiquitous in biology and bound to turn up with any attempt to sort biological organisms or phenomena into categories. Moreover, it may make sense to group heterogeneous elements, if doing so furthers a salient practical or scientific aim. The real issue here is the broad heterogeneity seen in “ASD” threatens two foundational goals of psychiatric classification and diagnosis—projectability and utility. Put roughly, projectability is a measure of inferential power and clinical utility is a measure of explanatory power. A highly projectable diagnostic category, then, is one that facilitates reliable predictions about the course of a condition, whereas the category’s clinical utility is commensurate with its usefulness in developing effective medical interventions. Importantly, the projectability and utility of a category tend to decrease as its extension grows broader, since reliable predictions and effective interventions are often based upon the category’s underlying biology.29 In other words, the heterogeneity of a diagnostic category is inversely proportional to its conduciveness to explanation, predication, and control. The most useful categories in medicine tend, therefore, to be based 29 This is one reason for the excitement over “precision medicine.” 60 upon discrete etiologies or biological mechanisms. For this reason, a number of conditions once defined at the functional level are beginning to be defined by lower-level biological features. For example, researchers have recently proposed that the diagnosis of Alzheimer’s disease be made on the basis of biomarkers rather than clinical presentation. Jack et al. (2018) found that as many as “30% of individuals clinically diagnosed as AD dementia by experts do not display AD neuropathologic changes at autopsy” (p. 538). On this new research framework “symptoms become the result of the disease, not the definition of the disease” (Hamilton, 2018), which is precisely what is needed to avoid the triviality problem raised above. A number of writers have convincingly made the point that diagnostic categories based upon biological features are more conducive to projectability and utility. Kendell (1989) states, “it is simply an empirical finding that the most etiologically based classifications are more useful” (p. 46). Boyd (1991) similarly notes that “successful induction and explanation always require that we accommodate our categories to the causal structure of the world” (p. 139). Expanding on Boyd’s work, Tsou (2016) has argued that psychiatry ought to ground its classifications in “networks of biological mechanisms” since these kinds more often “yield successful inductive inferences” (p. 413; see also Kendler, Zachar, & Craver, 2011). The empirical evidence regarding autism’s underlying biology, taken in conjunction with the view that diagnostic categories based upon causal, etiological, or mechanistic features tend to be more useful, suggests that biological heterogeneity threatens the projectability and utility of “ASD.” This threat appears to be borne out in experience. Consider, for example, the fact that parents of a young child receiving an ASD 61 diagnosis may be unsure whether their child will grow up to be a quirky but socially and professionally successful adult, or someone who is nonverbal and unable to live independently. This range of potential outcomes points to a serious, if obvious, shortcoming in our current scheme of classifying autism: there remains no reliable means of preventing or curing autism.30 And despite all the great work being done with applied behavioral analysis and related therapy techniques to mitigate behaviors considered disruptive or dysfunctional, few would argue that the interventions available for autism are sufficiently effective. A defender of the DSM’s approach to classification and diagnosis may simply concede the heterogeneity point while arguing that diagnostic categories like “ASD” should be seen as placeholders for more biologically robust categories, rather than fully developed empirical hypotheses.31 Understood as a stopgap in the face of developing science, it is no wonder that “ASD,” currently conceived, fails to pick out a discrete biological feature. It would be great if autism could be defined in terms of etiology or mechanism, we might think, but the science just is not there yet. While there certainly is much more to learn about autism’s biology, it is important to note that not all cases of ASD are considered to be idiopathic. Conservative estimates hold that “10-15% of individuals with ASD have a known condition that causes the disorder,” where “cause” is understood as sine-qua-non: “most of these [known conditions] are not common; however, they are especially relevant with regard to the prevention of 30 The notion of a “cure” for autism is anathema to those within the neurodiversity movement. I will leave this controversy aside for now, and assume that cure remains a desirable, if currently unattainable, goal for clinicians. See Chapters 4 and 5 for a discussion of this issue. 31 Inspiration for this kind of view can be found in Robins and Guze (1970). 62 ASD” (Balasubramanian et al., 2009). Among the “known etiologies” of autism are Fragile X syndrome, Rett syndrome, Cohen syndrome, Cornelia de Lange syndrome, Angelman syndrome, tuberous sclerosis, rubella, and phenylketonuria (Folstein & Rosen-Sheidley, 2001; Zafeiriou et al., 2007). Diagnosing ASD in two patients with distinct known medical causes seems akin to diagnosing “vomiting disorder” in both a pregnant woman and a child with food poisoning—it may reveal some overlap in their clinical presentation and suggest a basic symptomatic treatment, but does little to address the core medical issue. The point here is that, if “ASD” were a mere placeholder for a condition whose cause is unknown, then cases of known cause ought to be excluded. Yet the only exclusion criterion given in the DSM-5 is that the patient’s social deficits and restricted interests are not “better explained by intellectual disability” (APA, 2013a).32 This suggests a greater concern with the behavioral manifestations of autism than with etiology, which calls into question the notion that the DSM’s classification of autism is merely an unfortunate product of necessity. 3.6 Conclusion Characterizing autism in purely descriptive terms makes it trivially true that all and only those who satisfy the DSM criteria for ASD have autism. However, open questions surrounding certain ASD diagnoses impugn the necessary connection between the diagnostic criteria for ASD and the condition they are meant to identify. Avoiding the triviality problem demands that the behaviors outlined in the diagnostic criteria for ASD 32 The reason for this exclusion criterion is that a patient with severe intellectual disability may lack the mental capacity for social communication and a wide range of interests. 63 be seen as symptoms of an underlying disorder. Evidence suggests, though, that “ASD” is a biologically heterogeneous category, as is made clear by the fact that a number of cases of ASD are considered to be of distinct known causes. The upshot of the dilemma is that ASD as a diagnostic category is either trivial or unhelpful. 3.7 References Abrahams, B. S., & Geschwind, D. H. (2008). Advances in autism genetics: On the threshold of a new neurobiology. Nature Reviews Genetics, 9, 341-355. American Psychiatric Association (APA). (1980). Diagnostic and statistical manual of mental disorders: DSM-III. Washington, D.C: American Psychiatric Association. American Psychiatric Association. (2013a). Diagnostic and statistical manual of mental disorders: DSM-5. Washington, D.C: American Psychiatric Association. American Psychiatric Association. (2013b). DSM-5 Autism Spectrum Disorder fact sheet. Retrieved from http://www.dsm5.org/Documents/Autism%20Spectrum%20Disorder%20Fact%20 Sheet.pdf>. Balasubramanian, B., Bhatt, C. V., & Goyel, N. A. (2009). Genetic studies in children with intellectual disability and Autism Spectrum Disorders. Indian Journal of Human Genetics, 15(3), 103-107. Biklen, D. (2005). Autism and the myth of the person alone. New York, NY: New York University Press. Blumberg, S. J., Zablotsky, B., Avila, R. M., Colpe, L. J., Pringle, B. A., & Kogan, M. D. (2016). Diagnosis lost: Differences between children who had and who currently have an Autism Spectrum Disorder diagnosis. Autism, 20(7), 783-795. Boyd, R. (1991). Realism, anti-foundationalism and the enthusiasm for natural kinds. Philosophical Studies, 61(1/2), 127-148. Brokerick, A. A. (2011). Autism as rhetoric: Exploring watershed rhetorical moments in applied behavior analysis discourse. Disability Studies Quarterly, 31(3).= Carey, B. (2012, January 19). New definition of autism will exclude many, study suggests. New York Times. Retrieved from https://www.nytimes.com/2012/01/20/health/research/newautism-definition 64 would-exclude-many-study-suggests.html. Folstein, S. E. & Rosen-Sheidley, B. (2002). Genetics of autism: Complex aetiology for a heterogenous disorder. Nature Reviews Genetics, 2, 943-955. Finch, D. (2012). The journal of best practices: A memoir of marriage, Asperger syndrome, and one man’s quest to be a better husband. New York, NY: Scribner. Froehlich, W., & Fung, L. K. (2012). Autism spectrum and neurodevelopmental disorders. Psychiatric Times, 29(11). Geschwind, D. H. (2009). Advances in autism. Annual Review of Medicine, 60, 367-380. Geschwind, D. H., & Levitt, P. (2007). Autism spectrum disorders: developmental disconnection syndromes. Current Opinion in Neurobiology, 17, 103-111. Geschwind, D. H., & State, M. W. (2015). Gene hunting in Autism Spectrum Disorder: On the path to precision medicine. Lancet Neurology, 14, 1109-1120. Giles, D. C. (2014). ‘DSM-V is taking away our identity’: The reaction of the online community to the proposed changes in the diagnosis of Asperger’s disorder. Health, 18(2). Gnaulati, E. (2013). Back to normal: Why ordinary childhood behavior is mistaken for ADHD, Bipolar Disorder, and Autism Spectrum Disorder. Boston, MA: Beacon Press. Happé, F., Ronald, R. & Plomin, R. (2006). Time to give up on a single explanation for autism. Nature Neuroscience, 9(10), 1218-1220. Hamilton, J. (2018, April 10). Scientists push plan to change how researchers define Alzheimer’s. National Public Radio. Retrieved from https://www.npr.org/sections/health-shots/2018/04/10/600944750/scientists-pushplan-to-change-how-researchers-define-alzheimers. Horwitz, A. V., & Wakefield, J. C. (2007). The loss of sadness: How psychiatry transformed normal sorrow into depressive disorder. New York, NY: Oxford University Press. Hyman, S. E. (2007). Can neuroscience be integrated into the DSM-IV? Nature Reviews Neuroscience, 8(9), 725-732. Hyman, S. E. (2010). The diagnosis of mental disorders: The problem of reification. The Annual Review of Clinical Psychology, 6, 155-179. Jaarsma, P., & Welin, S. (2012). Autism as a natural human variation: Reflections on 65 the claims of the neurodiversity movement. Health Care Analysis, 20(1), 20-30. Jack Jr, C. R., Bennett, D. A., Blennow, K., Carrillo, M. C., Dunn, B., Haeberlein, S. B., ... & Liu, E. (2018). NIA-AA research framework: Toward a biological definition of Alzheimer’s disease. Alzheimer’s & Dementia, 14, 535-562 Kendell, R. E. (1989). Clinical validity. Psychological Medicine, 19(1), 45-44. Kendler, K. S. (2016). The phenomenology of Major Depressive Disorder and the representativeness and nature of DSM criteria. American Journal of Psychiatry, 173(8), 771-780. Kendler, K. S., Zachar, P., & Craver, C. (2011). What kinds of things are psychiatric disorders? Psychological Medicine, 41, 1143-1150. Lord, C. & Jones, R.M. (2012). Annual research review: Re-thinking the classification of Autism Spectrum Disorders. Journal of Child Psychology and Psychiatry and Allied Disciplines, 53(5), 490-509. Lord, C., Rutter, M., DiLavore, P., Risi, S., Gotham, K., & Bishop, S. (2012). Autism diagnostic observation schedule–2nd edition (ADOS-2). Los Angeles, CA: Western Psychological Corporation. Maenner, M. J., Rice, C. E., Arneson, C. L., Cunniff, C., Schieve, L. A., Carpenter, L. A., ... & Durkin, M. S. (2014). Potential impact of DSM-5 criteria on Autism Spectrum Disorder prevalence estimates. JAMA Psychiatry, March. Murphy, D. (2009). Psychiatry and the concept of disease as pathology. In M. Broome & L. Bortolotti (Eds.). Psychiatry as cognitive Neuroscience: Philosophical perspectives (pp. 103-117). New York, NY: Oxford University Press. Murphy, D. (2015). Philosophy of psychiatry. Stanford Encyclopedia of Philosophy. Retrieved from https://plato.stanford.edu/search/searcher.py?query=autism. Nugent, B. (2012, January 31). I had Asperger syndrome. Briefly. New York Times, January. Regier, D. A., Narrow, W. E., Kuhl, E. A., & Kupfer, D. J. (2009). The conceptual development of DSM-V. American Journal of Psychiatry, 166(6), 645-650. Robins, E., & Guze, S. B. (1970). Establishment of diagnostic validity in psychiatric illness: Its application to schizophrenia. American Journal of Psychiatry, 126(7), 107-111. Rutter, Michael, A. Le Couteur, & C. Lord. (2003). Autism diagnostic interviewrevised. Los Angeles: Western Psychological Services 29: 30. 66 Seltzer, M. M., Krauss, M. W., Shattuck, P. T., Orsmond, G., Swe, A., & Lord, C. (2003). The symptoms of autism spectrum disorders in adolescence and adulthood. Journal of Autism and Developmental Disorders, 33(6), 565-581. Shulman, L., D’Agostino, E., Lee, S., Valicenti-McDermott, M., Seijo, R., Tulloch, E., … Tarshis, N. (2019). When an early diagnosis of Autism Spectrum Disorder resolves, what remains? Journal of Child Neurology, 34(7): 382-386. Steinberg, P. (2012, January 31). Asperger’s history of over-diagnosis. New York Times. Retrieved from https://www.nytimes.com/2012/02/01/opinion/aspergers-historyof-over diagnosis.html. Stubblefield, A. (2013). Knowing other minds: Ethics and autism. In J. L. Anderson & S. Cushing (Eds.). The Philosophy of Autism (pp. 143-166). Lanham, MD: Rowman & Littlefield. Tsou, J. Y. (2011). DSM-5 and psychiatry’s second revolution: Descriptive vs. theoretical approaches to psychiatric classification. In S. Demanzeux and P. Singh (Eds.). The DSM-5 in perspective: Philosophical reflection on the psychiatric babel. Dordrecht: Springer. Tsou, J. Y. (2016). Natural kinds, psychiatric classification, and the history of the DSM. History of Psychiatry, 27(4), 406-424. Verhoeff, B. (2012). What is this thing called autism? A critical analysis of the tenacious search for autism’s essence. Biosocieties, 7(4), 410-432. Zafeiriou, D., Ververi, A., & Vargiami E. (2007). Childhood autism and associated comorbidities. Brain & Development, 29(5), 257-272. CHAPTER 4 THE IS AND OUGHT OF DISABILTY33 4.1 Abstract Much ink has been spilled in the debate over the extent to which disability is a social, rather than a medical, phenomenon. Some theorists understand the disadvantage often associated with disability to be the result of discrimination against individuals with impairments of various kinds; others place a greater emphasis on the impairments themselves. This kind of theorizing is meant to have some amount of normative force: If disability turns out to be (primarily or solely) a social phenomenon, then we should favor social change in response to disability. If, on the other hand, disability is more like a disease, then we should favor medical intervention. It is important to remember, though, that conceptual claims about disability do not entail normative positions. For example, one may reasonably accept the social model of disability while at the same time defending medicalization. If theorizing about disability is to have the intended effect, then, some argument must be made to bridge the “is and ought” of disability. In this chapter, I consider several such arguments, concluding that those appealing to identity and diversity are the most successful. 33 This chapter has been formatted for stand-alone publication. 68 4.2 Introduction Human difference is ubiquitous and multifarious: some people have red hair; some play the piano beautifully; some prefer sex with men; some struggle to do arithmetic in their heads; some are Episcopalian; some dislike cilantro. Certain differences are associated with harm or disadvantage, the source of which often varies. For example, the harm of having a disease may derive from morbidity and mortality, while the harm of belonging to an oppressed social group more likely derives from bias, discrimination, and institutional injustice. Disability theorists have long debated whether the disadvantage often associated with disability is more like that of disease or of belonging to an oppressed social group. Whereas some see disability as a bad-difference, others consider it to be a meredifference.34 Unsurprisingly, proponents of the bad-difference view have tended to favor individualized medical intervention as a response to disability; proponents of the meredifference view, on the other hand, have tended to favor social reform and have opposed the medicalization of disability. This suggests that theorizing about the concept of disability is meant to have some kind of normative impact—namely, arguments about the causes of the disadvantage of disability are supposed to support to certain responses to that disadvantage. It is important to remember, however, that, although views about the nature of disability naturally lend themselves to views about the proper response to disability, one does not entail the other. One could quite reasonably accept a mere-difference view of 34 This terminology comes from Barnes (2014), which I discuss in more detail in the next section. 69 disability, say, while endorsing an individualized medical response. Yet the notion of medicalizing what is seen as a social issue sits uneasily with many disability theorists and activists. The question, then, is what supports the intuition that is claims about the concept of disability should be aligned with ought claims about disability ethics. In this paper, I consider several arguments which might be given to bridge the gap between the “is and ought” of disability. 4.3 Models of Disability Contemporary debates surrounding disability have been informed by the distinction between social and medical models of disability. What has come to be called the social model of disability35 began to take shape in 1970s Britain with a group of disabled activists, the Union of the Physically Impaired Against Segregation, who proposed that disability be understood in sociopolitical terms: In our view, it is society which disables physically impaired people. Disability is something imposed on top of our impairments by the way we are unnecessarily isolated and excluded from society. Disabled people are therefore an oppressed group. (UPIAS, 1976, p. 4) According to the British social model, then, disability is the disadvantage of restriction of activity caused by a contemporary social organization which takes no or little account of people who have physical impairments36 and thus excludes them from participation in the mainstream of social activities. (UPIAS, 1976, p. 14) Here we see two key insights. First is the distinction between impairment—the 35 Mike Oliver (1981) introduced the term “the social model of disability” in The Handicapped Person: A New Perspective for Social Workers. 