Genomic Designation: New Kinds of People at the Intersection of Genetics, Medicine and Social Action

Genomic Designation: New kinds of people at the intersection of genetics, medicine and social action

Daniel Navon

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in the Graduate School of Arts and Sciences

COLUMBIA UNIVERSITY

2013

ABSTRACT

Genomic Designation: New kinds of people

at the intersection of genetics, medicine and social action

Daniel Navon

Genetics can do more than predict, explain or help treat medical conditions – it can create new ones. The social sciences have assumed that genetics must work in and through existing categories of human difference in order to inform clinical practice or social mobilization. By contrast, I go beyond the specter of reductionism and examine the emergence of new kinds of people at the intersection of genetics research, clinical practice and social action. For over fifty years, conditions like the XXX, Edwards, Fragile X and 22q11.2 Deletion Syndromes have been discovered, delineated and diagnosed strictly according to abnormalities in the genome, even in the absence of phenotypic coherence – a practice which I call ‘genomic designation’. This dissertation uses comparative historical methods, fieldwork and citation analysis to examine the history of genomic designation, its variable impact on practice and its implications for our understanding of the biosciences, medicine and social mobilization. I argue that genomic designation represents an important and growing practice that extends and challenges existing formulations of key concepts like ‘biosociality’, geneticization and the rise of a ‘molecular gaze’ in contemporary medicine. Furthermore, I show how it offers an opportunity to develop a typology of ways in which genetics can radically reconfigure medical classification.

However, over the course of its fifty-year history, genomic designation has varied enormously as a clinical and social phenomenon and therefore in the way it impacts lived experience. I show how, during the first few decades after genomically designated syndromes began to be delineated in the human genetics literature in 1959, they gave rise to very little by way of clinical protocols, practices or specialist centers and virtually no social or advocacy organizations. And yet, in recent decades, genomically designated conditions have emerged as bona fide categories of clinical practice and social mobilization. Drawing on Fleck, Foucault and

Haydu, I propose a framework of ‘reiterative facticity’ that aims to combine work from the sociology of science and medicine with a comparative-historical approach by analyzing the way that the very same genetic mutations take on divergent meanings and implications according to contrasting conditions of possibility, repertoires of collective action and the networks of research and advocacy organized around genomically designated conditions.

I discuss the way that genomically designated syndromes are often ‘leveraged’ as models in biomedical research, and how this can turn them into privileged sites of knowledge production, commercial investment and social mobilization. In particular, I analyze the intersection of genetic disorders and autism in order to understand the nosological conditions for genomic designation and the ‘trading zones’ in which genetic and psychiatric systems of classification can achieve a productive interface. Finally, I use historical and fieldwork material to examine the conditions and repertoires of collective action through which a complex network has been assembled around 22q11.2 Deletion Syndrome, turning it into what Hacking would call a new kind of person that can realign clinical judgment, treatment and care. In this way, a comparative study of genomic designation shows how biological abnormality must be mediated by historical conditions and prevailing modes of understanding and acting on human difference, but also mobilized by heterogeneous networks of actors working to interface with but also transform existing structures. By way of conclusion, I discuss the possible impact of new non-invasive prenatal genetic testing on genomic designation, summarize my findings and suggest fruitful lines of future research. !

Table of Contents

List of tables and figures ii

Acknowledgements iv

Introduction 1

Part I

1. From mutations to new medical conditions 21

2. The varieties of genomic designation: 52 Multiple pathways and an ideal-typical typology of syndromes

Part II

Introduction 85

3. Immobile mutations: 106 Nowhere to go in the 1960s and 70s and the exception that proves the rule

4. Of elves, warriors and autistics: 162 Leveraging abnormality in genetics research and advocacy

5. The trading zone of autism genetics: 217 Looping and the intersection of genomic and psychiatric classification

6. Assembling a new kind of person: 277 Mobilizing mutations and realigning illness

Conclusion 362

References 395

! i! !

List of tables and figures

Introduction Figure 1: FISH test indicating a 22q11.2 microdeletion 2

Chapter 1 Figure 1: Clinical Syndrome Referent 44 Figure 2: Genomically-Designated Syndrome Referent 44 Table 1: Genotype-phenotype relations in four medical conditions 45 Table 2: Preliminary list of genomically designated conditions 47

Chapter 2 Figure 1: Modularity analysis of research literature 65 Figure 2: Modularity analysis with and without 22q11-related papers 66 Figure 3: Citation networks for four key moments 68 Figure 4: Citation network and communities in 1992-1995 71 excluding articles about 22q11. Figure 5: Citation networks and communities in 1998-2002 and 2003-2009 73 Figure 6: Frequency of syndrome names in paper titles, 1996-2012 74

Chapter 3 Figure 1: The standardization of human chromosome numbers 107 agreed upon in Denver, 1960 Figure 2: New York Times, April 21 1968, p. 1 126 Figure 3: Articles published by year with XYY in their titles 141 Figure 4: KS&A Brochure from XYY Syndrome 157

! ii! !

Chapter 4 Table 1: Leveraging, alliances and resources in 207 MAOA, Williams and Fragile X Syndrome research

Chapter 5 Table 1: Disease genes reported in people diagnosed with 232 autism spectrum disorders Table 2: Recurrent genomic disorders/chromosomal abnormalities 232 associated with autism Figure 1: Articles on chromosomal disorders associated with autism 239 and rates of MR and ASD keywords, 1991-2002 Figure 2: Articles on chromosomal disorders associated with autism 240 and rates of ASD and Fragile X keywords, 1991-2012 Figure 3: Papers on chromosomal disorders associated with autism 241 and ASD, Fragile X titles by year, 1982-2012 Table 3: Histories of autism association in 10 genomically designated conditions 243

! iii! !

Acknowledgements

The first note of thanks has to go to all the families, advocates and biomedical experts who have taken the time to talk to a sociologist prying into their research, practice and lives. I look forward to more over the coming years.

The Binational Science Foundation, Mellon Foundation and Columbia’s Lindt and

Lazarsfeld Fellowships, its Institute for Social and Economic Research and Policy and the

Department of Sociology all provided crucial support while I was formulating, researching and writing the dissertation.

A number of scholars at Columbia and elsewhere have taken the time to read and comment on my work and provide sage advice over the last few years. I am especially grateful to Sarah Franklin, Herbert Gans, Dani Lainer-Vos, Shamus Khan, Bill McAllister,

Aaron Panofsky, Nikolas Rose, Chuck Tilly, Stefan Timmermans and Diane Vaughan.

Special thanks are due to John Chalcraft and Laleh Khalili, who generously took a disenchanted philosophy student under their wings in Edinburgh and helped him make the transition to the social sciences and humanities. Brendan Hart, Anne Montgomery, Roz

Redd, Aurora Fredriksen, Mattias Smångs, Hrag Balian, Jen Kondo, Des Fitzgerald, Nancy

Davenport and Onur Ozgode blurred the lines between friend and colleague in a way that made the whole experience more interesting and more fun. Matt Spooner and Dan Herbert were simply incredible friends throughout.

Sara Shostak and Alberto Cambrosio embraced me as a colleague as I took my first steps in a new subfield and wound up being fabulous external committee members. Uri

Shwed has been a wonderful friend, colleague and collaborator over the last few years, and

! iv! !

Chapter 2 of this dissertation can be traced back one of many cherished nights spent drinking beer in his Riverside Drive apartment. I had already completed my graduate coursework when I decided to write my dissertation about genetics, medicine and advocacy, and Nadia Abu El-Haj’s ‘Science and Society’ course was an invaluable introduction to my new field. More importantly, her expertise in science studies and genetics as well as her sage advice about my work as well as some of the trickier decisions of the last few years were invaluable. Peter Bearman helped a recovering philosophy undergraduate to begin to think like a sociologist, if an unconventional one. I try to not hold that against him. In his workshops and in our meetings he always helped me think though innovative research designs, including ones focused on the outliers and oddities, which could get at big questions. He helped me navigate grad school, the transition out of it and made a spirited case for the move towards this dissertation topic, for which I am very grateful.

I owe a truly extraordinary debt of gratitude to my advisor, Gil Eyal. Beginning with the two classes I took in my first year at Columbia, and especially after becoming my advisor when he returned from leave at the start of my third year, Gil has been unwaveringly generous with his time, his feedback on my work, and his advice on matters big and small all along the way. This dissertation would have been far poorer without his input, guidance and encouragement. It would take too many words to convey what a brilliant, wonderfully enabling advisor Gil has been, and so I will simply say that I hope to one-day approach being the kind of mentor he is.

I am grateful to my parents, Larry and Simone, and my brother, Joshua, for their understanding and encouragement throughout this process, and especially to my parents

! v! ! for instilling an enduring curiosity about the world. In particular, Larry read the dissertation after it was defended and caught an embarrassing number of typos that I was able to remove before depositing the thing, and feedback from both him and Joshua have given me confidence this work may be of some interest to non-specialists.

Finally, I can barely begin to express what Claire Edington has meant to me throughout this process. She read drafts of almost every chapter and helped to make each one more elegant and more cogent. Together, we were able to get through the last years of grad school and then write up, circulate and defend within a few days of each other. Most of all, from the time we met in the autumn of 2007 through to today, she has made a better, happier person.

! vi! !

! ! Introduction:!Beyond!Geneticization!

Abnormality at the level of the human genome can mean more than we realized.

This might seem like a surprising claim – after all, the decade since the publication of a draft of the human genome has demonstrated that very few important questions of health and illness find straightforward answers in our DNA (Hayden 2008; Lock 2005; Wade

2009). But what if, instead of helping to explain, predict or treat the conditions we already know and care about, knowledge about genetic mutations was used to carve out new populations of people and therefore novel programs of research, treatment and social mobilization spanning a wide array of fields? This dissertation examines the way genetics can accept its failure to explain existing forms of human difference and go beyond

‘geneticization’ to create new and otherwise unthinkable kinds of people. I want to explain how, for example, the absence of a bright dot on a stained chromosome seen through

Fluorescence in situ hybridization or FISH Test indicating a missing piece of DNA can distinguish between two clinically indistinguishable patients and provide entrée to a complex network of research, care and activism. This is not social science fiction, but a phenomenon with a fifty-year history, and one that helps us understand how, when and why genetics can fundamentally reconfigure biomedical classification, clinical practice and social action.

! 1! !

What does it mean, for example, when a FISH test indicates that someone has a microdeletion of genetic material at site 11.2 on the long arm of the twenty-second chromosome, as seen in Figure 1 below?

Figure 1. FISH test indicating a 22q11.2 microdeletion1 (Yagi et al. 2003:1369)

Today it means they have 22q11.2 Deletion Syndrome, the second most common genetic disorder after Down syndrome (Bassett et al. 2011). It means that they and their family have access to a growing network of biomedical research, clinical treatment, support and advocacy headlined by the International 22q11.2 Deletion Syndrome Foundation. It is a diagnosis that is thought to explain and encompass most of the incidence of older diagnoses like DiGeorge Syndrome and Velo-cardio-facial Syndrome (VCFS) as well as some cases of conditions ranging from schizophrenia and autism to constipation, malar

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 Although new high-throughput technologies are becoming increasingly standard, most of the patients I met during the course of my research were diagnosed using FISH tests that work by binding probes to particular genomic loci, thereby allowing the absence or doubling of a probe on one of a person’s chromosome pairs to indicate a missing or duplicated segment of DNA. The green arrows point to a control site while the yellow arrow points toward the 22q11.2 probe and, because it is missing on the left-most 22nd chromosome, this test is sufficient to diagnose someone with 22q11.2 Deletion Syndrome.

! 2! ! flatness and hypocalcaemia. 22q11.2 Deletion Syndrome (DS) can cause severe congenital heart defects and developmental delay or such a mild phenotype that the patient will not seek medical attention well into adulthood, or perhaps never at all (McDonald-McGinn et al. 2001). Finally, finding the microdeletion at 22q11.2 is increasingly likely to mean that parents face a dilemma about whether to terminate a pregnancy (e.g. Bretelle et al. 2010;

Signature Genomics). Numbers are rising rapidly, mostly in North America and Europe but also in India, Thailand, Chile and elsewhere, and 22q11.2DS is the focus of a growing network of specialist clinics, support groups and advocacy organizations.

22q11.2DS is one of a growing number of conditions, carved up according to genetic abnormalities but lacking the distinctive phenotype of a condition like Down syndrome, that have nevertheless become powerful categories of clinical practice and health advocacy. While statistics are not available, there can be no doubt that many thousands of people have been diagnosed with these novel disease categories and that millions are so diagnosable. But how did this come to pass? And what are the implications of this new way of delineating disease categories for the way we understand the impact of genetics research on medicine and social identity formation and, more generally, the social embeddedness and impact of knowledge production about human difference? This dissertation aims to explain how abnormalities in the human genome can give rise to powerful new categories of bioscientific classification, medical diagnosis, clinical practice, social mobilization and identity formation. Indeed, I show how they can do so even when the people who carry the mutation in question exhibit such wide phenotypic variation that they could never have been delineated as a population, even in principle, on the basis of their physical or psychological characteristics.

! 3! !

I also show how each of those categories does not imply the others: just because a researcher talks about a new ‘syndrome’ in the pages of a prestigious biomedical journal does not mean that it will ever inform clinical practice, never mind collective action and social identity. Rather, I want to explain how the observation of a genetic mutation can almost automatically be cast as a new category or syndrome in the esoteric field of human genetics research, and yet still require very particular conditions and years of painstaking work by a diverse set of actors in order to matter to a general practitioner, a behavioral psychologist, an educator, a biotech company, a concerned parent or the people who are so classified. In other words, I want to explain how the meaning of a genetic abnormality is a sociological phenomenon as well as a biomedical one. While the cases I study have rarely been taken up as objects of social scientific analysis, I will argue that they have important implications for the way we understand the much-discussed relationship between genetics and the social as well as the intersection of the sociology of science and medicine, the study of classification and historical sociology.

The remainder of this introduction will introduce the concept of ‘genomic designation’, outline the core of my argument and preview the chapter-by-chapter organization of the dissertation. A second introduction follows Chapter 2 and outlines the key questions and theoretical framework for Part II of the dissertation in greater detail.

! From geneticization to new kinds of people

The sociology of science and medicine has assumed that genetics has to work in and through existing categories of human difference in order to realize social significance.

For a time, this made sense. The ‘gene-for’ model of associating genetic test results with

! 4! ! diseases and traits was much-hyped in 1991 when Abby Lippmann coined the term geneticization – the idea that many categories of human difference would be “reduced to their DNA codes, with most disorders, behaviors and physiological variations defined, at least in part, as genetic in origin.” (Lippman 1991:19; see also 1991; 1998). The prospect of genetic reductionism raised important questions about medical practice, stigma, the experience of illness, and the capacity to situate health outcomes in social environments.

However, it turns out that genetic reductionism is extremely rare. With a handful of notable exceptions, the characteristics of our genomes do not tend to neatly line up with the medical conditions and traits we care about, and so while geneticization may remain an important discursive phenomenon it is limited by the general failure to reduce human difference to our DNA codes. In response, important sociological work has focused precisely on the failure of the ‘genes-for’ conditions like cystic fibrosis to neatly line up with clinical diagnostic criteria and the liminal cases of people produced by this discrepancy (A. Hedgecoe 1998; A. M. Hedgecoe 2000; Kerr 2000, 2004; Miller et al.

2006; Timmermans and Buchbinder 2011). But still, sociology assumes something like the

‘gene-for’ model, even if our main focus becomes its failure to work out neatly and the socio-technical work done to clean up the mess. As we will see in Chapter 1, even Paul

Rabinow’s enormously influential formulation of ‘biosociality’ only allowed for identity formation on the basis of genetically prescribed risk factors for existing conditions

(Rabinow 1992), while Nikolas Rose’s ‘molecular gaze’ (2007) does not consider the possibility that the contemporary biosciences could provide new bases for the classification of illness. In short, the sociology of science and medicine remains beholden to an outdated

! 5! !

‘gene-for’ model that only lets genetics impact social processes in and through existing categories of human difference.

Genomic designation

Ian Hacking has directed our attention to what he calls ‘kinds of people’ – the categories of persons that interact or ‘loop’ with the people who are so classified in a way that dynamically transforms the categories, expert practices and the people themselves

(Hacking 1998, 2006). Instead of reductionism then, can genetics give rise to what

Hacking would call new kinds of people? Can observations of the genome carve out novel categories of illness and bring about what Foucault (1973: 195) called a ‘syntactical reorganization of disease in which the limits of the visible and the invisible follow a new pattern’?

This dissertation argues that the answer to these questions is ‘yes’. Many biomedical experts, clinicians, patients and health activists have moved beyond geneticization. They have adopted a radical strategy for mobilizing the results of genetic tests that I propose we call ‘genomic designation’. Rather than rely on correlations with existing categories, they discover, delineate and diagnose disease strictly according to observations of the genome, sometimes in the face of enormous phenotypic heterogeneity.

To be clear, genomically designated conditions are not clinically diagnosable, and yet under certain conditions they can give rise to specialist clinics, further research, and social mobilization. Furthermore, they can be used to pry open far-reaching questions about human difference or, in other settings, realign clinical judgment. So here we have the core of this dissertation: what are the origins of this radical strategy for carving up human

! 6! ! difference at the level of the genome, how can it gain traction in medical practice and social settings, and what are its practical and theoretical implications?

It is worth pointing out that one could perhaps say that genomic designation also takes place when it comes to cancers (e.g. Venkitaraman 2002; for a sociological study see

Bourret, Keating, and Cambrosio 2011) and even the delineation of species of bacteria (see e.g. Rowan and Powers 1991; Zeaiter, Liang, and Raoult 2002). However, this dissertation focuses primarily on those genetic mutations that are associated with and in turn come to delineate new categories of developmental difference. The genomic designation of childhood abnormality, I argue, has a coherent history and very particular conditions of possibility. In other words, rather than susceptibilities, new categories of malignant growth or novel species of bacteria, I am interested in the way genetics has served to carve out new kinds of people over the last half-century.

A brief note on methods

Throughout this dissertation I shamelessly select on the dependent variable in the hope of seeing the transformative potential and theoretical implications of genomic designation. The point is not to mislead the reader: I acknowledge throughout that genomic designation remains a mostly marginal way of classifying human difference, though even in the last few years this has become less and less the case. Rather, I conduct mixed methods research on conditions, organizations and people who exemplify genomic designation and its status as a qualitatively novel way of classifying disease, illness and difference. Whether in the choice of conditions, the institutions I discuss or the

! 7! ! professionals I interviewed, the goal was to see something new emerging at the intersection of genetics, medicine and activism rather than to capture the whole field.

My primary method is qualitative comparative-historical research. Using the record of published biomedical research and other publically available resources alongside archival materials from the Wellcome Institute in London, I try to both outline the history of genomic designation as a way of classifying people and to compare outcomes across cases and sociohistorical contexts. As I discuss below and in Part II of the dissertation, I draw on Haydu’s (1998) discussion of how to ‘make use of the past’ in comparative- historical research and outline what I call ‘reiterated facticity’ as an approach that combines elements of Haydu’s reiterated problem solving and science and technology studies. Furthermore, in Chapter 2, I present the results of an ongoing collaboration with

Uri Shwed that uses citation analysis to model the impact of genetics on nosology. Those methods are discussed in detail in Chapter 2, where an initial analysis of the 22q11.2 deletion and the way it reconfigured the previous clinical nosology is presented. Finally, the entire dissertation draws on IRB-approved fieldwork conducted at several conferences and events for genetic disorders, primarily 22q11.2DS, during which I attended presentations, breakout sessions and social events. I also conducted approximately 30 interviews with biomedical researchers, clinicians and advocates and visited Elwyn

Services, Children’s Hospital Philadelphia and a meeting of the U.S. Department of Health and Human Services’ Secretary’s Advisory Committee on Heritable Disorders in

Newborns and Children on newborn screening. This research forms the basis of Chapter 6.

All of these methods are discussed in further detail in the relevant chapters and I plan to pursue all of them, and especially the fieldwork component, over the next few years.

! 8! !

Outline

This dissertation contains six chapters, but also two distinct parts. Part I introduces and defines the concept of genomic designation as a way of classifying disease, and provides a typology of ways in which genetics can reconfigure existing nosology as reported in the biomedical literature. Part II, by contrast, adopts a comparative-historical approach in order to examine the conditions and forms of collective action that can take the genomically designated conditions reported in the literature and turn them into powerful categories of practice spanning various fields of research, treatment, advocacy, commerce and self-practice.

Part I

Chapter 1 provides a working definition as well as a sociological and philosophical discussion of genomic designation as the practice of carving up human difference according to genetic mutations. I draw on the work of Saul Kripke (1980) to argue that this turn to ‘rigidly designate’ medical conditions to observations of the genome represents a qualitative shift in the delineation of disease categories such that it constitutes a fuller

Foucauldian realization of what Rose (2007) has called the ‘molecular gaze’ than has been previously recognized by the social sciences.

I take up the illustrative case of 22q13 Deletion Syndrome, now known as Phelan-

McDermid Syndrome. From the late-1980s several papers by Katy Phelan and others reported cases of a deletion at site q13 on the long arm of the 22nd chromosome (M. C.

Phelan, R C Rogers, and Stevenson 1988; M. C. Phelan et al. 1992; C. Prasad et al. 2000;

! 9! !

Mary C. Phelan et al. 2001). A series of parent meetings were organized and by 2002 the syndrome had been proclaimed and the 22q13 Deletion Syndrome Foundation had been officially formed. Thus in the fifteen years from 1988 the 22q13 microdeletion had gone from a quarter-page case report in The American Journal of Medical Genetics (Phelan et al.

1988) to a vibrant biosocial network. Today, the Phelan-McDermid Syndrome Foundation2 works on multiple fronts to facilitate biomedical research, care and support. All this takes place despite what Katy Phelan herself calls (Mary C. Phelan et al. 2001:98) “the lack of a recognizable phenotype.” In short, Phelan-McDermid/22q13 Deletion Syndrome represents a clear-cut case of genomic designation where characteristics of the genome take nosological precedence over symptoms and the anatomo-clinical observation of bodily systems. It is therefore in cases of genomic designation like 22q13, I argue, that we see Rose’s ‘molecular gaze’ realize the full import of Foucault’s seminal discussion of a

‘clinical gaze’ (1973) in the sense of a new ‘spatialization’ of illness.

In Chapter 2, I argue that the concept of genomic designation is usefully applied to cases that may not be as straightforward as 22q13DS, but where a condition is nevertheless carved out according to observations of the human genome. What do I mean by ‘less straightforward’? A case like 22q13DS allows for the alliterative phrase, ‘discovered, delineated and diagnosed’ strictly according to genomic anomalies, but in Chapter 2 I extend the concept to conditions whose historical origins lie elsewhere (usually in clinical medicine). I develop an ideal typical typology of pathways to genomic designation – i.e. the delineation and diagnosis of conditions according to genetic mutations – that extends to cases where clinical diagnostic criteria are replaced by the single criterion of a necessary !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2 Phelan-McDermid Syndrome Foundation: http://22q13.org/j15/

! 10! ! and sufficient genetic anomaly or test result. Insofar as this results in a re-delineation of the patients population, I argue that it is useful to think of in terms of genomic designation.

Using citation analysis, I examine the counterintuitive case of 22q11.2DS discussed above.

What makes 22q11.2DS counterintuitive in this context is not its broad clinical range but rather the fact that it subsumed several longstanding, clinical conditions like DiGeorge

Syndrome, Velocardiofacial Syndrome and several others. In other words, it is a genomically designated condition that at once absorbed the majority of those older conditions’ cases while also excluding some who did not have the 22q11.2 microdeletion and including many other people who did have the deletion but would not have met the clinical criteria of the older conditions.

I therefore suggest that genomic designation be treated according to an ideal typical typology or spectrum where cases like 22q13DS that emerge with no strong relationship to an existing diagnosis or de novo stand at one end, and rare cases of ‘geneticization’ at the other. In between we have unification or ‘lumping’ as seen in 22q11.2DS, splitting where conditions like autism are carved up into smaller genomically designated disorders, and recalibration where a clinical condition like Williams Syndrome becomes nosologically affixed to a genetic mutation and both excludes and includes significant numbers of patients on that basis. Of course these pathways are ideal types in Weber’s foundational sense of the term – they do not exist in reality but rather serve as useful categories for sociological analysis. What’s more, any given case of genomic designation could be conceived of according to more than one of these pathways depending on the analytic or empirical question at stake: for example we will see that even though 22q13 is a strong case of de novo genomic designation it is now treated as a subtype of autism, while even

! 11! ! the strongest cases of geneticization, like Down syndrome, tend to use genetic testing to arbitrate questionable cases and can therefore be thought of as weak cases of what I am calling recalibration. In sum, Chapter 2 both extends the concept of genomic designation and provides a toolkit for understanding the way that genomic observations can reconfigure clinical and potentially other forms of classification.

! Part%II% However, as a topic in the sociology of science and knowledge it is important to attend to genomic designation, not just as a an esoteric form of biomedical classification, but also as a consequential way of understanding and acting on human difference that is enabled and shaped by particular conditions of possibility. Most of the genomically designated conditions reported in the literature amount to little more than a journal article or two, and yet in recent years more and more have gone on to become bona fide categories of medical practice and social mobilization. In order to examine these divergent outcomes the second half of the dissertation takes the form of a comparative historical analysis of genomic designation over the last fifty-plus years. Drawing on historical research, fieldwork and interviews with biomedical experts, parents and activists, I analyze the processes of network formation and the repertoires of collective action undertaken to generate knowledge, resources and awareness about genomically designated conditions.

Having outlined the concept and nosological varieties of genomic designation in chapters 1 and 2, I provide an introductory section to Part II of the dissertation that lays out a theoretical framework for understanding how genomically designated conditions come to really matter in the world. I set up the study of genomically designated conditions as

! 12! !

‘micro-apparatuses’ in order to 1) draw on Foucault’s conception of an apparatus of knowledge/power and its emphasis on the way a ‘regime of truth’ combines with institutional logics to order diverse elements according to historically given needs, and 2) focus on Fleck’s problematization of specific disease categories or ‘scientific facts’, the way they achieve the status of taken-for-granted truths and how they come to shape practice and lay understanding. Furthermore, I propose a framework for studying genomic designation in comparative terms that nevertheless remains true to the specific historical conditions that shape the scope for genomic designation as well as the contingencies and processes of collective action that shape the way each such condition is understood and acted upon.

I therefore draw on Haydu’s (1998) comparative method of ‘reiterated problem solving’ that adopts a broadly narrative approach to causal explanation supplemented by a comparative analysis of the way problems or crises are dealt with by actors in different historical periods. Such an approach helps draw our attention to the historically specific conditions, contingencies and legacies from previous eras that shape strategies of action and outcomes in different periods. But what if, instead of similar groups of actors confronting what they would recognize as homologous problems (see Haydu, p. 355 and

Part II), we try to study the way that objects of knowledge (in this case genetic mutations) give rise to contrasting forms of research, clinical practice and social action in different historical settings? Alongside other strands of research in the contemporary sociology of science and medicine, I draw on this framework of ‘reiterated facticity’ in order to study the divergent historical conditions and different kinds of research and social mobilization that can turn a genetic mutation into different kinds of scientific facts and eventually even

! 13! ! a robust kind of person. In particular, I will examine the shifting conditions presented by other forms of classification, with a focus on the way the diagnostic expansion of autism has created a fecund new terrain for both researchers and advocates at the intersection of genetic and clinical classification.

Chapter 3 therefore traces the history of genomic designation back to its origins in the late 1950s and, surveying its status over the next few decades, examines its general failure to matter much outside of esoteric human genetics. I explain how the karyotype observations that were used to observe chromosomes and chromosomal abnormalities were standardized and how a first wave of genomically designated syndromes were delineated accordingly in this early period. However, I also show how these first cases of genomic designation attracted minimal interest from researchers in other fields, clinicians seeking better treatment regimes for their patients or advocates who might seek to mobilize those conditions to attract resources or shape self-identity. But why? I argue that the conditions of possibility – institutional, cultural, technological, nosological and so on – for genomic designation to matter outside of human genetics were not yet in place. Furthermore, I discuss the remarkable case of XYY or ‘Supermale’ Syndrome, where the roughly one in a thousand men who have an extra Y chromosome were thought to be prone to criminality, aggression, ‘mental subnormality’, increased stature, acne and so one, and argue that it represents the exception that proves the rule. Finally, I gesture towards the remaining chapters by showing how several of the genomically designated conditions discovered in the early 1960s have been taken up as objects of clinical practice and social mobilization in the contemporary period. They have done so in the context of new conditions of possibility that include, among other factors, the new repertoires of collective action and collaboration

! 14! ! that see genomic designation enjoying a kind of resurgence today. Indeed I show how even

XYY Syndrome has reemerged amidst this new wave of genomically designated syndromes, but with a clinical profile that is both dramatically different and suitable for the times. In so doing, I set up an analysis that treats genetic mutations, in part, as a sociological phenomenon and seeks to explain how they come to be mobilized as meaningful categories through which to understand and act on human difference.

In Chapter 4 I discuss the contrasting ways in which genomically designated conditions attract enormous levels of attention from biomedical researchers interested in more general categories of human difference. Because they are fixed according to specific genetic anomalies, genomically designated conditions can serve as biological models for the diseases and traits with which they overlap – an opportunity to start with a known genetic etiology and develop the gene-protein-physiology-phenotype/disease pathways that might shed light on otherwise intractable questions. I refer to this increasingly common strategy in human genetics as ‘leveraging genomic abnormality’, and discuss how it can bring otherwise unimaginable levels of funding and interest to genomically designated conditions. Crucially, I show how genetic mutations can attract such attention outside of bioscientific research and also how advocates concerned with genetic disorders can in turn leverage interest in their condition to advance their own distinct ends. I therefore discuss the way genetic anomalies often come to serve as ‘boundary objects’ that facilitate convergence and collaboration on the part of diverse scientific disciplines and other actors.

I draw on three case studies: first, the use of Brunner Syndrome and the MAOA-L gene variant to understand variable proclivities for aggression; second, the use of Williams

Syndrome and the 7q11.2 microdeletion to study language, musicality and sociability;

! 15! ! finally, the way Fragile X Syndrome is used as a genetic model for autistic spectrum disorders. Situating this leveraging strategy within the longstanding use of abnormal people to examine more general biological phenomena helps us understand how patient advocacy organizations and commercial biotechnology concerns have transformed the scope for mobilizing abnormality in ‘postgenomic’ research. However, I argue that by asking the question ‘boundary objects for whom?’ we can analyze the divergent biosocial circuits that shape genetics research and its implications. The extraordinary success of the

Fragile X Syndrome/autism interface indicates that the most successful projects based on leveraging genomic abnormality today are likely to be ones where the genetic mutation in question is mobilized to coordinate a wide variety of actors’ interests and speak to salient categories of illness that can attract state and private capital investment.

Chapter 5 develops the analysis of the way genomically designated syndromes like

Fragile X intersect with autism research and advocacy. However, it also extends the story both backwards and forwards. My starting point is the threefold role of diagnostic categories as coordinating devices, identities and sites of looping processes. While this might not be very controversial as a framework in the sociology of science and medicine, in this case at least it allows us to advance novel findings. Looking backwards to the conditions of the intersection, I show how the ‘geneticization’ of autism contributed to its diagnostic expansion because 1) it discursively helped to destigmatize autism by emphasizing that it is a genetic rather than a psychogenic disorder, and 2) the evidence from both twin studies and cohorts of people with genetic disorders pointed towards a broader autism phenotype. I also argue that the high degree of overlap seen today between genetic disorders and autism is contingent upon the latter’s diagnostic expansion. Indeed,

! 16! ! by using genomically designated conditions as a strategic research site I argue that we can see changes in diagnostic practice transforming the genetic makeup of the autism population. Genomically designated conditions are, after all, tautologically coextensive with genetic mutations, and therefore significantly increased rates of autism in probands, often from zero to very high, represents the incorporation of new mutations into the ranks of that autism population. Autism’s widely recognized genetic heterogeneity, I argue, is partly the unintended consequence of its geneticization and subsequent diagnostic expansion. Finally, I discuss the way that cooperation between autism researchers and advocates and their counterparts interested in genomically designated conditions works in practice. While some have argued that the genetic classification of disease should replace existing nosologies, and others insist that knowledge about genetics must be translated in order to speak to clinical categories, I show how the intersection of genetic and psychiatric classification is negotiated in practice. I find that, rather than incommensurability and conflict, diverse actors are sometimes able to establish what Gallison described as ‘trading zones’ where actors exchange resources and objects of knowledge even in the absence of shared goals or frameworks of understandings. This chapter therefore demonstrates how the conditions for genomic designation are decisively shaped by the shifting distributions of illness with which they can hope to interface.

Finally, in Chapter 6 arrives at an analysis of genomic designation today. Drawing largely on interviews and fieldwork with 22q11.2DS researchers, clinicians, advocates and parents, I examine how recent decades have seen some genetic mutations go from objects of esoteric research to kinds of people in the full sense of Hacking’s use of the term. I argue that, in order to really matter in the world, heterogeneous networks need to be

! 17! ! assembled around genomically designated conditions based on their ability to interface with other forms of classification (clinical, psychiatric, educational etc.) and translate the interests of actors from fields that are far removed from genetics, including patients and their advocates. As I show with 22q11.2DS, these networks can work to increase and broaden ascertainment, realign clinical judgment, redirect the provision of care and provide a new basis for identity formation and understanding difference. What’s more, as the network grows in symbolic and material resources it becomes ever easier for new patients and their families to organize and interpret their challenges and treatment regimes under the rubric of a genomically designated syndrome. My analysis of the way 22q11.2DS is being assembled as a new kind of person, alongside an overview of the changing historical conditions and other cases of genomic designation, allows us to see how genetic mutations have been established and constantly reiterated as facts over the last fifty years.

By way of conclusion, I take stock of the progress made by genomically designated conditions in recent decades and consider the likely trajectories and implications of genomic designation moving forward. For while it may seem as though the dissertation paints an unwavering arc towards the formation of networks of care and support for people with certain genomically designated conditions, there is no teleology that makes their newfound status a secure one and others certain to follow in their footsteps. Looming on the horizon, I argue, is a new and unprecedented wave of prenatal genetic testing that has the potential to turn the knowledge created by those very networks into a contributory factor in the steep reduction of genomically designated populations of people.

! 18! !

! Conclusion! We are on the cusp of an enormous proliferation of data about our genomes. In

2001, it cost 100 million dollars to sequence the three billion base codes of an entire human genome; in late 2007 it was 10 million; today it can be done for less than ten thousand dollars.3 Researchers are running whole genome and exome tests on large cohorts and finding a far greater range and volume of abnormalities than expected, even in seemingly normal people (e.g. The 1000 Genomes Project Consortium 2010), and there is a strong consensus that mass consumption whole genome tests are not far away. But with the limited success of the ‘gene-for’ model, what will we do with all this knowledge about our surprisingly anomaly-ridden genomes?

It remains unclear whether genomic designation will fade into obscurity, lead to improved health outcomes for millions or become what we might call, borrowing from

Troy Duster (2003), a ‘trapdoor to eugenics’. While I resist the temptation to indulge in futurology or get caught up in a discussion of the potential ethical implications of genomic designation throughout most of the dissertation, in the Conclusion I address a looming possible future where prenatal ascertainment of genomically designated conditions increases by orders of magnitude, creating vexed dilemmas for would-be parents and eugenic outcomes at the population level. For as much as conditions like 22q11.2DS and

Fragile X have made enormous strides as networks of research, treatment and advocacy in recent decades, they still remain dependent upon broader conditions whose impact they can only hope to partly control.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 3 For NIH data see: http://www.genome.gov/sequencingcosts/

! 19! !

In sum, we will see how the sociology of science and medicine needs to go beyond geneticization and think about the emergence of new kinds of people, forged at the interface of biomedicine and collective action, carved out at the level of the human genome.

Just as geneticists leverage genomic designation to get at more general questions, as sociologists we can use it as a vista on the novel biosocial intersections that are reshaping what it means to be abnormal. We need to analyze the way knowledge about abnormality is decisively shaped by social processes and institutions, even when the mutation or biological lesion that designates the category of abnormality is held constant. Not only does it take a heterogeneous network of actors and particular conditions of possibility to make knowledge about a genetic mutation matter to lived experience, but that network and those conditions also shape the kinds of research and the very clinical profiles that come to be associated with geomically designated syndromes. This dissertation therefore examines the history and present of those medical conditions delineated by human genetics in the half-century since it became possible to see and count our chromosomes. In so doing, we will come to understand how a missing piece of DNA can mean so much.

! ! !

! 20! !

Chapter 1: From mutations to medical conditions1

…this order of the solid, visible body is only one way – in all likelihood neither the first, nor the most fundamental – in which one spatializes disease. There have been, and will be, other distributions of illness.

– Michel Foucault, The Birth of the Clinic (1973, p. 3)

From the clinical standpoint, [22q13 deletion] syndrome may be underdiagnosed because of the lack of a recognizable phenotype that would lead the clinician to request studies to rule out this specific chromosome 22 deletion.

– Phelan et al., American Journal of Medical Genetics (2001, p. 98)

For almost fifty years, but with increasing salience, genetics experts have delineated kinds of people whose members are united, not by observations of their condition or predictions of their fate, but by observations made at the level of the genome.

This chapter introduces this biomedical practice – the ‘genomic designation’ of medical categories – and provides a working definition to guide the sociological analysis that follows.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 Most of this chapter is based on a previously published paper (Navon 2011). However it also revises several of that paper’s central claims and removes most of its preliminary historical analysis and the discussion of the relationship between genomically designated syndromes and autism, taking them up in a different form and in much greater detail in Chapters 3-6.

21! !

What does it mean to be a bearer of a genetic marker or mutation? As the sociological literature on genetics makes clear, it can mean many different things (e.g.

Callon and V. Rabeharisoa 2003; El-Haj 2007; Lippmann 1991; F. A. Miller et al. 2005;

Rabinow 1992). For people with a microdeletion of genetic material at site 13.3 on the long arm of the twenty-second chromosome, it means something quite unlike anything recognized in that literature: it means they have 22q13 Deletion Syndrome (22q13DS), now known as Phelan-McDermid Syndrome (PMS). PMS is a case of genomic designation because, unlike Down Syndrome which has been synonymous with chromosome 21 trisomy since 1959, or genomic markers which have shown people to be at greater risk of acquiring conditions like breast cancer or muscular dystrophy, it could only be discovered and made practicable after the identification of the 22q13.3 microdeletion. Genomic designation entails the delineation and diagnosis of medical conditions on the basis of observations made at the level of the genome2, be it whole chromosome aneuploidies observed through karyotype analysis or tiny mutations detected by contemporary genomics technologies. When the discovery of a genetic mutation or marker does not line up with an existing diagnostic category or phenotype and instead comes to designate a kind of person that had never existed before, and may have lacked the phenotypic coherence to have been recognizable, even in principle, we have a case of genomic designation.

There is something remarkable, even perplexing about the discovery and delineation of clinical entities like 22q13DS: the microdeletion was the subject of focused scientific investigation by 1988, some thirteen years before the syndrome named after it !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2 By ‘genomic’ I do not mean observations particular to the field of genomics, but observation of the characteristics of the genetic material contained in our chromosomes that constitutes the human genome. Genomic designation could and did occur through karyotype analysis (below).

22! ! was proclaimed and the 22q13 Deletion Syndrome Foundation was formed. Today, 22q13

(pronounced 22-q-1-3) Deletion Syndrome and the renamed Phelan-McDermid Syndrome

Foundation constitute a powerful network of clinical and social-identity practices whose members and phenotypic properties are assembled on the basis of genomic diagnosis. By instantiating this novel sequence of genotype-phenotype association, genomically designated syndromes like PMS represent a new way of categorizing human difference.

To date, social scientific literatures have focused on social action undertaken on the basis of genetically determined risk (Callon and V. Rabeharisoa 2003; Hacking 2006a;

Rabinow 1992; N. S. Rose 2007) or genetic correlations with racial and ethnic identities

(Duster 2003; El-Haj 2007; Nelson 2008; Fullwiley 2007). In PMS and its analogs we see the emergence of medical and social identities on the basis of genomic observations, not as a proxy in the form of a predisposition or an essentializing explanation, but as the core referent of a ‘kind of person’ or ‘human kind’ – Ian Hacking’s terms for the categories delineated and studied by the ‘human sciences’, and thereby subject to processes that mutually transform people and knowledge of them (Hacking 1995, 1998, 2006b). While this form of practice establishes an unusual level of certainty at the level of genetic etiology, it does so at the expense of phenotypic coherence; in most cases it is unlikely that diagnosis could, even in principle, take place on a clinical basis. Thus genomic designation subverts the ‘clinical gaze’ (Foucault 1973; N. Rose 1998) – the nosological relationship between the observed and the unobserved that privileges localized anatomo-clinical observation – which has dominated modern nosology since the nineteenth century. The sociological interest in genomic designation is not therefore limited to the study of just a bunch of new disease entities or ‘human kinds’, but rather extends to the emergence of a

23! ! new kind of human kind and the way it reconfigures the way we understand and act on human difference.

Although barely recognized in the biological or social-scientific literatures, the practice of genomic designation is now over half a century old. As we will see in greater detail in Chapter 3, it can be traced to the immediate aftermath of Lejeune et al.’s landmark

1959 discovery that trisomy of the 21st chromosome accounted for Mongolian Idiocy (now

Down Syndrome), which powerfully illustrated that a characteristic of the human genome could account for a kind of person (Lejeune, Turpin, and Gautier 1959; see Harper 2006).

By the end of 1959, two logics for bridging the gap between genotype and phenotype had become available. On the one hand, genetic analysis works according to an easily grasped logical sequence: take some phenotypic state like Down Syndrome, breast cancer, manic depression and so on and look for genomic correlates that might help account for some or all of their incidence. On the other hand, genomic designation inverts this logical relationship by taking a characteristic of the genome – ranging from aneuploidies to tiny microdeletions – and looking for phenotypic correlates that occur in some or all cases.3

Though marginal and largely unrecognized until recent years, its logic has been with us since the study of our chromosomes got underway (Edwards et al. 1960; P. Jacobs, Baikie, et al. 1959; below). What’s more, we will see that genomic designation is increasingly

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 3 As we will see, ascertainment bias means that certain populations –particularly people with developmental differences and congenital abnormalities – are far more likely to be given genetic tests. Similarly, there is a strong empirical basis for the idea that gross autosomal genetic mutations tend to lead to intellectual impairment, however variable. While this means that an overwhelming majority of people diagnosed with any given genomically designated condition might have, for example, developmental delay, the symptoms and traits that lead to referral for genetic testing remain wholly insufficient for diagnosis with a genomically designated disorder. Still, the often- complex profiles that are assembled around genomically designated conditions are indeed tabulated post-factum even if certain signs come to serve as pathways to testing.

24! ! informing diagnostic, clinical and social practice in our ‘postgenomic’ era, “saturated by genetic discourses” of the self (Taussig, Rapp, and Heath 2003), in which the life sciences have continued to make great technical strides in classifying genomes even as they have disappointed those who sought straightforward genetic correlates for complex phenotypic conditions (Lock 2005).

Social science approaches to genetics and medical classification

Where does genomic designation fit in to the existing sociological literature on genetics and medicine? While that question will be taken up, in various ways, throughout the dissertation, for now I want to situate genomic designation with respect to work that adopts a broad perspective on the relationship between genetics, classification and human difference. Social scientists are increasingly finding fecund terrain at the intersection of the genetic and the social, with many studies attempting to chart a productive middle-ground between genetic essentialism and unbridled criticism of genetic explanations of human difference (for a pertinent special issue, see Bearman 2008). Sociological explorations of genetic/evolutionary imbrications with the social have been long advocated by scholars like E. O. Wilson (1980) and, more recently, Jeremy Freese (2008). Others such as Troy

Duster (2003), Nikolas Rose (2007), Nadia Abu El Haj (2007), Sarah Franklin (2003) and

Michel Callon and Vololona Rabeharisoa (2003) have critically explored the societal implications, changing understandings of race, ethnicity and illness, and forms of identity inclusion/exclusion arising out of contemporary genetic knowledges and practices on the part of both experts and activists (for reviews, see Freese and Shostak 2009; Fujimura,

Duster, and Rajagopalan 2008). Similarly, Heath et al. (2004) have explored to the

25! ! complex networks of actors – spanning experts, activists, patients, politicians and businesspersons – and argued that ‘genetic citizenship’ has emerged. Elsewhere, the same authors (Taussig et al. 2003:59) have attended to the “forms of embodiment and subjectivity emerging from relations between biomedical experts and lay health advocates in an era when genetic explanations … appear to be proliferating throughout U.S. public culture.” Later in the dissertation we will deal with a number of the issues raised in this literature, but for the initial conceptual task of defining and situating genomic designation it is important to bore down into the specific literature on human genetics and medical classification.

As I discuss in more detail in Chapter 2, Miller et al. (2005) found that the identification of single gene mutations associated with Huntington’s disease, Cystics

Fibrosis and tuberous sclerosis had very different nosological implications in each case, sometimes being used to ‘rule in’ or ‘rule out’ borderline cases of a diagnosis and sometimes being subordinated to clinical judgment. More recently Timmermans and

Buchbinder (2010, 2012) found that newborn testing for rare genetic disorders produced an indeterminate class of ‘patients in waiting’ who either had the marker but not the characteristic symptoms or vice versa. As a result, experts were compelled to engage in

‘bridging work’ in order to reconcile this unexpected lack of alignment between biomarker and the symptomatology of the medical condition it was supposed to indicate.

Furthermore, Bourret, Keating and Cambrosio (2011) have examined the way knowledge about genomic markers is produced through extended collaborative networks or ‘clinical collectives’ and can impact classification and clinical judgment in oncology. Finally, as I discuss in greater detail in Chapter 6, Rabeharisoa and Bourret (2009) outline the

26! !

‘bioclinical collectives’ composed of various experts working at the intersection of biological/genetic and clinical practice in cancer and psychiatric genetics. By

“simultaneously producing the clinical relevance and the biological significance of mutations” (p. 693) contemporary genetic-clinical practice realizes neither the reductionism of geneticization nor the irrelevance to medical practice that some critics suggest. Rather, bioclinical collectives continually debate evidence and the “models that give meaning to mutations and to their complex relations with heterogeneous elements that may be involved in the development of different kinds of known and unknown diseases or syndromes.” Thus in reporting what they call the ‘clinic of mutations’, Rabeharisoa and

Bourret recognize something deeply akin to genomic designation.4 To their work this dissertation adds an analysis of the way those multivalent genetic mutations can become powerful bases for the formation of heterogeneous networks composed of advocates, researchers, clinicians, various protocols and technologies, commercial enterprises and others that transcend any given setting. They can therefore take on a far broader medical and social significance that actually embraces their phenotypic uncertainty.

Genomic designation therefore challenges and extends two key concepts in social scientific studies of genetics: ‘biosociality’ and ‘geneticization’. As we saw in the introduction, Lippman’s introduction of the concept of ‘geneticization’ helped spark an explosion of interest in the “ongoing process by which differences between individuals are reduced to their DNA codes, with most disorders, behaviors and psychological variations

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 4 Furthermore, Rabeharisoa and Bourret’s research on genetics in clinical work pertaining to autism is of particular relevance to this paper, even going so far as to mention that ‘The tools of molecular biology have also led to other discoveries, and specialists now believe that many chromosomal or genetic syndromes could be associated with autism’ (2009: 698)

27! ! defined, at least in part, as genetic in origin” (1991:19). Particular attention has been paid the way in which this process of geneticization essentializes differences between persons, especially those that mark ‘abnormalities’, and tends actually to increase the stigma directed towards those with psychological disorders (see J. C. Phelan 2005; also Hedgecoe

2001 on “enlightened geneticization”). On the one hand, genomic designation can be thought of as an instantiation of geneticization in that it entails understanding human difference according genetics. On the other hand, as previously discussed, it does not simply reduce existing categories to genetic correlates but actually produces new categories of disease and difference.

Paul Rabinow’s (1992) discussion of biosociality was a formative moment in the history of social scientific engagements with genomics (Hacking 2006a). Rabinow explicated biosociality with respect to genomics on two levels. First, there was the potential for intervention at the level of DNA in pursuit of socially derived ends. Second, and crucially for the present discussion, was social action and network formation on the basis of a perceived genetic predisposition for or risk of developing some phenotypically delineated pathological condition. Rabinow (1992:244) noted, “it is not hard to imagine groups formed around the chromosome 17, locus, 16,256, site 654,376 allele variant with a guanine substitution” while maintaining that the role of the genomic observations would remain strictly limited to that of risk ascription as such observations “carry with [them] no depth … it has no meaning.” I will argue that cases like 22q13.3 deletion allow us to develop the concept of biosociality by moving beyond Rabinow’s framework.

Observations of the genome can acquire meaning through the networks of knowledge production and social mobilization assembled on the basis of genomically designated

28! ! syndromes. Furthermore, I will show how these networks shape the course of biomedical research and the knowledge produced about genetic mutations, rather just being set in motion by them.

Works like these help bring into view the complex, burgeoning field within which a condition like 22q13DS realizes its significance. However, none of their valuable social- scientific tools allow us to fully grapple with the emerging biomedical and sociocultural capacity to delineate and make practicable new categories of persons according to properties of their genomes, though they are invaluable to the project of studying their concomitant networks of scientific research and social action. When genetic mutations such as the microdeletion at 22q13.3 come to serve as more than genomic correlates or explanans for disease categories like autism, and instead come to serve as the essential referent of new medical conditions Phelan-McDermid Syndrome, we begin to see geneticization and biosociality writ large.

The ‘molecular gaze’

Genomic designation also represents perhaps the clearest realization of what

Nikolas Rose (2007; 2007) has called the ‘molecular gaze’: the transformation of the ‘style of thought’ (Fleck, 1979, cited in Rose, 2007a: 12) instantiated by medical practice through the incorporation of an array of techniques aimed at understanding the body at the molecular level. By turning to the molecular – the complex processes whereby the sequence of some 3 billion polynucleotide base pairs that make up our genomes leads to phenotypical expression – contemporary medicine has begun to supplement the ‘clinical gaze’ documented by Foucault (1973) in The Birth of the Clinic. Rose has noted how this

29! ! molecular gaze has subtly destabilized the clinical nosology outlined by Foucault whereby localized, observable manifestations of disease, both in symptomologies and anatomo- clinical analysis, serve as the basis for disease classification (Rose, 2007: 11-15). As a result, previously indistinguishable forms of breast cancer have been differentiated on genetic grounds and disjunct diseases have been found to share a genetic etiology, leading researchers to draw distinctions and see underlying unities that would not have been possible on a clinical basis.

Furthermore, the biological sciences are beginning to turn to the dynamical interactions between genotype, phenotype, environment and other mediating levels of analysis. Loscalzo et al. (2007:1) directly challenge the “Contemporary classification of human disease [which] dates to the late 19th century, and derives from observational correlation between pathological analysis and clinical syndromes.” The stakes are nothing less than “redefining human disease in this postgenomic era…” Under this new approach indistinguishable phenotypes can be treated as distinct phenomena, and divergent ones as deeply akin (see also Goh et al. 2007). In sum, the hegemony of anatomo-clinical classification is facing a serious assault from researchers armed with new genetics technologies.

Perhaps then, even without looking to genomic designation, we can begin to discern a nascent “syntactical reorganization of disease in which the limits of the visible and the invisible follow a new pattern,” (Foucault, 1973: 195) that constitutes an important amendment to modern nosology. That said, no matter what genomic differentiation and unification occurs, which genetic explanations are invoked and which molecular-level treatments are developed, the units of analysis discussed by Rose and Loscalzo et al.

30! !

(above) remain phenotypically delineated. It is genomic designation – where the very object of analysis, the medical and social category, is delineated on a genomic basis, while the phenotype is decentralized and tabulated post-factum – which can be said to instantiate a bona fide molecular gaze in the Foucauldian sense, just as it represents a new basis for biosociality. Rather than localized anatomo-clinical observation (Foucault 1973, pp. 3-4), in genomic designation disease is classified according to genetic anomalies found in every cell of the body. Ludwig Fleck, whose framework for understanding scientific facts like particular disease categories will be taken up later on, presciently explained all the way back in 1935 (Fleck 1981:21) how “the modern concept of disease entity, for example, is an outcome of [historical] development and by no means the only logical possibility. As history shows, it is feasible to introduce completely different classifications of diseases.”

Genomic designation is just such an alternative classification of disease, though my discussion of it will be geared towards its capacity to interface with, rather than supplant, prevailing clinical and other forms of classification. As Foucault (1973: 3) himself put it,

“this order of the solid, visible body is only one way – in all likelihood neither the first, nor the most fundamental – in which one spatializes disease. There have been, and will be, other distributions of illness.”

Genomic designation: Definition and origins

Because the genomic designation of syndromes is not properly recognized as a distinct form of practice, and must be identified on a case-by-case basis, ascertaining its scope and scale is extremely difficult. My operational definition is straightforward.

Genomically designated syndromes, ideal typically, fulfill the following conditions:

31! !

1. Reports of cases of a particular characteristic of the genome and its association

with some phenotypic properties occur prior to or simultaneous with the initial

delineation of the associated syndrome.

2. Observation of a specific genomic abnormality is necessary and sufficient for

diagnosis, whether as visible characteristics of chromosomes using karyotype or

FISH analysis or through automated systems like comparative genomic

hybridization (CGH) arrays that can scan the entire genome and produce

increasingly fine-grained results.

3. The phenotypic properties of the syndrome are tabulated post factum to the

identification of the genetic marker or mutation and on the basis of those diagnosed

through genomic testing.

4. The postulated syndrome gives rise to clinical guidelines and/or specialist

centers as well as advocacy organizations and/or support groups, making it a bona

fide medical and social category.

This ideal type characterizes at least 22 syndromes (see Table 2 below), ranging in size from a few hundred to many thousands of subjects. Tracing their history through the biomedical literature and online associational activity confirms all four conditions.

Countless other syndromes do not fulfill all conditions perfectly, as we will see below with respect to16p11.2 deletion and Williams Syndromes, but are useful to think of in terms of genomic designation: once we recognize its clearest cases, genomic designation becomes useful as a broader conceptual framework for investigating the impact of contemporary

32! ! genetics research. A typology of genomically designated conditions will be outlined in

Chapter 2.

Down Syndrome serves as a useful point of reference and as a precursor to genomic designation’s emergence as a way of delineating medical categories. Down

Syndrome is often diagnosed through karyotype analysis – the visual inspection of the chromosomes of a cell, often extracted prenatally through amniocentesis – and underwrites a substantial social and clinical network. However, the Down Syndrome phenotype was first characterized on the basis of hospital observations in England by Dr. John Langdon

Down in publications in 1862 and, more prominently, in 1866.5 The latter, ‘Observations on an ethnic classification of idiots’ (Down 1866), described Down’s clinical observations and delineation of ‘Mongolian idiocy’ and sought to establish that among the “large number of idiots and imbeciles” observed by Down, “a considerable portion can be fairly referred to one of the great divisions of the human race other than the class from which they have sprung.” Down had discovered and delineated a syndrome on the basis of a clinically observed phenotype even as he posited a genetic explanation for ‘Mongolian

Idiocy’ at the level of heredity.

Perhaps the idea that ‘Mongolian Idiocy’ represented some kind of error of inheritance provided the impetus for Lejeune and his colleagues to look at people with

Down Syndrome as they sought to establish cytogenetics’ relevance to medicine, and indeed we will see in Chapter 3 that a chromosomal explanation for Mongolian Idiocy had been advanced years earlier. Still, even though Down Syndrome is now identified with !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 5 Earlier descriptions of what was probably the same phenotype were made by Jean Etienne- Dominique Esquirol in 1838 and Edouard Séguin in 1846 (see Roubertoux and Kerdelhué 2006), but lacked the diagnostic specificity, influence and longevity of Down’s work.

33! ! trisomy of the 21st chromosome, it is clearly not very useful to think about it in terms of genomic designation: a characteristic of the genome was not necessary for its discovery and delineation as disease category and it is not necessary for its diagnosis today. As a review of the early cytogenetic findings in The Journal of Pediatrics put it at the time:

“Exceptions to the association of an extra chromosome 21 and Mongolism have been recorded, indicating that from a cytogenetic point of view, there are at least two forms of

Mongolism.” (Rappoport and Kaplan 1961:428; see also Polani et al. 1960)6 Down had employed the clinical gaze, not genomic designation, and Mongolism continued to be delineated and diagnosed on the basis of its phenotype even once the association with trisomy of the 21st chromosome had been established.

The identification of Down Syndrome with chromosome 21 trisomy using karyotype analysis was nevertheless a major achievement in cytogenetic analysis and made it clear that an observation at the level of the chromosome (more fine-grained observations were not yet possible) could be correlated directly with a human kind. Within months,

Turner Syndrome (Ford et al. 1959) and Klinefelter Syndrome (P A Jacobs and Strong

1959) were associated with missing or extra chromosomes. Crucially, further aneuploidies were quickly found in people who were not diagnosable with a shared clinical condition.

These abnormal chromosome complements were therefore used to delineate new

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 6 However it should be noted that Polani et al. (1960) and other similar findings of a Down syndrome phenotype without 21st chromosome trisomy immediately began to postulate and search for other aberrations related to the 21st chromosome (e.g. partial duplication or translocation) that could account for the phenotype. Indeed separate studies were conducted on two populations who had already been distinguished by clinicians: those with classic mongolism and older mothers, and those with a somewhat different profile and young mothers. It was claimed that these corresponded to full 21 chromosome trisomy and 21 chromosome transolcations respectively (see Harper 2006:68)

34! ! syndromes: first came ‘super female’ or XXX Syndrome (P. Jacobs, Baikie, et al. 1959), closely followed by 18 trisomy or Edwards Syndrome (Edwards et al. 1960) and 13 trisomy or Patau Syndrome7 (Patau et al. 1960) appearing as back-to-back papers in the same issue of the Lancet in April 1960. In other words, as soon as cytogenetic techniques became capable of drawing distinctions among human genomes, nosology became subject to the molecular as well as the anatomo-clinical differentiation of persons. Further such syndromes would be delineated over the coming years, including ones fixed to chromosomal abnormalities much smaller than full aneuploidies like the 18p and 5p minus syndrome (De Grouchy et al. 1963; De Grouchy, Rossier, and Joab 1967; Lejeune et al.

1963). However, in its first decades genomic designation had a halting, uneven, and comparably marginal impact on clinical practice and social action – a history that will be taken up in subsequent chapters. For now, let’s fast-forward to a recent, clear-cut case in order to clarify what makes genomic designation a distinct form of classification.

22q13 Deletion Syndrome

On the end of the long arm of the twenty-second chromosome, at locus q13.3 where the SHANK3 gene rests, a microdeletion of between 130,000 and 9 million base pairs has been detected in no more than a few thousand people worldwide, though its true incidence is thought to be at least 1 in 20,000. It is usually detected using fluorescent in situ hybridization or FISH analysis, which attaches visible probes that bind to DNA

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 7 While it has been claimed that 13 trisomy is coextensive with a clinical phenotype first described by Danish physician Thomas Bartholin in 1657 (see M. Warburg and Mikkelsen 1963 for what I believe is the first claim to this effect) and described clinically by Warburg (1960), it seems quite clear that Patau et al.’s discovery did lead to the delineation of a new medical condition.

35! ! sequences at specific sites on different chromosomes in order to see if there are missing segments of DNA. The resulting haploinsufficiency, when the deletion includes the

SHANK3 gene, is associated with a complex set of physical and psychic symptoms from autistic behavior and delayed speech to fleshy hands and poorly formed ears (Manning et al. 2004; M. Phelan 2008).

Initial studies reported the phenotype associated with a deletion at 22q13 (see

Gustavson et al. 1986; Hinkel et al. 1997; Phelan et al. 1992; Phelan, Rogers, and

Stevenson 1988; Schmitt et al. 1994; Wong et al. 1995), and by 1997 a paper that included

David Ledbetter and Helen McDermid as coauthors was specifically discussing the

‘22q13.3 deletion syndrome’ (Wong et al. 1997). In 2000, Prasad et al. hypothesized that the 22q13 deletion syndrome “may represent a recognizable phenotype” and in 2001

22q13 Deletion Syndrome was delineated in a paper by Phelan et al. (M. C. Phelan et al.

2001; C. Prasad et al. 2000). The 22q13 Deletion Foundation was founded shortly thereafter. Today, the renamed Phelan-McDermid Syndrome Foundation includes over 900 families worldwide8, holds an annual conference that brings together many of the members, hosts support groups, scientific conferences, regional and international subgroups and parent forums, sells merchandise, and coordinates with the relevant scientists and clinicians.9 They are hiring staff and increasingly working to facilitate and redirect biomedical research on 22q13. In Chapter 5, we will see how the foundation is thriving like never before as a genetic disorder associated with autism (Cohen et al. 2005;

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 8 See http://22q13.org/j15/index.php?option=com_content&view=article&id=277:2013-01-01- newsletter&catid=126:2013-03-30-23-41-36&Itemid=243, accessed 4.30.2013.

9 See http://22q13.org/j15/, access 4.30.2013.

36! !