36 UPIAS eventually expanded their view to include sensory and cognitive impairments, which is now standard for models of disability (Barnes, 2012, p. 14). 70 physical limitations resulting from biological features of individual bodies—and disability—the disadvantage people with impairments face as a result of discrimination and social exclusion.37 Second is the idea that people with disabilities constitute an oppressed group, since they are disabled by unjust social arrangements rather than impairments per se. The British social model thus understands disability as a social phenomenon in the very deepest sense: The point is not merely that disability should be understood relationally, or that the disadvantages of disability are exacerbated by contingent and unjust social arrangements. The point is rather that disability is the oppression and discrimination people with impairments experience, just as racism is the oppression and discrimination experienced by people belonging to disadvantaged racial groups (Morris, 1991). Michael Oliver, an architect and leading proponent of the social model, puts the idea succinctly: “disability is wholly and exclusively social” (1996, p. 35; emphasis added). Over time, the social model has seen a number of iterations, which diverge from the British social model in a variety of ways. Some theorists (Bickenback, 1993; Crow, 1992; French, 1993; Oliver, 1996; Williams, 1999) argue that the British social model is too strong in that it underplays the significance of impairment in the experience and classification of disability. These concerns led to the emergence of the social relational model (Gustavsson et al., 2005; Thomas, 1999; Thomas, 2004), according to which disability results from the interaction of bodies and environments, both natural and social. 37 In the early days of American disability studies, it was more common to use “disability” to refer to physical limitations (“impairment” in the British model) and “handicap” to refer to social disadvantage (“disability”’ in the British model). The British terminology is now the international English standard for disability studies. 71 On this view, the disadvantage of disability involves both impairment and social exclusion, which are intimately linked and often difficult to distinguish in practice (Anastasiou & Kauffman, 2013; Martiny, 2015; Watson, 2002). Others (Cole, 2007; Davis, 2002; Shakespeare, 2006; Tremain, 2002) argue conversely that the model is too weak, in at least one sense: by assuming a value-free, biological notion of impairment, the British social model undersocializes disability, since “the idea of bodily impairment itself must be understood in social, political, and cultural terms” (Cole, 2007, p. 165).38 Amundson, for example, holds that impairment designations rely upon “implicit social judgments about the acceptability of certain kinds of biological variation,” rather than objective biomedical standards (2000, p. 33). Other notable versions of the social model include the minority group model (Hahn, 1985), which, like the British social model, emphasizes discrimination and exclusion as the primary sources of disability’s harm and advocates a civil rights approach to addressing such harm; and the human variation model, which understands disability as “an extension of the variability in physical and mental attributes beyond the present—but not the potential—ability of social institutions to roundly respond” (Scotch & Schriner, 1997, p. 155). What these various models have in common is their opposition to the medical model of disability,39 which characterizes disability as biologically based, individually located, and inevitably harmful, thus seeking “to reduce the complex problems of disabled people 38 Shakespeare notes that this move mirrors Butler’s (1993) criticism of the sexgender distinction. 39 The medical model is also known as the individual or functional limitation model. 72 to issues of medical prevention, cure or rehabilitation” (Shakespeare, 2006). Though rarely defended in the disability literature, the medical model “is often adopted unreflectively by health care professionals, bioethicists, and philosophers who ignore or underestimate the contribution of social and other environmental factors to the limitations faced by people with disabilities” (Wasserman et al., 2016). Despite the social model’s predominance among disability theorists and activists, there remains some disagreement about its scope. Some regard it, more modestly, as a political strategy for disability advocacy; others consider the social model to be a way of understanding the concept of disability itself. This disagreement arises from the fact that discussions about the social model involve what I see as three distinct, but easily confused, controversies. First is a theoretical disagreement over the valence of disability, or how best to understand the relationship between disability and well-being. This is how Barnes characterizes the debate when she distinguishes mere-difference views of disability and bad-difference views of disability. Mere-difference views, often endorsed within social models, hold that “disability is not, by itself, something bad, harmful, or suboptimal” (2014, p. 89). Bad-difference views, closely aligned with the medical model, maintain that, “not only is having a disability bad for you, having a disability would still be bad for you even if society was fully accommodating of disabled people” (Barnes, 2014, p. 89).40 Accordingly, on this view, the harm of disability amounts to the harm of the underlying impairment. 40 Barnes sees some version of the bad-difference view as “the dominant characterization of disability in analytic philosophy” (2009, p. 337) 73 Second is an empirical disagreement over the cause of disability’s disadvantage (see, for example, Oliver, 1990, p. 11). Even those who adopt disability-positive positions acknowledge that disabled people often face significant disadvantages. Strong versions of the social model maintain that the cause of such disadvantage is primarily or solely social. Impairment may be a precondition for disability—since disabled people are oppressed and excluded on the basis of their impairment—but the impairment does not cause the disability any more than certain racialized characteristics cause racial discrimination. Strong versions of the medical model, on the other hand, see the disadvantage facing disabled people as a direct result of the inherent and necessary harm of impairment. Moderate views of disability, such as the social relational model, understand the disadvantage of disability as involving a complex web of causes, both biological (e.g., functional impairment, pain, and chronic illness) and social (e.g., oppression, discrimination, and exclusion). Third is a normative disagreement over the proper response to disability.41 One way to address the disadvantage of disability is to modify individual bodies so that they may navigate existing natural and social environments; another is to make such environments navigable by a wider range of bodies. Which approach is to be preferred, or to what extent each is to be emphasized, is a point of contention between medical and social models of disability. Indeed, some see the normative component of the social model as more central to the view than its theoretical or ontological positions. It is, Cole states, “at root, a political model, and at its centre is a conception of active citizenship, of the dimensions of society 41 There is a question as to what, exactly, “appropriate” means here. Is this meant as a moral or prudential claim? To whom is its prescription meant to apply? Policy makers? Researchers? Clinicians? Insurers? Disabled individuals? This is an interesting and important issue, but I will put it aside for now. 74 in which all members are entitled to participate to the fullest extent” (Cole, 2007, p. 174). Debates over models of disability, then, involve both descriptive and prescriptive disagreements. However, there is a pervasive assumption in the literature that settling the descriptive disagreements will lead straightforwardly to prescriptive conclusions. The thought seems to be that if we can show that disability is a mere-difference whose harm is caused by contingent and unjust social arrangements, say, then it will follow that social responses are to be preferred to medical ones. In the next section, I highlight this assumption by canvassing disability’s conceptual space, noting that, although certain models of disability naturally lend support to certain ethical and political responses to disability, there is a gap between is claims about disability (e.g., disability is primarily a social construct) and ought claims (e.g., people with disabilities are owed such-and-such accommodation).42 It is important that we mind the is-ought gap since, presumably, one of the central aims of theorizing about disability is to inform our views about normative considerations such as how disabled persons should be treated and what they are owed. 4.4 The Is-Ought Gap The social and medical models of disability, in their various forms, seek to characterize the interaction between disability, impairment, and well-being. As we have seen, social models tend to emphasize the causal role of exclusion and discrimination, which result from contingent social arrangements, in disadvantaging persons with 42 Though I focus most of my attention on the social model, it is worth noting explicitly that the is-ought gap is equally present in the medical model: Even if the medical model is correct that the harm of some disability is solely or primarily biological, it does not follow that medical intervention is to be the preferred course of action. 75 impairments. Medical models, on the other hand, emphasize the harm of impairment itself, which results from biological features of individuals. In what follows, I address the assumption that disability ethics will fall out of this debate. Wasserman et al. note that, although “the different models of disability…seem to favor different responses to disability…claims about the causes of disadvantage do not always yield straightforward prescriptions for their remediation” (Wasserman et al., 2016). The fact that disability is a natural phenomenon, or a social one, or a result of the interaction between the two, entails nothing about what ought to be done about it. Samaha (2007) puts the point nicely: Despite the apparent connection between the social model and social change, there just is no necessary relationship there…One can accept the model’s insight regarding causes of disadvantage without committing to a particular response, even if one believes that disability is simply or importantly the result of people’s attitudes…Scholars should stop moving so quickly from assertions about social construction to arguments for social reconstruction. (pp. 1252-1253) Given that there is no necessary link between is claims and ought claims about disability, let us now canvass the conceptual space in which they interact. Table 1 maps the four views I will discuss. What I am calling the treatment view stems from a standard interpretation of the medical model of disability. It holds that the primary cause of the harm or disadvantage associated with disability is biological and that the appropriate response to such harm is medical intervention. Table 1 The Conceptual Space Medical Phenomenon Social Phenomenon Medical Intervention Treatment View Enhancement View Social Change Accommodation View Diversity View 76 While medical intervention seems a natural response to biologically caused harm, consider an alternative, which we might call the accommodation view. The accommodation view agrees with the treatment view that the harm of a disability is caused (solely, primarily, or importantly) by biological factors, but insist that social change is the proper response to such harm. Take Down syndrome, for example. Though it is uncontroversial to treat its various co-morbidities (e.g., congenital heart disease) with medical intervention, there has been significant resistance to using noninvasive prenatal testing (NIPT) to eradicate Down syndrome altogether (Belkin, 2010; Downpride, 2015). Presumably, a medical cure (or “cure”) for Down syndrome, if and when developed, would meet similar (or stronger) resistance.43 Important for the purposes of this paper is that one could quite reasonably resist calls for prevention and cure of Down syndrome while also holding that its harms or disadvantages are caused by a third copy of chromosome 21. Despite the purely biological cause of Down syndrome, one might think that people with Down syndrome should be celebrated and accommodated rather than fundamentally changed. The diversity view, related to a traditional understanding of the social model of disability, holds that disability is socially caused and ought to be addressed via social change. In its strongest form, the diversity view denies any causal role to impairment, and insists that (certain) disabled people are disadvantaged in precisely the same way as members of other socially disadvantaged groups. Defenses of strong versions of the 43 I say “stronger” because in addition to the eradication concern would be a concern about changing foundational features of persons with established social identities. See Chapter 5 for discussion of this issue. 77 diversity view can be found within Deaf culture and the neurodiversity movement (Ortega, 2009).44 More moderate positions acknowledge the role of impairment, but emphasize the social aspects of disability. An alternative to the diversity view is found in what I will call the enhancement view, which acknowledges the social nature of certain problems or sources of disadvantage but nonetheless advocates medical interventionist solutions. Charges of “medicalizing” disability are typically aimed at this kind of view. Defenses of it can be found in the bioethics literature on psychopharmaceuticals. Kramer (1993), for example, defends the practice of “cosmetic pharmacology”—prescribing antidepressants to patients who may be suffering from social alienation rather than pathological depression. More recently, Ray has argued for the moral permissibility of “using stimulants as cognitive-enhancing drugs to solve the social problem of poorly educated students in inadequate schools” (2016, p. 29). Such views can be defended on a number of grounds. Most compelling, perhaps, is that medical intervention is the best we can do in the face of the stubborn intractability of certain social problems. The purpose of outlining the above positions is to note the complexity surrounding the conceptual, empirical/ontological, and normative claims made by various models of disability. Such complexity should lead us to question the assumption that a given model’s normative conclusions will follow from its positions on the cause or valence of disability. In the next section, I propose and evaluate several arguments which might be given to bridge the is-ought gap seen within models of disability, focusing on the social model. 44 See Chapters 4 and 5 for discussion of the neurodiversity movement. 78 4.5 Bridging the Gap Thus far, I have argued that there is no entailment relation between the disadvantage of disability being socially caused and the claim that disability ought to be addressed through social change. There is, however, a strong intuitive pull to the idea that social problems warrant social solutions, whereas medical problems warrant medical solutions. What can be said in support of this intuition? In addressing this question, I attempt to avoid fundamental disagreements over the degree to which the disadvantage of disability is socially caused by considering the case of homosexuality. The analogy between disability and homosexuality, of course, is a controversial one.45 But I will adopt it here following the influential work of Barnes (2014). The thought behind the analogy is that, while homosexuality involves disadvantages (e.g., housing and employment discrimination), it is a mere-difference rather than a baddifference.46 Furthermore, most of us would agree that responding to the disadvantage of homosexuality through medical intervention on individual gay people—were such interventions effective47—would be (at least potentially) morally problematic. If we think there is good reason to modify society, rather than individuals, in the case of homosexuality, perhaps the same can be said of disability, or at least of certain disabilities. 45 Less controversial, though still fraught, analogies might appeal to sex or race. Some naturalistic accounts of disease maintain that homosexuality is in fact a disease insofar as it reduces evolutionary fitness (though this feature of fitness-based disease concepts is often viewed as a reductio ad absurdum). I will put this issue aside and assume that any harm homosexuals experience is social rather than biological. Readers who do not share this assumption are encouraged to substitute the example with one they find more agreeable, e.g., the disadvantage of being identified as belonging to a certain racial category. See Chapter 1 for discussion of disease concepts. 47 Though not effective, conversion therapy is alive and well. The state of Utah, for example, was set to ban conversion therapy in its 2019 legislative session, but ended up with a diluted piece of legislation that protected “therapists.” 46 79 In theory, there are at least two ways to mitigate the disadvantages of homosexuality. One is to change homophobic attitudes and institute public policies that discourage discrimination against gay people. Another is to do away with homosexuality itself by changing the desires and behaviors of individual gay people through medical intervention. Were both strategies practicable and available to us, the former clearly seems preferable. But why? Let us consider some possibilities. • Efficacy Argument: One obvious concern is that “treating” homosexuality would be either ineffective or dangerous. The cause of disadvantage is relevant insofar as it indicates the response most likely to produce the desired result. A social approach to mitigating the disadvantages of homosexuality should therefore be preferred to a medical approach on the grounds of efficacy. This is certainly a reasonable argument, and one that applies to disabilities for which no safe and effective medical treatments exist. Were efficacy the sole concern, however, then the view that social change is preferable to medical intervention would be merely contingent on the state of medical technology. Yet I take the opposition to medical “treatment” for homosexuality to go deeper than this: even if a perfectly safe and effective treatment were available, social change would be preferable to medical intervention. So, again, why? • Compensatory Justice Argument: Since the disadvantages of homosexuality result from social injustice, society should bear the cost of reparation. Favoring medical intervention shifts the burden from society to individual gay people. Therefore, a social approach to mitigating the disadvantages of homosexuality should be preferred to a medical approach on the grounds of justice. 