Goizet et al. 2000; Schaefer and Mendelsohn 2008). The history of 22q13DS exemplifies the typical pattern of a genomically designated syndrome: it clearly became the focus of significant clinical and social action well after the initial studies into the genetic mutation upon which it is based. All of this takes place even as the core scientists note that “[f]rom the clinical standpoint, this syndrome may be underdiagnosed because of the lack of a recognizable phenotype” (M. C. Phelan et al. 2001:98)

Prior to her delineation of 22q13DS in a 2001 paper, Mary C. Phelan had been conducting scientific and associational work on the 22q13 deletion. She published one of the first studies on the deletion, ‘a de novo terminal deletion of 22q’ (Phelan et al. 1988), in a quarter page report in the American Journal of Human Genetics that reported

“physical features [including] a central notch in the maxillary alveolar ridge, an abnormal palmer flexion crease on the left, four digital arches, minimal proximal cutaneous syndactyly of toes 2 and 3 bilaterally, and generalized hypotonia” (p. A118). More papers would follow by Phelan and others.

Crucially, the Phelan-McDermid Syndrome Foundation10 notes that Phelan was the

“telephone buddy” to the 15 families who formed the core of the subsequent group and foundation. What is more, semi-annual conferences have been held since 1998, when 20 of the 23 families with children known at that time to have the deletion met in Greenville SC near the genetics research start-up where Phelan worked. At the third conference in 2002

(immediately after the syndrome was delineated in the 2001 paper) the families founded the 22q13 Deletion Syndrome Foundation, having made extensive organizational efforts in

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 10 See http://22q13.org/j15/index.php?option=com_content&view=article&id=97&Itemid=113, accessed 4.30.2013.

37! ! the preceding years. The executive board met in 2003, agreeing to add the name Phelan-

McDermid or PMS, mainly because it made the syndrome easier to relate to outsiders.

That is, when the syndrome was delineated in Phelan and her colleagues’ 2001 paper a core network of families already was in place, and these families moved almost immediately to establish a foundation and a social network on the basis of the 22q13 microdeletion.

The support group now says fascinating things about phenotypic characteristics, in stark contrast to the very limited phenotype reported by Phelan et al. in 1988. For example, the foundation’s website gives the following account of the behavioral characteristics of

PMS:

There is not as much specific data on the behavioral aspects of individuals with deletion 22q13 as there is in other areas. Many of the behaviors listed below were brought up in parent sharing sessions. One parent would explain a particular behavior and many others would say they have observed it with their child too. Due to this, there is only one behavior with a known percentage showing what portion of the deletion population exhibit it.

Some behavioral characteristics:

• Chewing on non food items (clothing, bedding, toys) - 70% • Teeth grinding • Tongue thrusting • Hair pulling • Aversion to clothes Dr. Desmond Kelly compiled this list of attributes also saying many children outgrow the behaviors. Avoidance Strategies:

• Show anxiety in social situations • Sometimes flap arms and hands/ repetitive movements • Scream when excited • Self stimulatory behavior, rocking • Bite and/or hit themselves

38! !

• Sleep problems (many require fewer than normal hours of sleep, many don't sleep through the night) • Enjoy TV, music, movies that are repetitive Many children fall somewhere in the autism spectrum in terms of their behavior. There is debate as to whether these children are autistic and have a chromosome abnormality or if having deletion 22q13 negates an autistic diagnosis. "Autistic like" traits that the deletion population may exhibit are poor eye contact, tactile sensitivity, and communications issues, as well as some listed above and others…

Finally, individuals with deletion 22q13 are able to express a wide range of emotions including joy, happiness, and love.11

Similar, though better quantified results are reported for physical characteristics and they report developmental delay and speech impairment, in varying degrees of severity, in almost all cases. Thus what it means, phenotypically, to have PMS is determined by interaction and dialogue between biomedical experts on the one hand and, on the other, subjects and parents who are in many ways uniquely knowledgeable about the condition.

The contrast with the phenotype described in Phelan et al (1988; above) is illustrative of the way genomically designated syndromes’ phenotypic profiles change considerably over time as the mutation is observed in new cases and biomedical experts, patients and their parents work together, often through organizations like the Phelan-McDermid Syndrome

Foundation, to produce knowledge and care strategies. As we will see, a number of other genomically designated syndromes have followed a similar trajectory. The various fields within which genomically designated syndromes can take root and consolidate their status as independent social and clinical entities will be examined in Chapters 3-6.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 11(http://22q13.org/j15/index.php?option=com_content&view=article&id=61:behavioral&catid=56 :characteristics&Itemid=71, accessed October 4th 2010)

39! !

What’s in a name?

To capture what is conceptually new and important about cases like

PMS/22q13DS, it is useful to consider the meaning of the terms that denote syndromes. A syndrome is an association of several features or symptoms that tend to afflict subjects simultaneously, and the clinical identification and observation of some symptoms can help point to other diagnostic indicators. A syndrome is by definition multivalent – it denotes a number of phenotypic characteristics that tend to occur in some combination with (in theory) a single underlying biological unity. Indeed it is precisely this variability in phenotypic presentation that has led many to invest such hope in biomarkers, particularly observations of the genome, as a more objective and efficient basis for diagnosis (e.g.

Collins et al. 2003). However, we saw that when biomarkers are identified and implemented in screening programs they often generate incongruent results, requiring frontline medical professionals to engage in ‘bridging work’ to reconcile the marker, the people who have it and the medical condition in question (Timmermans and Buchbinder

2012). In genomic designation, however, the failure of genetic observations to line up with existing categories of illness is resolved by simply delineating entirely new categories of illness at the level of the genome. The resulting combinations of phenotypic characteristics are usually so non-specific and diffuse that they could never constitute diagnostic criteria for a medical category. We therefore need to grapple with the way genomic markers can come to serve as the essential referent of terms like ‘x syndrome’ rather than just a biological explanation for what those persons are like; in doing so we see a qualitatively new form of human kind come into view that can marry phenotypic complexity and rigid genomic designation.

40! !

Following Saul Kripke’s (1980) highly influential work in the philosophy of language, Naming and Necessity, we can say that genomically designated syndromes are

‘rigidly designated’ by terms that refer precisely to people bearing specific characteristics of the genome. For Kripke, names do not refer to descriptions, but to referents established by the causal force exercised by a “community of speakers”. When a characteristic of the genome is designated by a community of speakers – clinicians, molecular biologists, parents and so on – as the essential referent for a syndrome, a degree of rigidity is realized such that what Kripke calls an a posteriori12 necessity is established: the statements

‘bearers of genomic mutation n’ and ‘syndrome x’ assume a relationship of ‘rigid designation’ or logical equivalence. I would not seek to defend the idea of rigid designation in the context of a more general discussion of human kinds: Kripke strives for rigidity and logical necessity between names and their referents whereas a major advantage of the human kinds rubric is its focus on the dynamism and fluidity of categories of people.

In stark contrast, Hacking has described ‘human kinds’, such as multiple personality disorder and autism, as ‘moving targets’, and most complex disease designations are not

‘rigid’ in Kripke’s sense but subject to change (see also Parens, Chapman, and Press 2006

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 12 The relationship between the 22q13 deletion and 22q13DS is of course actually closer to one of a priori necessity. There was no category that was discovered through experience to be the same as a 22q13 deletion in the way water, to use Kripke’s example, was discovered to be H2O. However, some less ideal typical genomically designated syndromes result from the finding that an already extant syndrome is equivalent to a characteristic of the genome, such as the case of Williams Syndrome and the 7q11.23 microdeletion. What makes this a case of genomic designation rather than just geneticization, I will argue in Chapter 2, is the way the category is changed when it is rigidly designated by a genetic marker and cases are included and excluded on the basis of genomic observations. In any case, whether through a priori or a posteriori necessity, PMS is rigidly designated by the 22q13.3 microdeletion and phenotypically identical children can therefore receive different diagnoses depending on the results of genetic tests, just as Kripke argues we would distinguish between H2O as water and a functionally equivalent substance on another planet composed of XYZ as something else.

41! ! on the difficulties this raises for behavioral genetics). However, when we consider its implications with respect to cases like PMS, Kripke’s framework helps us see the important shift in the meaning of medical categories entailed in genomic designation.

Consequently, instead of referring to a phenotypic profile, PMS refers precisely to bearers of deletions at q13.3 on the 22nd chromosome that include the SHANK3 gene.13

That is, a relationship of rigid necessity – of designator to designated or signifier to signified (de Saussure 2011) – obtains between ‘genetic anomaly n’ and ‘syndrome x’, while the set of phenotypic correlates remains inchoate and subject to change and wide divergence. While the discovery of people with the Down syndrome phenotype but without

21 chromosome trisomy would create a nosological conundrum with an uncertain outcome14, discovering people with the paradigmatic characteristics of PMS, but without a deletion at 22q13.3, would not be problematic in the slightest: they are not diagnosed with

PMS or admitted into its network and community. Conversely, the (unlikely) discovery of a person with chromosome 21 trisomy but without the Down phenotype would probably not lead to a diagnosis of Down Syndrome, while discovering a person with a 22q13.3 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 13 This reference has changed over time as the specificity of the genomic marker has increased. Still, the syndrome has always referred to deletions at 22q13.

14 This has in fact been an issue since the outset of the ‘Mongolism’-trisomy 21 association. While Lejeune et al. published the first paper on the association in 1959 (ibid), we will see in Chapter 3 that several other teams were on the verge of reporting the same finding. A much larger study in Edinburgh by Jacobs et al. (1959) examined the chromosomes of around 40 patients diagnosed with Mongolism, but found six who had the normal chromosome complement of 46. Jacobs claimed (over forty years later) that when they had the leading British expert on Mongolism, Lionel Penrose, come up from London he re-diagnosed the patients without knowledge of their karyotypes both to the satisfaction of the institution where they had been previously diagnosed and in precise accordance with the trisomy theory of Down syndrome. Similarly, Warkany (1960:415) showed how it was almost immediately employed in research environments to help adjudicate clinically contested cases. Nevertheless, in postnatal cases especially the observation of trisomy or translocation 21 is often used to confirm a clinical diagnosis that can be made on the basis of phenotypic criteria (e.g. Devlin and Morrison 2004).

42! ! microdeletion and a phenotype atypical of 22q13DS would simply expand the profile of the syndrome. In the case of 22q11.2 Deletion Syndrome, one of the most prevalent genomically designated conditions,15 we will see below how parents lacking any of the major phenotypic characteristics have been diagnosed when they have tested positive for the microdeletion following the diagnosis of a child; the phenotypic profile of the syndrome changes accordingly to include “mild phenotypic presentations” (Miller, 2008;

Bales et al, 2010). The scope for phenotypic variation and divergence in genomically designated conditions (particularly findings in ‘control’ populations) remains an open question.

The vast majority of medicalized conditions, indeed almost all human kinds, are associated with a phenotype – the result of interactions between genetic, epigenetic, environmental, developmental and other factors amounting to how a person appears to clinicians, themselves and others. Increasingly, distinctions and connections may come to be drawn between phenotypes on genomic lines (above), and I will argue that just such scenarios are key incubators of genomic designation. However, insofar as the categories themselves are not designated on genomic lines, and most are not, the human kind itself as a nosological, identity or other kind of category remains intact. The counterfactual discovery of multiple genomic correlates does not necessarily raise any nosological

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 15 The standard prevalence estimate is 1 in 2000-4000, though we will see that many experts think it is likely to rise considerably as ascertainment improves. We will also see, in Chapter 2, how 22q11.2 Deletion Syndrome is less ideal typical than 22q13DS in that fairly specific clinical categories (velo-cardio-facial syndrome, DiGeorge syndrome and several others) preceded it and were brought together after genomic analysis identified the deletion from the late 1970s (see de la Chapelle et al., 1981). However it is rigidly designated according to the deletion at 22q11.2 such that the population it delineates is very different from any of those older conditions (or their combination) and it can safely be considered a case of genomic designation (see Miller, 2008).

43! ! conundrums for phenotypically designated categories (though they do present empirical challenges): the syndrome name refers to a more-or-less stable phenotypic profile whose genomic correlates can change without impacting the relationship that obtains between syndrome name and phenotype: ‘y syndrome’ refers to a set of clinically observable phenotypes ‘a’, while genomic correlate(s) ‘n’ can be discovered, debated and discarded.

Figures 1 and 2 indicate the relationship between syndromic categories, genetic markers and phenotypic characteristics in ‘clinical’ and genomically designated syndromes respectively. In Figure 1, we see that clinical syndrome ‘y’ refers to phenotypic diagnostic criterion ‘a’, which in turn mediates any relationship with subsequently identified genomic correlates. In Figure 2 we see the inverse relationship in the case of genomically designated syndromes, where the syndrome name ‘X’ refers to some genomic marker ‘n’, even though its phenotypic correlate may remain vague, with subsequent refinements of the phenotypic meaning mediated strictly through the genetic marker.

Fig 1. Clinical Syndrome Referent

Y Syndrome ====== Phenotype a ⇑ ⇓ Genomic correlate(s) n

Fig 2. Genomically-Designated Syndrome Referent

X Syndrome ====== Genomic anomaly n ⇑ ⇓ Phenotypic Characteristics

44! !

Table 1 attempts to situate PMS along with previous syndromes that have been ascribed a genetic etiology. It shows, respectively, a standard case where a syndrome was delineated on phenotypic grounds and later shown to be more-or-less coextensive with a genetic mutation, one where an established syndrome is related to multiple markers, one whose status remains in flux and, finally, one that is genomically designated.

Table 1: Genotype-phenotype relations in four medical conditions

Genomic 16p11.2 Trisomy 21 Multiple 22q13.3 Deletion Marker Deletion Autism/ 16p11.2 Phelan- Syndrome Down Autistic deletion McDermid/22q13 Syndrome Spectrum Name syndrome Deletion Syndrome Disorder Year of first 2007 1866 1943 2001 syndrome (Ballif et al. (Down) (Kanner 1943) (Phelan et al.) publication 2007) Year of first 1992 (Gustavson al. genotype- 1959 1970s (Buiting et al. 1985; see also (Lejeune et al.) phenotype 1992) Phelan 1988) publication Coextensive Coextensive; Complex, syndrome; still diagnosis if and Relationship Coextensive; with dozens treated as a only if 22q13.3 between genetic of implicated candidate deletion is genotype and correlate of mutations etiology of observed; established conferring syndrome’s autism; found in syndrome features syndrome various risk phenotype control are delineated post factors population factum Clinical/social Yes Yes No Yes networks?

In Down Syndrome and autism we see the usual temporal sequence from phenotype to corresponding genotype: either a single genetic mutation is found to be coextensive with the syndrome and therefore ascribed causality, as with Down, or multiple

45! ! genetic correlates are found, making for a intractably complex relationship between genotype and phenotype, as is the case for autism.

It remains to be seen whether 16p11.2 will follow the path of 22q13. An award- winning New York Times article used 16p11.2 as its headline case of “genetic kinship”, even though 22q13 was also discussed in the article (Harmon 2007). However, a study finding both 16p11.2 duplication and deletion in control populations may make its consolidation as an independent clinical entity less likely (Weiss et al., 2007).

(Interestingly, there has still never been a large-N control study of 22q13.3 deletion, though there was a small one in 2000 (Goizet et al) for a paper interested in it as a potential cause of autistic disorders.) The future of the 16p11.2 deletion appears open to different paths, and both social action by clinicians, parents and others and the biomedical evidence brought to bear will play a key role in determining whether it becomes a bona fide genomically designated syndrome. As we will see, the same applies to the dozens of genotype-phenotype correlations which have been the subject of scientific papers reporting new and phenotypically diffuse syndromes, but have not formed the basis of significant social organization (e.g. Simovich et al (2008) on ‘proximal 3q microdeletion syndrome’).

22q13 deletion, representing genomically designated syndromes, is unique in this table: As with the 16p11.2 microdeletion, it was first delineated in a paper discussing a previously reported geneotype-phenotype relationship. Unlike 16p11.2, however, 22q13.3

Deletion Syndrome/PMS has established a clinical profile and a growing network of support and advocacy. There are at least 22 such syndromes, with prevalence estimates as high as ~1:2000 live births in the cases of 22q11.2 Deletion Syndrome (Bales et al.: 2010),

XXX and XYY Syndromes and most exhibiting clear upward trajectories as ascertainment

46! ! expands. They all marry an otherwise untenable lack of phenotypic specificity with rigid designation according to a characteristic of the genome, and all have become bona fide categories of clinical and social practice. A preliminary list is outlined in Table 2 below.

Table 2: Preliminary list of genomically designated conditions

Syndrome Name/ Notable Foundation or Key/Illustrative Papers Locus Organization (Shapira et al. 1997; Gajecka, 1p36 Deletion Mackay, and Lisa G Shaffer 1p36 Deletion Support & Awareness Syndrome 2007; D’Angelo et al. 2010) 5p Deletion/ (Lejeune et al. 1963; Cri du Chat Overhauser et al. 1994; J.-S. 5p- Society; fivepminus.org Syndrome Fang et al. 2008) (Alfi et al. 1973; Hauge et al. 9p-/Alfi’s Syndrome Chromosome 9p- Network 2008; Barbaro et al. 2009) (Aglan, Kamel, and Helmy 2008; Distal Trisomy 10q J. Davies, Jaffé, and Bush 1998; Distal Trisomy 10q Families Yunis and Sanchez 1974) 13q Deletion (Allderdice et al. 1969; S. Brown World Wide Chromosome Deletion Syndrome et al. 1995; Quelin et al. 2009) 13q Support Network (Rineer, Finucane, and Simon 15q Duplication IsoDicentric 15 Exchange, 1998; Moeschler et al. 2002; Syndrome Advocacy & Support (IDEAS) Hogart et al. 2009) (A. C. Smith et al. 1986; Potocki Smith-Magenis/ Parents & Researchers Interested in et al. 2003; Gropman, Duncan, 17p11.2 Deletion Smith-Magenis Syndrome (PRISMS) and A. C. M. Smith 2006) Potocki- Lupski/17p11.2 (A. Brown et al. 1996; Potocki et Potocki-Lupski Syndrome Outreach Duplication al. 2000, 2007) Foundation, Inc Syndrome (David A Koolen et al. 2006; D A 17q21.31 Deletion Koolen et al. 2008; Grisart et al. Chromo17 Kids Syndrome 2009) Edwards/ (Edwards et al. 1960; Trisomy 18 Trisomy 18 Foundation Boghosiansell et al. 1994) Syndrome 18p-/DeGrouchy (de Grouchy, Bonnette, and C. Chromosome 18 Registry & Syndrome Salmon 1966; Turleau 2008) Research Society

18q- Syndrome/ (Kline et al. 1993; Feenstra et http://ulf.org/18q-syndrome; Leukodystrophy al. 2007) C18 Registry & Research Society

47! !

(Herva, Saarinen, and Ring Chromosome Leikkonen 1977; Alpman et al. Ring Chromosome 20 Foundation 20 Syndrome 2005; Elghezal et al. 2007) (Scambler et al. 1991; D. A. International 22q11.2 Deletion 22q11.2 Deletion Driscoll et al. 1992; Hall 1993; Syndrome Foundation Syndrome Bassett et al. 2011) (www.22q.org) Phelan-McDermid/ (M. C. Phelan et al. 1988; Mary Phelan-McDermid Syndrome 22q13 Deletion C. Phelan et al. 2001; Dhar et Foundation (22q13.org) Syndrome al. 2010) (Kirk et al. 2009; Ramocki et al. MECP2 Duplication 2009; Ramocki, Tavyev, and S. MECP2duplication.com Syndrome U. Peters 2010) (Lubs 1969; R. J. Hagerman et National Fragile X Foundation; Fragile X Syndrome al. 1986; McLennan et al. 2011) FRAXA Research Foundation (P. Jacobs, Baikie, et al. 1959; XXX Johnston et al. 1961; N. R. KS&A (genetic.org) Tartaglia et al. 2010) (Ellis et al. 1961; Parker et al. The XXYY Project/ XXYY 1970; N. Tartaglia et al. 2008) xxyysyndrome.org (Sandberg et al. 1961; Patricia XYY A. Jacobs et al. 1965; Lancet KS&A (genetic.org) 1966; J. L. Ross et al. 2012) (Patau et al. 1960; D. W. Smith Patau/13 trisomy et al. 1960; Baty, Blackburn, Trisomy 13 Support & Resources and J. C. Carey 2005) (Ensenauer et al. 2003; 22q11.2 Duplication Mukaddes and Herguner 2007; International 22q11.2 Foundation Syndrome Portnoī 2009) ! Table 2 therefore provides a preliminary list of genomically designated conditions in the strong sense that they are 1) discussed in the literature as bona fide medical conditions, rather than just genetic findings; 2) delineated strictly according to observed genetic mutations; 3) the object of a social organization, whether it is a staffed foundation or a small support group. Excluded, therefore, are: the dozens of mutations that have been labeled ‘syndromes’ in the literature but about which little or nothing has been written by way of clinical associations and management; conditions that are still diagnosed clinically or about which there remains considerable debate or uncertainty about the status of a single mutation as a necessary and sufficient condition for diagnosis (e.g. tuberous sclerosis);

48! ! cases of what I will call ‘recalibration’ (see Chapter 2) where a clinical condition, for example Williams Syndrome, is fixed to a genetic mutation and delineated anew, in the process both excluding and including substantial numbers of patients as vs. the clinical criteria; and finally the many conditions like 16p11.2 Deletion Syndrome, discussed above, that have not given rise to social action that has, in turn, coalesced into formal organizations (I exclude blogs, Facbook pages and so on).

Conclusion

Why try to marry Hacking’s conception of diagnostic categories or ‘human kinds’ as fluid, historically contingent ‘moving targets’ with Kripke’s account of names as ‘rigid designators’ that seeks to establish a high level of referential rigidity? I think the tension is worth pursuing because I think Kripke is right to tell us (ibid, p. 106) “reference actually seems to be determined by the fact that the speaker is a member of a community of speakers who use the name.” Yet despite deferring to a sociological phenomenon realized by ‘communities of speakers’, Kripke gives no sense of the sociological contingency that entails. In short, in the quest for a posteriori necessity Kripke loses sight of the sociological contingency and historical fluidity found in different sequences of declaring, characterizing and delineating natural and, a fortiori, human kinds. That Kripke’s argument works so well for genomically designated human kinds, but so poorly for complex and fluid categories of people like autism and intellectual disability that we will see act as the primary ‘surfaces of emergence’ (Foucault 2002:41) for genomic designation, is indicative of how radically different genomic designation is as a way of categorizing people with developmental differences. While Hacking’s rich conception of kinds of people as ‘moving targets’ (2006)

49! ! is prima facie anathema to rigid designation, in the case of genomic designation the combination is productive. The unholy alliance of Hacking and Kripke allows us to see how genomic designation gives rise to human kinds that are simultaneously more phenotypically diffuse than we usually see in what Hacking (2006c) calls the ‘human sciences’ (including “a good deal of clinical medicine”), even as they are fixed with an unusual rigidity according to genomic observations. Furthermore, we will see in Chapter 6 how that rigidity enables rather than forecloses the realignment of clinical judgment and treatment regimes for people with genomically designated conditions. Studying genomic designation therefore offers an opportunity to study the looping processes at play when there is an unusual level of fixity as to who can be diagnosed, but nevertheless significant scope for shaping and reshaping the meaning and implications of diagnosis.

In sum, this chapter has reported and provided a working definition of a kind of biosocial fact – genomic designation – whereby genetic mutations can give rise to new, otherwise unthinkable medical conditions. It therefore raises two related lines of inquiry.

The first, to be answered in the next short chapter, concerns various pathways to genomic designation: how can observations from genetics reconfigure existing categories of medical classification and therefore what kinds of conditions will it be useful to think about in terms of genomic designation? The second, which occupies the bulk of the dissertation, asks the more sociologically complex questions: how do some genetic mutations give rise to genomic designation in the strong sense outlined above with all its attendant clinical and social implications, while others remain squarely at the level of esoteric objects of biomedical research even when they are reported as ‘syndromes’ in the literature? How do they gain traction and how are productive points of exchange and

50! ! interface established between genomically designated conditions and categories based on other systems of human classification? Finally, how has genomic designation remained constant and how has it been transformed over its fifty-year history, and what might we expect of it in the years ahead?

51! !

Chapter 2 – The varieties of genomic designation Multiple pathways and an ideal typical typology of syndromes1

We have seen a clear-cut case, 22q13DS, where the discovery of a genetic mutation served to delineate a novel medical condition. It was almost as though a new kind of person emerged from nothing – or de novo – upon the discovery a newly observed genetic anomaly. But to equate genomic designation with limit cases like 22q13DS would obscure the many other ways in which observations of the genome can be used to carve out new categories of medical classification and social action. To reiterate, genomic designation is about the delineation and diagnosis of conditions according to observed genetic mutations.

Therefore, to the extent that existing nosological categories are reconfigured strictly on the basis of observations made at the level of the genome, and thereby sideline clinical diagnostic criteria, it is useful to think about them as cases of genomic designation as well.

In other words, 22q13DS helped us see genomic designation in stark relief and begin to grapple with some of its theoretical implications for the social studies of science and medicine. However, we need to take stock of the multiple pathways to genomic designation. This chapter proceeds in two steps. First, I use quantitative citation analysis supplemented by fieldwork and qualitative historical research to examine a particularly complex and counterintuitive such pathway – the way knowledge about a genetic mutation can unify fields of clinical research and ‘lump’ or subsume several longstanding clinically !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 Most of this chapter is based on a paper with Uri Shwed (Navon and Shwed 2012), who implemented the network analyses and produced Figures 1-5.

! 52! ! delineated disorders. Second, I build on that case to outline five ideal typical pathways to the rigid designation of medical conditions according to genetic mutations. !

22q11.2 Deletion Syndrome and the unification of medical conditions

22q11.2 Deletion Syndrome demonstrates that observations from genetics can be used to reconfigure medical classification in more complex ways than was seen in the case of 22q13DS. Nevertheless, 22q11.2DS is a case of genomic designation: a condition that is diagnosed and delineated on the basis of an observable characteristic of the genome that does not line up with any previously recognized category of person or indeed with any clinical diagnostic criteria. However, we will see that there is a crucial difference between

22q11.2DS and our initial case study of genomic designation, 22q13 Deletion Syndrome: while 22q13DS was never thought to have anything approaching a one-to-one relationship with an existing disorder, 22q11.2DS was not cut from whole cloth. Rather, the 22q11 deletion was initially posited as an etiology for DiGeorge Syndrome and then several other rare conditions. In its complex reconfiguration of biomedical subfields, nosology and medical populations, the 22q11.2 microdeletion suggests that we should pay close attention to the multiple pathways to the genomic designation of medical conditions. Even though

22q11.2DS has strong relationships with the older clinical disorders that it has mostly subsumed, its phenotypic range is actually much broader than 22q13, bringing together patients with serious heart malformations, cleft palate, intellectual disability, ear infections, schizophrenia, anxiety disorder, constipation, autism, and around 180 other symptoms and traits alongside incidentally ascertained people with few or none of the above. As Donna

McDonald McGinn, one of the leading biomedical experts and advocates for 22q11.2DS,

! 53! ! put it in an interview (2011) when pushed on its nosological status, “22q is 22q… it's a spectrum from no symptoms to every malformation under the sun.”

Now, it is true that the 22q11.2 deletion began its career as a case of what the social studies of science and medicine would probably call ‘geneticization’ – Lippmann’s aforementioned term for the idea that categories of human difference would be “reduced to their DNA codes” (Lippman 1991:19; 1991). After all, it first came to medical relevance as a genetic explanation for the rare clinical disorder, DiGeorge Syndrome. However, we will see below that even though it may have been first mobilized in the 1980s as a reductive explanation for an existing, clinical disorder, the 22q11.2 deletion did not behave itself in the years that followed. Instead, some people diagnosable with DiGeorge Syndrome had a full 22nd chromosome even as the deletion began to crop up in other, previously unrelated conditions like Velocardiofacial Syndrome and Opitz Syndrome. Rather than embrace the idea of a complex relationship between abnormalities in the genome and medical classification, however, some people began to think of the 22q11.2 deletion as delineating a syndrome that did not line up with existing diagnostic categories but should nevertheless subsume and replace them. Instead of geneticization and the reduction of human difference, we therefore need to examine how research on the 22q11.2 microdeletion was productive of a new medical condition.

I examine 22q11.2DS by attending to the development of three closely related, but analytically distinct kinds of biomedical phenomena: clinical disorders, genetic abnormalities, and genomically designated syndromes. The current case starts with a number of fairly rare clinical disorders, primarily DiGeorge Syndrome, VCFS and Opitz

G/BBB Syndrome, with their own symptomatologies and independent origins as medical

! 54! ! conditions. Then there is the genetic abnormality that came to be associated with those clinical conditions, the 22q11.2 microdeletion: the observation of cases, through genetic testing, of a deletion of DNA from site 11.2 on the long arm of the 22nd chromosome.

Finally, a new category emerged that subsumes the above clinical conditions – what is now called 22q11.2 Deletion Syndrome. Analyzing the dynamics among these objects of knowledge in the case of 22q11.2DS, this chapter presents a mechanism – the unification of fields – whereby genetic mutations can be mobilized to reconfigure medical categorization.

To examine this process of unification my colleague Uri Shwed and I used novel citation analysis techniques that he developed, supplemented by qualitative historical and

IRB-approved fieldwork research, to show how 22q11.2DS emerged as an object of knowledge. I begin with a review of the literature on 22q11.2DS and related conditions, followed by a review of the pertinent social scientific literature. Then, I describe our citation analysis strategy, extending the technique of Shwed and Bearman (2010) for mapping the structure and community salience of scientific literatures. I then present the results and argue that the 22q11.2 deletion was more than just an etiological finding: it was the actant or ‘boundary object’ (Star and Griesemer 1989) that unified otherwise disjunct fields of research and made possible the hybrid field in which 22q11.2DS could emerge as a qualitatively new medical condition. I therefore provide a targeted account of

22q11.2DS’s conditions of possibility (Foucault 1973: xix) by modeling its emergence from older fields of biomedical research. In so doing, I hope to contribute to our understanding of the way that observations from genetics can reconfigure categories of medical classification and produce new ways of understanding human difference.

! 55! !

Literature(s) Review

22q11.2 Deletion Syndrome has been the subject of over four hundred biomedical papers, while many more have investigated the medical implications of the microdeletion at 22q11.2. However 22q11.2DS has never been a subject of social scientific analysis despite its growing prevalence and, as I will argue, its capacity to speak to key issues in the social studies of genetics and medicine. This section will therefore review two disparate literatures – the extensive bioscientific literature on 22q11.2DS and the social scientific literature that has neglected it – in order to suggest that each has something to learn from the other.