80 This argument appeals to something like Aristotle’s notion of corrective justice or Nozick’s rectification principle, though the case of homosexuality is concerned with reparation for harm rather than unjustly acquired resources. Important for our purposes is the idea that the causal story about disadvantage is relevant to the normative conclusion insofar as it identifies the party responsible for rectifying past (and preventing future) disadvantage. The problem with the justice argument is underdetermination: it establishes only that society has an obligation to address the disadvantage it causes, not the means of fulfilling the obligation. One could then grant the social causes of the disadvantages facing gay people while maintaining that medical intervention is the best solution, so long as it is free, widely available, quick, painless, and so on. Ensuring access to medical treatment that eliminates x is arguably one way for society to fulfill its obligations to those it has disadvantaged on the basis of x. Consider an analogous case. Capitalist economies disadvantage those with little capital or means of acquiring it. While dismantling capitalism is one way to address such economic disadvantage, another is to provide economic resources—as well as providing education, job training, and other means of securing equality of opportunity—to the disadvantaged. This latter approach embraces an individual solution to a social-structural problem, but is an available (and perhaps, in some cases, preferable) solution nonetheless.48 It appears, therefore, that while the social nature of homosexuality’s disadvantage establishes a social responsibility to address the disadvantage, it does not, per se, favor a 48 In Chapter 5, I make the point that preventing autism (and other disabilities) may be preferable, if regrettable, in certain nonideal social circumstances. 81 policy or political approach over a medical one. Yet the intuition that medically intervening upon homosexuality is morally problematic persists. This suggests that the issue goes deeper still. Consider a third argument: • Suspect Norms Argument: “Treating” homosexuality medically implies that the gay individual, rather than society, is in need of change. This supports the suspect norm that gay people are somehow defective, which in turn promotes stigma and injustice. This argument takes its name from Little, who writes that “we should have special, and deep, concerns about medicine participating in practices that reflect and reinforce certain suspect norms” which “ends up contributing to a broad and unjust system of constraining pressures and forces” (1998, p. 176). Though made in the context of cosmetic surgery and norms of appearance, it is easy to see how Little’s argument applies to the case of homosexuality: medicalizing homosexuality reinforces the very attitudes and policies that makes homosexuality disadvantageous to begin with. A closely related argument, known as the expressivist objection, is levied against the practice of preventing, or “selecting against,” disability on the grounds that doing so expresses disparaging and hurtful views about existing disabled people. Alternatively, we may think of this as an argument from corrective justice, which would hold that the best way to correct for injustice is to change wrongful attitudes, rather than to change people to fit those wrongful attitudes. This style of argument gets closer to explaining the intuition that medically intervening upon homosexuality is morally problematic. It is interesting to note, though, that some have favored medicalization as a way of cutting against suspect norms and 82 stigma. Disease models of addiction and obesity, for example, are often presented as a way of pushing back against the attitude that such conditions are character flaws. Rethink Mental Illness, an English mental health charity, likewise stresses the idea that “mental health problems are illnesses akin to physical illnesses, in the sense that they occur for reasons outside a person’s control (which arguably reduces blame), can be treated, are not to be feared, and are not a sign of weakness” (Rashed, 2018, p. 14). Unfortunately, while medicalization may have been effective as a challenge to antiquated notions that people with mental illness are cursed or morally deviant, empirical work suggests that “treating mental illness as an illness like any other” does not actually reduce the stigma. In fact, for conditions like schizophrenia, biomedical explanations of the condition been shown to be “positively related to perceptions of dangerousness and unpredictability, and to fear and desire for social distance” (Read et al., 2006, p. 303). Beyond that, medicalization may simply be offensive, as the case of homosexuality— which was classified as a mental disorder by the American Psychiatric Association until the DSM-III (APA, 1980)—clearly shows. Note, however, that advocates of a medical approach to homosexuality need not claim that gay people are defective; they need claim only that there is a poor fit between gay people and their social environment and that a medical solution is simply the most effective, expedient, or cost-effective means of addressing the discrepancy.49 Whether or not this turns out to be the case is an empirical issue, and one that seems to miss the moral core of the judgment that there is something wrong with medical approaches to addressing 49 See Chapter 5 for a related discussion of the expressivist objection, in which I argue that selecting against disability may express more about social structures than disability per se. 83 homosexuality. Resistance to medialization goes deeper than a concern about what kinds of interventions are likely to work; it is rather a concern about what kinds of interventions are permissible. The personal identity argument seems to do better on this count: • Personal Identity Argument: Ridding people of homosexuality with medical intervention harms them by eliminating an important part of their identity. Social change addresses the disadvantage of homosexuality while avoiding this harm. Social change should therefore be preferred to medical intervention. The idea here is that, if we can address the disadvantage of homosexuality without thereby threatening the identities of individual gay people, then we ought to on the grounds of avoiding unnecessary harm. Given the social cause of the disadvantage facing gay people, social change exists as an alternative that avoids the harm in question. One shortcoming of this argument is that it does not rule out the (theoretical) practice of prenatal screening or gene editing against homosexuality, which is (at least potentially) morally problematic despite the fact that prevention is, by definition, antecedent to identity.50 Alternatively, the following argument applies to both treatment and prevention: • Group Identity Argument: Ridding people of homosexuality with medical intervention, or preventing homosexuality through preimplantation genetic screening (and related technologies), threatens the identity of the homosexual community. Social change addresses the disadvantage of homosexuality while preserving this valuable group identity. Social change should therefore be 50 A way around this shortcoming is to note that practices which intentionally reduce the number of gay people—whether through medical intervention or genetic selection—express contempt for an important aspect of the identities of those who are gay at present. This response, though, seems just to amount to the suspect norms argument. See Chapter 5 for a discussion of issues related to prevention. 84 preferred to medical intervention and genetic selection. This argument seems to get closer to the mark, but it raises several questions: What constitutes a group identity? Which, if any, disability communities possess this kind of group identity? What gives group identity moral significance? Answers to these questions will provide a potential grounding for the normative side of the social model of disability. The group identity argument is closely related to the final argument I will consider: • Diversity Argument: We, qua society or moral community, would be losing something of value by doing away with homosexuality. Social change addresses the disadvantage of homosexuality while preserving this valuable diversity. Social change should be therefore be preferred to medical intervention. There are a couple of things to note here. First is that it may not apply cleanly to all cases of disability. Not all diversity is equally valuable, after all. Some forms of disability may be quite valuable in this respect, while others are less valuable, or not valuable at all. Second is a worry about this kind of argument: To the extent that it localizes the impetus for tolerance in the advantages to society, we might think that the diversity argument gets things backwards, by failing to appreciate the moral value of the individual. That is, it incorrectly assumes that the individual is in some sense responsible for justifying their differences by proving how others—namely, “we qua society or moral community”— benefit from them. Third is that we must be careful about promoting social value over the choices of individuals. Even granting that something like homosexuality or a particular disability contributes valuable diversity to our society, we may be hesitant to claim that medical 85 intervention is impermissible when requested by a gay or disabled individual. In other words, we should be hesitant to foist the burdens of gradual social change on the shoulders of individuals who face immediate disadvantages. Keeping respect for autonomy in mind, the diversity argument may function only to point out the benefits of favoring social change over medical intervention, should individuals choose to forgo medical intervention. I discuss this issue in greater detail in Chapters 5 and 6. This consideration becomes particularly relevant in the face of misinformation about disability in the context of prenatal screening. While it may well be the prerogative of a pregnant woman to terminate a pregnancy on the grounds that the fetus has screened positive for some kind of disability—since it may well be the prerogative of a pregnant woman to terminate a pregnancy for any reason whatsoever—such a significant decision should not be made on the basis of grim prognostics made by clinicians operating under pernicious ableist assumptions. The task, then, is to determine what kind of diversity is valuable, on what grounds, and which disabilities qualify. 4.6 Conclusion The claim that disability is a mere-difference, or that the disadvantages associated with disability are purely social, entails nothing about the proper response to disability. Yet there is strong intuitive pull to the notion that we ought not medicalize social ills. I have argued that the most promising explanation for this intuition involves considerations of identity and diversity, rather than considerations of efficacy, justice, or suspect norms. However, I have not attempted here to establish that disability is indeed valuable on the grounds of identity or diversity. This, I submit, is the project disability advocates ought to 86 pursue, with the aim of making disability theory relevant to disability ethics. 4.7 References American Psychiatric Association (APA). (1980). Diagnostic and statistical manual of mental disorders: DSM-III. Washington, D.C: American Psychiatric Association. Amundson, R. (1992). Disability, handicap, and the environment. Journal of Social Philosophy 23, 105-118. Amundson, R. (2000). Against normal function. Studies in History, Philosophy, Biology and Biomedical Science 31(1), 33–53. Anastasiou, D. & Kauffman, J. M. (2013). The social model of disability: Dichotomy between impairment and disability. Journal of Medicine and Philosophy 38(4), 441–459. Barnes, C. (2012). Understanding the social model of disability: Past, present and future. In N. Watson et. al. (Eds.), Routledge handbook of disability studies, 12-29. New York, NY: Taylor and Francis. Barnes, E. (2009). Disability, minority, and difference. Journal of Applied Philosophy 26(4), 337-355. Barnes, E. (2014). Valuing disability, causing disability. Ethics 125(1), 88-113. Belkin, L. (2010). Should Down syndrome be cured? New York Times: January 11. Bickenbach, J. (1993). Physical disability and social policy. Toronto: University of Toronto Press. Butler, Judith. (1993). Bodies that matter: On the discursive limits of “sex". New York, NY: Routledge. Cole, P. (2007). The body politic: Theorising disability and impairment. Journal of Applied Philosophy 24(2), 169–176. Crow, L. (1992). Renewing the social model of disability. Coalition, July, 5–9. Davis, L. J. (2002). Bending over backwards: Disability, dismodernism & other difficult positions. New York, NY: New York University Press. Downpride. (2015). Appeal to the United Nations to stop discriminatory use of prenatal genetic screening aimed at eradication of people with Down syndrome and other 87 groups. Published online: https://downpride.com/appeal-to-the-united-nations/. French, S. (1993). Disability, impairment or something in between. In J. Swain, S. French, C. Barnes & C. Thomas (Eds.) Disabling barriers, enabling environments (pp. 17-25). London, UK: Sage. Gustavsson, A., Sandvin, J., Traustadóttir, R., & Tossebrø, J. (2005). Resistance, reflection and change: Nordic disability research. Lund, Sweden: Studentlitteratur. Hahn, H. (1985). Toward a politics of disability: Definitions, disciplines, and policies. Social Science Journal 4(93), 87-195. Kramer, P. D. (1993). Listening to prozac: A psychiatrist explores antidepressant drugs and the remaking of the self. London, UK: Penguin. Little, M. O. (1998). Cosmetic surgery, suspect norms, and the ethics of complicity. In E. Parens (ed.) Enhancing human traits: Ethical and social implications (pp. 162176). Washington, D.C.: Georgetown University Press. Martiny, K. M. (2015). How to develop a phenomenological model of disability. Medicine, Health Care and Philosophy 18(4), 553–565. Morris, J. (1991). Pride against prejudice. London, UK: Women’s Press. Oliver, M. (1981). A new model of the social work role in relation to disability. In J. Campling (Ed.) The handicapped person: A new perspective for social workers, (pp. 19-32). London: RADAR. Oliver, M. (1996). Understanding disability: From theory to practice. Basingstoke, UK: Macmillan. Ortega, F. (2009). The cerebral subject and the challenge of neurodiversity. BioSocieties 4, 425-445. Rashed, M. A. (2018). In defense of madness. Journal of Medicine and Philosophy. Open Access Online. Ray, K. S. (2016). Not just “study drugs” for the rich: Stimulants as moral tools for creating opportunities for socially disadvantaged students. The American Journal of Bioethics 16(6), 29-38. Read, J., N. Haslam, L. Sayce, & E. Davies. (2006). Prejudice and schizophrenia: A review of the ‘mental illness is an illness like any other’ approach. Acta Psychiatrica Scandinavia 114(5), 303–18. 88 Samaha, A. M. (2007). What good is the social model of disability? The University of Chicago Law Review 74(4), 1251-1308. Scotch, R. K., & K. Schriner. (1997). Disability as human variation: Implications for policy. The Annals of the American Academy of Political and Social Science 549(1), 148–159. Shakespeare, T. (2006a). Disability rights and wrongs. London, UK: Routledge. Shakespeare, T. (2006b). The social model of disability. In L. J. Davis (ed.) The disability studies reader (pp. 197-204). New York, NY: Routledge, Taylor and Francis. Thomas, C. (1999). Female forms. Buckingham, UK: Open University Press. Thomas, C. (2004). Developing the social relational in the social model of disability: A theoretical agenda. In C. Barnes & G. Mercer (Eds.) Implementing the social model of disability: Theory and research. Leeds, UK: The Disability Press. Tremain, S. (2002). On the subject of impairment. In M. Corker & T. Shakespeare (Eds.) Disability/postmodernity: Embodying disability theory (pp. 32–47). London, UK: Continuum. UPIAS. (1976). Fundamental principles of disability. London, UK: Union of the Physically Impaired Against Segregation. Wasserman, D., Asch, A., Blustein, J., & Putman, D. (2016). Disability: Definitions, models, experience. Stanford Encyclopedia of Philosophy. Watson, N. (2002). Well, I know this is going to sound very strange to you, but I don’t see myself as a disabled person: Identity and disability. Disability and Society 17(5), 509–528. Williams, S. J. (1999). Is anybody there? Critical realism, chronic Illness, and the disability debate. Sociology of Health and Illness 21(6), 797–819. CHAPTER 5 A CRITIQUE OF NEURODIVERSITY51 5.1 Abstract The neurodiversity view makes both a conceptual claim about what kind of thing autism is and a political claim about how it ought to be addressed. Conceptually, the neurodiversity view holds that certain neurological differences currently classified as disorders—autism, most notably—are best understood as forms of diversity. Politically, it holds that, rather than being medicalized and “treated,” neurodiversity ought to be respected in the way other human differences—such as differences in race and sexual orientation—are respected. In this chapter, I challenge the arguments given in support of neurodiversity’s conceptual claim, while defending its political aims of respect, inclusion, and accommodation. Although neurodiversity provides a compelling challenge to the status quo, I conclude that it falls short of establishing that autism is not a disorder. Fortunately, given the is-ought gap discussed in Chapter 4, the fact that autism is a disorder does not entail that medicalization is the only course. My view is ultimately that social change and medical intervention need not be considered mutually exclusive. 51 This chapter has been formatted for stand-alone publication. 90 5.2 The Neurodiversity Movement The neurodiversity movement began to take shape online in the early 1990s among a group of autistic self-advocates. Responding to the pro-cure and prevention rhetoric of some parent and medical groups,52 members of the community campaigned for the rights and acceptance of autistic people on the grounds that atypical ways of thinking, communicating, and behaving are not necessarily disordered (Ortega, 2009). This general idea came to be referred to as “neurodiversity,” a term credited to the autistic sociologist Judy Singer (1999).53 Singer proposed that neurodiversity be understood as just another form of human difference, arguing that “the ‘neurologically different’ represent a new addition to the familiar political categories of class/gender/race” (p. 91). Though it began with a group of “high-functioning”54 autistics, the neurodiversity movement has expanded to encompass the broader Autism Spectrum, in addition to a number of other neurological differences, including attention deficit hyperactivity disorder (ADHD), bipolar disorder, schizophrenia, developmental dyspraxia, dyslexia, epilepsy, psychopathy, sociopathy, Tourette’s syndrome, oppositional defiant disorder, dyscalculia, paracusia,55 and the umbrella category of “madness”56 (Anton, 2013; Fenton & Krahn, 52 Such as the AUTISM List and the Cure Autism Now Foundation. Around the same time, the term appeared in Jane Meyerding’s “Thoughts on Finding Myself Differently Brained." 54 I use scare quotes here to note the pervasive misconception that autism falls on a simple high-to-low functioning continuum. Functioning is complex and multifaceted, and people may be high-functioning in one regard without thereby being high-functioning in another, and vice versa. Hacking (2009) argues that referring to autism as a spectrum is problematic, since “spectra are linear and autism isn’t.” He continues: “Spectrum is a metaphor from optics; if we are to use a metaphor from the sciences, I would prefer to speak of an autistic manifold.” 55 See the Hearing Voices Movement. 56 See Mad Pride. 53 91 2007; Mackenzie & Watts, 2011; Rashed, 2018).57 As the movement grows, neurodiversity has begun to attract attention in the disability literature, with a groundswell of support from disability theorists wishing to push back against what they see as the medicalization of social ills, including stigma, discrimination, and exclusion (see Disability Studies Quarterly’s special issue “Autism and the Concept of Neurodiversity”). Generally speaking, the neurodiversity view makes both conceptual and political commitments. Conceptually, the view holds that “being neurodiverse or neurotypical are just different ways of existing as humans” (Jaarsma & Welin, 2012); and that neurodiversity involves merely “different ways of socializing, communicating and sensing that may not necessarily be disadvantageous” (Ortega, 2009). Wolbring (2007) provides a similar, if broader, characterization, stating “neurodiversity is defined as the whole of human mental or psychological neurological structures or behaviors, seen as not necessarily problematic, but as alternate, acceptable forms of human biology” (p. 17). Politically, the neurodiversity view is committed to the idea that neurodiversity “should be respected and supported, even celebrated, rather than eliminated” (Dinishak, 2016, p. 12). Indeed, neurodiversity advocates claim that curing autism would be “a par with curing a gay, black, [or] left-handed…individual: if the one were legitimate, so would be the others. But none are pathological conditions, only ways of being” (Ortega, 2009, p. 426). In this sense, the neurodiversity movement takes the form of a “struggle for the civil rights of all those diagnosed with neurological or neurodevelopmental disorders” (Jaarsma 57 This list is neither exhaustive nor uncontroversial, and thus should not be taken as definitive. As I note later, there is disagreement within the movement over who belongs. 92 & Welin, 2012). In Harmon’s (2004) profile of an “Aspie”58 school in Catskills, one of the students sums up the view nicely: “We don't have a disease," said Jack, echoing the opinion of the other 15 boys…“So we can't be ‘cured.’” (Harmon, 2004, p. 4). The claims of the neurodiversity movement are worth taking seriously for a number of reasons, not least of which is medicine’s troubling history of pathologizing difference, particularly in the field of psychiatry. Well-known examples of socially “undesirable” characteristics or behaviors classified as medical disorders include drapetomania (which caused enslaved persons to flee captivity), masturbatory insanity (which caused, well, masturbation), and homosexuality (which was considered a mental disorder by the American Psychiatric Association until 1973). More recently, critics have argued that mental disorder is attributed far too broadly, from those grieving the deaths of loved ones with depression (who may be diagnosed with depression; Horwitz, 2015) to children with late birthdays in early elementary classrooms (who are more frequently diagnosed with ADHD; Layton et al., 2018). These concerns are enough to motivate the project of neurodiversity, which seeks to challenge this tradition of medicalization. Before moving onto the arguments, however, I would like to note two areas of ambiguity that raise concerns for the view. First, despite the laundry list of conditions associated with the neurodiversity movement, proponents tend not to specify what, exactly, falls under the neurodiversity banner. This is unsurprising, given that the whole ethos behind the movement is one of inclusion and 58 “Aspie” is a common abbreviation for “Asperger’s syndrome.” which was the diagnostic category for “high-functioning” autistics in the DSM-IV. The category was subsumed by “Autism Spectrum Disorder” in the DSM-5, to the exasperation of many selfdescribed Aspies (see Giles, 2014). 93 acceptance. But this ambiguity poses a potential dilemma: If the claim is that there are no cases of neurodisorder, then the position faces what seem to be clear counter-examples— namely, those with severely impairing neurological differences. Beyond appearing straightforwardly implausible, this position risks ignoring the experiences of individuals with autism, and their families, who have found their neurological differences to be significant challenges they would rather live without. If, on the other hand, the claim is that there is some distinction between neurodiversity and neurodisorder (just not the distinction drawn by, say, the DSM-5), then the view needs to come up with a way of drawing this distinction that is not subject to its own criticisms.59 One common strategy for distinguishing between neurodiversity and neurodisorder, in the context of autism, is to distinguish between high- and low-functioning autistics (Jaarsma & Welin, 2012), a distinction typically made by identifying those with and without comorbid intellectual disability.60 However, this kind of distinction appears to run afoul of the very arguments given in support of neurodiversity. For one thing, determinations of intellectual disability are arguably arbitrary (for those that appeal to a threshold of a 70 IQ, for example) and insufficiently pluralistic (IQ tests are often denounced for being biased and neglectful of alternative forms of intelligence; see Gould’s The Mismeasure of Man), two criticisms commonly levelled against the suggestion that autism is a disorder. These worries aside, there is a further question about how neurodiversity can reject the idea that (say) impaired 59 A further concern is that if neurodiversity is simply challenging the disorder concept employed in contemporary psychiatry, then it loses novelty, since this debate has been playing out in the philosophy of medicine literature for decades. 60 Approximately 30% of those diagnosed with ASD have no intellectual disability (see Matson & Shoemaker 2009). 94 social-emotional reciprocity makes one disordered while endorsing the idea that impaired intellectual capacity makes one disordered. The risk is that the view will end up hoisted by its own petard. A second ambiguity is that it is unclear how strong neurodiversity’s ethico-political claim about medical intervention is meant to be. One way of interpreting the view is as claiming that it is impermissible to address neurodiversity with medical intervention,61 just as we tend to think it is impermissible to intervene medically upon one’s, or one’s dependent’s, racial identity or sexual orientation (supposing that such interventions were available and effective).62 An alternative interpretation would take the claim to be that medical intervention ought not be the default, expected, or mandated course of action. For example, whereas we might think that parents who fail to consent to treatment for their child’s treatable cancer are doing something morally objectionable, and ought to be subject to coercive state action, neurodiversity would maintain that parents should be given a choice as to whether or not to pursue medical treatment for their child’s neurodiverse condition. The second interpretation, of course, is much weaker than the first; and the plausibility of the neurodiversity view will depend in part on which interpretation it espouses. With these issues in mind, let us move onto the arguments meant to cast doubt on the distinction between neurodiversity and neurodisorder. For the sake of simplicity, I will now refer to “neurodiversity,” assuming that the neurodiversity view will be able to 61 Medical intervention may include both treatment and prevention, an issue I take up in Chapter 5. 62 See the expressionist objection, suspect norms argument, and genocide arguments (Ortega, 2009, pp. 428-429). See Chapter 5 for a discussion of these and related arguments. 95 distinguish between conditions it wishes to defend as mere-differences, and those it regards as disordered. 5.3 Arguments and Critique In this section, I consider four arguments for the claim that neurological difference should be understood as diversity rather than disorder: (1) The Argument from Natural Variation, (2) The Argument from Positive Attributes, (3) The Argument from Normativity, and (4) The Argument from Social Cause. 5.3.1 The Argument From Natural Variation The Argument from Natural Variation holds that neurological differences are a natural and ubiquitous form of biological variation, and there is no principled means of distinguishing differences that are abnormal or disordered from those that are merely atypical. Among proponents of this kind of view is Baron-Cohen (2015), who, referring to autism, states If in the population people simply vary in terms of the number of autistic traits they have, in what sense can we call someone “normal” and someone else “abnormal”? The very notion collapses as people without a diagnosis differ from each other as much as people with a diagnosis. (p. 1760) The argument expressed here notes that neurodiversity traits fall on a continuum, with people manifesting them in varying numbers and to various degrees (see also Amundson, 2000; Armstrong, 2015; Broderick, 2008; Fenton & Krahn, 2007). In light of this variation, one could argue, it is arbitrary and inaccurate to designate some people on the continuum disordered and others not. This being so, neurodiversity ought to be considered a meredifference rather than a disorder. An alternative formulation of this argument appeals to 96 natural kinds, which are discrete groupings found in nature: Since neurodiversity traits fall on a continuum, with people manifesting them in varying numbers and to various degrees, neurodiversity does not involve natural kinds. And if we think that disorders are best understood as natural kinds, we should consider neurodiversity to be a mere-difference rather than a disorder. I certainly do not wish to deny the factual claim that variation is ubiquitous in human biology; what I wish to deny is that the fact of natural variation is sufficient grounds for throwing out the distinction between difference and disorder, even if that distinction is less than clearly defined. As we know from the continuum fallacy and the Sorites paradox, that the distinction between difference and disorder is a blurry one does not mean that there is no distinction to speak of. Furthermore, if biological variation were sufficient to undermine disorder designations in the context of neurological difference, then it would seem to have the same result in the context of other kinds of physical difference. Consider, for example, biometric measures like white blood cell count, blood pressure, and visual acuity, all of which fall on a continuum. While reading binary categories like “immunosuppressed,” “hypertensive,” and “vision impaired” off of such continua may involve a certain amount of arbitrariness, we should not conclude from this arbitrariness that no one is immunosuppressed or hypertensive or vision impaired. The same can, and should, be said of neurological difference. This means we cannot conclude from the observation that neurological variation is a ubiquitous and natural part of human biology that there are no neurological disorders. It seems equally implausible, though, to claim that all atypical forms of neurological 97 difference are disordered.63 Unsurprisingly, this leaves us with the view that some neurological differences are disordered, while others are mere-differences. The question, then, is how to distinguish the two. As Glannon (2007) notes, the neurodiversity debate “forces us to ask what counts as a mental disorder, and whether certain mental traits that deviate from those of the general population should be characterized as differences rather than disabilities” (p. 26). Once again, this places the burden on the neurodiversity view to defend a distinction between difference and disorder that is not subject to its own criticisms. 5.3.2 The Argument From Positive Attributes The Argument from Positive Attributes can be gleaned from statements like this one from Armstrong (2015): Many psychiatric conditions come with strengths as well as weaknesses. These strengths may have been evolutionarily advantageous, which explains why they are still in the gene pool. We should replace the ‘illness’ paradigm with a ‘diversity’ paradigm. (p. 348) Barnbaum (2013) holds a similar view about autism in particular: The neurodiversity movement holds that ASD is characterized not merely by impediments, but also by normal functioning, as well as cognitive strengths. ASD, on this account, may be best understood as a cognitive difference, not as a shortcoming. (p. 133) The argument at play here appeals to the premise that neurodiversity involves both impairments and positive attributes (see also Baker, 2006; Glannon, 2007; Happé, 1999; Treffert, 2009). Treating neurodiversity as a disorder, then, might fail to recognize these positive attributes, focusing instead solely on the impairments. This approach is either 63 For there are some forms of neurological atypicality, such as having an IQ of 130, that we tend to think are not disordered (indeed, the opposite), despite being uncommon. 98 arbitrary (since it focuses on the negative aspects and ignores the positive) or inaccurate (since it provides an incomplete characterization of neurodiversity), which leads to the conclusion that neurodiversity ought to be considered a mere-difference rather than a disorder. Positive attributes play an especially prominent role in popular narratives surrounding autism, which often emphasize things like “savant skills in maths, music and drawing, and islets of ability in visuospatial tests and rote memory” (Happé, 1999). We must be cautious, though, to avoid Rain Man-inspired stereotypes of autistic people as savants, since, as Hacking (2009) notes, although “a few savants are indeed autistic,” “hardly any autistic people are savants.” Treffert (2009) confirms this, estimating that less than 10% of autistic people demonstrate “special skills.” Howlin et al. (2009) put the number slightly higher, at 28.5% for “savant skills or exceptional cognitive skills.”64 In any case, it is worth noting that the positive attributes of autism are at risk of being overstated. A more serious problem with the argument from positive attributes is that it proves too much. There appear to be a number of disorders that involve positive attributes, which suggests that the fact that a condition comes with a positive attribute is not sufficient to establish that the condition is not a disorder. Examples include Williams syndrome, which involves desirable personality traits such as sociability and congeniality; sickle cell anemia, which leaves one less susceptible to malaria; Huntington’s disease, which makes one less susceptible to certain cancers later in life; cystic fibrosis, which makes one less susceptible 64 The discrepancy here might be a result of different criteria for “special skills” and the fact that Howlin et al. replied upon parental reports rather than laboratory observation. 99 to cholera and tuberculosis; neurosyphilis, which can result in feelings of euphoria, excitement, and relaxation; secondary polycythemia, which results in increased stamina; Marfan syndrome, which gives one long and flexible fingers (and was thought to have benefitted Paganini on the violin and Rachmaninoff on the piano); and lymphoedema distichiasis syndrome, which can result in Liz Taylor-esque eyelashes. If any one of these examples involves a paring of a condition rightly considered to be a disorder and an attribute rightly considered to be a positive one, then it must be the case that the positive attributes of neurological differences are insufficient to establish that they are not disorders. 5.3.3 The Argument From Normativity The Argument from Normativity can be derived from the work of those who have challenged purely biological conceptions of normal function. Amundson, for example, states: The disadvantages experienced by people who are assessed as ‘abnormal’ derive not from biology, but from implicit social judgments about the acceptability of certain kinds of biological variation. (2000, p. 33) The idea here is that an account of normal functioning cannot be rooted in biology alone, in which case accounts of abnormal functioning, and thus medical disorder, contain normative content (in addition to any empirical content). Accordingly, the claim that neurodiversity is a disorder involves a normative judgment about the condition, rather than the application of some purely objective standard—and negative normative judgments about neurodiversity are precisely what the view is challenging. So, the argument goes, those who claim that autism is a disorder are either falsely asserting that the condition is biologically dysfunctional, or begging the question against the neurodiversity view, which 100 denies that being autistic is an unacceptable form of variation. There are two ways to respond to this argument. First is to deny that accounts of normal and abnormal functioning are value-laden by defending a naturalist or objectivist view.65 I will not take up this response, however, since value-free accounts of functioning and disorder face serious conceptual problems. In light of such problems, many theorists have opted to defend hybrid views: “The tendency in recent philosophy has been to see disease concepts as involving empirical judgments about human physiology and normative judgments about human behavior or well-being” (Murphy, 2015, p. 1). This debate aside, it is simply not necessary to defend a naturalist account of functioning and disorder in response to the argument from normativity. A more promising response is to grant that concepts of function and disorder involve normativity, while noting that the argument again proves too much. If the normativity of judgments about neurodiversity render them illegitimate, then so too would the normativity of other disorder designations, in which case there would be no possibility of disorder at all, psychiatric or otherwise. Insofar as we think there are at least some conditions rightly considered disorders, this is an unacceptable result. While it is true that there are no purely biological grounds on which to base the claim that a condition like autism is a disorder, if the argument from normativity is right, then there are likewise no purely biological grounds on which to base the claim that a condition like pancreatic cancer is a disorder. A more promising approach may be to acknowledge the normativity of disorder, while arguing that certain normative judgments legitimately come into play (perhaps those 65 See Chapter 1 for a discussion of disease concepts. 