The biomedical literature on 22q11.2DS

Reading the biomedical literature one is struck by the way 22q11.2DS is often treated as synonymous with other diagnoses, in particular DiGeorge syndrome and

Velocardiofacial (VCFS)/Shprintzen syndrome. The literatures on DiGeorge Syndrome and VCFS may be traced back to papers in 1968 (DiGeorge 1968) and 1978 (R J

Shprintzen et al. 1978) respectively, and other syndromes like Opitz G/BBB Syndrome, conotruncal anomaly face syndrome and Sedlakova syndrome similarly predate

22q11.2DS and are often considered to be subsumed by it. As a set of practical guidelines for 22q11.2DS recently published in The Journal of Pediatrics put it:

Although clinically under-recognized, 22q11DS is the most common microdeletion syndrome (MIM#188400/#192430), with an estimated prevalence of 1 in 4000 live births. However, the actual occurrence may be higher because of variable expressivity… The 22q11.2 deletion is the second most common cause of developmental delay and major congenital heart disease after Down syndrome, accounting for approximately 2.4% of individuals with developmental disabilities and approximately 10% to 15% of patients with tetralogy of Fallot. 22q11.2 deletions have been identified in most patients

! 56! !

with DiGeorge syndrome, velocardiofacial syndrome, and conotruncal anomaly face syndrome and in a subset with autosomal dominant Opitz G/BBB syndrome and Cayler cardiofacial syndrome. Although this list of associated disorders may appear quite perplexing, it is understandable because the diagnoses were originally described by clinicians concentrating on their particular areas of interest. After the widespread use of FISH, however, patients with a deletion became collectively referred to by their chromosomal etiology: the 22q11.2DS. (Bassett et al. 2011:2)

Similarly, a review in Newborn and Infant Nursing Reviews (K. A. Miller 2008:e11) tells us that 22q11.2DS “encompasses” what “were once thought to be different conditions with different diagnoses.” One could cite many similar passages. Indeed since Wulfsberg et al.’s

1996 paper ‘What’s in a name?’ (E. A. Wulfsberg, Leana-Cox, and Neri 1996; see also E.

Wulfsberg, LeanaCox, and Neri 1997) the situation has often been likened to the parable of the blind men studying different parts of an elephant, sometimes even with a cartoon ‘22’ elephant to illustrate the point (D M McDonald-McGinn, E H Zackai, and Low 1997:247).

But what are we to make of this nosological framework? Does 22q11.2DS

‘encompass’ these older syndromes because the deletion simply helped biomedical experts to see a clinical syndrome that they were previously blind to? The situation is far more complex. First, the proportion of people who are diagnosable with the clinically delineated syndromes listed above who also have a 22q11.2 microdeletion varies, but none approaches 100%. Thus in contrast to Down syndrome, this is not a straightforward case of a genetic etiology being discovered for an already-extant diagnostic category. Second,

22q11.2DS’s clinical profile consists of more than 180 phenotypes, many of which are observable in only a small minority of cases and were not part of the profile of the clinical conditions with which the 22q11.2 deletion came to be associated. Finally, it is not necessary for a subject to be diagnosable with one of those longer-standing syndromes, or indeed with any clinically diagnosable condition, for them to be diagnosed with

22q11.2DS.

! 57! !

Consider this passage from a review in Genetics in Medicine, which follows a discussion of the multiple syndromes now associated with 22q11.2 deletions:

Although the deletion is identical in most patients studied, the phenotype varies greatly. Goodship et al. report a case of monozygotic twins with 22q11DS where one twin’s phenotype is more severe, showing that genotype alone does not account for the presence or absence of various features of 22q11DS. More than 180 clinical findings have been associated with 22q11DS… Both the number of organ systems involved and severity of involvement vary. Severe cases may result in neonatal death, whereas mildly affected individuals may remain undiagnosed, even as adults. (Bales, Zaleski, and McPherson 2010:135)

An observed microdeletion at 22q11.22 is both necessary and sufficient for diagnosis with

22q11.2 Deletion Syndrome, even though the population of persons with the deletion do not line up with any pre-existing clinical category or coherent clinical diagnostic criteria.

What’s more, patients with the paradigmatic phenotype of, say, DiGeorge syndrome do not necessarily have 22q11.2DS while ‘patients’ lacking any of the most strongly associated clinical symptoms can have 22q11.2DS. As caseloads increase and existing patients age,

22q11.2DS’s already expansive phenotypic profile is likely to grow further still. To describe 22q11.2DS as ‘encompassing’ or ‘explaining’ the older clinical diagnoses obfuscates an important shift in the relation between genotype, phenotype and the delineation of medical categories.

Social scientific literature

As we have already seen, a range of social scientific work has examined the impact of genetics on medical classification and social identity formation. From Paul Rabinow’s

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2 The deletion is usually around three million base pairs long and, although deletion size can vary considerably, existing studies have found no meaningful correlation between deletion size or break points and phenotypic severity or range (Philip and Bassett 2011; McDonald-McGinn et al. 2001). There is some debate about the status of mutations on the key gene TBX1.

! 58! ! discussion of the way genetic information could lead to ‘biosocial’ group formation (1992) and Heath, Rapp and Tausig’s (2004) discussion of ‘genetic citizenship’ to Nikolas Rose’s investigation of the myriad implications of biomedicine’s turn to the ‘molecular gaze’

(Rose 2007), seminal work has attended to the impact of genetics on nosology, biomedical practice and social identity (Freese and Shostak 2009). Of particular relevance to this chapter is Lippman’s previously discussed warning of the danger of ‘geneticization’ whereby traits and conditions would be reduced to their genetic correlates, in turn prescribing the scope for human development for the people who are so understood (1991).

We have also seen how, more recently, Rabeharisoa and Bourret (2009) and Timmermans and Buchbinder (2010, 2012) have both shown how the results of genetic tests are often ambiguous in their clinical implications, requiring extensive interpretive work on the part of medical practitioners in order to reconcile genotype, phenotype and diagnosis. In short, genetics is both reshaping the way many existing conditions are understood, diagnosed and treated and producing complex cases that problematize the idea of geneticization

(Hedgecoe 2003, 2001; Kerr 2004; F. A. Miller et al. 2005, 2006). Still, a condition like

22q11.2DS does not fit comfortably into any of those theoretical frameworks for understanding genetics and medical classification. While more complex in relation to existing nosology than 22q13DS, it is most definitely a case of genomic designation: it is a medical condition that is delineated and diagnosed strictly according to observable characteristics of the genome even though it lacks the phenotypic coherence to be diagnosed clinically. While we will see that the 22q11.2 microdeletion entered the biomedical literature in what we might call an attempt to ‘geneticize’ DiGeorge Syndrome,

! 59! ! its ultimate nosological outcome was the delineation the a qualitatively new, genomically designated 22q11.2DS.

Beginning with Star and Griesemer’s (1989) seminal study of the Berkeley

Museum of Vertebrate Zoology, the concept of ‘boundary objects’ has helped science and technology studies understand cooperation and translation between fields (see also Star

2010; for applications to biomedical research see Fox 2011). As Star and Griesemer put it

(1989, p. 393): “Boundary objects are objects which are both plastic enough to adapt to local needs and the constraints of the several parties employing them, yet robust enough to maintain identity across sites” and key to “developing and maintaining coherence across intersecting social worlds.” On the one hand, this chapter examines the way that the

22q11.2 microdeletion served as a boundary object in a network comprised of divergent bioscientific and clinical disciplines. On the other hand, we will see how the 22q11.2 microdeletion did not simply facilitate exchange between already-connected communities, but actually served to create a new, hybrid research subfield and delineate a novel medical condition. As such I also adopt a broader actor-network theoretic approach in order to examine the role of the 22q11.2 microdeletion as an ‘actant’ or non-human actor in a process of knowledge production, translation and application (Latour 1988; Law 1992). In so doing, my quantitative and historical analyses of 22q11.2DS contributes to the mostly qualitative literature on the way genetics can impact medical classification.

Methods

This chapter examines 22q11.2DS as an object of knowledge and therefore focuses on the biomedical literatures within which it emerged. However it is also part of two larger

! 60! ! studies: the first uses mixed methods to understand 22q11.2DS not only as an object of biomedical research, but also as an object of medical practice and social action and is developed in Chapter 6; the second is an ongoing collaborative project with Uri Shwed that uses citation analysis to model the impact of genetics on bioclinical literatures. Following the example of Cambrosio et al. (2004) in their analysis of innovation in the development of antibody reagents, I employ quantitative network analysis and qualitative historical and fieldwork methods recursively to inform one another, with this report focusing on the former. In other words, the quantitative analysis in this chapter was informed by and interpreted with respect to ongoing historical research and IRB-approved (10.12.2010) fieldwork conducted at 22q11.2DS conferences and meetings as well as interviews with parents, activists and biomedical experts throughout 2011 and 2012. Similarly, that ongoing fieldwork has been informed by the citation analysis presented in this chapter and will form the empirical basis of Chapter 6 where I analyze the forms of clinical practice and advocacy associated with 22q11.2DS.

In order to model the fields of research from which 22q11.2DS emerged I represent the relevant literature as networks of papers linked by citations. Since Price (1965), many have exploited the descriptive power of network graphs to observe underlying social structures in collaborative networks (e.g. Cambrosio et. al. 2004) or substantive meaning in semantic and co-citation networks (e.g. Bourret et al. 2006). Our approach is different: focusing on the role of a specific object of knowledge rather than the social structures creating it, we observe citations as associations between articles. The entire scientific literature on an issue can thus be described as a directed, unvalued network of articles tied by citations. Such networks require far less analysts’ discretion regarding thresholds for tie

! 61! ! strength and similar decisions. We then supplement the descriptive power of networks with the analytical power of Leicht and Newman’s modularity algorithm (2008) that finds meaningful groups within the network and evaluates their contribution to the network’s overall structure.

Using Thompson Reuter’s ISI Web of Science database I extracted all peer- reviewed articles about 22q11.2DS, the older syndromes associated with it, and articles about the microdeletion at 22q11.2 that reference one of the associated conditions in their keywords or abstract.3 The resulting dataset includes 1418 such papers and the citations between them. We then use characteristics of the data to create a dynamic periodization that slices the 1418 papers into overlapping networks of all the papers published in a period of two to seven years, as well as the papers cited by those papers. The exact number of years, or observation width, represents the scope of relevant scientific discussion as determined by the median age of citations made from each year (see Shwed and Bearman

2010). For example, in 1991 citations went relatively far back in time and their mean age was 6. The network labeled “1991” contains all papers published between 1985 and 1991.

In 1992, median citation age was 4 and the observation width shrunk to include only 1988-

1992. Although restricted to peer-reviewed papers, the analysis is comprehensive and

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 3 Our search string: “(TI=("Di George syndrome" OR "DiGeorge syndrome" OR "DiGeorge anom*" OR "velo-cardio-facial" OR velocardiofacial OR "Shprintzen syndrome" OR "conotruncal anomaly face syndrome" OR "Opitz G syndrome" OR "GBBB Syndrome" OR "BBB Syndrome" OR "Sedlackova syndrome" OR "22q11 deletion syndrome" OR "22q11.2 deletion syndrome" OR 22q11.2DS OR 22q11DS OR "22Q11.2 DS" OR "22Q11 DS")) OR (TI=(22q* OR TBX1 OR "CHROMOSOME-22*") AND TS=("Di George syndrome" OR "DiGeorge syndrome" OR "DiGeorge anom*" OR "velo-cardio-facial" OR velocardiofacial OR "Shprintzen syndrome" OR "conotruncal anomaly face syndrome" OR "Opitz G syndrome" OR "Opitz syndrome" OR "GBBB Syndrome" OR "BBB Syndrome" OR "Sedlackova syndrome" OR "22q11 deletion syndrome" OR "22q11.2 deletion syndrome" OR 22q11.2DS OR 22q11DS OR "22Q11.2 DS" OR "22Q11 DS"))”.

! 62! ! symmetrical in that selection is explicitly defined by the search string and periodization is independent of analysts’ discretion.

The core of our analysis rests on finding meaningful divisions in these temporal networks. This too is symmetrical: divisions are made by the modularity algorithm and based solely on citations patterns. The algorithm also evaluates the salience of those divisions for the network structure. It defines salience as the fraction of within-community ties beyond the expected fraction conditional on nodes’ connectivity and community sizes

(Leicht and Newman 2008). This algorithm is the most popular for detecting meaningful divisions in networks (Fortunato 2010), but we exploit it further by examining community salience regardless of their specific structure. Changes in this property reveal turning points in the history of the case and scaling it for network size allows for the comparison of different periods (Shwed and Bearman 2010). Specifically, an increase in scaled modularity indicates that internal divisions between communities have become more salient, while a decline in scaled modularity indicates convergence around a common core.

A scaled modularity score below 0.1 indicates a unified literature where communities are not very important. After identifying crucial turning points using the modularity values of the temporal networks we draw complete graphs for those years with nodes marked by the communities generated by the algorithm (communities are analyzed for each temporal slice independently).

We then extend the quantitative analysis with qualitative content analysis:

Examining the top cited papers in each community, I assign substantive meanings to those communities, e.g. a community primarily interested in clinical aspects of DiGeorge

Syndrome, another in the 22q11 deletion and so on. The substantive meaning as well as the

! 63! ! salience of the divisions are not constant across time; in periods of high modularity and a fragmented literature these communities delineate distinct objects of knowledge, such as

VCFS and Opitz G/BBB Syndrome, while in periods of convergence and low modularity communities could display substantive overlap and represent other factors that affect citation patterns like ‘invisible colleges’ (Crane 1969).

Finally, to formally test the role of the 22q11.2 deletion we constructed variance and confidence intervals. As the network is not sampled, we employ a pseudo-jackknifing procedure that compares the modularity dynamics of the full data with those of a subset of the data excluding papers with 22q11.2, TBX1 and other terms related specifically to the deletion at 22q11.2 in their title or keywords. We then generated 100 random subsets of the same size from the same data, randomly deleting from every year the same number of papers as the number of papers mentioning the deletion. Thus for each year we now have

102 scaled modularity estimations: for the full data, for the subset absent the deletion, and for 100 randomly generated subsets of the same size as the subset without papers related to

22q11.2. This allows us to determine when removing the deletion increases network modularity in a non-random, statistically significant manner. All analyses were conducted in the R software environment with the igraph package (Csardi and Nepusz 2006) with the exception of the Leicht-Newman calculation, obtained from the authors.

Results

Figure 1 presents scaled modularity scores, which were shown by Shwed and

Bearman (2010) to index networks’ coherence (low modularity) or fragmentation (rising modularity), alongside logged network size for temporal slices of the complete data. With

! 64! ! the first paper reporting VCFS (Shprintzen 1978) modularity begins a moderate rise, reaching a plateau for the 1985-9 period. Between 1989 and 1995 modularity rapidly declines and from 1995 to 2002 there is another mild increase in modularity. These trends suggest the following dynamics: Prior to 1989 there are several small, disjunct literatures on a few rare clinical disorders. 1989-1995 is a period of convergence and coherence around a common core, which I argue is the deletion at 22q11.2. Then until 2002 there is some re-fragmentation or specialization in the literature, which has remained relatively stable since. Below, we trace these dynamics by mapping the literature in these specific moments: 1979, 1989, 1992, 1995, 2002 and 2009.

But how can we be sure that it was the 22q11.2 deletion that enabled this unified field of research? We formally test this hypothesis in Figure 2, which replicates Figure 1

! 65! ! with three changes. First, the dotted line represents scaled modularity for the same networks, excluding all papers making reference to 22q11.2 and its cognates. Second, the grey area represents the mean and six standard deviations of the 100 jackknifing draws described above to provide a robust confidence level of P<0.002. When the dotted line is outside of the grey area, the modularity of the network excluding the marker in that year is significantly different from a random exclusion of the same size. Finally, the dashed line represents the size of the 101 counterfactual networks as a fraction of the full network

(referring to the right hand side axis). This leaves us with: 1. The full literature (solid line);

2. The literature without papers concerning 22q11.2 (dotted line); 3. A pool of jackknifed literatures to ensure that the difference between 1 and 2 is not an artifact of removing a large proportion of the networks’ nodes.

! 66! !

The decline in modularity seen in Figure 1 is apparent in Figure 2 for the 100 simulated networks, which are not significantly different from the full network until 1997 when the full network is 332 papers compared to only 130 in the jackknifing and no- marker networks. Even when most of the network is deleted the confidence interval echoes the trends of the full data. The nonrandom draw excluding papers concerned with 22q11.2, however, is radically different: It exhibits a notable increase in modularity – or fragmentation – between 1992 and 1995 and a stable, high level of modularity since. This difference is statistically significant from 1992 when only 20% of the networks’ papers were removed, precisely the moment when a link with 22q11, already well-established for

DiGeorge Syndrome, was found for VCFS. By the time “22q11.2 Deletion Syndrome” first appears in 1995 the scaled modularity of the real literature is well below 0.1, which Shwed and Bearman argue is a threshold for a unified literature (2010), while that of the literature excluding papers about the deletion remains high. This indicates that the core around which the network converged in the early 1990s is the genetic mutation at 22q11.2. When omitted, the dynamic of internal cohesion and declining modularity are completely absent.

We further analyze these trends by looking at the structure of the literature in specific years. Figure 3 presents the story in a nutshell, Figure 4 demonstrates how one of these networks looks without papers about the genetic marker, and Figure 5 shows the literature in the twenty-first century. Figure 3 describes four key moments: First, the network of papers published prior to 1979, with one connected component of papers about

DiGeorge Syndrome, one dyad and 11 isolates not cited or citing any other paper in this literature, including Shprintzen’s paper delineating VCFS. We have removed communities’ labels because in such small networks a glance provides more information

! 67! ! than the calculation. By 1989 (top right), before the convergence of the field apparent in

Figure 1, small but coherent fields of research have been established for several distinct clinical syndromes: A clear community demarcates VCFS research, another Opitz G/BBB, and a main component on DiGeorge Syndrome contains three overlapping communities of clinical research and a more cohesive community focusing on the 22q11 deletion.

Although one report in the supplemental pages of The American Journal of Human

Genetics (R. Goldberg et al. 1985) noted clinical overlap between VCFS and DiGeorge

Syndrome, it created no ties between these disjunct literatures through formal citations.

By 1992 (bottom left), the network’s modularity was in the middle of a precipitous decline (Fig. 1). During this period Fluorescent In Situ Hybridization or ‘FISH’ testing was

! 68! ! refined and widely disseminated, making the detection of 22q11.2 deletions far more practicable (Larson and Butler 1995). Several papers now unite the previously distinct literatures on DiGeorge Syndrome and VCFS: Two papers by VCFS researchers (Stevens,

Carey, and Shigeoka 1990; A. H. Lipson et al. 1991) noted the syndromes’ phenotypic overlap and cited, but were not cited by, papers concerned with DiGeorge Syndrome.

Crucially, several others (see esp. P. J. Scambler et al. 1992; Deborah A. Driscoll et al.

1992) noted that the deletion at 22q11 that had been discovered years earlier in most cases of DiGeorge Syndrome was also observable in many cases of VCFS. These genetics papers were, by 1992, cited by papers interested in VCFS and DiGeorge Syndrome.

An interview with Peter Scambler (2011), one of the medical geneticists who established the association between the 22q11 deletion and VCFS, confirms that some degree of clinical overlap with DiGeorge Syndrome informed the initial studies on 22q11 and VCFS. However, he also tested a number of other rare conditions for similar reasons and, when 22q11 deletions were not observed, they remained unaffected by the field of research discussed in this paper. In short, it was 22q11 research that bridged the VCFS and

DiGeorege Syndrome communities. This is confirmed by Figure 2 which demonstrates that the deletion had a statistically significant role in uniting the field. Further papers pertaining to the 22q11.2 microdeletion solidified that unification in the years that followed.

This unification on the basis of the shared interest in the 22q11 microdeletion had almost immediate nosological consequences. In 1993 an editorial by pediatrician Judith

Hall introducing a special issue of the Journal of Medical Genetics on 22q11 recounted its initial association with DiGeorge Syndrome before remarking, “To everyone's surprise and delight what had been thought to be a very separate disorder, the Shprintzen or

! 69! ! velocardiofacial syndrome, has been found also frequently to have a 22q11 deletion.” (Hall

1993:801) She noted further associations with non-syndromal conotruncal anomalies and advocated a shift towards the new, all-encompassing term ‘CATCH 22’ after which the editorial was named (801-2):

The acronym CATCH 22… not only helps to remember the features of concern (Cardiac, Abnormal facies, Thymic hypoplasia, Cleft palate, and Hypocalcaemia) and the 22nd chromosome but also suggests the complexity which exists in 1993 with regard to defining the clinical features and understanding the genetic ramifications of a deletion from a specific area of a specific chromosome. What fun it is to find that what were thought to be separate syndromes are now merging into one group, CATCH 22, while other patients who were thought to belong to those same previously separated syndromes are being excluded from that catchy acronymous group because they must have different mechanisms... CATCH 22 is a wonderful model for what is to come over the next 10 years of human genome work. We are very likely to end up with more questions to answer than when we started and with a more appropriate name, such as a DNA sequence designation. But for now the acronym CATCH 22 will help to focus on this fascinating problem.

Hall was right that the name CATCH 22 would be short lived. First, a letter in response the following year (A. Lipson et al. 1994) noted that the range of observed phenotypes was such that “the unifying feature appears to be the presence of a 22q microdeletion,” (741) rather than any combination of the acronymous ‘CATCH’ symptoms (above). Having a name that referenced symptomatology had already proven unfeasible for the 22q11 population. Second, patients, parents and others objected to the derogatory implication of a

‘no-win’ situation described in Joseph Heller’s satirical novel about WWII airmen. As a result, the geneticist who first proposed the term CATCH 22 eventually wrote an article named after Heller’s less well-known sequel, Closing Time, which sought to remove the term from circulation (Burn 1999).

By 1995 the network’s modularity plummets to a level that indicates community divisions are far less important (see Fig. 1). In Figure 3 (bottom right panel) we see that only the VCFS and the Opitz G/BBB communities are distinguishable in their network locations, while the DiGeorge and genetics communities are intertwined. VCFS research is

! 70! ! well integrated with the remainder of the literature via the genetics communities. The Opitz

G/BBB community in 1995 resembles the VCFS community in 1992: it is connected by the genetics communities but still clearly demarcated. Comparing this graph to Figure 4

(below), which represents the network minus the papers pertaining to the deletion, the contrast is clear: the full network in 1995 is for the most part densely connected, while the network without the marker remains a single component by virtue of just a handful of papers. With the deletion excluded the network shrinks from 198 papers to 100, the number of isolates rises from 7 to 18, the Opitz G/BBB community is disconnected from the main component and the VCFS community remains connected by only three papers. In other words, the deletion is the thread that holds this literature together.

! 71! !

Crucially, 1995 is also the year in which ‘22q11 Deletion Syndrome’ first appears in the literature in a paper by Lynch et al. (1995), marked by a square in Figure 3 (bottom right) and situated in the VCFS community near its interface with the genetics communities. It is worth pointing out that several of its authors were involved in the key work on 22q11.2 in the early 1990s and have gone on to establish specialist centers and become central advocates working closely with the International 22q11.2 Deletion

Syndrome Foundation. While it would be some time before ‘22q11.2 Deletion Syndrome’ attained predominance, the bridging of clinical fields on the basis of a shared genetic etiology was increasingly giving way to a unified, coherent field of research on a new genomically designated condition.

By 2009 a dense network of papers about 22q11.2DS, its phenotypic implications and clinical predecessors, and the genetic mechanisms associated with the 22q11.2 microdeletion is visible. Figure 5 presents the literature in 2002 and 2009. Squares mark papers referring to 22q11.2 Deletion Syndrome – 10% of the literature in 2002 and 35% in

2009. Of the papers actually published between 2003 and 2009 (eliminating those cited by them) a solid majority refers to the syndrome. By 2009 these dominate one community, and are found throughout the literature. Again, since the mid 1990s, low modularity levels suggest that community divisions are not salient and their graphic proximity confirms that finding.

! 72! !

Modularity and community analyses demonstrate, with minimal analysts’ discretion, how several disconnected literatures about rare clinical conditions were brought together by research on the 22q11 microdeletion. As that literature grew and the deletion was observed in a growing variety of patients a new kind of condition emerged: the genomically designated 22q11.2DS. 22q11.2DS is increasingly supplanting and encompassing its phenotypically delineated predecessors in the biomedical literature.

Figure 6 represents the frequency of DiGeorge Syndrome and VCFS (by far the most prominent such syndromes) and their synonyms on the one hand and 22q11.2DS on the other in the titles of biomedical publications. Within ten years of its first appearance in the

! 73! ! title of a biomedical paper in 1995 the name ‘22q11.2 Deletion Syndrome’ had overtaken its two major predecessors, and it has since gone on to dominate the field.

Fig.%6.%Frequency%of%syndrome%names%in%paper%titles,%1996:2012%

80!

70!

60!

50! DGS! 40! VCFS! 30! 22q11.2!DS! 20!

10!

Attending an International 22q11.2 Foundation conference or visiting its website and the page devoted to their ‘Same Name Campaign’4, one is struck by the efforts to both explain and move beyond the lingering existence of older syndrome names. We will return to this issue, and the social mobilization around 22q11.2DS more generally, in Chapter 6.

For now, the key point is that even as a minority remains committed to the VCFS label – a source of considerable acrimony between some advocates and experts – its clinical origins notwithstanding they now consider it to be coextensive with the 22q11.2 deletion (see the

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 4 See http://www.22q.org/index.php/what-is-22q/same-name-campaign (accessed 3.12.13).

! 74! !

VCFS Educational Foundation; Shprintzen’s 1998 paper, ‘The Name Game’).

Furthermore, while some biomedical experts are careful to distinguish between VCFS,

DiGeorge etc. as clinical diagnoses and 22q11.2DS as an overlapping genetic one, most now treat the terms as synonymous (Scambler interview, 2011). My interviews and fieldwork confirm that, at least for many in the field, the shift to a genomic diagnosis is taken to be self-evident. The continued disagreement about nomenclature therefore belies a consensus about the nosological significance of the 22q11.2 microdeletion, despite its enormous phenotypic variability.

Although phenotypic overlap between DiGeorge Syndrome and VCFS was noted in 1985 and again in the early 1990s, we saw that it was research on the 22q11.2 deletion in the early and mid-1990s that united the literatures on VCFS, Opitz G/BBB and

DiGeorege Syndromes into a single field of knowledge production and created the conditions for the delineation of a qualitatively new medical category. In what follows I consider some of the implications of this process for our sociological understanding of the interface of genetics and medicine then present a working typology of the pathways to genomic designation.

Discussion

My analysis demonstrates that, as what Latour (1988) and others have called an

‘actant’ or a non-human actor, the 22q11.2 microdeletion was essential to the formation of a hybrid population of patients and field of knowledge production that was previously invisible to clinical research. I showed how multiple medical literatures with independent origins were united into a single, hybrid subfield by a boundary object from another

! 75! ! discipline: genetics. When the microdeletion that was initially associated with DiGeorge

Syndrome was found in people with VCFS and then other conditions, the research communities interested in those older conditions were unified on the basis of a shared etiology. The meaning of the 22q11.2 microdeletion is not conceptualized in the same way across all the fields for which it is an object of knowledge and classification, and yet it is robust enough to unite them and coordinate action.

However, the 22q11.2 deletion is more than a boundary object that alleviates tension in the intersection of social worlds – it is also a two-fold boundary-making object.

First, it is not simply an epistemic object that is able to circulate and maintain coherence across already-connected fields of scientific and social practice, but the very thread that ties those fields together. Furthermore, rather than just facilitate communication and exchange between subfields the 22q11.2 microdeletion served as an actant that created the conditions for biomedical experts to see one genomically designated condition in lieu of several clinical conditions. Second, this actant also performs ‘boundary work’ (Gieryn

1983): because the 22q11.2 microdeletion is not observable in 100% of those who had been diagnosed with the pertinent clinical syndromes, the minority that did not carry the deletion were excluded from the new category of 22q11.2DS. Furthermore, other conditions that exhibit clinical overlap with DiGeorge Syndrome and VCFS were excluded from the new genomically designated category when studies found no association with the microdeletion. In short, the 22q11.2 microdeletion served to create a qualitatively new biomedical subfield and then a new kind of medical condition.

By 1993, when the field was undergoing a process of unification on the basis of the

22q11 deletion, researchers began to see a new medical condition despite the absence of

! 76! ! diagnostic criteria that could have been derived from what Foucault called the ‘clinical gaze’ (1973). As an ever-greater range of clinical features were observed in 22q11.2 deletion probands, researchers drew on the longstanding (though recently named) practice of ‘genotype-first diagnosis’ (David H. Ledbetter 2008) or what I am calling genomic designation and delineated a new medical condition – first named CATCH 22 and then

22q11.2DS – whose coherence is found at the level of the genome. Despite considerable and often acrimonious debate about the name for the condition delineated by the 22q11.2 microdeletion (see e.g. Wulfsberg et al. 1996), there was remarkably little debate about the move to delineate a syndrome on that basis. The relative ease with which the 22q11.2 microdeletion unified the field and led to a new ‘causal’ diagnosis, even in the face of so much clinical variability, is suggestive of privileged status genetic mutations can have as actants in contemporary biomedical classification.

Like other cases of genomic designation, 22q11.2DS is not simply a question of molecular evidence ‘ruling in’ or ‘ruling out’ cases of existing disease categories (Miller et al. 2005), but a matter of delineating new and otherwise inconceivable medical conditions according to molecular observations. However, 22q11.2DS is not only far more prevalent than the case of genomic designation, 22q13DS, discussed in Chapter 1; it also demonstrates how genomic designation can occur through a very different pathway.

22q13DS could be described as emerging de novo: the deletion at 22q13 was never thought to have anything approaching a one-to-one relation with a clinical condition until it attained (tautologically) a strong relationship with 22q13DS, delineated in 2001 despite the

“lack of a recognizable phenotype” (Phelan et al. 2001: 98). By contrast, research on the

22q11.2 deletion unified several existing clinical conditions. While the discovery of the

! 77! !

22q11 deletion in most cases of DiGeorge Syndrome could be seen as an attempt at

‘geneticization’, the ultimate result of research on 22q11 deletions cannot. Rather, it produced a new category of human difference. 22q11.2DS therefore challenges our understanding of both geneticzation and genomic designation, demonstrating that genetics research can produce powerful intermediary ways of understanding illness.

A typology of pathways to genomic designation

Again, 22q11.2DS also helps us to extend the concept of genomic designation and suggest that the path of declaring a new syndrome cut from whole cloth is just one extreme, with geneticization being the other, of the ways in which genetics can reshape nosology. By systematically examining a case where existing syndromes were brought together into a new and otherwise unthinkable condition, we can confidently begin to think about the other kinds of intermediary, ideal typical pathways to genomic designation.

Following Weber (2011) , I use the term ‘ideal type’ to indicate that these pathways are intended as useful abstractions to guide further sociological analysis and the formulation of causally and semantically adequate explanations, rather than distinct phenomena that are found in reality. As we will see, any given case will likely instantiate several of these ideal types in different ways and to varying degrees, but I will nevertheless argue that they represent the necessary conceptual toolkit for a thorough understanding of genetics’ intersection with and realignment of existing, and especially clinical, classificatory systems.