101 related to a conception of wellbeing or flourishing), while others are illegitimate (perhaps those based upon biased or false beliefs). The question, though, is whether this distinction can be made such that it, one, avoids being ad hoc and, two, gets the result neurodiversity wants—namely, that the legitimate normative judgments appealed to when calling pancreatic cancer a disorder do not apply also to autism. 5.3.4 The Argument From Social Cause The Argument from Social Cause is motivated by the social model of disability, which, recall, understands disability as “the disadvantage or restriction of activity caused by a contemporary social organization which takes no or little account of people who have…impairments and thus excludes them from participation in the mainstream of social activities” (UPIAS, 1976, p. 30). According to strong versions of the social model, impairments are “mere-differences,” which become disabling only in exclusionary ableist social environments. If this is the correct way of thinking about neurodiversity, then we can see the neurodiverse as we do any other group of oppressed people. And we know that to be a member of an oppressed social group is not to have a disorder because belonging to the group would not be harmful or disadvantageous in a world without bias and discrimination. Designating a condition disordered, though, assumes endogenous harm. The argument goes that if the harm of neurodiversity is exogenous, and the harm of disorder is endogenous, then it is simply false to call neurodiversity a disorder, and ought to be called a mere-difference instead. This argument, as I see it, is the most compelling so far considered. Even so, it faces 102 a significant difficulty: What does it mean to assert the truth of the statement “if x were the case, then y would be the case” in a world where x is not, and never has been, the case? Though neurodiversity proponents insist that the harm of neurodiversity derives primarily, or solely, from neurotypical bias, opponents will insist on the significant role of inherent harm. Given that both of these claims are counterfactual, and that reasonable people may disagree about what holds true in possible worlds, how are we to adjudicate this kind of dispute without begging the question? While counterfactual arguments may be convincing in easy cases, they are a nonstarter for controversial cases like neurodiversity. What is not under dispute, however, is that there are things we can do in this world to make its navigation easier for people with neurological differences. In the next section, I will make the case that the primary political aims of the neurodiversity movement can be pursued without relying on flawed arguments that attempt to dissolve the distinction between neurodisorder and neurodiversity. 5.4 Preserving the Political; or Having and Eating Cake Thus far, I have argued that the neurodiversity view fails to provide convincing arguments for its conceptual claim that neurodiversity is a mere-difference rather than a disorder. In this section, I consider its political claim that neurodiversity ought to be respected and accommodated, rather than medicalized and “treated.” I argue that, although a strong interpretation of the political claim depends upon the soundness of the conceptual claim, the aims of a more moderate interpretation can be pursued without it. There are two ways of interpreting neurodiversity’s political claim. The first, which I will call the impermissibility interpretation, rests directly upon the conceptual claim that 103 a condition like autism is not a disorder: Since autism is a form of diversity rather than a disorder, and diversity is not an acceptable target of medical intervention, autism is not an acceptable target of medical intervention. This argument reflects the contention that the medicalization of neurodiversity “illustrates denial and intolerance toward differences, as well as the promotion of eugenic and genocidal policies”66 (Ortega, 2009, p. 7). Given that neurodiversity’s conceptual claim acts as a premise for this argument, my rejection of the conceptual claim amounts to a rejection of the impermissibility objection. However, there is a second interpretation of neurodiversity’s political claim that operates independently of the conceptual claim. What I will call the autonomy interpretation holds that, while a focus on medical intervention may reflect misguided priorities and ignore the social factors contributing to the disadvantage associated with neurodiversity, medical intervention ought to be available to those who choose it. This interpretation maintains simultaneously that (i) autism is a disorder; (ii) we ought to be more inclusive and accommodating of people with autism; (iii) and prevention, treatment, and cure may not be the best response to all cases of autism. On this interpretation, then, we can pursue neurodiversity’s political aims of social change, accommodation, and inclusion despite the flaws of conceptual claim. Furthermore, even if one remains convinced of neurodiversity’s conceptual claim, there may be strategic reasons for keeping the political claim independent. Consider Diamond and Rosky’s (2016) argument that advocacy for the rights of “sexual minorities” should not hinge on empirical claims about the immutability of sexual orientation, since these claims are controversial, may turn out to be false, and simply are not necessary for 66 See Chapter 5 for a discussion of the Impermissibility Interpretation. 104 the political agenda. A more stable foundation for these rights can be found in widelyshared cultural values like liberty and equality. The real reason for respecting the rights of sexual minorities is not that they were “born this way,” it is rather that we value selfdetermination, think “love is love,” and believe consenting adults should be afforded sexual freedom. It seems that we can say the same kind of thing about neurodiversity: Neurodiversity advocacy should not hinge on a conceptual claim about the differencedisorder distinction, since these kinds of claims are controversial, may turn out to be false, and simply are not necessary for the political agenda. We should be respectful and inclusive of those with neurological differences not because such differences are not disorders, but because respect and inclusion are deeply held values we wish to promote. A nice model for this approach comes from Down syndrome (Ds) advocacy, which seeks to further the interests of people with Ds without claiming that Ds is not a disorder. Activists in the Ds community, which includes both people with Ds and their allies, have been involved in a multifaceted campaign for acceptance, involving rhetoric (using disability-positive, yet tongue-in-cheek, terms like “Up syndrome,” “chromosomallyenhanced,” and “homie with an extra chromie” on social media); lobbying groups (such as DownPride, which challenges the practice of assumed pregnancy termination of the basis of a positive trisomy 21 screen in Europe);67 and Ds achievement centers (like GiGi’s Playhouse, whose mission is “to change the way the world views Down syndrome and send a global message of acceptance for all”). These strategies encourage the world to adapt to people with Ds, rather than insisting that people with Ds change to fit the world, without 67 See Chapter 5 for a more detailed discussion of this controversy. 105 denying the fact that Ds is a disorder—specifically, a chromosomal disorder associated with a number of health problems and lifelong challenges. This makes clear that the way a condition is classified has no direct bearing on the way people with that condition ought to be treated, as I argued in Chapter 4. Of course, it must be acknowledged that medical diagnoses, especially psychiatric diagnoses, carry with them real consequences as a matter of fact. As mentioned, there is the risk of stigmatization, labelling effects, diagnostic pessimism, threats to identity, inter alia. At the same time, we must recognize the fact that many benefit from their diagnosis, whether it be through access to services like therapy and education assistance, empathy from those around them, or simply having the conceptual resources with which to makes sense of their lived experience. The fact that a diagnosis of a condition like autism involves a range of effects underscores the autonomy interpretation of neurodiversity’s political claim, which defers to autistic individuals and their families about what they regard to be the best response to the case of autism with which they are faced. A further consideration is the possibility that the conceptual claim of the neurodiversity movement may actually cut against its political aims: If we think that people with medical disorders are owed certain things that people without medical disorders are not—e.g., workplace accommodations, additional resources for medical care, reduced responsibility for harmful actions—then lumping neurodiversity in with other meredifferences would seem to undermine the justification for special consideration for the neurodiverse, at least given the system in which we currently operate. The upshot here is that claims about which differences are disorders do not entail anything about disrespect or medical intervention. If current practice suggests otherwise, 106 that reflects a shortcoming of current practice, rather than of critiques of neurodiversity. What I am suggesting is that we can have our cake—by making accurate claims about the distinction between neurodiversity and neurodisorder—and eat it too—by advocating respect and acceptance of neurodiversity and resisting knee-jerk medical intervention. The range of experiences reported by those in the neurodiversity community indeed suggests that no single response suits all cases. Medical intervention would seem insulting and problematic for someone like Temple Grandin, who states, “If I could snap my fingers and be nonautistic, I would not. Autism is part of what I am” (2009, p. 34). Others, though, hold different attitudes towards their neurodiversity. Gwendolyn Kansen (2016), for example, says this: Some advocates feel they wouldn’t be the same person without autism. I’ve never felt that way. I still feel autism keeping me from achieving my potential…If autism is a double-edged sword, as they like to say, then it’s the sharpest one I’ve ever seen. (p. 4) Given this range of experience, I conclude that we ought to promote choice and respect autonomy for the neurodiverse, while also fostering a more respectful and inclusive society. Doing so does not require throwing out the distinction between difference and disorder. 5.5 References Amundson, R. (2000). Against normal function. Studies in the History and Philosophy of Biology & Biomedical Science, 31(1), 33-53. Anton, A. L. (2013). The virtue of psychopathy: How to appreciate the neurodiversity of psychopaths and sociopaths without becoming a victim. In C. D. Herrera & A. Perry (Eds.). Ethics and neurodiversity (pp. 111-130). Newcastle, UK: Cambridge Scholars Publishing. Armstrong, T. (2015). The myth of the normal brain: Embracing neurodiversity. AMA 107 Journal of Ethics, 17(4), 348-352. Baker, D. L. (2006). Neurodiversity, neurological disability and the public sector: Notes on the autism spectrum. Disability & Society, 21(1), 15- 29. Baron-Cohen, S. (2015). Neuroethics of neurodiversity. In Clausen and N. Levy (Eds.). Handbook of neuroethics. New York, NY: Springer Science. Barnbaum, D. R. (2013). The Neurodiverse and the neurotypical: Still talking across and ethical divide. In C. D. Herrera & A. Perry (Eds.). Ethics and neurodiversity (pp. 131-145). Newcastle, UK: Cambridge Scholars Publishing. Broderick, A. A. (2008). Autism as metaphor: Narrative and counter-narrative. International Journal of Inclusive Education, 12(5), 459- 476. Diamond, L. M., & Rosky, C. J. (2016). Scrutinizing immutability: Research on sexual orientation and US legal advocacy for sexual minorities. The Journal of Sex Research, 53(4-5), 363-391. Dinishak, J. (2016). The deficit view and its critics. Disability Studies Quarterly. Retrieved from http://dsq-sds.org/article/view/5236/4475. Fenton, A., & Krahn, T. (2007). Autism, neurodiversity and equality beyond the ‘normal’. Journal of Ethics in Mental Health, 2(2), 1-6. Giles, D. C. (2013). ‘DSM-V is taking away our identity”: The reaction of the online community to the proposed change of Asperger’s disorder. Health, 0(0), 1-17. Glannon, W. (2007). Neurodiversity. Journal of Ethics in Mental Health, 2(2), 5-7. Gould, S. J. (1996). The mismeasure of man. New York, NY: WW Norton & Company. Grandin, T. (2009). Thinking in pictures: My life with Autism. NewYork, NY: Bloomsbury. Hacking, I. (2009). How we have been learning to talk about autism: A role for stories. Metaphilosophy, 40(3-4), 499-516. Hacking, I. (2009). Humans, aliens & autism. Daedalus, 138(3), 44-59. Happé, F. (1999). Autism: Cognitive deficit or cognitive style? Trends in Cognitive Sciences, 3(6), 216-222. Harmon, A. (2004, December 20). How about not ‘curing’ us, some autistics are pleading. New York Times. 108 Horwitz, A. V. (2015). The DSM-5 and the continuing transformation of normal sadness into depressive disorder. Emotion Review, 7(3), 209-215. Howlin, P., Goode, S., Hutton, J., & Rutter, M. (2009). Savant skills in autism: Psychometric approaches and parental reports. Philosophical Transactions of the Royal Society B: Biological Sciences, 364(1522), 1359-1367. Jaarsma, P., & Welin, S. (2012). Autism as a natural human variation: Reflections on the claims of the neurodiversity movement. Health Care Analysis, 20(1), 20-30. Layton, T. J., Barnett, M. L., Hicks, T. R., & Jena, A. B. (2018). Attention Deficit– Hyperactivity Disorder and month of school enrollment. New England Journal of Medicine, 379(22), 2122-2130 Mackenzie, R., & Watts, J. (2011). Including emotionality in tests of competence: How does neurodiversity affect measures of free will and agency in medical decision making? AJOB Neuroscience, 2(3), 27-36. Matson, J. L., & Shoemaker, M. (2009). Intellectual disability and its relationship to Autism Spectrum Disorders. Research in Developmental Disabilities, 30(6), 1107-1114. McMahan, J. (2005). Causing disabled people to exist and causing people to be disabled. Ethics 116, 77–99. Meyerding, J. (1998). Thoughts on finding myself differently brained. Retrieved from http://mjane.zolaweb.com/diff.html. Murphy, D. (2015). Philosophy of psychiatry. Stanford Encyclopedia of Philosophy. Retrieved from https://plato.stanford.edu/entries/psychiatry/. Ortega, F. (2009). The cerebral subject and the challenge of neurodiversity. BioSocieties, 4, 425-445. Rashed, M. A. (2018). In defense of madness. Journal of Medicine and Philosophy, 44(2), 150-174. Singer, J. (1999). Why can’t you be normal for once in your life? From a “problem with no name’ to the emergence of a new category of difference. In M. Corker & S. French (eds.). Disability Discourse. Buckingham/Philadelphia: Open University Press. Treffert, D.A. (2009). Savant syndrome: An extraordinary condition. Philosophical Transactions of the Royal Society: Biological Sciences, 364, 1351–1357. UPIAS. (1976). Fundamental principles of disability. London: Union of the Physically 109 Impaired Against Segregation. Wolbring, G. (2007). Neurodiversity, neuroenhancement, neurodisease, and neurobusiness. Retrieved from http://innovationwatcharchive.com/choiceisyours/choiceisyours2007-04-30.htm, CHAPTER 6 NEURODIVERSITY IN AN IMPERFECT WORLD68 6.1 Abstract The neurodiversity view holds that certain neurological differences, like autism, are mere-differences (rather than bad-differences) and ought to be respected (rather than medicalized). Efforts to prevent autism can be criticized on the grounds that they incorrectly assume autism is a bad-difference, and they express something negative about autistic people. In this chapter, I provide a limited defense of autism prevention by arguing that these criticisms rely upon an equivocation between ideal and non-ideal circumstances. The claim that autism is a mere-difference is a counterfactual appeal to an ideal society: If the social world were more inclusive, being autistic would not be disadvantageous. A decision to select against autism, by contrast, is made in the non-ideal circumstances in which we find ourselves. This being so, prevention efforts may be said to express something about the social world rather than something about autism per se. 6.2 Introduction Generally speaking, there are two ways to reduce the incidence and prevalence of a medical condition: (1) treat existing cases, and (2) prevent cases from occurring in the 68 This chapter has been formatted for stand-alone publication. 111 first place. Treatment typically involves identifying individuals with the condition and providing them with medical care, with cure being the goal. Prevention takes a number of forms. Public health and safety initiatives seek to address environmental and behavioral causes of a condition in the population; vaccinations promote resistance; check-ups and screening identifies those at higher risk of developing a condition so as to intervene before it manifests; genetic testing of prospective parents identifies those at risk of passing on genetically transmitted conditions, which may lead some to forgo having children, or to have children through alternative means (e.g., sperm/egg donation or adoption); prenatal diagnostic procedures (e.g., amniocentesis and chorionic villus sampling) reduce the number of babies born with certain conditions; preimplantation genetic diagnosis (PGD) following in vitro fertilization (IVF) makes it possible for prospective parents to select only embryos that test negative for certain conditions. Taken together, treatment and prevention further the aims of clinicians, health scientists, and health policy makers alike. When it comes to conditions like lung cancer and Tay Sachs disease, such measures are generally considered to be unequivocally beneficial and worth significant investment. For some other conditions, however, talk of “cure” and “eradication” is much more controversial. Recent reports of the increased use of abortion in response to Down syndrome (Thiessen, 2018), for example, have prompted backlash from lawmakers (Rogers, 2019), ethicists (Kaposy, 2013), and disability advocates (DownPride, 2019), many of whom generally hold strongly pro-choice positions. The trend in Iceland, where Down syndrome has been effectively eliminated through prenatal screening and selective abortion, has been described provocatively by some as “genocide” (Will, 2018) and a system of “Nazi-like eugenics” (JLF, 2017). 