I therefore propose a five-fold typology of ideal typical paths:

! 78! !

First, we have our old friend ‘geneticization’ – the much-discussed but rare finding of a near one-to-one correlation between a genotype and a medical condition with established clinical diagnostic criteria (e.g. Down syndrome).5

Second, we have a kind of genomically designated recalibration. This is akin to geneticization, except that the condition becomes fixed according to a characteristic of the genome in such a way that both the population and the clinical profile undergo a considerable shift. In some ways, many cases of ‘geneticization’ in the literature more closely match this model of genomically designated recalibration. As Miller et al. (2005) have shown it is one nosological solution among several when a proposed genetic etiology turns out to be a less than perfect correlate with the clinical disease category it is supposed to indicate and explain. Still, for our purposes it is important to distinguish this kind of recalibration from the ‘reduction’ represented by geneticization: when a genetic marker is adopted as both necessary and sufficient for diagnosis of a condition, and in so doing brings new people into the population and excludes others, genetics has delineated a new and different population than the one picked out by clinical presentation. I will therefore treat medical conditions that are rigidly designated by observations of the genome, even though they inherit the name and history of a clinical category, as weak cases of genomic designation.

Take the case of Williams Syndrome, which we will return to in greater detail in

Chapters 4 and 5. It was first delineated in the 1960s and was associated a very distinctive !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 5 Of course Lipmann’s seminal discussion of geneticization was about far more than the identification of genomically specific etiologies, focusing more on the reduction of human difference and the foreclosing of possibilities for development on the basis of a presumed genetic basis for disease. Indeed it did not even require an etiological finding. What’s more, we will see in Chapter 6 how more recent work has further complicated ideas about geneticization.

! 79! ! phenotypic profile – elfin facial features, high verbal ability, sociability and musicality alongside developmental delay and heart defects. However, after becoming fixed to a

7q11.23 microdeletion in the mid-90s (Nickerson et al. 1995), some who matched the clinical criteria were excluded and others were brought into the fold. Early studies show that a small minority of ‘classic’ Williams Syndrome patients did not have the deletion, and neither did a majority of ‘uncertain WS patients’ (LOWERY et al. 1995). And yet, as the Williams Syndrome Association itself puts it: “It is important to stress that WS is a genetic diagnosis and an individual who does not have the gene deletion does not have

Williams syndrome (i.e. a person who was clinically diagnosed with WS but was later found not [to] have a deletion in fact, does NOT have WS).”6 So even while Williams

Syndrome is considered to have one of the most consistent and distinctive phenotypes among the disorders discussed in this dissertation, affixing it to a genetic mutation required nosology boundary work in both directions. Not only were some clinically diagnosed cases excluded, as seen above, but atypical cases in whom a clinical diagnosis could not be made have in fact been diagnosed by virtue of testing positive for a 7q11.23 deletion. As a study conducted when the diagnostic implications of the deletion were still in question put it:

In this small pilot study, all five cases in whom a confident diagnosis had been established, were found to have a submicroscopic deletion at 7ql 1.23 with FISH analysis. More interestingly, out of the five atypical cases, in whom a definitive diagnosis could not be made on clinical features alone, three were found to have the submicroscopic deletion at the elastin locus. Establishing a diagnosis in the neonatal period and in infancy can sometimes be difficult. Burns et al. noted that many patients with WS are not diagnosed until they are old enough to show the characteristic personality and facial changes. Moreover, a number of these changes are subtle and might not be present in all affected subjects. In our study, patient 3 had reached 15 years of age before the diagnosis of

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 6 http://www.williams-syndrome.org/diagnosing-williams-syndrome/diagnosing-williams- syndrome

! 80! !

Williams syndrome was excluded. Another atypical case, patient 5, who was found to have the deletion, had only been suspected of having the condition while this study was in progress. She had reached her ninth birthday and other possible diagnoses that had been considered by the referring doctor were Down's and Albright's syndromes. (Borg, Delhanty, and Baraitser 1995:694–5)

In short, even when it comes to the most strikingly characteristic of phenotypes, genomic designation can still impact diagnosis, delineation and therefore the clinical profile of a disorder.

Third, conditions can be unified or ‘lumped’, as we just saw with 22q11.2 DS. The fact that this most radical and counterintuitive reconfiguration of existing nosology was shown to have occurred and to be very well established in the literature and in practice helps to situate the more well established pathways discussed above and below.

Fourth, conditions can be split or fragmented according to genetic subtypes. This is well understood when it comes to cancers, and especially the role of mutations in the

BRCA genes (e.g. Venkitaraman 2002; see Bourret, P. Keating, and Cambrosio 2011), but less well understood when it comes to broader surveillance categories like intellectual disability, x-linked mental retardation and autism. As we will see, when people with autism or XLMR are found to have a sufficiently large trinucleotide repeat at Xq27.3, a new and otherwise indistinct condition known as Fragile X Syndrome is diagnosed, with its own clinical profile and expansive biosocial network (Blomquist et al. 1985; Reiss and

Freund 1992).7 Similarly, as we will examine in Chapters 4 and 5, Fragile X and many other genomically designated conditions are often thought of and mobilized as genetically specific subtypes of autism.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 7 The National Fragile X Foundation: http://www.fragilex.org/html/home.shtml

! 81! !

Finally, we have the extreme cases of genomic designation, like Phelan-McDermid

Syndrome, where new conditions are delineated ‘de novo’ – the genotype in question was never thought to explain an already recognized illness, but simply posited as the basis for a new one.

Again, these idealtypical pathways are not mutually exclusive. From the perspective of autism or intellectual disability, Williams Syndrome, 22q11.2DS and

Phelan-McDermid Syndrome all represent cases of splitting, even though I explained how their nosological histories are perhaps most usefully framed according to the second, third and fifth types discussed above. Nor are they ever perfectly instantiated: even Down

Syndrome has undergone a degree of recalibration as the test for trisomy 21 is used to arbitrate clinically questionable cases, while 22q11.2DS has also substantially recalibrated the diagnoses of DiGeorge Syndrome and VCFS for many of those who still use those terms. In short, this conceptual toolkit is precisely that: a set of tools for making sense of genomic designation, not a way of parsing it up that works in strict accordance to the often messy reality of genetic, medical and other forms of classification.

Conclusion

Increasingly, biomedical researchers are following the US National Human

Genome Research Institute’s ‘Grand Challenge II-3’ in its call for “a new molecular taxonomy of illness [to] replace our present, largely empirical, classification schemes,”

(Collins et al. 2003:841; see also Hayden 2008; Loscalzo, Kohane, and Barabasi 2007), which in the field of developmental disability has begun to coalesce under the rubric of

‘genotype-first’ discovery and diagnosis (Saul and Moeschler 2009; David H. Ledbetter

! 82! !

2008; Cody 2009; D H Ledbetter 2009; David H. Ledbetter 2009; see Conclusion). At the same time, recent studies have identified a far greater than anticipated volume and incidence of potentially pathogenic genomic variants in humans, including tens of thousands of structural variants like CNVs (copy-number variations) and chromosomal abnormalities as well as millions of smaller mutations (see The 1000 Genomes Project

Consortium 2010). With genetic testing becoming ever more rapid, precise and affordable, and efforts to aggregate data about mutations and their phenotypes more expansive and sophisticated, how will all this knowledge about our genomes impact our understanding of human difference more generally?

This chapter indicates that knowledge about genetic mutations can reconfigure existing categories of biomedical research and lead to the emergence of qualitatively new diagnostic categories. I analyzed a biomedical literature over a thirty-five year period, its internal divisions, and trends in the salience of those divisions. In so doing, I demonstrated that the microdeletion at 22q11.2 served as an actant that fostered enduring ties between several small, previously disjunct subfields of medical research, creating a densely connected literature that brought together an otherwise incoherent set of patients, expertise and clinical observations. Once this unified, hybrid field had been created on the basis of a genetic mutation, biomedical experts were able to delineate a new biomedical category,

22q11.2DS, which was qualitatively different from its predecessors. My analysis of

22q11.2DS suggests that we need to go beyond our existing social scientific toolkit in order to grapple with the empirically variable ways in which the growing tide of observations from genetics can reconfigure medical classification. Only by tracing the role

! 83! ! of genetic mutations as actants in networks of biomedical research and practice will we be able to understand the multiple pathways to genomic designation.

! 22q11.2 Deletion Syndrome is among the most prevalent cases of genomic designation, encompassing a diverse, growing population and an expansive range of phenotypic presentations. What’s more, it is the subject of a sizable biomedical literature and, as we will see in Chapter 6, a complex network of clinics, advocacy groups, protocols and many other actors. However the precise pathway taken from the observation of a genetic mutation at 22q11 to the unification of biomedical subfields and then the delineation of a qualitatively new medical condition complicates our existing understanding of genomic designation in particular and the nosological implications of genetic mutations more generally. Modeling the development of 22q11.2DS as an object of knowledge therefore demonstrates how the observation of genetic mutations, which is likely to occur with far greater frequency in the coming years, can reconfigure categories of medical research and classification. Finally, it allowed us to set up a five-fold typology of ideal typical pathways whereby disease categories can become rigidly fixed to genetic mutations in order to help guide us through the rest of the dissertation.

! 84! !

Part II

So far we have focused on uncovering and defining genomic designation as a form of human classification, its relationship to existing work in the social studies of science and medicine and the multiple pathways through which the observation of a genetic mutation can lead to the delineation of a new medical condition. But this just scratches the surface of genomic designation as a sociological phenomenon. After all, thousands of genomic abnormalities have been reported in the human genetics literature over the last fifty years, but only some have been cast in that literature as new, genomically designated syndromes and even fewer have gone on to inform and impact clinical practice, advocacy, social identity formation and the like. What’s more, genomic designation came into view over fifty years ago, almost as soon as it became possible to observe chromosomal aneuploidies, and yet it is only the last two decades that have seen many of them achieve social and clinical traction.

How then are these paths from genetic abnormalities, to novel ‘syndromes’ in the literature, to real medical and biosocial categories actually taken? Citation analysis helped us think about the changing structure of scientific fields that can lead to genomic designation, but that is only one side of the story. To put it more bluntly, so what if some geneticists talk about new syndromes in their journals if no one does the work to imbue them with clinical and social salience? Just because 22q11.2DS exists in biomedical journals does not mean that it automatically becomes an object of clinical practice and

! 85! ! social action. Consider again for a moment Phelan-McDermid, i.e. 22q13 Deletion

Syndrome. In 1993 Katy Phelan reported another chromosomal deletion – 2q37 (M. C.

Phelan, R C Rogers, and Byrd 1993; see also Mary C. Phelan et al. 1995) – that has only recently been subject to genomic designation (see Galasso et al. 2008) and is yet to achieve anything approaching the medical and social significance of 22q13. But why? When I asked Phelan about the divergent outcomes of the 22q13 and 2q37 microdeletions she paused, and said that it had very little to do with the physiological implications of those pieces of chromosome and more to do with the fact that the parents of the first few 2q37 kids were satisfied with their existing diagnoses. By contrast, the 22q13 parents began to communicate and organize. (K. Phelan 2011)

A sociological analysis of genomic designation must therefore go beyond the fact that it can give rise to new medical conditions in the literature and the various ways in which it can reconfigure nosology and examine the conditions and processes that make those new categories matter to the way we understand and act on human difference. It must go beyond the field of human genetics, for whom we will see genetic classification came as an obvious and unproblematic turn once they had the sociotechnical capacity to observe abnormal chromosome complements, and examine how genomic designation has been taken up by researchers in other fields, clinical practitioners, commercial enterprises, media, advocates and the diagnosed themselves as a useful way of classifying people. This section outlines my framework for thinking about the historical conditions, processes of network formation and forms of collective action that can turn a genetic mutation into a new kind of person, with all that means for the diagnosed and their families.

! 86! !

Studying micro-apparatuses

This shift in our level of analysis is analogous to Foucault’s move from archaeology to genealogy – from the study of epistemic formations to an analysis of the way a particular truth regime combines with networks of power and practice to shape lived experience. We need to understand how one group’s ‘enunciative modality’ or ‘grid of specification’ became compelling for other disciplines, lay experts and broader constituencies of stakeholders. In other words, we need to explore the conditions of possibility for genomic designation to form what Foucault would call an apparatus or

‘dispositif’ that combines a certain ‘regime of truth’ with the practices and institutional forms that make it a bona fide site of knowledge/power (Michel Foucault 2010, 2002). Of course, Foucault tended to deal with matters much broader in scope and scale than a relatively marginal phenomenon like genomic designation. However, my aim is still to:

pick out… a thoroughly heterogenous ensemble consisting of discourses, institutions, architectural forms, regulatory decisions, laws, administrative measures, scientific statements, philosophical, moral and philanthropic propositions–in short, the said as much as the unsaid. Such are the elements of the apparatus. The apparatus itself is the system of relations that can be established between these elements. Secondly, what I am trying to identify in this apparatus is precisely the nature of the connection that can exist between these heterogenous elements… Thirdly, I understand by the term “apparatus” a sort of– shall we say–formation which has as its major function at a given historical moment that of responding to an urgent need. (Michel Foucault 1977)

My goal then is to arrive at an account of the way that genomic designation reconfigures the relations between heterogeneous elements as well as the need to which it responds. Of course, genomic designation still functions within much larger discursive and institutional systems – clinical medicine, liberal governance, modern biopower, just to take a few dealt with by Foucault himself (2009, 1973, 2010). At a discursive level we have seen that, although it certainly does not represent a threat to any prevailing, epochal ‘episteme’ in the

! 87! ! sense outlined in Foucault’s earlier work (2001), genomic designation does represent a novel ‘formation’ (Michel Foucault 2002). What’s more, we will see how on both the

‘discursive’ and ‘institutional’ levels, genomically designated conditions must work within existing relations even as they seek to reconfigure them. Genomic designation, in other words, is a rupture in the typical organization of elements for making scientific statements about human difference and requires novel relations between heterogeneous elements in order to matter at the level of practice. It is therefore worth analyzing in Foucauldian terms.

We can perhaps say that genomic designation is an embryonic dispositif, and one that is unlikely to ever supplant or even break free of its gestational environment.

Nevertheless, as we saw in Chapter 1 it does represent a fundamental realignment of the relationship between the visible and the invisible – a new ‘spatialization of disease’ – and one that requires a reorientation of clinical diagnosis and practice. It involves the philosophical commitment to give the genome a kind of ontological primacy – natural enough for human geneticists perhaps but now adopted by, for example, parents who want to understand their child’s developmental difference; it responds to the need for personalized medicine and care, the need to understand difference according to biological lesions and the need to find common cause, mutual aid and support for people whose difference does not fit neatly into established categories. None of this followed seamlessly from the capacity to detect genomic abnormalities – what Rabinow calls ‘biosociality’ was not simply instigated by geneticists’ findings – but rather it required certain conditions of possibility and the work of actors who sought to reshape research, care and advocacy according to this new way of carving out kinds of people. Even today, experts and

! 88! ! advocates work within heterogenous contexts that, one way or another, are geared towards clinical classification. To grossly paraphrase: people classify human difference, but they do not do so just as they please, under self-selected circumstances, but under circumstances existing already, given and transmitted from the past.

Genomic designation is a way of delineating categories of disease or difference. It is less a new paradigm being fought over by competing groups or necessitated by what

Foucault called a ‘strategic’ need than an aggregate of individual categories born of a novel kind of human classification. Even though Foucault did come to examine the interplay between multiple discursive formations (and the apparatuses of which they were part), his work is less useful in trying to understand the establishment and circulation of particular statements or facts. For that it is more useful to invoke another, complementary theoretical framework: Ludwig Fleck’s seminal work on the origins and development of scientific facts (1981[1935]). How, for example, did objects of knowledge from researchers in one field – human genetics – were able to not only achieve acceptance by what Fleck calls ‘exoteric’ fields and publics (1981:113), but actually become objects of knowledge production and practice in those fields. Fleck discusses the particular path from

‘journal science’ to ‘vademecum science’ and eventually ‘popular knowledge’, at which point (p. 125) a “fact becomes incarnated as an immediately perceptible object of reality.”

In order to become both apodictic and consequential, genomically designated syndromes therefore had to gain traction beyond the vanguard of cytogeneticists who studied them.

Take Fleck’s account of a bacteriological examination (pp. 113-5): He describes the extended, technical report one esoteric expert would give another – an idealization of a painstakingly standardized and often more vexed laboratory observation, itself based on

! 89! ! years of varyingly successful basic research – and how unsatisfactory it would be to a general practitioner. Rather, the report to the GP would simply read: “The microscopic specimen shows numerous small rods whose shapes and positions correspond to those of diphtheria bacilli.” As Fleck puts it, “This finding is specially written to suit the general practitioner, but it does not represent the knowledge of the expert. It is vivid, simplified, and apodictic. The general practitioner can rely upon it.” (113) Eventually, the fact becomes more apodictic still and “thought-constrained proofs disappear” as we move to the sphere of the ‘generally educated’, such that “the mother of the child whose throat swab had been examined is simply informed: ‘Your child has diphtheria.’” In the language of actor-network theory, the diagnosis has undergone ‘punctualization’ (Law 1992) or been black-boxed.

However, while Fleck insists that “Popular exoteric knowledge stems from specialized esoteric knowledge” (ibid: 113), that does not make for a unidirectional relationship between the two. The mother’s understanding of diphtheria, it follows, also impacts the bacteriologist and their expertise: “Certainty, simplicity, vividness originate in popular knowledge. That is where the expert obtains his faith in this triad as the ideal of knowledge. Therein lies the general epistemological significance of popular science.” –

Fleck (1981[1935]: p. 115) In short, popular knowledge stabilizes its esoteric originator and “reacts in turn upon the expert.” (113) I want to draw on this insight from Fleck and examine the way that genomically designated syndromes as scientific facts emerge quite unproblematically in esoteric medical genetics, but often remain halting, insecure findings as they struggle to gain traction in exoteric fields and lay constituencies while others come to be adopted and thereby transformed by broader fields and publics.

! 90! !

Thus while Fleck’s most cited concept, the ‘style of thought’ that enables and constrains knowledge production, is frequently compared to Foucault’s ‘episteme’ or

Kuhn’s ‘paradigm’1, his real empirical and analytic work focuses on the more fine grained issue of the way a very particular scientific fact, namely the biochemical isolation and delineation of syphilis according to the Wassermann reaction, developed within a particular ‘thought collective’ and led to the emergence of serology as a field. Decades before Foucault’s groundbreaking articulation of much the same point (see Chapter 1),

Fleck recognized that “the modern concept of disease entity, for example, is an outcome of

[an historical development] and by no means the only logical possibility.” (p. 21)

Nevertheless, insofar as Fleck engages with disease classification more generally, it is in order to show how a style of thought functioned and fluctuated in the overlapping fields that sought to define and devise a test for syphilis. Where Fleck’s analysis is at its strongest and its most relevant is his discussion of the ‘active’ and ‘passive’ or background elements of knowledge that characterize a style of thought, the research cul de sacs, the shifting disciplinary boundaries and the changing cultural imperatives that characterize the history of a particular disease entity. He recounts how his work in the venereal section of a large hospital convinced him “that it would never occur even to a modern researcher, equipped with a complete intellectual and material armory” to isolate the disease entity ‘syphilis’ from the totality of their cases (p. 22). He continues: “Only through organized cooperative research, supported by popular knowledge and continuing over several generations, might

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 Fleck briefly discusses the frequent occurrence of “’mutations’ in thought style” (ibid: p. 26), even citing the shift to relativity theory in modern physics (one of Kuhn’s seminal paradigm ‘revolutions’). However, his focus lies elsewhere: he is more concerned with tracing the history of a specific scientific fact over the course of multiple thought styles than he is with styles of thought themselves and epochal transformations from one to the next.

! 91! ! a unified picture emerge, for the development of the disease phenomena requires decades.”

While genomic designation upends the sequence of developing a disease entity – after all, it often begins with the isolation of a medical condition according to a biomarker – we will nevertheless see how genomically designated syndromes also require long processes of sociotechnical refinement and development before they can function as powerful categories of difference.

Despite both recognizing the historical contingency of modern disease classification, the contrasting theoretical arsenals of Fleck and Foucault respectively help us to grapple with very different aspects of genomic designation. Foucault helps us see how a new discursive formation that both conditions and depends upon particular institutional practices becomes possible, exigent and powerful in a given time and place.

Fleck, while concurring that scientific research is embedded in historically specific and often unarticulated background rules for producing knowledge, is more useful when it comes to a comparative-historical investigation of specific scientific facts like disease entities. This combination allows me to historically trace the development of genomically designated conditions across disciplines and into ‘popular culture’, while also paying attention to the way they function within institutional arrangements, as forms of self- practice and as responses to historically specific needs. If Foucault helps us think about the formation of an apparatus of knowledge, practice and power, Fleck helps us conduct such an analysis at the level of individual objects of knowledge. In short, we begin to see how a genetic mutation can form the basis of its own micro-apparatus.

! 92! !

Reiterated facticity

How then, can we bring together comparative-historical sociology, science and technology studies and Fleck’s problematization of the establishment of scientific and especially biomedical facts across different kinds of fields – esoteric and exoteric experts and publics alike? Furthermore, how can we hope to explain the contrasting fate of the same mutations as facts in different historical periods? To a significant degree, the logic of this dissertation is similar to what Haydu (1998) calls ‘reiterative problem solving’: a tool in historical sociology that follows narrative and path dependency approaches in taking historically specific conditions and contingencies seriously in a causal analysis, but departs from them by examining the way that actors respond to similar problems or crises in different historical periods. Such an approach allows for a comparative analysis of different historical settings, the explanation of different outcomes and careful attention to the way that settlements in one period both constrain and enable actors working to solve a homologous problem later on. For our purposes the traditional comparative-historical logic is fatally undermined by the fact that even our comparison between cases are far from independent or equivalent (W. H. Sewell 1996) – many researchers and advocates are aware of and can often be found working with multiple genomically designated syndromes, while the physiological implications of the different mutations themselves are often far from equivalent. Furthermore, when we are analyzing the contrasting meaning and mobilization surrounding the same mutations over time, a conventional comparative logic is clearly a non-starter. Finally, the variable-centered approach adopted in traditional comparative-historical work is antithetical to the STS literature drawn on throughout this

! 93! ! dissertation. Reiterative problem solving, with its focus on the narrative and actor-centered explanation of related historical episodes, therefore holds great promise for the diachronic comparison of genomic designation over the last half-century offered in the second part of this dissertation.

However, we cannot really say that we are “connecting events between periods through sequences of problem solving.” (p. 349) Why? Haydu insists that in defining and delimiting ‘recurring problems’, “one criterion must be the social actors’ own understandings.” (p. 355) As such:

… there must be some correspondence between the observer’s conception of a recurring problem and the social actors’ experiences of confronting common obstacles and devising ways to surmount them… this requirement compels the researcher to build into his or her methodological strategy the agency of social actors as they define problems, devise solutions, and take action.

Now in the case of genomic designation there certainly is continuity across contrasting periods – indeed in a handful of cases the very same genetic anomalies are at stake over time – wherein we see actors engaging in very different forms of action. That said, while

Haydu proposes to investigate “contrasts in how social actors constructed the problem and what solutions appeared to be realistic within each historical context,” (p. 355, my emphasis) we are interested, by contrast, in how actors construct an object of knowledge and the realistic ways of mobilizing it across historical contexts. The consistent element that guides casing is the objects themselves rather than actors and their conceptions of what to do with them.

Consider Haydu’s casing of US industrial relations regimes, where he writes:

“Making 1900 the cutoff thus provides for more ‘controlled’ comparisons between periods and trains attention on what both observers today and social actors in 1900 would

! 94! ! recognize as common dilemmas of management.” (361) This encapsulates why reiterative problem solving requires some amendment for our purposes: it is not the case that human geneticists in the 1960s and a parent advocate today would ‘recognize a common dilemma’ or share experiences of confronting common obstacles (above), even if their interest lay in precisely the same genetic mutation. Furthermore, even today, we find incommensurable formulations of the matter at hand. I am not referring simply to divergent interests with respect to the same problem – Haydu’s analysis of employer and union conflict and settlement over labor relation is of course designed for just those kinds of ‘problems’.

Rather, we will see that forging points of interface and frameworks of commensurability with new groups of actors and reframing existing understandings of the matter at hand, i.e. problems, is central to the rise of genomically designated conditions. We do not want to treat the work of advocates, genetic counselors and others as ‘exogenous’, as Haydu frames state intervention during the formulation of a new regime of labor relations in the

1930s (p. 364). Rather, our casing has to do with a particular form of classification, based on a certain kind of observation, the more longstanding examples of which allow us to compare very different periods.

What is constant across the time periods examined here is not the goals of a particular set of actors, but rather the mutations themselves. Our ‘dependent variable’ – the way knowledge about those mutations is produced and the practices it gives rise to – is explained in large part by changes in the networks of actors, not just the problem solving of any analytically privileged group. Actors and the goals of their action cannot therefore be the basis for historical casing, but rather an explanation of change over time when we hold the mutations constant. So while reiterated problem solving “puts social actors at

! 95! ! center stage” (357), my approach puts objects of knowledge center stage and explains their meaning and the uses to which they are put in terms of the changing concatenations of actors who mobilize around them. Drawing on Fleck’s discussion of the shifting status of scientific facts as they gain traction in ‘exoteric’ fields, I therefore call this approach one of

‘reiterative facticity’ in order to draw attention to the comparative logic of tracing the same facts across time periods, the changing networks built up around them and the divergent forms of knowledge and practice they give rise to. Reiterated facticity can therefore apply the kind of causal explanation advocated by Haydu to many of the topics addressed in the science studies literature by facilitating the comparative analysis of actor-network formation across historical periods.

In sum, the problematization here is less about a homologous group of actors pursuing similar goals across different historical contexts than it is about a scientific fact developing within and across different spheres of knowledge production and practice. It is not about how the same kind of groups, like labor relations boards in Haydu’s case study, learn from the past and adapt to the present in pursuit of similar ends. For while the fields of human and medical genetics certainly exhibit a significant degree of contiguity over the half-century covered in this dissertation, we cannot remain focused on them as the central actors or assume that they were really, in the early period, aiming for anything like the kind of networks organized around genomically designated syndromes seen today. In short, there was no homologous ‘problem’ as conceived of by a consistent set of historical actors over time that we can treat as our unit of analysis. What remains constant, rather, are the mutations themselves. In other words, we want to undertake a comparative study of, as

Callon (1995:54) puts it, the way an actant’s “identity depends on the state of the network

! 96! ! and the translations under way, that is on the history in which they are participating.

Society and nature fluctuate like the networks that order them – existence precedes essence.” Hence the need to look at reiterated facticity: the ways in which the same kind of actant – in this case genetic mutations – can take very different form as scientific facts and give rise to strikingly divergent forms of knowledge and practice across historical contexts and according to the work done by contrasting networks of actors (human and otherwise) over time. Thus while I do compare cases within time periods, reiterated facticity allows us to see the developing conditions of possibility, repertoires of collective action and processes of network formation that can turn esoteric knowledge about a genetic mutation into a new kind of person.

This framing implies, correctly, that there is a prehistory to this dissertation, viz. the many thousands of years where these mutations have existed in human populations without being recognized as such. As Hacking explains (2006a, 1998, 2006b, 1995), there undoubtedly were people who, had they lived today, would have been diagnosable with autism or multiple personality disorder long before either became available as diagnostic categories. It is perhaps even more clear-cut to say that there were people in the nineteenth century who would have been diagnosable with 22q11.2 Deletion Syndrome or XYY

Syndrome were they alive today – after all, their diagnostic criteria are quite straightforward even if the kind of people they delineate are many-varied. Granted, there may have been fewer as medical advances have, for example, allowed the significant minority of 22q11.2DS with infant heart malformations to survive, and an increase in average maternal age will have made a substantial contribution to genomically designated populations. Hacking’s key point, however, is that classifying people matters: it creates

! 97! ! expectations, ways of seeing otherwise obscure forms of difference and repertoires of expert and self practice that act upon and change the person who has been so classified.

That is the point of dynamic nominalism: human kinds may be ‘constructions’ that do not exist independently of our formulations of them, but they make and remake both themselves and the people who are so classified. The arc of this dissertation shows how genetic mutations were first observed and used to carve out esoteric ‘syndromes’ in human genetics, and only much later and in a small minority of cases were assembled into robust facts such that they constituted kinds of people in Hacking’s sense of the term. While there is clearly less scope for looping in terms of who can be diagnosed with genomcially designated syndromes (though there certainly is in terms of who is likely to be given the relevant test), we will see that there is plenty when it comes to their phenotypic profiles and the clinical and social implications of a diagnosis. Thinking about reiterated facticity therefore helps us explain how genetic mutations can be enrolled in the heterogeneous networks that can turn them into the essential referent of a new kind of person.

An important part of the answer, I will argue, is to be found in the work of the many advocacy organizations associated with genomically designated conditions. As I discuss in detail in Chapter 6, in many respects these resemble and actually draw on the repertoires of collective action adopted by the many disease advocacy organizations that have proliferated and grown in influence in recent decades. To be sure, the literature on disease-based advocacy points towards important findings that certainly apply to the cases

I discuss – the way advocacy groups raise awareness, lobby and fundraise for treatment and research, collaborate with experts to facilitate knowledge production and so on.

However, as important as this new brand of disease advocacy is for genomically

! 98! ! designated syndromes like 22q11.2DS today, it is only part of the story. For one thing, genomic designation depends on a particular set of symbolic, institutional, technological conditions related to genetics and medicine. Furthermore, we will see how repertoires of mobilization have developed that are quite specific to genetic disorders like Fragile X and

22q11.2DS and how actors working with genomically designated syndromes must creatively redirect existing expert practice and material structures in order to make genetic mutations really matter in diverse settings. Doing so helps us to think about the relations between elements that a genomically condition must develop in order to go from an esoteric category in human genetics to an apparatus of knowledge and power that profoundly shapes practice and lived experience. Furthermore, it allows us to chart not only the reach of genomic designation with respect to what Fleck calls exoteric fields, but also the way that actors seeking to advance the cause of genomically designated conditions have to work within but also creatively transform existing structures (see W. H. J. Sewell

1992) for understanding and acting on human difference in order to realize their full potential as novel kinds of people.

Classification and the translation of interests

The insight that classification – the ways in which we carve up the natural world into different categories – is a product of collective, social processes goes back to at least

Durkheim ([1912] 2001). In recent years, considerable social scientific attention has been devoted to the importance of standardization (see Timmermans and Epstein 2010 for a reivew) in general and to medical classification in particular (Armstrong 2011; Bowker and

Star 2000; Timmermans and Berg 1997). Similarly, many scholars have followed Latour

! 99! !