112 Further complicating matters is the contention of some disability scholars and advocates that a number of conditions currently classified as disorders are, in fact, not disorders at all. While few would deny that Down syndrome is a medical disorder,69 other conditions—such as congenital deafness and achondroplasia—are argued to be neutral characteristics whose disadvantages stem primarily (or entirely) from social exclusion rather than biological dysfunction. If preventing a disorder like Down syndrome raises ethical concerns, then preventing differences mistaken for disorders may be even more problematic and deserving of provocative comparisons. One particularly compelling challenge to the status quo of popular opinion and medical practice comes from the neurodiversity movement (introduced in Chapter 5), which holds that certain neurological differences—most notably autism—are best understood as mere-differences (rather than bad-differences) and ought to be respected (rather than medicalized). Taking these claims seriously raises ethical questions about efforts to treat and prevent autism. There are two primary objections to treating autism in existing individuals. The first relates to the commonly held view that autism is central to personal identity, in which case attempts to cure autism would seem to threaten the identities of autistic persons. The second appeals to justice as it relates to the idea that autism is a mere-difference rather than a disorder. From this view, it would seem unjust to impose the costs of medical intervention on a group of people simply because they fail to conform to an arbitrary social standard of normalcy.70 69 Although some do. The advocacy group Down Syndrome Uprising, for example, states explicitly that “Down syndrome is not a disease.” I will not engage this view here. 70 See Section 4.5 for a discussion of closely related arguments. 113 Interestingly, however, neither of these concerns pertain to prevention efforts, since prevention occurs either prior to the manifestation of the condition or prior to the existence of any particular individual.71 On what grounds, then, might they be opposed? In what follows, I consider a pair of closely related arguments, which I will call the epistemic objection and the expressivist objection, meant to establish the impermissibility of selecting against autism. These objections are typically levelled against PGD and selective abortion, rather than other means of prevention.72 Despite the fact that preventing autism through such means remains a hypothetical possibility, autism advocates have already begun to worry about the implications. The epistemic objection opposes (certain) prevention efforts on the grounds that they assume that autism is a bad-difference, which is false in light of the claims of the neurodiversity view. The expressivist objection opposes (certain) prevention efforts to not because they cause direct harm—to embryos, fetuses, or future persons—but because they cause indirect harm to existing autistic people by expressing demeaning attitudes about them, which risks reinforcing exclusionary and discriminatory social practices. The aim of this chapter is to provide a limited defense of autism prevention, while granting neurodiversity’s mere-difference claim and avoiding both the epistemic and expressivist objections. I argue that, in our imperfect and ableist world, a decision to select against autism need not assume that autism is a bad-difference nor express something 71 As is the case for prevention following prenatal and preimplantation diagnosis (given the stipulation that the fetus/embryo is not a subject of moral concern qua individual). 72 This raises the question of how one can oppose prevention by means of PGD and selection abortion while permitting other means of prevention, such as public health initiatives. I consider this question briefly in Section 6.4. 114 negative about autistic people. Instead, such a decision may be based upon judgments about the nonideal circumstances in which we find ourselves. Although some instances of negative selection may be subject to the epistemic or expressivist objections, I conclude, these objections cannot justify a blanket ban. The structure of the chapter is as follows: In Section 6.3, I provide an overview of the neurodiversity movement and its ideological forebear, the social model of disability. In Section 6.4, I outline some of the ethical concerns related to the treatment and prevention of autism and disability more generally. In Section 6.5, I present my argument and consider the objection that my view implies the permissibility of selecting against characteristics like race and sexual orientation. 6.3 Neurodiversity and the Social Model of Disability The neurodiversity movement, recall, advocates for the rights and acceptance of the “neurologically different,” insisting that atypical ways of thinking, communicating, and behaving are not necessarily disordered (Ortega, 2009). Judy Singer, credited with coining the term, argues that neurodiversity ought to be understood as just another form of human difference: “the ‘neurologically different’ represent a new addition to the familiar political categories of class/gender/race” (1999, p. 91). Jami Anderson, a philosopher and parent to an autistic child, characterizes neurodiversity this way: Neurodiversity means regarding autistic individuals as fully persons rather than as broken beings in need of repair. It also means that rather than regarding autistic neurological structures as “defective” or “disordered,” one should regard autistic neurology as worth valuing because each neurological structure contributes to the collective variety of human neurological diversity, in much the same way that each human culture contributes to cultural diversity and each of the hundreds of human languages makes a valuable contribution to human linguistic diversity. (2013, p. 127) 115 Though it began in the autism community, the neurodiversity movement has come to encompass a number of other neurological differences, including attention deficit hyperactivity disorder, bipolar disorder, schizophrenia, developmental dyspraxia, dyslexia, epilepsy, psychopathy, sociopathy, Tourette’s syndrome, oppositional defiant disorder, dyscalculia, paracusia (see the “Hearing Voices Movement”), and the umbrella category of “madness” (see Anton, 2013; Fenton & Krahn, 2007; Mackenzie & Watts, 2011; Rashed, 2018). As the movement grows, neurodiversity has begun to attract attention in the disability literature, with a groundswell of support from academics wishing to push back against what they see as the medicalization of social ills (see Disability Studies Quarterly’s special issue “Autism and the Concept of Neurodiversity”). Proponents of the neurodiversity movement advance both conceptual and political claims. Conceptually, the view holds that “being neurodiverse or neurotypical are just different ways of existing as humans” (Jaarsma & Welin, 2012, p. 21); and that neurodiverse conditions involve “different ways of socializing, communicating and sensing that may not necessarily be disadvantageous” (Ortega, 2009, p. 426). Politically, the view is committed to the idea that “atypical neurological development…should be tolerated and respected in the same way as other human differences” (Griffin & Pollak, 2009, p. 27), rather than regarded as “a disease to be treated and, if possible, cured” (Ortega, 2009, p. 426). The neurodiversity movement can be seen as an outgrowth of the broader social model of disability—discussed in Chapter 4—which understands disability as a sociopolitical phenomenon in which ableist social arrangements actively disadvantage those with atypical bodies and brains. The social model is contrasted with the medical model of 116 disability,73 which characterizes disability the result of individually located impairments to be addressed with the tools of medical science. A more recent description of the social/medical divide comes from Barnes (2014), who distinguishes between “mere-difference” and “bad-difference” views of disability. Mere-difference views hold that “disability is not, by itself, something bad, harmful, or suboptimal” (p. 89). Bad-difference views, on the other hand, hold that “not only is having a disability bad for you, having a disability would still be bad for you even if society was fully accommodating of disabled people” (p. 89). The fundamental disagreement here is over the interaction between disability and well-being. Put roughly, the idea is that, while bad-difference views see the disadvantage associated with disability as being inherent and biological, mere-difference views see it as a contingent byproduct of an unjust social arrangement. Barnes acknowledges the complexities involved in clarifying the mere-difference view in light of disparate theories of well-being. In an attempt to accommodate this range of theories, she characterizes the mere-difference view as denying that (i) “disability is something that is an automatic or intrinsic cost to your well-being;” (ii) “even if we eradicated ableism, disability would still have a negative impact on well-being;” and (iii) “being disabled always makes your life go worse for you” (2009, pp. 91-92). The view need not, however, deny that “the actual well-being of disabled people is, on average, lower than that of nondisabled people” (given our current ableist social arrangement), nor that “disability always involves the loss of some goods” (2009, p. 90). Barnes (2014) fleshes 73 model. The medical model is also known as the individual or functional limitation 117 out the mere-difference view by distinguishing between local quality of life and overall quality of life: the view can admit that disability involves certain local harms (it often makes a person’s life harder in certain respects) while denying that it makes life worse all things considered (and thus maintain that disability is a mere-difference, rather than a baddifference). When it comes to autism, the debate surrounding the social and medical models, and mere-difference and bad-difference views, plays out well beyond the academic literature. One high profile instance of the divide involves the ongoing conflict between Autism Speaks—whose work pursues scientific research, increased screening, and medical intervention—and the Autism Self Advocacy Network (ASAN)—whose work is deeply rooted in the disability rights movement and focuses on neurodiversity activism. While Autism Speaks (2019) sees itself as being “dedicated to promoting solutions, across the spectrum and throughout the life span, for the needs of people with autism and their families,” the ASAN accuses the organization of promoting “offensive and outdated rhetoric of fear and pity, presenting the lives of autistic people as tragic burdens on our families and society” (ASAN, 2014). The stakes of the debate as framed are difficult to overstate: Is our investment the treatment and prevention of autism a noble quest for health and well-being or a form of eugenics meant to eliminate a group of people merely for being different? Seen this way, it is no wonder people have strong feelings about the issue. With these theoretical and practical considerations in mind, let us take a more detailed look at the ethical issues raised by the possibility of treating and preventing conditions like autism. 118 6.4 Causing, Treating, and Preventing Autism In 2002, Sharon Duchesneau and Candy McCullough, a deaf couple, created a stir in both the bioethics literature and popular press when they announced that they had taken steps to have a deaf child rather than hearing child (Spriggs, 2002). Specifically, they selected a deaf sperm donor with a family history of congenital deafness to increase the likelihood that their future child would be deaf, which indeed he turned out to be. In a similar case, a couple with achondroplasia demanded the use of PGD in order to select only embryos testing positive for the achondroplasia gene (Sanghavi, 2006; Smolensky, 2008). When the IVF physician resisted, the couple informed him that, if he refused PGD, they would simply abort any fetus without the gene and try again to ensure their child would have dwarfism. The practice of selecting for disability sits uneasily with many, including some within the disability community. In response to the Duchesneau-McCullough case, Nancy Rarus of the National Association of the Deaf stated, “I can’t understand why anybody would want to bring a disabled child into the world” (BBC, 2002). Bioethicists have likewise objected to the practice, on a number of grounds. Opportunity-based objections build on Feinberg’s (1980) account of a “right to an open future.” Davis (1997), for example, argues that the decision to create a deaf child, “confines her forever to a narrow group of people and a limited choice of careers,” and thereby “violates the child’s right to an open future” (p. 14). Welfarist objections take a slightly different tack by contesting the purported neutrality of disability. According to Glover (2006), deafness involves “the loss of a whole dimension of enriching experience: the sounds of rivers and waterfalls, of male and female voices, of laughter, of tractors and birds, of coffee bubbling, of the baby’s first 119 cry, of the whole of music,” and is thus “an obstacle to human flourishing” (p. 23; see also Murphy, 2009). Though such critics enjoy an intuitively plausible position, objections to the practice of creating disabled people face a deep philosophical challenge, known as the nonidentity problem (Parfit, 1984): How can bringing a disabled child into existence be said to harm the child, since not doing so would mean the child would not have existed at all?74 And if it does not harm the child, how can it be said to be wrong? The nonidentity problem, along with the advent of PGD, has provided fertile ground for bioethics research. Notice, however, that the question under consideration in this chapter is different than the question of causing disability. It is one thing to claim that it is permissible to cause the existence of a disabled child, as in the deafness and dwarfism cases; it is quite another to claim that it is impermissible to cause the nonexistence of a disabled child (qua disabled child), which is the contention of those who oppose selecting against disability. The latter claim is stronger insofar as it proscribes an action rather than permits it. Furthermore, whereas the claim that it is permissible to cause disability is aided by the nonidentity problem, the problem cuts against the claim that it is impermissible to prevent disability: If an action prevents the existence of a disabled person, to whom is it harmful? And if it harms no one, in what sense can it be said to be wrong? Importantly, however, the remainder of the discussion will pertain to the impermissibility of disability prevention, rather than issues related to the permissibility of causing disability or the complexities of the nonidentity problem. I take my aims to be relatively modest: I hope to show that not all 74 Assuming the disability in question is not so severe that it makes life not worth living. This, though, raises another complication: how can one coherently compare the value of nonexistence with the value of existence? 120 instances of selecting against autism fall prey to the epistemic or expressivist objections. Before considering their merits, though, we must get clear on what exactly these kinds of objections are objecting to. Consider the following hypothetical methods of autism prevention: • Selective Abortion: Researchers identify biomarkers for autism, detectable prenatally. At a visit with her OB-GYN physician, a patient, who is 12 weeks pregnant, elects to test for the biomarkers with amniocentesis. The test yields a positive result. She decides to have an abortion, specifying that she would have carried the pregnancy to term had the test been negative. • Selective Implantation: Researchers develop a genetic test for autism, detectable in IVF embryos prior to implantation. A couple in the process of assisted reproduction is told they have two viable embryos, one of which has tested positive for the autism gene. They decide to have the embryo that tested negative implanted and the embryo that tested positive discarded, specifying that they would have had both embryos implanted had they both tested negative. • Water Supply: After recording a precipitous spike in the number of cases of autism in a small town, researchers discover that the town’s water supply contains an uncommon chemical recently determined to cause autism. Fortunately, there is an effective and inexpensive method to remove the chemical from the water supply. Officials institute the method at the municipal water treatment plant and, over time, the rate of new cases of autism in the town returns to the national average. 121 • Vitamin G: Researchers discover that insufficient prenatal levels of vitamin G dramatically increase the likelihood that a child will develop autism. In response, the National Health Administration launches a public awareness campaign urging pregnant women to take vitamin G supplements. The campaign is a success and the rate of new cases of autism drops precipitously. As made clear above, disability advocates have explicitly objected to prevention efforts resembling Selective Abortion and Selective Implantation. Yet, as far as I know, none have resisted disability prevention efforts taking the form of Water Supply and Vitamin G. One question, then, is whether this is a consistent position. Parens and Asch (2000) consider the closely related examples of taking folic acid to reduce the likelihood of spina bifida and avoiding medications that are known to interfere with fetal development. While these modes of prevention are generally seen as protecting the health of a developing fetus, Parens and Asch note, disability critics tend to see selective abortion differently: “Abortion does not protect the developing fetus from anything. It prevents disability by simply killing the fetus” (p. 16). Similar sentiments are expressed about embryo selection on the basis of PGD. It is difficult to see how one can make this distinction, however, without smuggling in an assumption about the wrongness of abortion or destruction of embryos per se, which is supposed to be off the table for the purposes of this discussion. This seems to present a serious challenge to those who oppose selecting against disability. For now, let us assume that there is a principled means of distinguishing selective abortion and implantation from other ostensibly permissible means of disability prevention. What, then, does this mean for autism? Once again, it is important to note that, 122 although a small number of therapies have been shown to diminish the behavioral manifestations of autism in some, there exists no “cure” for existing cases, or any precise means by which to identify future cases. Furthermore, as Anderson (2013) notes, “a conversation about genetic screening for autism is purely theoretical because, although it is generally agreed that there is a strong genetic basis for autism, one’s genetics do not entirely determine one’s autism” (p. 125). This means that predicting autism faces two distinct challenges. First is that our best science simply has not revealed enough about autism’s underlying biology to make prediction currently feasible.75 Second is that, even if the science progresses significantly, complex gene-environment (or gene-gene) interactions will likely render any prediction probabilistic at best. Even so, there is optimism among some, and fear among others, those that autism may eventually go the way of Down syndrome in Iceland. Dr. Joseph Buxbaum’s 2005 prediction that researchers would develop a prenatal test for autism by 2015 was met with both elation and the launch of an “Autism Genocide Countdown Clock.” While Buxbaum’s prediction has not yet materialized, it is worth thinking carefully about the ethical issues raised by the mere possibility of treating and preventing autism. The general idea behind the disability rights critique of eradication efforts is this: Taken at face value, this critique condemns any effort to eliminate disabilities through medical interventions, genetic or otherwise…For if taken literally the slogan “change society, not individuals” does not merely insist that we try to make the social world more accessible to those whose impairments cannot be corrected; it would require accommodating those with impairments rather than using medical science to prevent or correct impairments. (Buchanan et al., 2001, p. 265) Of course, as expressed earlier, treatment and prevention raise distinct ethical 75 See Sections 2.2.3 and 3.5 for discussion of autism’s underlying biology. 