(1988, 1993, 2007) and Callon (1986) in examining the way that seemingly stable formations of knowledge and practice can be viewed as heterogeneous networks geared towards the translation of actors’ interests, while also offering important amendments and extending ‘actor-network theory’ (ANT) to new empirical domains (e.g. Bockman and

Eyal 2002; Mol 2002; Singleton and Michael 1993). By bringing these frameworks together I hope to demonstrate the dual projects of knowledge production and social action that can allow genomically designated conditions to effect that kind of translation and become powerful categories of clinical practice and social mobilization. Throughout, I pay particular attention to the conditions of possibility for those projects and their increasing success as well as the forms of social action that work to transform those conditions and make genomic designation of a powerful mode of human classification.

Foucault wrote, “Classificatory thought gives itself an essential space, which it proceeds to efface at each moment. Disease exists only in that space, since that space constitutes it as nature; and yet it always appears rather out of phase in relation to that space, because it is manifested in a real patient, beneath the observing eye of a forearmed doctor.” (1973: 9) When a system of classification – of disease or whatever else – is at its most powerful there is a prima facie acceptance or ‘taken-for-grantedness’ about the way it carves up observation and experience. Whether symptoms, anatomical observation, psychiatric characteristics or genetic test results, the observations we privilege in the classification of disease are historically variable and representative of dominant modes of thought and institutional organization. As Armstrong writes in a piece drawing on both

Durkheim and Foucault, “underlying classificatory principles – whether of mobile symptoms or anatomically fixed lesions – both constitute and reflect the very nature of

! 100! ! identity.” (Armstrong 2011:802) In this way, classificatory systems can withstand the

‘effacement’ or resistance mentioned by Foucault (above; see also Armstrong (ibid) on the manifold challenges of achieving commensuration between the clinical nosology of the

ICD and the symptom-centered complaints dealt with by general practitioners). Genomic designation depends, in part, on the discursive power of genetics and ideas about the genome and identity (see Taussig, et al., 2003, p. 59;) but also on the non-functionalist work of actors seeking to transform classificatory principles, medical practice and other structures that are not geared towards genetics in general or genomic designation in particular. Classificatory systems have to find symbolic and institutional structures that are amenable to their way of ordering reality, but they must also be taken up by actors who seek to organize the world in their image. Absent that, any unit of observation and especially a kind of person will remain sociologically thin – merely a paper in a journal rather than a category for organizing various forms of practice and experience.

However, even the most established forms of medical classification must overcome continual discordance or effacement in any given practical setting. As we will see in

Chapter 6, Timmermans and Berg’s (1997) discussion of the way standards must achieve

‘local universality’ is highly applicable to conditions like 22q11.2DS. At every given site where a parent may seek to bring a genomically designated syndrome to bear – a hospital, a GP, a school, the psychologist’s office, etc. – they may find institutional structures and actors who are more or less amenable. Just as Armstrong (2011) shows how even the dominant classificatory system of clinical nosology represented by the ICD faces obstacles and resistances in general practitioners’ frequently symptoms-oriented practice, the objects of genomic classification studied here are continually confronted by the recalcitrance of the

! 101! ! many sites where they seek to gain traction. Genomically designated syndromes must be assembled as networks of expertise (Eyal 2013) one doctor, advocate, clinic, testing kit, protocol and IEP at a time, and with wide variation in amenability both historically and across sites. That said, when faced with a doctor who is unwilling to let knowledge about a genetic mutation shape their clinical judgment and practice, parents can seek to enroll the larger network in their struggle by bringing prestigious journal articles, allied experts and other resources to bear.

From illness, to disease and on to etiology

A final necessary starting point for a sociological analysis of genomic designation is the idea that it is not just the experience of illness, or even the classification of disease, that is contingent upon prevailing cultural schemas, activist repertoires, institutional structures, technological capacities and so on; it is also the frequency2, ascertainment and construction of etiological – in this case genetic – findings. The sociology of science and medicine has already made the highly productive turn from a prescribed focus on the experience of illness and the social factors associated with it – the focus of medical sociology from Parsons’ 1951 conceptualization of illness as deviance (1991:288–322) onwards – to an engagement with disease itself (see esp. Mol 2002; Timmermans and Haas

2008). However, this dissertation demonstrates that we need to delve further into what is

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2 This might seem like a surprising claim, but advances in medical care can increase the size of a population (e.g. with 22q11.2DS given the high frequency of infant heart malformations) while the high national and institutional variation in prenatal testing and termination rates for genetic disorders ensure that the incidence of certain genetic mutations vary considerably according to time and place.

! 102! ! perhaps uncomfortable biomedical terrain and grapple with issues related to biological etiology.

The point is not that the ‘lesions’ that cause disease are ‘social constructions’. Far from it: in the present case, some genetic mutations do have consistent phenotypic outcomes and even more consistently have some kind of phenotypic effect. However, genetic mutations intersect with different classificatory systems and institutional structures, or what Foucault would call the ‘surfaces of emergence’ (Michel Foucault 2002:41) for genomic designation. As we will see, only certain populations are likely to be referred for genetic testing. As those populations, the ways in which we classify and treat them, and the institutions through which we do so change, so too does the meaning of etiological findings. I will show how even the same genetic mutation can therefore be associated with divergent symptomatologies and comorbid conditions, take on strikingly divergent meanings and lead to different forms of treatment, prospect horizons, social action and cultural repertoires in different times and places. In short, affixing a disease entity to a genomic abnormality does not magically obviate the social processes that mediate the way that medical classifications are understood and acted upon. We are still left with the question: ‘etiology of what?’ The prevailing nosological conditions – i.e. the kinds of people we care to explain – will be shown to decisively shape the meaning of genetic mutations even as those mutations are in turn mobilized to realign clinical judgment and practice. By extending the reach of medical sociology not just from illness to disease but all the way to etiology, we therefore get a fuller account of the way social structures and cultural schemas mediate biological abnormality.

! 103! !

The rest of this dissertation will therefore draw on fieldwork, publically available information and comparative historical research in order to explain why some genetic mutations become powerful rubrics for understanding and acting on human difference, while others remain sociologically stillborn in the pages of genetics journals. Furthermore,

I will compare the kinds of social action and cultural meaning associated with genomically designated syndromes in the 1960s and 70s to those of recent decades, providing both an explanation for genomic designation’s resurgence in recent years and a targeted analysis of the shifting terrain upon which biology, medicine and social action intersect to shape the experience of illness and developmental difference.

To preview: I will argue knowledge about any given genomic abnormality and what it means to have it is, at least in large part, a sociological phenomenon. Genomic designation is as much a question of collective action, repertoires of mobilization and institutional infrastructures as it is one of genes, proteins and physiological pathways.

Therefore, to see how genomically designated conditions like 22q11.2DS and 22q13DS achieve biosocial salience, we need to take a comparative historical perspective. What we learn is that even when human kinds are delineated at the level of the genome, they are still done so largely according to the features we care about and within prevailing structures for acting on human and especially childhood developmental difference. We will see that there are many ways to make genomically designated conditions matter, but also pay especially close attention to a model of collaborative research and advocacy that has emerged most powerfully in recent years. It is only by creating successful alliances and facilitating productive points of interface between genomic designation and other forms of classification that knowledge about a genetic mutation can transform knowledge

! 104! ! production, clinical practice and lived experience. However, we also see that some collectives of actors working with genomically designated conditions are able to transform those structures and reconfigure both clinical judgment and personal identity. In short, we will see how they are able to turn sociologically thin facts about genetic mutations into fully-fledged kinds of people.

! 105! !

Chapter 3 – Immobile mutations Nowhere to go in the 1960s and 70s and the exception that proves the rule

During 1968 a special branch of the science of genetics was drawn from the quietude of laboratories and professional journals and exposed to the glare of courtrooms and newspaper headlines.

– ‘Report on the XYY Chromosomal Abnormality’, NIMH Center for Studies of Crime and Delinquency (Shah 1970: 3)

Genomic designation has been with us now for over half a century. As we will see, almost as soon as it became possible to see, count and label chromosomes in the late

1950s, human geneticists began to find abnormal chromosome complements and delineate syndromes on their basis. However, save for one very notable exception, the first decades of genomic designation did not give rise to the kinds of clinical and social practice that would make it a suitable topic for a sustained sociological analysis. This chapter will chart that early history, delve especially into the case of XYY or ‘Supermale’ Syndrome as the exception that proves the rule, and set up a diachronic comparison with genomic designation as it functions today. In so doing we can begin to outline the epistemic, institutional and sociocultural conditions of possibility for genomic designation. We will see how, even once we have the technological capacity to observe genetic mutations, what they mean scientifically, clinically and more generally for their bearers is decisively mediated by historically and locally specific conditions of possibility as well as the networks of expertise and social mobilization formed around them.

! 106! !

The first cases of genomic designation

Once the number of chromosomes in a normal human was established, it did not take long for geneticists to discover people with abnormal chromosome complements.

After some debate throughout the 1950s as early cytogeneticists strained to observe, distinguish and count chromosomes under the microscope, it was in 1956 that a definitive study established ‘The Chromosome Number of Man’ as 46 (Joe Hin Tjio and Albert

Levan 1956). By 1960 a meeting in Denver had established an international system for numbering the twenty-two autosome pairs and identifying each sex chromosome (J.

Lejeune et al. 1960), though the 14 participants were apparently only able to reconcile the some six competing nomenclatures after a Chinese system of classification was proposed

(Harper 2006:142–6; see also Lindee 2008). This required the standardization of multiple numbering systems in use at that time, as seen in Figure 1 (J. Lejeune et al. 1960:1064):

Figure 1: The standardization of human chromosome numbers

agreed upon in Denver, 1960 (J. Lejeune et al. 1960:1064)

! 107! !

From Denver on, human geneticists would be able to refer to chromosomes and aneuploidies according to a standardized system, and we will see below how subsequent meetings laid the sociotechnical groundwork for making and reporting more fine-grained observations.

As noted in Chapter 1, the 1959 discovery by Lejeune et al. that Mongolian Idiocy or Mongolism (now of course known as Down syndrome) was strongly associated with chromosome 211 trisomy constituted a seminal moment in the nascent field of medical genetics. As a review of ‘Chromosomal aberrations in man’ published in The Journal of

Pediatrics in 1961 put it, “For the pediatrician, the association of an extra chromosome with most cases of Mongolism has been one of the most dramatic medical discoveries in recent years.” (Rappoport and Kaplan 1961:428) T. C. Hsu, an important researcher in the field, even wrote in his retrospective review of cytogenetics (1979: p. 39), “if it were not for Jérôme Lejeune’s discovery of a trisomic condition associated with mongolism, human cytogenetics would probably have died soon after the correct diploid number was determined and the chromosomes characterized. Lejeune’s report… created a new field of medicine – medical cytogenetics.” Even though several studies, beginning as early as 1932

(Waardenburg 1932, cited in Hsu 1979) had specifically postulated that Mongolism was likely due to a chromosome abnormality, Lejeune et al.’s 2 findings gave human

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 It should be noted that Lejeune et al. did not use the as-yet unavailable Denver system for number chromosomes, and at that stage only different sized groups could be confidently distinguished in any case. Indeed the autosome chromosomes are numbered according to size, with the exception of 21 and 22 which are reversed because their true size was not determined until after the 1960 conference in Denver and it was felt that ’21 trisomy’ was too entrenched to be revised (Harper 2006:146).

2 Hsu’s counterfactual, however, is a weak one. Lejeune et al. did make and report the discovery first, but other teams reported similar and indeed more systematic findings the same year (Harper 2006; P. Jacobs, Court Brown, et al. 1959).

! 108! ! cytogenetics traction in medical research and diagnosis.3 However, it also helped to direct the attention of human geneticists as their field began to transition to what we now call medical genetics. Hsu notes (ibid: p. 41) how the trisomy 21 finding “hit the scientific community like a storm” and led others to pursue similar research on people with congenital disorders, especially those characterized by multiple anomalies, or in other words: “‘funny looking’ kids became the prime targets of inquiry.”

In fact some were excited by the prospect of the new chromosomal marker pointing towards a broader phenotype. Warkany summarized a National Association for

Retarded Citizens Conference on the Etiology of Mongolism from October 1959: “Only a few months had passed since the first publication reporting the existence of an extra chromosome in Mongols. It is clear that this required a reorientation of medical thinking.”

(1960: 417) He discussed a clinically contentious case that was found to have 47 chromosomes, noting that the development of the child in question “should contribute to the question of the diagnostic value of chromosome studies.” (415) Summarizing Dr.

Jervis’ discussion of the impact of chromosomal findings on clinical thinking about

Mongolism, Warkany writes (417):

if these data are confirmed, the clinical understanding of Mongolism may be considerably enlarged since it will be based on an exact cytologic diagnosis and not only on the evaluation of a constellation of physical traits. It may be possible to solve the old problem of the “formes frustes” of the disease, a problem which recurs so often in the medical literature. Perhaps the nosologic concept of Mongolism will be extended to include forms of mental retardation heretofore eluding classification.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 3 Furthermore, three reports in the Lancet the following year demonstrated that a variant of Down syndrome could be caused by a chromosome translocation and actually inherited from asymptomatic parents (Carter et al. 1960; Penrose, Ellis, and Delhanty 1960; Polani et al. 1960), spurring interest in prenatal diagnosis (see Harper 2006:68–9).

! 109! !

In other words, even when it came to finding the ‘gene-for’ (or in this case an ‘aneuploidy- for’) a medical condition, human geneticists were immediately open to the possibility that it would substantially recalibrate its delineation. While the association with trisomy 21 has in fact only marginally impacted the delineation of Down syndrome – genetic testing is used mostly to rule out clinically ambiguous cases – this is only because of the unusually close alignment of mutation and clinical syndrome. Subsequent findings would require more ‘nosologic’ innovation.

Indeed as soon as Mongolian Idiocy was associated with chromosome 21 trisomy by Lejeune et al. (1959), with Ford et al.’s identification of a missing sex chromosome in cases of Turner Syndrome (Ford et al. 1959) and Jacobs et al.’s identification of an XXY sex chromosome composition in Klinefelter Syndrome coming the same year (P A Jacobs and Strong 1959), the search was on for phenotypes associated with monsomy or trisomy of the remaining chromosomes. This quickly led to the identification of a “super female” with three X chromosomes (P. Jacobs, Baikie, et al. 1959) and cases of 18 trisomy

(Edwards et al. 1960) and 13 trisomy (Patau et al. 1960; D. W. Smith et al. 1960). ‘Triple-

X’ Syndrome, Edwards Syndrome and Patau were thus delineated as new clinical entities and therefore represented the first genomically designated kinds of people (see Chapter 1).

While it was correctly hypothesized that many aneuploidy conditions would be incompatible with life4 (e.g. Patau et al. 1960:792), a lead article in The Lancet nevertheless speculated:

Every one of the twelve chromosome pairs of that old friend of the materia medica, Datura stramomium [a common weed], can exist in trisomic form, and !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 4 The examination of spontaneously aborted fetuses would soon become a significant component of medical cytogenetics.

! 110! !

every one has its characteristic phenotype, or, as one calls it in medicine, its characteristic clinical syndrome. Perhaps we will soon recognize equally characteristic syndromes for each of our own twenty-two autosome pairs. (Lancet, 1960)

By 1961 Lejeune and Turpin would write in a review of ‘Chromosomal Aberrations in

Man’ that, in the 22 months since they published their ‘discovery of the mongolian

Trisomy’, ‘the number of observations obtained in many laboratories is so vast that at least

23 different aberrations have been collected … [a] rate of about a syndrome a month’

(Lejeune and Turpin, 1961: 175). Crucially, they observe (p. 176): ‘The other trisomics actually described concern clinical entities less well defined than mongolism, for the very reason that in these individualization has resulted from the chromosomal findings.’ Clearly then, genomic designation was visible at the very outset of endeavors to link human difference and characteristics of the genome.

As cytogenetic techniques improved, meetings in London (1963), Chicago (1966), and Paris (1971 and 1975) would establish conventions for banding chromosomes5, allowing abnormalities smaller than whole deletions or duplications to be observed, labeled and then circulated and deemed equivalent, related or different to mutations found in labs around the world (Shaffer 2009). Beginning in 1963, a series of reports by de

Grouchy et al. (J De Grouchy et al. 1964; J de Grouchy, Bonnette, and Salmon 1966;

Thieffry et al. 1963) focused on deletions at 18p, the short arm of the eighteenth chromosome, while a team led by Lejeune discovered several cases with a deletion at 5p (J

Lejeune et al. 1963). In 1967, de Grouchy et al. proclaimed (1967:221, my translation):

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 5 Each chromosome was divided into short (‘p’ for petit) and long (‘q’, which simply follows p) arms and then labeled by region (first digit) and band (second digit), later followed by decimal points. Thus 22q13 refers to the twenty-second chromosome’s long arm, region 1, band 3.

! 111! !

“The deletion on the short arm of chromosome 18 is the first autosomal deletion described in humans.” It would not be long before these mutations would achieve the status of fully- fledged syndromes in the medical genetics literature: 5p deletion, Lejeune’s or Cri du Chat syndrome (Dumars, Gaskill, and Kitzmiller 1964) and de Grouchy/18p- or 18p deletion syndrome (De Grouchy, ibid) respectively. When Lejeune and De Grouchy turned to genomic abnormalities smaller than chromosomal trisomy or full monosomy, they were pursuing the logic inherent in the Lancet editorial quoted above: if we have the technical capacity to draw genomic distinctions between humans we should seek to construct post factum clinical categories on their basis. By entering our nosology these chromosomal syndromes, along with Edwards Syndrome and the other trisomic syndromes, represent early instances of a molecular ‘gaze’ in Foucault’s sense of a new spatialization of illness.

Genomic designation was beginning to function on the basis of abnormalities smaller than whole chromosome aneuploidies – a trajectory of development that would allow for the proliferation of new syndromes with every advance in cytogenetics or genomic testing

(Ledbetter 2008).

In 1966 Victor McKusick, widely regarded as the ‘founding father’ of medical genetics, began publishing his Mendelian Inheritance in Man catalogs of genetic conditions. Those catalogs grew precipitously over the years until the twelfth edition in

1998 (McKusick 1998), and increasingly reported disorders associated with observable mutations. Today, the catalog lives on as the Online Mendelian Inheritance in Man or

OMIM database – perhaps the leading resource for information about genes, mutations and genetic disorders. Already in 1966, McKusick stridently advocated the turn to what I am calling genomic designation:

! 112! !

In medical genetics there is little place for expressions such as “spectrum of disease,” “disease A is a mild form, or a variant, of disease B,” and so on. They are either the same disease, if they are based on the same mutation, or they are different diseases. Phenotypic overlap is not necessarily any basis for considering them fundamentally the same or closely related. (McKusick 1966:x)

Furthermore, a number of important textbooks and atlases began to be published that specifically addressed cytogenetics and chromosomal abnormalities (Digamber S.

Borgaonkar 1975; W. M. C. E. al Brown 1964; Jean de Grouchy and Turleau 1977; John

L. Hamerton 1963, 1971; Lamy and Jean de Grouchy 1967; Turpin and Jérôme Lejeune

1965, 1969; Yunis 1977); meanwhile, both new studies and review pieces continued to be published regularly in human genetics and medical research journals.

In these catalogs, atlases and reviews we see how genomic designation had gone beyond the esoteric sciences represented in specialist journals and entered the sphere of what Fleck called Vademecum science (1981:118–20): the solidification or conversion of provisional knowledge produced by individual researchers, often representing a ‘vanguard’ on a certain subject (p. 124), into the impersonal, certain knowledge of vademecum science via “the migration of ideas throughout the collective.”6 (p. 119) Genomic designation was employed in hundreds of journal articles during the 1960s and 70s – ranging from case !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 6 Fleck’s distinction between journal and vademecum science is an important one, and it offers an important rejoinder to Daston and Galison’s recent work on objectivity that used such atlases as their main source material to chart the emergence of the modern concept of objectivity in Western science (Daston and Galison 2010). As Fleck points out, esoteric journal articles tend to be far more hesitant and circumspect, embracing the fallibility of the authors pending validation by the scientific collective. My reading of the last half century of human and medical genetics suggests that, if “objectivity is the suppression of some aspect of the self, the countering of subjectivity,” (Dalston & Galison, 2010, p. 36) then much of the published work in contemporary science is more open to subjectivity and uncertainty than Daston and Galison would have us believe. Written some 75 years earlier, Fleck’s account of modern science allows us to say that by working with atlases, Daston and Galison were, in effect, selecting on their dependent variable: atlases belong to the realm of vademecum science and their very role is to present an objective, disembodied and purified account of the more fluid, guarded and circumspect representation of knowledge that is esoteric or journal science. In genetics at least, most science is less ‘objective’ than they imply.

! 113! ! studies to systematic reviews – and those findings were in turn integrated into the kind of vademecum texts that, as Fleck puts it, submits the more tentative knowledge claims of journal science to the collective scrutiny of the field and allows for the expression of more certain, objective knowledge that is suitable for broader consumption. Without question, a plethora of genomically designated conditions reached this level of collective significance as medical categories.

For geneticists then, carving out new disease entities on the basis of abnormal genomes came quite quickly and unproblematically when it became technically possible to do so. Being able to see an extra or a missing chromosome required years of sociotechnical groundwork and made for a major finding. Delineating a new ‘syndrome’ on the basis of an aneuploidy, by contrast, was unproblematic for human geneticists, as seen in

McKusick’s strident declaration quoted above. However, the syndromes they discovered were not translated beyond the level of basic science: they did not gain traction as objects of clinical research and practice or give rise to biosocial identity practices of the kind that concern theorists like Hacking, Rose and Rabinow. One finds no records of advocacy organizations, specialist clinics or even clinical guidelines for genomically designated conditions and rates of diagnosis remained low. In some sense, this is hardly surprising:

Triple-X Syndrome tends to be very mild phenotypically 7 while Edwards Syndrome is so severe that it almost always leads to in utero or infant death. Meanwhile, 18p Deletion and

5p- Syndrome are very rare, making awareness on the part of clinicians or organization on

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 7 It is worth noting that, even though Triple-X Syndrome usually causes no phenotypic symptoms (though it is associated with increased height and risk of developmental abnormalities), its observation through prenatal testing regimes has led to a significant number of terminated pregnancies, as I will discuss below.

! 114! ! their basis impracticable in the 1960s and 70s.8 Nevertheless, we will see that they have all become subjects of patient/parent advocacy and clinical practice in recent years, though other genomically designated conditions are far more advanced in this regard. Why then did precisely the same disorders – defined, after all, according to genetic mutations – fail to gain traction when they represented important scientific discoveries in the early 1960s, only to begin doing so decades later? In other words, how do we get from biomedical categories to robust clinical and biosocial ones?

Lindee (2005: esp. 92-119) has outlined the early successes and the socio-technical work done, particularly the standardization of chromosomal terminology, to make medical cytogenetics possible and hers is just about the only work in the human sciences to notice what I am calling genomic designation. Before going on to cite the Lejeune et al. (1961) article discussed above, Lindee notes (but does not develop) that reports of chromosomal aberrations in these years “included links between chromosomal aberrations and some poorly defined clinical entities that the chromosomal condition basically defined: the chromosomal difference permitted some conditions to be extracted from the undifferentiated mass of things causing mental retardation” (Lindee 2008:103, my emphasis). It is hardly surprising that Lindee does not develop this observation because no one other than medical geneticists found this differentiation to be very practicable. There simply was not much to be done with it. There was very little scope for differentiation, on genetic or other lines, to gain traction: Most of the ascertained cases were institutionalized with mental retardation, and in those settings a chromosomal abnormality mattered little to !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 8 As we will see in later chapters, contemporary communications technologies are a vital resource for the organization of research and community on the basis of rare genomically designated conditions.

! 115! ! the patient, caregiver or the usually absent families. As we saw, medical geneticists published numerous papers, textbooks and atlases on chromosome abnormalities, establishing genomic designation in what Fleck calls ‘esoteric’ and ‘Vademecum’ scientific domains. So while geneticists took to institutions for the mentally retarded and the like in search of abnormal genomes, we will see in Chapter 6 how it is only in recent years that institutions geared towards people with developmental disorders began to take an interest in the geneticists’ findings and adopt genomically designated syndromes as objects of practice.

But what scope was there at this juncture for genomic designation to impact what

Fleck calls ‘popular knowledge’ and the way abnormality is understood and acted upon in practice? In 1964 London’s Eugenics Review reprinted the annual Galton lecture entitled

‘Some mechanisms of chromosome variations and their relation to human malformations’ that dealt almost exclusively with the first years of medical cytogenetics – trisomy 21, Cri du Chat and so on as well as a series of translocations with unclear pathological implications (Jan A. Böök 1964).

IN MEDICAL GENETICS we are concerned with genetical material that can be transmitted from parents to children and cause individual deviations of development, differentiation and function of such a significance that conventionally we would call them diseases or defects. This common denominator in basic etiology, very broadly speaking, exerts its important influence on the individual as well as on the population. Phrased differently, the two foci of interest and concern are the patient with his personal medical problem and the population with its epidemiological problems presented as a broad spectrum of genetical diseases or defects. (Böök, 1964, p. 151, my emphasis)

In passages like this we see how genomic designation could have become an important object of early twentieth century eugenics research and policy. However, medical genetics did not become an important tool of post-WWII biopower, despite Böök’s emphasis on the

! 116! !

‘two foci’ of medical genetics – individual and population health (Foucault 1990, 2003). In part, it was the very association with eugenics that made such a broad deployment of medical genetics impracticable (see e.g. Finucane et al. 2003 on the challenge posed to genetic counseling’s adoption in schools by its association with eugenics). Instead, rates of diagnosis were low and there was very little scope for mobilizing genetic difference to inform treatment, self-understanding or social mobilization.

Even though they could be reliably observed and described, genetic mutations did not find a ‘surfaces of emergence’ or social-institutional circuitry through which they could be mobilized as a meaningful basis for understanding human difference. No matter how much they captivated the esoteric field of human genetics, chromosomal anomalies could not, for the most part, translate the interests of clinicians, patients and their families or other potential stakeholders. While a framework for making standardized, commensurable observations of chromosomes had been established by and for geneticists, it would take years of work and, I will argue, new conditions of possibility in order to make genomically designated conditions into the kind of nosological categories that could inform clinical practice, social organization and identity formation or be useful in channeling material and symbolic resources. In short, even though it gave rise to numerous

‘syndromes’ in the biomedical literature, genomic designation was not a practice that could create considerable commensurability between other fields of knowledge production and practice.

Nevertheless, there are two important exceptions to genomic designation’s relative impotency as a form of classification that actually mattered to people in the 1960s and 70s:

XYY or ‘Supermale’ Syndrome and prenatal genetic testing. In both, we see how the

! 117! ! imperatives of early twentieth century eugenics, which had been present since the birth of both medical genetics (Kevles 1985) and molecular biology (Kay 1996), were very much active in genomic designation’s early years and, despite vast differences, remain visible through to the present.

XYY Syndrome

To reiterate, during its first few decades genomic designation did not substantially inform treatment, identity formation or social mobilization, nor did it gain much traction in the public sphere. There was, however one very notable exception: males who have an extra Y chromosome or XYY Syndrome. As we will see, it may be the exception that proves the rule.

When I quoted Hsu’s retrospective account on the role of ‘funny looking kids’ as the primary target of cytogenetic investigations, I intentionally omitted the rest of the sentence and the two immediately following:

… but inmates of mental institutions and criminal institutions also contributed heavily to our knowledge of human cytogenetics. Because of their association with abnormalities with sex chromatin, several kinds of congenital maladies received early attention. As a result, the relationships between the abnormalities in sex chromatin and sex chromosomes were clarified. (ibid: 41)

Indeed our overview of the early chromosome abnormalities showed that, while children with congenital disorders proved the most fecund terrain for early cytogeneticists, people with sex disorders like Turner and Klinefelter Syndromes were a close second. However, if it seems as though the transition between the first and second sentences is nonsensical or almost cryptic it is likely because Hsu was writing at a time, the late 1970s, when cytogenetics found itself at the center of a major controversy that grew out of precisely this

! 118! ! pairing of research subjects drawn from prisons and psychiatric institutions with the analysis of sex chromosome abnormalities. Tracing the history of that controversy and the way XYY Syndrome became perhaps the most famous and certainly the most infamous of genomically designated conditions, helps make the broader point about the limited scope for genomic designation to matter outside of human genetics during its first decades.

Shortly after the association of sex chromosome aneuploidies with Turner and

Klinefelter Syndromes, Jacob et al. (1959) reported cases of women with an extra X chromosome and genomically designated a new ‘Super Female’ or XXX Syndrome. Only two years later the equivalent extra-Y chromosome syndrome was reported by Sandberg et al. in a Lancet ‘Letter to the Editor’ (1961). Simply titled ‘An XYY Human Male’,

Sandberg et al.’s letter described (p. 489) a “forty-four-year-old white man of average intelligence and without physical defects, despite the 47 chromosomes found in his marrow and blood, started work at seventeen after two years of high school.” However, “Each of his [two] wives gave birth to at least one abnormal child” and each had one miscarriage: one eighteen-year-old daughter found to have no breast development or, following an operation, internal sexual organs, though she had a normal XX sex chromosome complement; further, a 22-month-old daughter from the second marriage was a “typical

47-chromosome mongoloid.” Five other children were “apparently living or well,” though because they were adopted it was not possible to study their chromosomes. (ibid) The

XYY subject “claim[ed] to have a normal libido, and results of examination were unremarkable.” In other words, he was not sexually abnormal despite the extra Y chromosome. They continued: “He has somewhat large facial features, is obese and weighs

287 lb., and has a neurodermatitis, an umbilical hernia, and, in the left mandible, has had a

! 119! ! cystic lesion for many years. The buccal-mucosa cells were chromatin-negative.” Finally,

Sandberg et al. discuss the appearance of the extra chromosome in relation to the recently standardized nomenclature, arguing that it looks more like the Y chromosome than the similar-looking 21 and 22 autosomes, while also pointing to the fact that “Absences of the severe mental and physical stigmata regularly associated with autosomal trisomy is further evidence” for the conclusion that “This case may therefore be considered as an XYY male.” (ibid) Other reports of XYY males with a variety of abnormalities followed in leading journals over the next few years (T. Hauschka et al. 1962; Ricci and Malacarne

1964; Avery A. Sandberg et al. 1963; Vignetti, Capotorti, and Ferrante 1964), but the clinical findings were scattered and only a handful of cases were identified in total.

The ‘criminal chromosome’

XYY started to gain significant scientific and popular attention when the leading human geneticist and first author on the aforementioned XXX paper, Patricia Jacobs, and her team reported a remarkably high prevalence of XYY males in Scottish prison and asylum populations (Patricia A. Jacobs et al. 1965). Their study was the result of a hunch

(interview with Patricia Jacobs in CD supplement to Harper 2006): Previous studies had found that around 1% of males in institutions for the ‘mentally sub-normal’ were chromatin-positive, indicating the presence of an extra X chromosome. However, a study of criminal and difficult-to-manage men of low intelligence in Sweden had found a rate of

2% while an unpublished study in England communicated to the authors had found a rate of 2.2% in 942 men tested in similar institutions and, crucially, seven of the 21 chromatin- positive men had an XXYY chromosome complement. By contrast, a study of 2,607

! 120! !