123 issues, since treatment occurs in existing persons or existing cases and prevention does not. One prominent objection to the treatment and cure of autism concerns personal identity. Jim Sinclair, an autistic writer and early defender of the “anti-cure” movement, expresses the point eloquently in the piece “Don’t Mourn For Us”: It is not possible to separate the autism from the person. Therefore, when parents say, ‘I wish my child did not have autism,’ what they’re really saying is, ‘I wish the autistic child I have did not exist and I had a different (non-autistic) child instead.’ (1993, pp. 1-2) Anderson (2013) provides a pointed ethical analysis of this sentiment: If it really was [sic] possible to cure autism using gene therapy, the implications of undertaking such a treatment program are grotesque, given that the identity of a person is intimately connected to their brain structures. Indeed, the autistic person’s very being—who they are (their memories, their desires, wishes, hopes, plans) and what they can become—would be altered by this so-called cure. Surely such a neurological treatment is deeply unethical.” (p. 123) A second objection to treatment and cure appeals to social justice. If the neurodiversity movement is correct in its conceptual claim that neurodiversity is a meredifference, then it would seem unjust to impose the burdens of treatment upon individuals. Disadvantages caused by exclusionary and discriminatory social arrangements, we tend to think, ought to be addressed through social change, rather than by changing individuals to conform with arbitrary standards of normalcy. And there are indeed burdens that come along with medicalization: In addition to worries about personal identity, and about the safety and effectiveness of hypothetical interventions, existing therapies such as applied behavioral analysis (ABA) therapy can be extremely time76 and resource77 intensive. 76 “ABA is most successful when intensely applied for more than 20 hours a week and prior to the age of 4.” See Psychology Today. 77 The CDC (2018) estimates that “intensive behavioral interventions for children with ASD cost $40,000-$60,000 per child per year,” costs which may or may not be covered by insurance. 124 Moreover, such therapy may involve other burdens upon its recipient, such as emotional distress or loss of self-esteem. Arguments related to personal identity and social justice are certainly worth careful consideration. Interestingly, however, they do not bear directly on the issue of prevention, since prevention occurs either prior to the manifestation of the condition or prior to the existence of any particular individual. The practices of abortion and embryo disposal are, of course, ethically fraught in their own right. And it would follow trivially from the impermissibility of these actions that it is wrong to terminate a pregnancy or discard a human embryo on the basis of some diagnosis. The more interesting position is that which holds that abortion and embryo disposal are generally permissible, but are impermissible when done for the purposes of preventing a condition like autism. The two objections I want to consider together suggest that selecting against autism is both morally and factually suspect. The expressivist objection claims that prevention efforts harm existing disabled people and is therefore ethically problematic. The epistemic objection claims that prevention efforts rest upon the false claim that disability is a baddifference. Let us consider both in turn. “The expressivist objection,” Parens and Asch (2000) state “has been raised against selective abortion on the grounds that the practice “expresses discriminative attitudes” that “send a hurtful message to people who live with those same traits” (p. 13). McMahan (2005b) likewise notes the concern that “claiming that there are reasons to prevent the existence of disabled people may be expressive of a demeaning and hurtful view of the status of existing disabled people, a view that may encourage discriminatory attitudes towards and treatment of the disabled” (p. 142). 125 Writers with disabilities provide heartfelt renderings of the objection. Saxton (1997), for one, states, “the message at the heart of widespread selective abortion on the basis of prenatal diagnosis is the greatest insult: some of us are ‘too flawed’ in our very DNA to exist; we are unworthy of being born” (p. 12). Wendell, who has chronic fatigue immune dysfunction syndrome, states “the widespread use of selective abortion to reduce the number of people born with disabilities . . . sends a message to children and adults with disabilities, especially people who have genetic or prenatal disabilities, that ‘we do not want any more like you’” (1996, p. 153). Cokley, who has achondroplasia, asks us to consider “the message that society’s fear of the deviant—that boogeyman of imperfection—says to disabled people: ‘We don’t want you here. We’re actively working to make sure that people like you don’t exist because we think we know what’s best for you’” (2017). Addressing autism in particular, Kurchak (2015) writes emphatically that the existence of autistic people like me is not a ‘tragedy.’ Yet many autism awareness narratives insist it is because they prioritize the feelings of neurotypicals (non-autistic people) and dismisses the rest of us as little more than zombies. And when people buy into this idea, it actively hurts autistic people. The second objection, which I am calling the epistemic objection, contends that the practice of selecting against autism rests upon the false view that autism is a bad-difference (i.e., a disorder or negative difference-maker). Barnes (2009) expresses this kind of objection about disability generally: The only motivation for negative selection seems to be bound up in false views about disability—particularly, in the construal of disability as a negative differencemaker rather than just a difference-maker. The choice for negative selection relies on the notion that disability is somehow sub-optimal, making an able-bodied child preferable to a child with a disability. But this view of disability is, I’ve argued, a profound misunderstanding. (pp. 349-350) Whereas the expressivist objection is fundamentally ethical, in that it is concerned 126 with harm, this objection is fundamentally epistemic, in that is concerned with justification. In the next section, I will present my argument for the claim that selecting against autism does not, necessarily, express negative views, or rest upon false assumptions, about disability. However, in doing so, I wish to grant both neurodiversity’s conceptual claim that autism is a mere-difference, and the import of the expressivist and epistemic objections. If preventing autism always involves false assumptions and harmful attitudes, then the practice would indeed be impermissible. But, I will argue, it does not. 6.5 Perfect and Imperfect Worlds In Section 6.3, I noted the claim made by the neurodiversity view that autism is a mere-difference. In Section 6.4, I described the expressivist and epistemic objections, which oppose preventing autism on the grounds that doing so expresses something negative about autistic people and false about autism. The upshot so far, then, is that prevention efforts both assume that autism a bad-difference and reinforce the very social arrangement that leads people to this assumption. Parens and Asch (2000) put the idea this way: According to the disability rights critique of prenatal testing, if people with disabilities were fully integrated into society, then there would be no need for the testing. In the world they seek to create, if a given health status turned out to be a handicap, that would be because of societal, not personal, deficits; the appropriate response would be to change society so that the person could live a full life with a range of talents, capacities, and difficulties that exist for everyone. In a society that welcomed the disabled as well as the non-disabled, there would be no reason to prevent the births of people with traits now call disabling. (p. 23) This statement reflects the views of the social model of disability, which (recall) locates the harm or disadvantage associated with disability in the social exclusion of, and discrimination against, disabled people, rather than some inherent defect, deficit, or dysfunction. How, then, do we sort out whether a given characteristic is a mere-difference 127 or a bad-difference? The method most commonly employed seems to be counterfactual reasoning of the following sort: Consider some characteristic, C, that carries with it some kind of disadvantage or harm, and how society treats people with C. If there is some feasible means by which to modify the way that society treats people with C that eliminates or significantly mitigates C’s harm, then C is a mere-difference. If not, then C is a baddifference. This method of counterfactual reasoning seems to get the right result in most cases. Consider the characteristics of being a woman or a person of color. At this place and time, being a woman or a person of color carries with it certain disadvantages, given the robust phenomenon of gender-based discrimination, white privilege, and related injustices. Yet we can easily imagine a world in which such injustice does not exist. Surely, in such a world, there would be nothing disadvantageous about being a woman or person of color. These characteristics are therefore mere-differences.78 Consider alternatively the characteristic of having lung cancer. While we can imagine improved oncological treatments and better access to such treatment, there seems to be no amount of social rearrangement that would make lung cancer in no way harmful or disadvantageous. Even in the best-case scenario, lung cancer would be a nuisance to be addressed. We can conclude, therefore, that lung cancer is a bad-difference. Unsurprisingly, the neurodiversity view takes being autistic to be more like being 78 Interestingly, the counter-factual here may need to be “a world in which such injustice never existed” rather than “a world in which such injustice were eradicated,” since there is evidence to suggest that trauma is subject to epigenetic transmission (see Youssef et al., 2018). If so, then the harm of being a member of an oppressed social group may go beyond mere social harm to include genetic harm, even if of social origin. Thanks to Natalia Washington for raising this important point. 128 a person of color than having lung cancer. Despite the fact that our current social arrangement—with its unjustly ableist and neurotypical design—makes it the case that autistic people face consequential challenges and disadvantages, it is easy to imagine a world without, or with significantly fewer, such challenges and disadvantages. Although I have reservations about the purported neutrality of autism,79 I would like here to grant the point arguendo. It is important to note, however, the counterfactual form the claim takes: Autism is a mere-difference, since, in the right kind of world, it would involve no significant disadvantage. My contention is that this claim is perfectly compatible with the claim being expressed by efforts to prevent future cases of autism: In this world, autism involves significant disadvantages. In other words, we can agree with neurodiversity’s conceptual claim while denying that selecting against autism expresses anything negative about autistic people. What the practice of prevention really expresses is a concern about the fit between current social structures and certain kinds of atypical neurologies.80 The inevitable objection to the above argument is this: If it is permissible to select against autism in virtue of disadvantaging social arrangements, then it would be permissible to prevent other mere-differences, such as gender, ethnicity, and sexual orientation. But since it is impermissible to prevent these kinds of mere-differences, it must also be impermissible to select against autism on these grounds. My response to this objection is a qualified avowal that it may indeed be 79 See Chapter 4 for my critique of neurodiversity’s conceptual claim. A further worry about the expressivist objection comes from JL Nelson (2000), who argues that “considered as a practice, abortions prompted by prenatal testing are not semantically well-behaved enough to send any particular message.” 80 129 permissible to select against these kinds of mere-difference as well. To be clear, I think there are many cases in which such prevention efforts would be deeply problematic, just as I think there are cases in which preventing autism would be problematic. But we require more information to determine the moral status of this kind of action. As Nelson (2000) notes: Someone could, of course, harbor disrespectful thoughts about people with disabilities and be motivated to seek out testing and termination on the basis of those thoughts; legislators might vote to fund genetic counseling for the same kind of reason. Similarly, a person might take folic acid, or consent to prenatal surgery, out of contempt for disabled people; the NIH might allocate funds for research designed to eliminate the impact of disabilities because key policy makers have the same view. (p. 226) To emphasize this point, consider the following scenarios: • Sperm Bank 1: A White couple has decided to use a sperm donor for their IVF process. Prior to the selection process, they have decided to rule out donors of color, because they believe that their child will be more intelligent with a White donor. • Sperm Bank 2: A White couple has decided to use a sperm donor for their IVF process. Prior to the selection process, they have decided to rule out donors of color, because the fear they are unequipped to raise a child of color given their life of White privilege. It seems clear that these scenarios are not on a moral and epistemic par. (1) expresses negative, demeaning, and false views about persons of color, views which contribute to the culture that imposes the disadvantage they face. The couple in (1) acts wrongly. A moral assessment of (2), though, is likely to be much different. This couple may lament the fact that we live in a world in which race plays such a prominent role in 130 one’s life experience and, in the right circumstances, would love to have a child of color. But, given our imperfect world, they choose to have a White child. This decision seems to express not racist attitudes, but rather a sensitivity to the racial complexities of our nonideal circumstances. Consider two more scenarios: • PGD 1: A couple in the process of IVF has decided to screen their viable embryos for spina bifida. One of the embryos tests positive. They decide not to implant the embryo because they “don’t want a retarded child.” • PGD 2: A couple in the process of IVF has decided to screen their viable embryos for spina bifida. One of the embryos tests positive. They decide not to implant the embryo because they do not have health insurance and know that children with spina bifida often require multiple medical procedures. They also note that they live in a jurisdiction that does not require accessibility for those with limited mobility. Here again we see same action with very different motivations. In (1), the couple believes that children with spina bifida are “retarded,” which assumes falsely that the condition necessarily affects cognitive function and expresses the view that existing people with spina bifida are “retarded.” The couple in (1) acts wrongly. In (2), the couple’s decision is based upon resource limitations and environmental obstacles, which I take to be legitimate considerations. Plausible claims about the need for universal healthcare and accessibility in public spaces seem to have no bearing on the ethical assessment of (2). In her condemnation of selecting against disability, Barnes (2009) considers McMahan’s (2005a) suggestion that “a couple that decides to abort a foetus when told it will have a disability isn’t claiming that disability is sub-optimal per se; they’re simply 131 claiming that having a child with a disability is sub-optimal for them.” She responds that “the couple may well only be demonstrating that having a child with a disability is suboptimal for them—but that isn’t an appropriate attitude for them to have!” (p. 351). Interestingly, though, Barnes includes this in a footnote from the section: There may, of course, be extrinsic factors that could affect such a decision — e.g. a couple without health insurance in a society without universal healthcare learning that they will have a child that requires substantial medical care. But their choosing to negatively select wouldn’t be any judgement against disability per se — it would simply be a manifestation of other (largely unrelated) social injustices. So that couple isn’t claiming that having a child with a disability would be of less value to them; they might well love to have a child with a disability, but simply can’t because of their economic circumstances. (2009, p. 355) But the fact that negative selection “wouldn’t be any judgment against disability per se” is precisely the point I wish to make: Efforts to prevent autism need not express anything negative about people with autism, or assume falsely that autism is a baddifference. Rather, a given decision to select against autism may simply be a reflection of the difficulties associated with neurodiversity in the imperfect and ableist circumstances in which we find ourselves. There may well be something regrettable about the practice of selecting against disability—namely, that our social conditions have given rise to a perceived need for the practice in the first place. There is a difference, however, between noting that an action is regrettable and claiming that it is impermissible. We may lament the need for cars, while acknowledging the permissibility of driving in places with poor public transit and limited infrastructure for alternative modes of transportation. We may find it regrettable that so many people feel the need to conform to arbitrary and oppressive standards of beauty, while recognizing their right to do so in a society, like ours, in which conventional attractiveness brings tangible advantages. Nonideal circumstances often beget nonideal actions. But one 132 should be hesitant to condemn the actions themselves. Such condemnation places the burden of social change on the shoulders of individuals who may relish the opportunity to sell their cars and get rid of their makeup, but are waiting for a different world. 