“ordinary mentally sub-normal males” found that only two of 28 of the chromatin-positive cases were XXYY, leading Jacobs et al. “to wonder whether an extra Y chromosome predisposes its carriers to unusually aggressive behavior.” They thus set out to survey rates of XYY chromosome complements in a “survey of mentally sub-normal male patients with dangerous, violent or criminal propensities in an institution where they are treated under conditions of special security” (p. 1351) near Edinburgh.

Of the 197 samples taken twelve had a chromosome abnormality, including seven cases of XYY and one case of XXYY. This finding led Jacobs et al. to suggest that, while

“Very little is known about the XYY male, as only a few cases have so far been described… from whose description no clear picture of the XYY male has emerged.”

However, it was clear that:

the finding that 3.5 per cent of the population we studied were XYY males must represent a marked increase in frequency by comparison with the frequency of such males at birth. On theoretical grounds XYY males at birth must be less common that XXY males, who form approximately 0.2 per cent of the new-born male population. We have examined 266 randomly selected new-born male babies without finding an XYY individual, and we have also examined 209 randomly selected adult males, again without finding one with an XYY constitution. In addition, we have examined the chromosomes of approximately 1,500 males for a variety of reasons and only one of these was found to have an XYY sex chromosome constitution. (p. 1352)

Furthermore, those seven were, on average, over six inches taller than their XY counterparts (67.0 vs. 73.1 inches) with men over 73 inches tall having “an approximately

1 in 2 chance of having an XYY constitution.” It was these associations – institutionalization in mental-penal settings and above average height – that would drive both research on and, crucially, ascertainment of XYY as it became a major topic of research over the following decade.

! 121! !

A leading article in the Lancet (1966) lauded the discovery of this “new and very interesting syndrome,” (p. 583) which they titled (along with the article) ‘The YY

Syndrome’. While previous studies “seemed to present no very convincing picture,” they enthusiastically repeat Jacobs et al.’s 1965 finding of seven XYY cases out of an initial study of 197 men in a high-security Scottish institution as vs. only one in 2000 male controls. Discussing a paper in that issue of the Lancet that built on Jacobs et al. to make it

9/315 patients (Price et al. 1966), they summarize: “No new diagnostic features emerge. 8 of the 9 are high-grade mental defectives. Their type of delinquency was not clearly different from that of other patients in the hospital, though most were aggressive and violent.” (ibid) Physically, above-average height was the only notable feature (six of nine were over six feet tall), prompting the Lancet to proclaim: “The height is so striking a feature that it has been said that half of all men over 6 feet in special-security institutions are YY – probably an exaggeration, but unlikely to be a very gross one.”9 In contrast to other chromosome anomalies, which are characterized by “largely non-specific mental defect,” in XYY “we seem to have a fairly specific mental disorder associated with a !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 9 One letter by a pathologist published in response to this point (Park 1966:1468) noted that this is “a matter clearly relevant to current events and, as you suggested, one of much psychiatric and medicolegal significance: for example, is a YY psychopathic criminal (to be) regarded in law differently from an XY psychopathic criminal? I wonder whether our psychiatrist colleagues could tell us what the chances are of restoring a YY psychopathic individual to full social insight and responsibility by therapy of whatever type. Should the chances be slender or nil, much of the time directed towards rehabilitation is in effect time wasted, and this, if true, would be a pity, for a great deal of sympathetic thought and care are devoted to these and other delinquent persons; psychiatrists, prison staffs, and others may be blaming themselves unnecessarily for failure to achieve better results than they do. No doubt the implications of aneuploidy vis-à-vis the criminal population are well known to Ministers of the Crown, commissions appointed to report on prison affairs, and other students of penology. So, at any rate, we should hope, for they greatly matter.” While more extreme and simplistic than most subsequent discussions, this letter would be the first of many proposals relating to XYY and the management of criminal populations. As we will see, these ranged from using XYY as grounds for innocence by way of insanity (even with normal psychiatric evaluations) to systematic abortion.

! 122! ! highly specific lesion – a matter of considerable psychiatric importance.” (my emphasis)

This condition, they explained, was of immense ‘chromosomological’ interest, joining as it did the ‘testis-evoking gene’ and the ‘Indian hairy-ears gene’10 as exceptions to the idea that the Y chromosome is genetically inert (ibid). Not only was this chromosomal lesion seen to have a distinctive psychiatric phenotype – a finding that held the promise of making medical genetics highly relevant to what Rose has called the ‘psy-disciplines’

(Rose 1998) – but that phenotype was such that “XYY must make its small but significant contribution to the country’s delinquent population.” (ibid)

Similar reports followed (Patricia A. Jacobs et al. 1968; Price et al. 1966; Price and

Whatmore 1967; Telfer et al. 1968) and XYY Syndrome came to be increasingly associated with the profile of a ‘Super-Male’, characterized by acne, increased stature, aggression, criminality, sexual deviance and mental ‘subnormality’. Its prevalence was estimated at the time to be around 1 in every 2000 men. There followed a decade of intense interest from geneticists, but unlike other genomically designated syndromes in the 1960 and 70s XYY also gained traction in criminology, medical-penal institutions, courts, popular culture and major media outlets.

In April 1968 a New York Times editorial addressed some of the dilemmas raised by XYY and the way they “reviv[ed] the old argument as to whether it is nature or nurture that creates a criminal.” (New York Times, April 23 1968: 46) After reviewing the state of

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 10 A letter in response later that year (Collier 1966) stated (p. 1036): “The condition of hypertrichosis pinnae auris (H.P.A.), the Y-linkage of which is still debated, has also been reported in Italians, Israelis, and Singhalese; I report here its occurrence in a native of the British Isles who attended this hospital.” Even though the Lancet had not reported an association between said trait and XYY, the author thought it prudent to conduct a karyotype nonetheless. While no abnormality was detected, he did find that “the Y chromosome was at the upper limit of normal, being always larger than the no. 21 and in one case equal to the no. 17.”

! 123! ! knowledge about XYY the Times writes, “Should such persons be held responsible for their crimes, or treated as victims of conditions for which they are not responsible, on a par with the criminally insane?” They cautioned against hasty conclusions, noting “many XYY males are neither criminals nor mental defectives,” and speculated that people with XYY may have problems in sexual development during adolescence, exacerbating the challenge of growing and therefore engendering deviance. Finally, they suggested a “fruitful line of research may stem from early identification… which perhaps would enable preventive measures” to help people with XYY. However, while early identification and intervention may be the gold standard when it comes to genetic disorders today, we will see that precisely such a line of research would be the undoing of XYY as a major topic of inquiry and criminal management.

XYY also began to be invoked by defense teams in criminal trials, but with mixed results. One man standing trial for murder in Australia was acquitted partly on the basis of his XYY karyotype. However, a jury in France rejected the XYY-defense in the trial of

Daniel Hugon (Garrison October 15 1968: 5). The same report noted that a prominent

French biologist suggested that people with chromosome imbalances should have to carry cards naming their defect, but the proposal met opposition from the leading communist newspaper L’Humanité who countered that it would stigmatize thousands of people as would-be criminals and amount to a form of racism. Jerome Lejeune, who established the association between Mongolism and trisomy 21, was one of two medical experts to testify.

He said that “the born criminal does not exist,” but that people with chromosome abnormalities have a 30% greater chance of becoming criminals and Hugon had been doomed to be a sick man whose hereditary affliction precluded the exercise of normal

! 124! ! responsibility. The prosecution simply noted that many XYY men lived normal lives and that an extra Y chromosome could therefore be no more than a contributing factor. The jury delivered a guilty verdict in less than 40 minutes (ibid), but it appears that the XYY diagnosis gained Hugon a significantly diminished sentence of seven years (N.A.

1968:892) The first such case in the US saw one Sean Farley of Brooklyn seek immunity from prosecution for the rape and murder of a 49-year-old woman in Queens. Although psychiatrists had declared him competent to stand trial, a cytogenetic examination was undertaken (possibly because Farley was 6’8”) and he was found to be XYY (New York

Times, Oct 18th 1968, p. 18). However, a geneticist giving expert testimony argued that

XYY alone could not cause criminal behavior and Farley was convicted on the charge of murder and given a 25 year to life sentence (N.Y. Times, April 24, 1969, at 53, col. 1). A number of similar cases followed, but it is not clear if the use of XYY by defense attorneys was ever successful in gaining acquittals for XYY men.

However, the trials did help gain plenty of attention for XYY with a story even appearing on the front page of the New York Times as seen in Figure 2 below (Lyons

1968a). Coverage reached new heights when it was reported that the notorious Chicago serial killer, Richard Speck, had XYY Syndrome. Stories appeared in several leading newspapers and magazines (Anon 1968a, Anon 1968b, Anon 1968c; Lyons 1968b; Stock

1968), and partly as a result XYY left a lasting imprint in popular culture. Thus even though it turned out that both a leading human geneticist, Hideo Sato, and Speck’s head lawyer were incorrect in ascribing him an XYY chromosome complement, the episode remains one of the most important in the XYY story. When one of the leading U.S. XYY researchers, Mary Telfer, was asked about the case she said, “if I had to pick anyone who

! 125! ! fit the XYY pattern, I would have chosen Mr. Speck.” His features and record, she averred, were good evidence that he was XYY.

Figure 2: New York Times, April 21 1968, p. 1

A letter by Patricia Jacobs and her boss, William Court Brown (Court Brown,

Price, and P. A. Jacobs 1968:513) expressed their mixed feelings about the use of XYY in criminal trials as well as the coverage it was beginning to receive in the popular press:

The defence in both cases might have been different had it not been for the discovery in 1965 and 1966 of the high incidence of XYY males in the British maximum security hospitals, an incidence so high that it was clearly not due to chance, nor were there any grounds for suspecting it due to biased sampling. This discovery has led to a world-wide search for these males in maximum security institutions, in prisons, among juvenile delinquents, and among the psychiatrically disturbed. At the same time the discovery has itself been discovered by the popular news media, and a good deal of nonsense has been written and broadcast about the "born criminal" and about the "criminal " chromosome.

! 126! !

Crucially, they pointed out two “disturbing features” of present knowledge about XYY:

First, “the bulk of our information on XYY's [sic] is based on the examination of adults found from the surveying of groups of men which by definition consist of men the great majority of whom, if not all, have criminal records.” Second, there were no good estimates of the incidence of XYY either at birth or in the general population, making comparisons with the selected group impossible. In their own research they had seen a great range of cases, “from the apparently normal through those with a mild personality defect to those who are severe psychopaths,” and they wanted to at once stand stridently by the significant overrepresentation of XYY caseloads in mental-penal institutions while pointing out the wrongheadedness of determinist ideas about a ‘criminal chromosome’. What was needed was large-scale research to determine the true prevalence and developmental implications of an XYY chromosome complement. As Court Brown put it in an extensive review of

XYY that same year, “In the end there can be no substitute for an extensive and prolonged study of newborn children” (Court Brown 1968:357).

However, by decade’s end the association between XYY and criminality began to be more decisively called in to question. At the Bar Harbor genetics course – an annual event sponsored by the March of Dimes and run by Victor McKusick that brought together geneticists and interested parties from other disciplines, the media and various stakeholders to promote the value of medical genetics – it was reported that XYY was likely far more common than previously thought. Evidence from Edinburgh, New Haven, Pittsburgh and

London Ontario were presented that suggested a prevalence rate of as high as 1 in 300 men based on examinations of 3,700 newborns (roughly half male). As Dr. Park S. Gerald of

! 127! !

Harvard put it in a presentation to a group of science writers reported in the New York

Times and elsewhere, this meant there could be one third of a million XYY men in the late

1960s US. Gerald said it was unclear if XYY ‘doomed’ its bearers to a life of crime and violence, but if so “’we have to deal with these people differently from the ordinary criminal. They should be restrained, not punished.’” He called for a long term prospective study of 50,000 male infants and a 10-15 year follow up of the XYYs. Such a study would cost around $750,000, he said, and was going to go ahead at Harvard “’until the Vietnam war cut off our funds.’” (quoted in Brody 1968:34; see also Black 1968; Writer 1968)

Just months later the Nobel-prize winning biologist Joshua Lederberg took to the

Washington Post to outline the ‘considerable discussion’ generated by XYY at that year’s

American Society of Human Genetics (1968). He pointed to the now widely-recognized need to obtain good population estimates for XYY, but also the wide variation in existing estimates from newborn studies. Thus while he was clear that there “is enough evidence to support the concept that XYY males are abnormally predisposed to… the most serious difficulties in their social adjustment,” he stressed that “we have no idea about the roots of the problem or just what fraction of XYYs will make such miserable failures of their own lives and inflict so much harm on others.” However, Lederberg also foresaw some of the dilemmas related to newborn studies: Our lack of understanding of “the biology of violence” meant that it was “of the greatest importance to follow the development of

[neonatally identified XYY] children through adulthood.” At the same time, he argued that

“it would be a tragic injustice to identify them” because doing so might bias both their development and the evidence gathered. Finally, he wrote:

! 128! !

One line of thought would lead naturally to the sacrifice of XYY newborns as a more humane disposition than allowing them to live out their predisposed fates. To say this is to reject it in favor of the alternative: to learn how to restore to them the possibility of membership in the human community. Indeed we should remember that even a single Y chromosome already conveys a strong disposition to violence in our species.

In sum, a more complex story about XYY was now widely accepted, as was the need for unbiased prevalence and incidence studies. However, even in its more liberal iteration, that story still carried with it the idea that the increased maleness conferred by the XYY chromosome complement brought significant risks to both its bearer and society more generally.

By 1969 it had become even clearer that XYY’s incidence was considerably higher than previously thought, at more like 1 in 500-1000 rather than 1 in 2000, calling into question the strength of its association with criminality and other traits observed in prisons and asylums. There could be little doubt that there were many men with an extra Y chromosome who had not been brought to the attention of criminal justice or psychiatric institutions. In other words, it seemed inevitable that there were plenty of ‘normal’ XYYs out there. A paper in the New England Journal of Medicine by Sergovich et al. (1969) conducted a karyotype analysis of 2159 consecutive newborns, with 1066 successful cultures from newborn males obtained. They found four XYY males suggesting a prevalence of ~1:250, which although they acknowledged could be exaggerated by statistical chance, indicated that the prevalence of XYY “is far higher than the previous estimate of 1 in 2000 male births.” This, they argued, challenged existing knowledge about the profile of men with XYY:

Our finding has particular relevance to the interpretation of data derived from the studies of the chromosome complements in highly selected populations of neurologically inadequate male criminals. Although it is clear, on the basis of information presently available, that a

! 129! !

strong correlation exists between the XYY complement, peculiar behavioral abnormalities and tall stature, the association may not be so strong as has previously been implied… It may be that chromosome anomalies contribute to a spectrum of neurologic deficit that may become manifest in some persons as overt mental handicap and in others determine a more subtle form of deviation, reflected perhaps only as antisocial behavior. Consequently, we urge that one be cautious in accepting the interpretation that XYY is specifically associated with criminal behavior and particularly so with reference to the medicolegal validity of these arguments. (ibid: 854)

Thus they suggested, “In no other syndrome is the necessity for prospective study more evident than in the XYY genotype.” (ibid)

The paper received coverage in the New York Times under the title ‘Scientists

Doubt “Criminal” Genes’ (April 18 1969), noting “there may be well-adjusted persons with an XYY make-up, casting doubt on its so-called predisposition to criminal tendencies.” Sergovich is quoted saying that XYY “does influence behavior, but it probably isn’t ann [sic] all-or-nothing phenomenon. There’s too much diversity among the

XYY’s we have studied to say outright that an XYY man is inevitably going to have abnormal behavior.” In reference to the ascertainment bias of previous studies, Sergovich said, “’if you look only at abnormal populations, you will find only abnormal XYY’s [sic].

But unless you study the population at large, you can’t say how many XYY men may be functioning reasonably normally.” An article in the Times the next month (May 6 1969:

93) reported that 200,000 men may have the XYY ‘genetic make-up’ but that, quoting a speech from the American Psychiatric Association meeting that year, “there is growing evidence that many XYY individuals are stable, law-abiding citizens… It appears that the

XYY male has been falsely stigmatized.”

However, these more attenuated findings about the XYY-criminality association did not undermine the biomedical and public fascination with it. Hence an NIH-funded

! 130! ! study in the American Psychologist could note that only a small (if wildly disproportionate) minority of violent crimes were likely committed by XYY males, but then go on to write:

Yet, quite apart from questions of law and ethics, the XYY genotype may have importance exceeding by far its numerical impact in contributing to our understanding of aggressive behavior. As previously noted, the Y chromosome is the male determining chromosome; therefore, it should come as no surprise that an extra Y chromosome can produce an individual with heightened masculinity, evinced by characteristics such as unusual tallness, increased fertility (although most XYYs do not have children, some have produced as many as 10)11, and powerful aggressive tendencies. The XYY genotype may be seen as highlighting the association between maleness and violence. (Jarvik, Klodin, and Matsuyama 1973:679–80)

Furthermore, Jarvik et al. hoped that “With the XYY syndrome as a research tool, it is conceivable that advances can be made in our understanding of the complexity of factors that together tend to either produce or curb aggressive responses.” In an era concerned with overpopulation and rising rates of violent crime, this was no small matter: “as our population continues to spiral upward at an alarming rate and our planet becomes crowded to its very limits.... Disaster will follow if we do not make a concerted effort to understand ourselves now. Hopefully, the XYY genotype can contribute to such an understanding.”

(ibid, p. 80) In sum, the emerging probability that most men with an extra Y chromosome were ‘normal’, socially adapted people did not halt the rationale for continuing research on

XYY Syndrome and criminality. However, amidst the newly recognized uncertainty of outcomes, there was an emerging consensus among both critics of the XYY-criminality nexus like Sergovich and promoters like Gerald and Jacobs (above): prospective studies were the only way to get to the bottom of the XYY conundrum.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 11 Today, infertility is more likely to be investigated in XYY Syndrome (Wong et al. 2008).

! 131! !

Even if the XYY-criminality association turned out to be looser than previously thought, it did not halt public, state and biomedical fascination with the extra Y chromosome. Newsweek was yet to publish its XYY piece entitled ‘Congenital Criminals?’

(Newsweek 1970) and while the tenor of published work on XYY became more nuanced the basic object of enquiry remained. Against this backdrop a two-day conference was hosted by the National Institute of Mental Health’s (NIMH) Center for Studies of Crime and Delinquency in June of 1969. In attendance were thirteen researchers from leading

American universities, at least two of whom would be embroiled in the episode of public contention outlined below, researchers and heads of several other NIMH centers, and representatives from the Department of Justice and the National Institutes of Health (NIH).

The resulting report, written by the Center for Studies of Crime and Delinquency’s ‘Chief’,

Saleem Shah (1970), was intended for a general audience, researchers and professionals, but (p. 1) “especially addressed to persons who are most likely to be confronted with important questions and decisions pertaining to the topic discussed, e.g. lawyers, judges, administrators of correctional, mental health and related programs, research administrators, legislators and policy makers.” Shah captured the exceptionality of XYY as a genomically designated condition well (p. 3; my emphasis): “During 1968 a special branch of the science of genetics was drawn from the quietude of laboratories and professional journals and exposed to the glare of courtrooms and newspaper headlines.” Given the existing knowledge about XYY, the aforementioned legal issues and the “many press and broadcast stories [that] have appeared about the allegedly antisocial propensities of persons with the

XYY chromosome abnormality,” there were “some important medical-legal-ethical questions” that needed to be addressed (p. 4).

! 132! !

However, in the absence of biologically proven pathways from XYY to behavior or sufficiently large-scale, unbiased prevalence or incidence estimates to ground comparisons with institutionalized caseloads, it was not possible to accurately address the behavioral implications of having an extra Y chromosome. Thus a “pervasive note struck throughout the conference was the need for more knowledge obtained through extensive and meticulous research,” (p. 34) with an emphasis throughout on the need for large, unbiased samples. In attendance were Park Gerald and Stanley Walzer of Harvard and Children’s

Hospital Medical Center in Boston, whose subsequent NIMH-funded study into the incidence and developmental course of boys with sex chromosome abnormalities we will turn to presently. First, it should be noted, that participants at the NIMH conference realized the need to balance the “scientific as well as societal need to conduct further research” on XYY with the “important social values which require that the rights, welfare, confidentiality and privacy of research subjects be safeguarded in such research endeavors.” (ibid, p. 4) Nevertheless, the two major studies funded following the conference attracted a backlash from which the XYY-criminality nexus would never recover.

Opposition and the undoing of the XYY-criminology nexus

The aforementioned conference led to the NIMH funding at least two significant

XYY studies at Harvard and Johns Hopkins that Shah hoped would yield better prevalence and incidence data (Shah, quoted in Bauer et al. 1980:3). Both, however, would attract the ire of activists concerned about children’s rights and the lingering specter of eugenics.

! 133! !

The first was led by Shah’s frequent collaborator (see below), Digamber

Borgaonkar of Johns Hopkins University, and received $300,000 over three years. The project, approved by Maryland’s Departments of Health and Juvenile Services and federal agencies, planned to test some 15,000 boys for XYY – around 6,000 from juvenile jails and 7,500 drawn overwhelmingly from poor black families enrolled in a free medical program at Johns Hopkins. The initial protocol did not include obtaining consent from parents on the grounds that the XYY tests were to be conducted on the same blood samples used for routine anemia checks, nor did it include informing them of results unless requested. However, they did plan to give the results to the juvenile correctional institutions involved and, according to the director of Juvenile Services, those would

“probably be passed on to the courts for whatever use they can make of it.” (cited in Bauer

1972:342) Groups like the ACLU as well as concerned parents, doctors and lawyers rose up in opposition to the study, both for its lack of a consent procedure and for the potential abuse of knowledge about the XYY status of juvenile detainees who might be given harsh sentences even for minor offences on the grounds that they were genetically predisposed to aggression, violence and criminality. By February of 1970 a lawsuit was filed against both

NIMH and Johns Hopkins, and Shah temporarily suspended the study pending the revision of consent procedures.

By May of 1970, however, tests were resumed. Consent was now sought from parents, but without any information about XYY or which abnormalities were being tested for and with the form noting that “The chromosome study ordinarily costs about $100…

[but] no fees will be charged to you.” While Borgaonkar and his colleagues continued to publish on XYY over the coming years, it is unclear whether they ever completed the full

! 134! ! survey. At the 1972 Annual Meeting of the American Society of Human Genetics in

Philadelphia, Borgaonkar (1972:13a) reported on progress “in an ongoing chromosome survey,” the ‘socio-legal issues’ related to consent and the results of “correspondence with approximately 15,000 parents.” Chromosomal analysis was successfully performed on around 3,300 boys and XYY results were as follows: 2 of “1,800 institutionalized juveniles”; 2 of “300 mentally disturbed institutionalized juveniles”; 1 in “200 boys in a private center for emotionally disturbed children”; none in “1,000 normal boys.” However, the discussion of samples in Borgaonkar’s subsequent papers on XYY strongly suggests that the survey of institutionalized Maryland boys did not obtain chromosome analyses or follow-up research on anything approaching the proposed volume of cases.12

The other project referenced by Shah was led by two participants of the 1969 XYY conference, Stanley Walzer and Park Gerald, and conducted at Harvard and the Boston

Hospital for Women. Their goals were threefold: 1) by screening a large, unbiased sample of newborn boys at Boston Hospital for Women (the plan was actually to screen every baby born at the hospital over several years) they hoped to provide a baseline incidence rate for XYY (and XXY) against which cases ascertained in particular institutional settings could be measured (similar studies were underway in Canada and Scotland); 2) conduct a

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 12 A 1974 paper by Borgaonkar’s team (John Money et al. 1974:372) included two XYY cases from the private center for disturbed youth and three from a larger survey of institutionalized inmates “legally classified as ‘defective delinquents’ and over 6 ft (183 cm) tall.” Another paper the following year examined” Three [subjects] located through screening at correctional institutions, three from screenings at institutions for the behaviorally and emotionally disturbed, and two in screenings at a school for behaviorally and emotionally disturbed children” as well as “Five subjects were referred by clinicians, one by an attorney, and one by a parent.” (J Money, A. Franzke, and D S Borgaonkar 1975:1537) This very interesting paper argued that social class strongly mediated the type of treatment behaviorally disturbed XYY men were likely to receive – correctional facilities like jails and reformatories for lower and lower-middle-class subjects and mental hospitals for the upper-middle-class.

! 135! ! longitudinal examination of the ones found to have a sex chromosome abnormality in order to provide and a detailed evaluation of the developmental course of XYY from infancy through adolescence and into adulthood; 3) provide early diagnosis and intervention to avoid or mitigate the adverse outcomes associated with sex chromosome abnormalities, especially XYY. This research received an eight-year, $465,000 grant from the NIMH

Center for the Study of Crime and Delinquency.

Not long after the study got underway, a group of professors at Harvard led by biologist Jonathan Beckwith raised grave concerns about the study and what they called the ‘XYY myth’ in particular. They charged that it was not possible to obtain fully informed consent in this case. Furthermore, they argued that with all the publicity about the ‘criminal chromosome’ there was a serious risk of biasing expectations against boys found to have XYY, leading to a self-fulfilling prophecy, and finally that this kind of essentializing research diverted attention and resources away from the real, socioeconomic causes of violence and crime (Bauer et al. 1980:197; Jon Beckwith 1975; Jon Beckwith and Jonathan King 1974; R. Pyeritz, J Beckwith, and L Miller 1975). When internal review procedures including votes by faculty on the matter failed to halt the study, Beckwith, MIT biologist Daniel King and others began to publically oppose the study under the banner

‘Science for the People’, attracting coverage in leading media outlets in the process (e.g.

Brody 1974b; Knox 1974). Walzer agreed to have his work periodically reviewed and the study design potentially modified by a Harvard committee, calling it “a good and sensible request.” “’If it appears that the vast majority of XYY’s [sic] are developing normally, it may be better for them not to know, not to be identified at birth,’ Dr. Walzer said. ‘But right now we simply cannot answer this question. That’s what this study is attempting to

! 136! ! do.’” Beckwith called the committee a ‘whitewash’ and criticized the lack of public involvement, contravening Department of Health, Education and Welfare guidelines.

Walzer rejected criticism that the XYY ‘myth’, as Beckwith et al. described it, would have adverse outcomes for children. Instead, he argued, his careful counseling and the potential for early intervention and therapy if behavioral problems did arise far outweighed the risks.

(Brody 1974a)

Eventually, the study was discontinued. As the New York Times reported (Brody

1975:6), this was due to:

continued controversy over the study’s ethics and potential scientific value. The chief investigator, Dr. Stanley Walzer, a psychiatrist at Harvard University, said in an interview that while he had originally intended to stop screening in any case ‘sometime this year,’ he was also worn down by harassment, unrelenting controversy and the threat of further opposition to his work by groups supporting children’s rights. He said it was impossible to continue working in the atmosphere that had been created. [my emphasis]

The article notes that it was XYY and the issues raised by Science for the People that drove the controversy, despite reviews by five professional committees at various institutions that supported the study, and that the Children’s Defense Fund was about to begin a campaign against the research before it was announced that it would be halted.

Walzer reported receiving threatening phone calls at his home, and bemoaned: “These groups could make a difference. They could help an investigator, make him change his mind in some cases… I think more people should participate in decisions relevant to research, but not in the way it happened to me.” (Brody 1975:6; see also Culliton 1975;

Weiss 1975) Beckwith et al. were adamant, however, that the XYY study and indeed the entire research program of which it was a part was a grave danger to men and boys found to have an extra Y chromosome and part of a broader, insipidly eugenicist failure to seek

! 137! ! the social causes and remedies of crime (Jon Beckwith and Jonathan King 1974; R. E.

Pyeritz et al. 1977).

By 1977 the New York Times was discussing XYY in more critical terms.

Following a brief discussion of Lombroso’s late eighteenth century theories about inherited criminality and the phrenological characteristics of criminals, the author reports that “A more recent biological theory linked violent criminality with the so-called XYY defect in men.” She notes the findings from studies with criminal populations, and (re)quotes geneticist Fred Sergovich as saying, “if you look only at abnormal populations, you will find only abnormal XYY’s.” (Adams 1977: 188) She discusses a study conducted in

Denmark by a research psychologist from Princeton which found no evidence that XYY offenders are more likely to commit violent crimes than XY offenders and even suggested that higher conviction rates among XYY men “may reflect a higher detection rate than a higher rate of commission.” Having been effectively tied to the specter of eugenics, first by groups like Science for the People and then more generally, XYY was truly receding as a legitimate explanation of violence and criminality. Hence when Stephen Jay Gould added an epilogue to his chapter on ‘Measuring Bodies’ in his famous 1981 critique of race science, phrenology, eugenics and intelligence testing, XYY would be his main piece of evidence for the idea that “We live in a more subtle century, but the basic arguments never seem to change… The signs of innate criminality are no longer sought in stigmata of gross anatomy, but in twentieth-century criteria: genes and fine structure of the brain.”

(1996:173) Gould dismissed the XYY program as ‘a myth’ based on “elementary flaws of method” and “the singularly simplistic notion that since males are more aggressive than females and possess a Y that females lack, Y must be the seat of aggression and a double Y

! 138! ! spells double trouble,” arguing that we would do better to attend to the social ills and forms of oppression rather than “the determinist philosophy of blaming the victim” (ibid 174-5) if we want to understand modern crime and violence.

Even Walzer, Gerald and Shah, the key XYY researchers at Harvard and the

NIMH, conceded in a coauthored Annual Review of Medicine piece entitled ‘The XYY

Genotype’ that it may be “inappropriate to allude to an XYY syndrome. The terms syndrome implies a degree of symptom consistency that is not supported by the data available at this time.” (1978:568) They still spoke of the need for longitudinal, prospective studies in order to explain the “epidemiological data pertaining to the XYY genotype [which] suggest that there is a three- to fourfold overrepresentation of XYY individuals in mental and penal settings and a twentyfold overrepresentation in mental- penal (special security) settings,” though they quickly pointed out that the “reasons behind the risk for behavioral disability are not known at this time.” (569) In short, they found themselves back where they started when they first sought to conduct prospective studies:

Since much of the available information about the XYY sex chromosome complement is biased, more data is required before definitive statements can be made about the personality characteristics or intelligence of the vast majority of XYY men not appearing in social settings oriented towards behavioral deviancy. (569)

What’s more, they had acknowledged many of the criticisms from Science for the People, pointing to a study by Money et al. (1975) that found that XYY men from upper and upper-middle class families were less likely than their lower and lower-middle class counterparts to be institutionalized as “evidence that the social management of the behavioral deviancy seen in some XYY individuals may be class related.” (568) Walzer et al. were certainly not about to give up on the idea that an XYY genotype did in fact confer

! 139! ! behavioral traits like impulsivity, aggression and outbursts, summarizing another study thus: “XYY individuals living apparently normal lives in the community could be distinguished from non-XYY men on the basis of formal psychological assessment.” (568)

Nevertheless, it was clear that XYY Syndrome was not going to be viable as a major topic of research or a category of institutional practice or social management in the aftermath of the intense controversy of the mid-1970s.13 While articles and letters about the controversy itself were common and existing studies continued to get their results published over the next few years (though generally in far less prestigious venues), publications on XYY declined dramatically, as we see in Figure 2 below.