6.7 References Anderson, J. L. (2013). A dash of autism. In J. L. Anderson & S. Cushing (Eds.). The philosophy of autism (pp. 143-166). Lanham, MD: Rowman & Littlefield. Anton, A. L. (2013). The virtue of psychopathy: How to appreciate the neurodiversity of psychopaths and sociopaths without becoming a victim. In C. D. Herrera & A. Perry (Eds.). Ethics and neurodiversity (pp. 111-130). Newcastle, UK: Cambridge Scholars Publishing. Autism Self Advocacy Network (ASAN). (2014, January 6). Joint letter to the sponsors of Autism Speaks. Retrieved from https://autisticadvocacy.org/2014/01/2013joint-letter-to-the-sponsors-of-autism-speaks/. Barnes, E. (2009). Disability, minority, and difference. Journal of Applied Philosophy, 26(4), 337-355. Barnes, E. (2014). Valuing disability, causing disability. Ethics, 125(1), 88-113. BBC. (2002, April 8). Couple “choose” to have deaf baby. BBC News. Retrieved from http://news.bbc.co.uk/hi/english/health/newsid_1916000/1916462.stm. Buchanan, A., Brock, D. W., Daniels, N., & Wikler, D. (2001). From chance to choice: Genetics and justice. New York, NY: Cambridge University Press. Center for Disease Control and Prevention (CDC). (2018). Data & statistics on Autism Spectrum Disorder. Retrieved from https://www.cdc.gov/ncbddd/autism/data.html#references. Cokley, R. (2017, August 10). Please don’t edit me out. Washington Post. Retrieved from https://www.washingtonpost.com/opinions/if-we-start-editing-genes-people-likeme-might-not-exist/2017/08/10/e9adf206-7d27-11e7-a669b400c5c7e1cc_story.html?utm_term=.dcf3e6d374f0. Davis, D. S. (1997). Genetic dilemmas and the child’s right to an open future. Hastings Center Report, 27(2), 7–15. Downpride. (2019). Retrieved from https://downpride.com/574-2/. 133 Feinberg, J. (1980). The child’s right to an open future. In W. Aiken & H. LaFollette (Eds.). Whose child? Children’s rights, parental authority, and state power (pp. 124–153). Totowa, NJ: Rowman & Littlefield. Fenton, A., & Krahan, T. (2007). Autism, neurodiversity and equality beyond the ‘normal.’ Journal of Ethics in Mental Health, 2(2), 1-6. Glover, J. (2006). Choosing children: Genes, disability, and design. Oxford, UK: Oxford University Press. Griffin, E., & Pollak, D. (2009). Student experiences of neurodiversity in higher education: Insights from the BRAINHE Project. Dyslexia, 15, 23-41. Jaarsma, P., & Welin, S. (2012). Autism as a natural human variation: Reflections on the claims of the neurodiversity movement. Health Care Analysis, 20(1), 20-30. Jerome Lejeune Foundation (JLF). (2017). Patricia Heaton: Iceland is not eliminating Down syndrome, but killing those with it. Retrieved from https://lejeunefoundation.org/patricia-heaton-iceland-eliminating-downsyndrome/. Kaposy, C. (2013). A disability critique of the new prenatal test for Down syndrome. Kennedy Institute of Ethics Journal, 23(4), 299-324. Kurchak, S. (2015, April 30). Autistic people are not tragedies. My life has value. The Guardian. Retrieved from https://www.theguardian.com/commentisfree/2015/apr/30/autism-is-not-atragedy-take-it-from-me. Mackenzie, R., & Watts, J. (2011). Including emotionality in tests of competence: How does neurodiversity affect measures of free will and agency in medical decision making? AJOB Neuroscience, 2(3), 27-36. McMahan, J. (2005a). Causing disabled people to exist and causing people to be disabled. Ethics, 116, 77–99. McMahan, J. (2005b). Preventing the existence of people with disabilities. In D. Wasserman, J. Bickenbach, & R. Wachbroit (Eds.). Quality of life and human difference: Genetic testing, health care, and disability. New York, NY: Cambridge University Press. Meyerding, J. (1998). Thoughts on finding myself differently brained. Retrieved from http://mjane.zolaweb.com/diff.html. Murphy, T, F. (2009). Choosing disabilities and enhancements in children: A choice too far? Ethics, Bioscience and Life, 18 (Supplement 1), 43–49. 134 Nelson, J. L. (1998). The meaning of the act. Kennedy Institute of Ethics Journal, 8, 16582. Nelson, J. L. (2000). Prenatal diagnosis, personal identity, and disability. Kennedy Institute of Ethics Journal, 10(3), 213–228. Ortega, F. (2009). The cerebral subject and the challenge of neurodiversity. BioSocieties, 4, 425-445. Parens, E., & Asch, A. (2000). The disability rights critique of prenatal testing: Reflections and recommendations. In E. Parens & A. Asch (Eds.). Prenatal testing and disability rights. Washington, D.C.: Georgetown University Press. Parfit, D. (1984). Reasons and persons. Oxford, UK: Oxford University Press. Psychology Today. (2019). Retrieved from https://www.psychologytoday.com/us/therapy-types/applied-behavior-analysis. Rashed, M. A. (2018). In defense of madness. Journal of Medicine and Philosophy, 44(2), 150-174. Rodgers, B. (2019, February 6). Sponsor of bill on Down syndrome abortion ban says this year’s version sidesteps constitutional issues. Salt Lake Tribune. Retrieved from https://www.sltrib.com/news/politics/2019/02/07/sponsor-bill-down/. Sanghavi, D. M. (2006, December 5). Wanting babies like themselves, some parents choose genetic defects. The New York Times. Retrieved from https://www.nytimes.com/2006/12/05/health/05essa.html. Saxton, M. (1997). Disability rights and selective abortion. In R. Solinger (Ed.). Abortion Wars: A Half Century of Struggle, 1950-2000 (pp. 374-395). Berkeley and Los Angeles, CA: University of California Press. Sinclair, J. (1993). Don’t mourn for us. Our Voice, 1(3). Singer, J. (1999). Why can’t you be normal for once in your life? From a ‘problem with no name’ to the emergence of a new category of difference. In M. Corker, S. French (Eds.). Disability discourse. Philadelphia, PA: Open University Press. Smolensky, K. R. (2008) Creating children with disabilities: Parental tort liability for preimplantation genetic diagnosis. Hastings Law Journal, 60(2), 299-346. Spriggs, M. (2002). Lesbian couple create a child who is Deaf like them. Journal of Medical Ethics, 28, 283. Thiessen, M. A. (2018, March 8). When will we stop killing humans with Down 135 syndrome?” Washington Post. Retrieved from https://www.washingtonpost.com/opinions/when-will-we-stop-killing-humanswith-down-syndrome/2018/03/08/244c9eba-2306-11e8-badd7c9f29a55815_story.html?utm_term=.4a9dd43f1ec8. Wendell, S. (1996). The rejected body. New York, NY: Routledge. Will, G. F. (2018, March 14). The real Down syndrome problem is accepting genocide. Washington Post. Retrieved from https://www.v.com/opinions/whats-the-realdown-syndrome-problem-the-genocide/2018/03/14/3c4f8ab8-26ee-11e8-b79df3d931db7f68_story.html?utm_term=.35baee9c9505. Youssef, N., Lockwood, L., Su, S., Hao, G., & Rutten, B. (2018). The effects of trauma, with or without PTSD, on the transgenerational DNA methylation alterations in human offsprings. Brain sciences, 8(5), 83. 136 CHAPTER 7 CONCLUSION I recently received an email from the Utah Down Syndrome Association (UDSA) entitled “Make Your Voice Heard!” As a parent of a child with Down syndrome, I am often sent messages from disability organizations about local social events, awareness campaigns, support groups, and so on. This particular email concerned a bill that was making its way through the Utah Legislature, HB 166, which “protects unborn children in Utah from discrimination based on a diagnosis (or belief) that they have Down syndrome.” The UDSA was asking those of us in state’s Down syndrome community to call our representatives, march on the capitol, and speak up in favor of the bill. There is plenty to say about HB 166, and others like it81—about their constitutionality, about whether they are based upon a good faith concern for people with Down syndrome, about whether the conservative lawmakers supporting these bills would also support allocating state resources to improve services for disabled people and their families. What struck me most about the email, though, was not the content of the bill itself, but rather the presupposition that I and other members of the UDSA would support it. In one sense, I understand: After all, we are members of a group that had been brought 81 Similar bills have become law in Indiana, Ohio, North Dakota, Louisiana, and Kentucky. 137 together by a love for someone with Down syndrome and a shared desire to make lives of people with disabilities better. Why would we not support a legislative effort to recognize the value of people with Down syndrome, recognition long fought for and denied? In another sense, though, I was taken aback: why should the fact that I have a child with Down syndrome dictate my views about such a legally and morally nuanced issue? As both a disability advocate and someone who thinks carefully about philosophical issues related to disability, my personal and professional identities have become intertwined in complex ways. Usually, when asked why I chose to write a dissertation on autism, I say something about how it presents a fascinating test case for some contemporary debates in the philosophy of medicine, bioethics, and public policy.82 Truth be told, my work on autism—and disability generally—originated more from a personal interest than an intellectual one. The interest grew out of my relationship with my younger brother, Taylor, who has Autism Spectrum Disorder, Cornelia de Lange syndrome, and intellectual disability. Growing up with him made life complicated, funny, embarrassing, and beautiful all at once. This interest has continued to develop through conversations with my partner, Kendra, who diagnoses and treats children with developmental disabilities, and through the experience of parenting our daughter, Fern. As a nondisabled person, I would never presume to understand fully the disabled experience; but I can say confidently that I have become well acquainted with some of the joys and challenges associated with autism and certain other forms of disability. 82 See “Introduction.” 138 When it came time to write a dissertation, I was eager to apply the philosophical tools I had acquired in my training to a topic that has had such a profound presence in my life. And the personal connections I had with my academic work brought advantages: as someone with a stake in disability issues, I was motivated; as someone familiar with disability, I was able to draw on my own experiences; as someone with contacts in the disability world, I was able to do things like observe physicians in a comprehensive care clinic for autistic people. Yet my background with disability was not, I think, unequivocally beneficial. Throughout the writing process, I found myself engaging in motivated reasoning, which led me to favor conclusions before working through the arguments. At the same time, I found others expecting me to hold views about which I was deeply uncertain. Such was the case when I was asked to support HB 166: If I was going to Make My Voice Heard!, which voice would it be? Would it be the voice of a parent who thinks that people have no reason to fear Down syndrome? Or the voice of a philosopher who is skeptical of the notion that fetuses can be discriminated against, or that having an abortion on the basis of a trisomy 21 diagnosis is somehow more objectionable than having an abortion for any other reason? Similar questions arose in my writing: Did personal experience with disability contribute to the project, or was my thinking being clouded by sentimentality? Did my relationships with Taylor and Fern make me more akin to a situated knower or a surgeon operating on a family member? Thinking about this led me to posit a distinction between advocacy scholarship and scholarly advocacy. Advocacy scholarship, of the sort I found tempting, identifies itself with view or objective and subsequently provides argumentation to support it—the 139 advocacy drives the scholarship. Scholarly advocacy, by contrast, attempts to further a view or objective based upon scholarly activity, while embracing fallibility—the scholarship drives the advocacy. In this sense, the distinction resembles that between pseudo-science and science: whereas pseudoscience reinforces, science challenges. One of my goals for this dissertation, then, has been to capture the richness of my experience with disability while being intellectually honest and following the arguments where they lead. This approach has resulted in less-than-positive feedback from some (otherwise likeminded) disability advocates who expressed concern that my views are insufficiently progressive or disruptive. In light of such criticisms, all I can say is that I have done my best here to advance positions that seemed to me the most reasonable. My hope is that they will be judged by their content, rather than the extent to which they adhere to the standards of advocacy scholarship. With that, let us review. I began the dissertation by posing two general questions, one descriptive and one normative: (1) How should we understand autism? and (2) How should we respond to it? My aim was to address both questions while exploring the ways in which each bears on the other. In doing so, I examined autism through the conceptual lenses of disorder, disability, and diversity, and these concepts through the lens of autism. Chapter 2 served as an introduction, distinguishing between three levels of analysis for autism: biology, cognition, and behavior. It also outlined the minimal and strong interpretations of the medical model of psychiatry, as well as a number of positions in medical ontology. I argued that figuring out what kind of thing autism is matters insofar as it determines which people get diagnosed with autism. Moreover, diagnosis matters insofar as it determines which people receive which kinds of treatment, accommodation, and 140 judgment. The current gold standard for psychiatric diagnosis comes from the fifth edition of the American Psychiatric Association’s Diagnostic and Statistical Manuel for Mental Disorders, which includes diagnostic criteria for “Autism Spectrum Disorder” (ASD). In Chapter 3, I posed a dilemma for the DSM-5: If all it takes to have ASD is satisfying the DSM’s diagnostic criteria for ASD, then it seems that there is no possibility that the DSM gets ASD wrong, so long as the criteria are correctly applied. If, on the other hand, the diagnostic criteria for ASD serve as mere symptoms of autism, then the biological heterogeneity seen among autistic people threatens the usefulness of an ASD diagnosis in developing interventions and predicting the course of the condition. The upshot of this dilemma is that the category “ASD” is either trivial or unhelpful. Moving from disorder to disability, Chapter 4 introduced the social model of disability, which characterizes the disadvantages associated with disability as a result of discrimination against individuals with impairments, rather than the result of impairment per se. This model also endorses the idea that we should “change the world, not people” in response to these disadvantages. Claims about what disability is, though, entail nothing about what ought to be done about it. The question, then, is how proponents of the social model might bridge this is-ought gap. I argued that considerations of identity and diversity go the furthest in doing so. Chapter 5 addressed the closely related neurodiversity view, which holds that certain neurological differences currently classified as disorders are best understood as forms of diversity. Rather than being medicalized, the view holds, neurodiversity ought to be respected and accommodated. Despite neurodiversity’s merits, I concluded that it falls 141 short of establishing that autism is not a disorder. Fortunately, given the discussion in Chapter 3, the fact that autism is a disorder does not entail anything about whether it ought to be medicalized. This being so, I argued, we can pursue neurodiversity’s political aims of social change and accommodation without relying on implausible conceptual claims about disorder. Finally, in Chapter 6, I considered the issue of whether it is permissible to use medical intervention to prevent autism. The practice of selecting against disability—by means of selective abortion and preimplantation genetic diagnosis, for example—has been criticized on the grounds that it falsely assumes that disability is a bad-difference; and that it expresses something negative about, and thus causes harm to, existing people with disabilities. In considering these objections, I provided a limited defense of autism prevention by arguing that, since a decision to select against autism is made in the nonideal circumstances in which we find ourselves, it may be said to express something negative about the ableist social world rather than something negative about autism per se. Looking ahead, I am interested in expanding my thinking about autism to include the wider phenomenon of intellectual disability (ID). ID is most often mentioned in the philosophy literature as a way of testing intuitions about the limits of the moral community, often in the context of social contracts and moral agency. One issue that has yet to receive sufficient attention is the possibility of shared agency for those with ID, whether through supported decision making, assisted communication, relational autonomy, or some other form of “agency scaffolding.” Shared agency is in tension with the long-held notion of human beings as silos of rationality and islands of self-determination. The case of intellectual disability suggests that this picture does not pertain to at least some of us; and 142 I suspect that examining it more closely will reveal that it pertains to no one at all. Speculation aside, I will continue to work on issues related to disability, all while attempting to reconcile my various voices into one that is worth making heard.