XYY did linger on at the margins. As we will see below, XYY is easily and routinely picked up on prenatal test results and, when so observed, often leads to terminated pregnancies. One 1973 high school textbook, after discussing Down syndrome and the other sex chromosome conditions – Turner, Klinefelter and Trisomy X (noting that

“some are so-called ’super females’”) – turned to XYY: “Another abnormal condition results when a normal X-bearing egg is fertilized by a YY sperm, formed by non- disjunction during spermatogenesis. This produces an XYY male who is usually over six feet in height and very aggressive. Studies have revealed that a high percentage of inmates for the criminally insane have this chromosome abnormality.” (Otto and Towle 1973:185)

It served as the premise for a popular British novel and TV series called XYY Man (Royce

1973, 1977; Anon 1976) as well as making appearances on Law and Order and the like. In

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 13 This did not stop Alan Dershowitz, some thirty years later, from warning of the risks of using XYY “in the service of crime prevention or criminal responsibility,” (2007:274) while at the same time railing against the dangers of scientific censorship born of fear of finding that there really is a link between chromosomal abnormalities and violence. (Passages in the chapter suggest it was written much earlier).

! 140! !

David Fincher’s Alien 3 Ellen Ripley crash-lands on a ‘Maximum Security Double Y

Chromosome-Work Correctional Facility’ where the prisoners have taken over. Described by their leader as “thieves, murderers, rapists and child molesters… all scum,” our hero is therefore left to contend with both an alien killer and a group of genetically depraved criminals (Fincher 1992). Yet like other genomically designated conditions in the late

1970s and 80s, in the realm of non-fiction XYY Syndrome had little relevance as a

biosocial category.

Fig. 2: Articles published by year with XYY14 in their titles (Web of Science)

!!!!!!!!!

In sum, XYY turns out to be the exception that proves the rule. Ascertained in medical-penal institutions, mobilized beyond the lab as a tool of social control and deligitimated through social activism, XYY demonstrates the limited potential for genomic designation to achieve socio-cultural traction in the 1960s and 70s. !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 14 The actual search string used was “TI=(XYY* OR "YY SYNDROME" OR "DOUBLE Y CHROMOSOM*")” in order to capture the some 24 papers using the other two terms, almost all of them in the late 1960s

! 141! !

Discussion

So the turn to genomic designation was an easy one for human geneticists. The pertinent question then is: under what conditions can genomic designation matter, and to whom? How did this new form of classification gain traction beyond the field in which it made prima facie sense? Why is it that, in recent years, genomically designated conditions have taken on levels of clinical and social relevance that were unimaginable in the decades after 1959, even when in a few cases we are talking about the same syndromes? Answers to many of these questions will have to wait until we turn to the contemporary period in

Chapter 6. But for now, we can set up that analysis and begin to think about both the continuities and discontinuities seen in the practice of genomic designation over time.

Adopting Foucault’s terminology (2004[1979]: 19), the question then becomes how this particular ‘regime of truth’ that was so straightforward to human genetics was able to achieve an effective coupling with a ‘set of practices’ in other domains, be they clinical, advocacy, criminological or whatever. Only then would genomic designation “form an apparatus (dispositif) of knowledge/power that effectively marks out in reality that which does not exist and legitimately submits it to the division between true and false.” It should go without saying that ‘genomic designation’ is tiny in comparison with Foucault’s

‘dispositifs’ – psychiatry, clinical medicine, delinquency and liberal governance.

Nevertheless, it does allow us to study a nascent regime of truth as it penetrates various spheres of practice, reconfigures relations between them and accrues medical and sociocultural traction. On the level of the individual syndrome, the question can be formulated as one about ‘statements’ and the conditions necessary for their formulation,

! 142! ! repetition and circulation between various domains (Foucault 1972). Following Fleck, as discussed above, we can trace genomically designated syndromes as scientific facts from their initial, hesitant instantiation in a human genetics journal through their codification in the vademecum science of review articles, textbooks and atlases and on to the realization of significance at the level of institutional practice, popular knowledge and identity formation.

This framework helps us to see how genomic designation quickly achieved what

Fleck would call esoteric and vademecum reality, but mostly failed to attain the kind of traction at the level of popular knowledge that bestows upon a scientific fact fully-fledged certainty and givenness. In subsequent chapters we will see how a host of genomically designated conditions have taken on that kind of significance in more recent years, becoming micro-apparatuses of knowledge/power that increasingly shape clinical and social practice. They have realized the capacity to achieve a coupling or a practicable degree of commensurability with clinical nosology and to translate the interests of patients, their advocates and various institutional and commercial actors. But what can we learn about the conditions of possibility for genomic designation from our study of its early years and the exceptional case of XYY Syndrome?

One hypothesis is that XYY gained the traction it did, but then ultimately failed, because of the lingering but no longer legitimate imperatives of eugenic thinking. There is much to be said for this hypothesis: human genetics had grown fairly directly out of the field of eugenics (Kevles 1985) with, for example, important early work on genomic designation being performed at the Galton Institute in London and published in its Annals of Human Genetics, which had been the Annals of Eugenics until 1954. The language of

! 143! ! phrenology is endemic in the clinical descriptions of genomically designated conditions, and one sees frequent discussions and proposals relating to prenatal testing and termination of fetuses that, by virtue of their genomically designated condition, are deemed unlikely to develop into people capable of sustaining a life worth living. Finally, was it not precisely such a eugenic approach that drove scientific, state and popular interest in XYY Syndrome and criminality but, thankfully, also elicited the fatal backlash against it?

Unfortunately, this kind of straightforward answer does not bear the weight of further scrutiny. It was not simply a lag time that undid the debilitating specter of eugenics, but the work of actors ranging from biomedical researchers and genetic counselors to parent activists and others to rehabilitate the idea that genetics was a legitimate basis for shaping practice in fields like pediatric care and special education (Finucane et al. 2003;

Kerr, Cunningham-Burley, and Amos 1998). Furthermore, one can still trace a direct lineage from early twentieth century eugenics to contemporary medical genetics (Kevles

1985), and the charge of eugenicism still functions as a powerful rhetorical tool in scholarly and popular critiques of genetics and prenatal testing (Carlson 2012; Duster

2003; Shakespeare 1998). Nor was it a fundamental change in practice that obviated the homology with eugenics. We saw how Hsu (ibid) noted that ‘funny looking kids’ were ‘the prime targets of inquiry’ in the years after the trisomy 21-mongolism discovery, and it remains the case to this day that medical geneticists discuss ‘FLKs’, often comparing photographs and discussing the phrenological features (though the term is rarely used) that remain the only clinical signs that point towards genomically designated diagnoses with any reliability whatsoever (though it is still generally poor) (Morelle 2007; Shaw et al.

2003). What’s more, we will see in the conclusion how genomically designated conditions

! 144! ! are increasingly included in prenatal testing kits, often leading to the termination of pregnancies. Although proposals for mandatory testing and the elimination of genomically designated populations are no longer found in the literature, they were extremely rare in the 1960s and 70s. In sum, we cannot simply say that the charge of eugenicism has lost its discursive force or that there is no longer any basis upon which to levy it.

What then of the XYY genotype? Was not that an unfortunate episode of eugenicist thinking where much hope was placed in the use of genetic inheritance to manage a portion of the criminal population and perhaps gain insight into the remainder? It is true that the activists who mobilized against the XYY-criminality nexus tried to discursively tie it to eugenics, and with good reason. However, it is also the case that XYY research had moved on by the time that charge was mobilized. After all, Walzer, Park, Borgaonkar and Shah were by no means a bunch of crass eugenicists. They either always held or at least came to hold quite circumspect views on the implications of an XYY karyotype, qualms about the more sensationalist reportage associated with it and strongly held views against its employment in criminal justice systems before painstaking research could establish its true incidence and phenotypic implications. Indeed one of the findings of the Harvard study (S

Walzer and P. S. Gerald 1975) refuted an oft-speculated correlation between XYY and low socioeconomic status. Similarly, Borgaonkar’s team at Johns Hopkins published a paper showing that social class mediated the outcomes of XYY-related problems and the kind of treatment and institutionalization resulting from behavioral problems and police detention of XYY males (J Money et al. 1975; see above). Finally, it appears that none of Walzer’s families left the XYY study at Harvard, despite the freedom to do so at any time and the intense controversy that surrounded it.

! 145! !

In fact, we even see striking glimpses of the kind of parent-advocate discourse that drives genomically designated conditions today. While the Harvard controversy raged on, one mother wrote a letter to the New England Journal of Medicine in which she explained

(A. W. Franzke 1975:100–1):

I am the mother of a son who happens to be an XYY karyotypic male. Our son, now 21 years of age, is certainly not a criminal. He is a gentle, nice guy, somewhat inadequate occupationally and socially, but always doing his best to succeed. Since birth he has been developmentally atypical, and his behavior has been impulsive. He had behavioral, speech and learning problems in childhood. We were not aware of his genetic make-up until he was 16 years old. He was karyotyped at my request at Johns Hopkins Hospital, after I had learned about a possible behavioral effect of a supernumerary Y chromosome. In the Johns Hopkins XYY program, he has for the first time received informed and intelligent treatment for his condition. If we had known at an early age that he had the XYY genetic make-up we could have provided more adequate help for him instead of traveling through mazes of misdiagnoses and mistreatment for 16 years.

As we saw in the introduction and will delve into more deeply in Chapter 6, this passage evokes a key feature of contemporary parent advocacy with respect to genomically designated conditions: the frustration with what is now often referred to as a ‘diagnostic odyssey’ experienced by the families of children with multiple developmental complications, married with the belief that a genomically designated diagnosis would have provided the explanation and the basis for sound treatment.

It is this belief that the genetic cause of developmental difference provides the most sound basis for understanding and acting on it, even in the face of enormous variability, that drives the alliances of patients, advocates and experts working on genomic designation to this day. Again, this was already available for parents like Franzke:

The emotional damage of “not knowing” has been enormous for the child and his family. The best way to provide adequate help for a child with problems is to know the probable causes of these problems, whatever they may be. If the parents are denied the knowledge, and early treatment for XYY boys is thus not available, these boys may not have the chance for a normal life. (ibid. pp. 100)

! 146! !

This mother went on to ask, “Should this be taken from them just because of some person’s theory?” before excoriating Beckwith “and his ‘Science for the People’” for working to deny vital information to troubled children and their families. She had become something of a lay expert on XYY: “According to my own research and knowledge of the

XYY literature, no XYY boys have been located who have not displayed noticeable developmental deviancy in childhood, regardless of when labeled XYY. Early diagnosis appears to offer more benefit to these boys and their families than lack of a possible stigmatizing label.” (ibid, p. 101) That is, at least some of the parents of XYY children enrolled in the very studies that attracted such devastating oppositional activism were invoking the very same kind of arguments and displays of lay expertise that animate activism in support of genomically designated conditions today. What requires further examination then, are the conditions under which this way of thinking about genetics and human difference found a ‘surfaces of emergence’ or social circuitry in which to take root and begin to thrive.

In sum, even though Walzer and Park were supported by most of their colleagues at

Harvard and in the academy more generally when their work was put up for review, as well as at least some parents who displayed many of the characteristics and discursive tropes that drive genomic designation today, they failed to build a broader alliance. Instead, they found themselves confronted by activist groups and advocacy organizations concerned with children’s rights and wary of the stigmatizing and potentially eugenicist implications of XYY research. Once those activists mobilized against XYY research and the network of human geneticists, criminologists and related researchers and actors who

! 147! ! saw potential in it, the network dissolved and left the XYY classification bereft of the conditions it needed to really matter much to anyone. As a result, XYY Syndrome fell into crushing disrepute as a biomedical category. While longitudinal studies in Canada,

Scotland and Denmark did continue for a time, the public controversy that had derailed the

Harvard study was picked up in the UK by 1978, ensuring that XYY could give rise to little more than what Fleck called journal or esoteric science. Patricia Jacobs, whose 1965 study launched the XYY-criminality nexus, poignantly left Edinburgh for Hawaii to conduct research on spontaneously aborted fetuses – a population with extremely high rates of chromosomal abnormality but very little capacity to inform practice or attract controversy. The conditions and cultural repertoires that enable genomic designation to give rise to legitimate categories of clinical practice and social mobilization were simply not yet available. What might have transpired had XYY become visible during the heyday of early twentieth century eugenics remains nothing more than a disquieting counterfactual.

Continuities

Much of this chapter – the relative sociological impotence of genomic designation in its first generation or so of existence and XYY Syndrome as the exception that proves the rule – is geared towards setting up a contrast with genomic designation in recent years.

However, I want to briefly mention four important continuities that we might consider endemic to genomic designation to be explored in more detail later:

! 148! !

1. The related problems of systemic and often institutionally-channeled ascertainment

bias, post-factum assembling of clinical profiles and the absence of prima facie

anatomical or physiological bases to rule associations in or out (genetic mutations

are, after all, usually present in every cell of the body).

2. The emphasis, seen most clearly in the early period for XYY but almost

omnipresent for genomically designated conditions today, on early intervention for

affected people.

3. Phrenology: as we have already seen, craniofacial features were and remain

perhaps the most distinctive clinical signs associated with genomically designated

conditions.

4. Prenatal testing and the termination of genomically designated fetuses.

Again, all four of these continuities will be examined in greater detail in subsequent chapters. For now, a brief word on prenatal testing, which will in turn be picked up in the dissertation’s conclusion.

The potential to prenatally diagnose chromosomal abnormalities and terminate pregnancies on their bases was almost immediately recognized as perhaps the main form of intervention made possible by the newfound ability to perform karyotype analysis. Even before prenatal testing was technically possible, the capacity to inform recurrence risk was seen as one of the major accomplishments of the emerging field. For example, after reviewing all of the known chromosomal aberrations reported at the time, the early review of the new cytogenetics in Pediatrics cited above concluded: “For the pediatrician, the studies on Mongolism are of primary importance. The genetic aspects of the condition

! 149! ! have been clarified and a basis for the famlilial occurrence emphasizes the necessity of chromosomal analysis in providing information in regard to future pregnancies in young mothers with affected children.” (Rappoport and Kaplan 1961:434) Once amniocentesis came into view in the mid-1960s (Steele and Breg 1966), Down syndrome and a host of genomically designated syndromes could be used as the basis for the termination of pregnancies (see Harper 2006:162; Reinhold 1968). Thus when Bentley Glass delivered his presidential address to the American Academy of Arts and Sciences in 1970, published in Science the following year, he told the audience assembled in Chicago:

Human power is advancing with extraordinary rapidity in this realm of control over the genetic characteristics of the unborn. Perhaps, as Carl Becker so pregnantly stated, our race, far from having any aversion from power, will welcome this power too, will seek it, fashion it, and grasp it tenaciously. Unlimited access to state-regulated abortion will combine with the now perfected techniques of determining chromosome abnormalities in the developing fetus to rid us of the several percentages of all births that today represent uncontrollable defects such as mongolism (Down's syndrome) and sex deviants such as the XYY type. (Glass 1971:28; my emphasis)

Of course, not all genomically designated conditions are created equal: they are all variable, but some are more variable than others. Few people without a dogmatic objection to abortion would find the termination of fetuses diagnosed with Edwards (18 trisomy) or

Patau (13 trisomy) Syndrome to be ethically problematic. Indeed, surveying the literature on those conditions makes it clear that prenatal detection has been far and away the primary object of medical research on those conditions.

But what about XYY? Did abortions of XYY fetuses cease once it became clear that most males with an extra Y chromosome have minor physical and psychiatric problems at most? The answer is simple: no. An editorial in the British Medical Journal

! 150! ! entitled ‘What is to be done with the XYY fetus’ (1979) observed (p. 1519) that “in the sex chromosome aneuploidies there may be a wide range from near normality to severe handicap,” but that “in practice termination is usually requested… 47XYY, is a clear example of this problem.” They discuss the options, each of which were adopted by many clinicians, when XYY was found prenatally: do not tell the parents; tell the parents but emphasize the relatively small risks; tell the parents and explain “that the risks are high enough to justify abortion” (p. 1520). As they put it, “the dilemma remains” (ibid). In the

1980 retrospective conference on XYY discussed above, Stanley Walzer explained his approach to the dilemma in the frequent event that he received calls from women asking whether they should terminate a pregnancy on the basis of an XYY or XXY finding:

What I will do is sit down with the people, and I will go over with them in confidence the developmental data that I have available on my fourteen XYYs. I will show them photographs of these lovely children. I will talk to them about some of the problems the kids are having and how wonderful the other ones are doing. I will present the facts – all the facts that I have. And I will then tell them to go back home and make their own decision. (quoted in Bauer et al. 1980:10)

In other words, even parents with the capital to ‘sit down’ with one of the world’s leading experts are still left with a deeply uncertain dilemma. As we will see in the conclusion, rates of termination for even phenotypically mild conditions like XXX and XYY

Syndrome remain high, and new technological advances may soon make the termination of pregnancies upon the diagnosis of genomically designated conditions an extremely common phenomenon.

! 151! !

Coda

How does this account of the relative social and clinical impotence of genomic designation in the 1960s and 70s help us understand its increasing power and salience in recent years? I argue that, when viewed through the lens of what I called ‘reiterative facticity’, they provide comparative historical perspective on genomic designation as a form of human classification. In changed historical conditions, and with new repertoires of collective action at their disposal, several of the genomically designated conditions discovered in the 1960s have, in recent decades, assumed a different form and taken on a broader significance. Conditions like Cri du Chat/5p- Syndrome and de Grouchy/18p

Deletion Syndrome have become objects of real social action with groups like the Five P

Minus Society15 and the Chromosome 18 Registry & Research Society16 providing support, raising funds, promoting awareness, facilitating research and holding conferences for patients, parents, clinicians and researchers. We can therefore analyze how the categories born of genomic designation have been transformed as objects of knowledge and practice, even in those cases where we are talking about precisely the same genetic mutations.

Even Edwards Syndrome/Trisomy 18 has an active Trisomy 18 Foundation17 dedicated to ‘Support, Advocacy, Research’ and equipped with fundraising operations and family, medical and research advisory councils, a board of directors, staff and a network of volunteers and parent support. As a letter from the foundation’s founder, Victoria Miller, puts it on their website, the foundation is “embarking upon an era of new hope, an era in

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 15 http://www.fivepminus.org/, accessed January 12 2013.

16 http://www.chromosome18.org/, accessed January 12 2013.

17 http://www.trisomy18.org/, accessed January 12 2013.

! 152! ! which Trisomy 18 can become a preventable and treatable condition for future generations.” That outcome is, of course, remote given the severity of Trisomy 18 and its gross nature as a whole autosomal chromosome duplication. However Miller, who began working to establish a Trisomy 18 community after losing her own son to the condition eleven days after he was born, goes on to outline a more immediate function for the foundation:

Since our beginnings, over 10,000 mothers, fathers, grandparents, and family supporters have been served by our nationally-recognized programs. Those just learning about their child’s Trisomy 18 diagnosis can immediately access diverse community support from peers who have walked the same path as theirs. To families just learning about Trisomy 18, we welcome you to this caring community. Together we will help each other. This website was designed to provide you with the information, support and tools you need - no matter where you are on the path today. And we will be here, walking beside you, throughout you and your child’s journey with Trisomy 18.

While a cure for trisomy 18 may be an implausible goal, that does not mean the foundation cannot interface with biomedical researchers and healthcare professionals:

Trisomy 18 is fundamentally a health and medical crisis that affects those in their child- bearing years. It occurs in 1:3000 live births and is far more common than is generally understood. Families need more than just comfort while they cope and endure. They need immediate solutions for themselves and their children who struggle and die too soon. The Foundation is busy crafting those solutions in collaborative efforts with health professionals groups, international research leaders, and public health agencies cutting across many different disciplines and medical specialties. Your help is needed to get there! We must prevent and reduce how often Trisomy 18 happens in pregnancy and we have to improve the health outcomes for children who are born struggling with the consequences of having an extra 18th chromosome. In honor of all our children with Trisomy 18...yesterday, today and tomorrow.

Thus the first of the foundation’s five goals is to “insure [sic] that all appropriate physicians and health care providers are aware of the full range of health outcomes, have ready access to the latest information to insure more accurate diagnosis and prognosis, and

! 153! ! can easily make use of the resources they need to provide access to effective medical treatments and psycho-social care for the entire family.”

They also advocate for greater funding from the NIH and CDC, and seek to emulate the success other genetic disorders have had in this regard:

Many disorders effecting [sic] many fewer children per year than Trisomy 18 have this kind of public research support, but not Trisomy 18. The difference is that these other disorders have had an effective patient advocacy organization lobbying to make this happen. We can be effective at this too. But first we must combine our voices and direct them to lobby the key decision-makers and committees in this process. With your help, all this is possible! Please return to this area to keep abreast of developments in this area that you can get involved in. We have affected families in all 50 states and when we speak together, we will be heard in the halls of Congress.

In other words, Trisomy 18 may have been one of the first genomically designated syndromes, but it was not the first one to employ a new model that could make them matter as conditions and as ‘biosocial’ communities. They are, however, making a concerted effort to follow suit.

If Edwards Syndrome was the second genomically designated syndrome reported in the literature and one of the most consistently severe, what about the first and one of the most consistently mild: XXX or ‘Superfemale’ Syndrome? It too has been taken up as an object of contemporary research, care and advocacy. Founded in 1989, Knowledge,

Support & Action18 is increasingly working to promote awareness and research on sex chromosome aneuploidies, now in collaboration with the eXtraordinarY Kids Clinic at

Children’s Hospital Colorado. Triple-X Syndrome appears on popular health review pages like WebMD19 and CNN Health20 as well as the myriad new sites and repositories

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 18 http://www.genetic.org/, accessed January 12 2013.

19 http://children.webmd.com/triplo-x-syndrome, accessed January 12 2013.

! 154! ! dedicated specifically to genetic disorders. Finally, local support groups are being established around the country and their second national conference will be held in Denver in July 2013. They are explicitly trying to get sex chromosome aneuploidies like XXX and

XYY Syndromes recognized by general practitioners, with a series of brochures designed for pediatricians downloadable from their site.21 They have high hopes for these kinds of efforts:

Consider the possibilities: Imagine handing your child's pediatrician the generic brochure and the brochure that describes your child and then having this conversation: "Doctor, did you know that 1 in 500 individuals has an X or Y chromosome variation? It's true. So, it [sic] you have 2000 patients in your practice, that means that there are probably 4 children that you care for who have this condition. Are they all diagnosed?" Let's start a movement to enlighten every professional you encounter and help them understand - and diagnose - every child who has X and Y chromosome variations!”

In other words, we now see the parents of children with sex chromosome aneuploidies like

XYY advocating for early and universal diagnosis, evaluation and intervention.

XYY Redux

Today, after a couple of decades in the wilderness, XYY Syndrome is making a comeback, but in a form appropriate for the times.22 It receives the same kind of press as

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 20 http://www.cnn.com/HEALTH/library/triple-x-syndrome/DS01090.html, accessed January 12 2013.

21 http://www.genetic.org/Knowledge/Brochures.aspx

22 This is not to say that we are now immune to programs of research that seek the cause of violence and crime in the human genome. As we will see in the following chapter, one extremely rare genomically designated syndrome gave rise to a program of research as well as media coverage and a direct-to-consumer test for a ‘Warrior Gene’. Similarly, the NIH ran into considerable opposition and had to halt plans to hold a conference on ‘Genetic Factors in Crime: Findings, Uses and Implications’ in 1992 (Hilts 1992). Finally, recently announced plans to sequence the genome of the Newtown CT killer, Adam Lanza, eerily recall the association between

! 155! ! other genomically designated syndromes, like a piece in the UK’s second biggest newspaper, The Daily Mail23, about a former actress and the challenge of raising a son with learning difficulties (Giles 2008). Networks of support and advocacy are emerging as part of the Knowledge, Support & Action foundation (above) and the Unique24 charity for children with chromosomal disorders in the UK, and new programs of research are underway.25 Indeed the KS&A Foundation’s XYY leaflet captures how much has changed in the representation of what an extra Y chromosome means for a boy’s developmental

(see Figure 4 below). Clinical profiles and guidelines for XYY are appearing, but rather than deviance, criminality and ‘mental sub-normality’ it is now characterized by increased risk of ADD and autism and an average of 10-15 IQ points lower than unaffected siblings

(see e.g. J. L. Ross et al. 2012; Bishop et al. 2011; Cordeiro et al. 2012; Geerts, Steyaert, and Fryns 2003; Randi Jenssen Hagerman 1999:185–92; Ratcliffe 1999; N. R. Tartaglia et al. 2006).26 In short, it is being made suitable for our age.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! XYY and infamous serial killer Richard Speck forty years ago, though the move immediately elicited comparisons to XYY and the aforementioned ‘Warrior Gene’ as well as appropriate skepticism about the scientific value of scanning a single murderer’s DNA (Kolata 2012).

23 As any good consumer of the British press would surely agree, no contemporary newspaper would be more likely to latch on to the idea of genetically-determined criminals than The Daily Mail.

24 http://www.rarechromo.org/html/home.asp See esp. their Unique XYY Study Day Report: http://www.rarechromo.org/files/XYYStudyDayReport.pdf

25 See http://www.genetic.org/Knowledge/WhatareXamp;YChromosomeVariations/Tellmeabout47,XYY. aspx

26 For example, see XYY Syndrome’s WebMD entry: http://men.webmd.com/xyy-syndrome

! 156! !

Figure 4: KS&A Brochure from XYY Syndrome (excerpts from pp. 1 and 2)27

Conclusion

In the aftermath of the XYY controversy, the president of the American Society of Human

Genetics bemoaned the ‘genetic McCarthyism’ (J L Hamerton 1976: 109) of groups like

‘Science for the People’ and insisted XYY research was both important and sound. The battle, however, was lost: when geneticists confronted activists, they came away bemused

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 27 Available at: http://www.genetic.org/Portals/0/Public/Brochures/KS&A%20brochure%20green%20xYY%20W EB.pdf

! 157! ! and beaten. However, they did not stay down for very long. When Patricia Jacobs, perhaps the key figure in XYY research, gave a speech after receiving a career award at the

American Society of Human Genetics in 1982, she too insisted: “The finding of an excess of XYY males in medical-penal settings… should have provided a useful, objective tool in the rational study of human behavior.” (P A Jacobs 1982: 693) Discussing the activism that halted almost all XYY research, she said, “can we, with hindsight, try to insure [sic] that such an episode is never again allowed to sully human genetics?” Her answer in 1982 was not quite there yet: she just wanted the Society to be more of a public force and for geneticists to be wary of media appropriation of their language. Even as late as 2004 she seemed incredulous that XYY research had attracted such a backlash. After describing the process that, through her work with sex chromosome anomalies, led her to suspect that an extra Y chromosome might contribute to aggressive behavior, Jacobs said (cited in Harper

2006:90): “If you stop and think about it, 98% or some such number of the prison population are males. So you can’t say that the Y has got nothing to do with behavior…

And they try to tell me the Y has got nothing to do with it. Absurd!” The book in which this interview appears, while a fabulous resource, is essentially a hagiography of the key figures in early cytogenetics research. As such, we get no discussion of the ‘they’ to whom

Jacobs refers. It does not take much imagination, however, to deduce that it is the early critics of XYY research represented by Jon Beckwith et al. who had essentially obliterated the early programs of research and social management related to XYY Syndrome.

In other words, Jacobs and perhaps many of the other founding figures of medical genetics did not appreciate the alliances and sociotechnical work that would be required in order to make medical categories derived from human genetics research really matter to

! 158! ! the way we understand and act on human difference. But later in her speech (p. 696)

Jacobs pointed to a new syndrome she and others were working on, before suggesting, “X- linked mental retardation associated with a marker at Xq28 will be to the 1980s what

Down syndrome was to the 1960s.” Jacobs was more prescient than she could have known at the time. She was talking about what we now known as Fragile X Syndrome, which we will see has since gone on to pioneer a new model for the mobilization of genetic mutations.

That new model will be the subject of the following chapters. But for now we need to begin to think about how it came to be that, in recent years, it is precisely groups who

‘support children’s rights’ that now advocate for and work to advance research on genomically designated conditions. They often call for newborn studies and even national newborn screening despite the systemic ascertainment bias of current knowledge about most genomically designated conditions and their highly variable clinical and developmental implications; in other words, precisely the same kind of studies that attracted such ire in the case of XYY in the 1970s. Finally, even XYY has been recast as a syndrome, albeit one suitable for this new field of research, care and advocacy for people with clinically variable genetic disorders.

So here we have proof that the impact of genomic abnormalities is as much a sociological phenomenon as it is a biomedical one. Through the perspective provided by reiterated facticity, we see that genetic mutations have to be mobilized and translated through repertoires of collective action, institutions and cultural frameworks for managing abnormality and acting on human difference. This is how they attain their meaning as both objects of knowledge in the biomedical sciences and categories of social action and clinical

! 159! ! practice. The general failure of genomically designated conditions in the decades after

1959 to catch on outside of the sphere of esoteric journals, textbooks and prenatal testing, as well as the spectacular-but-fleeting exception of XYY Syndrome, therefore helps us better understand the resurgence of genomic designation in recent years. The standardization of chromosomal analysis was necessary, but far from sufficient: in order to really matter to lived experience, genomically designated ‘syndromes’ have to be able to translate the interests of actors outside of the field of human genetics, get taken up as objects of complex collaboration and achieve a degree of commensurability with clinical nosology.

Haydu’s focus on the way homologous groups of actors in different periods work to solve similar problems, for example human geneticists in the 1960s and 70s working with chromosomal abnormalities as versus their successors today, will not quite work for our purposes. Rather, by attending to the shifting networks of actors formed around genomically designated conditions and the way they fundamentally reframe the ‘problem’ or goals of action, reiterated facticity allows us to frame genomic designation in comparative-historical terms. As I will argue over the next three chapters, genomically designated syndromes have to be mobilized in networks and under historical conditions of possibility or ‘surfaces of emergence’ that are amenable to the way they realign the classification of people and redirect clinical judgment, even if they in turn work to transform existing structures according to genomic designation’s ‘grid of specification’

(Foucault 1982:41–2).

! 160! !

Before we turn, in Chapter 6, to a fieldwork-based study of the new model of expert-stakeholder collaboration and advocacy animating genomic designation in recent years, it is important to further examine a broader condition of possibility: the variable ways in which genomically designated conditions intersect with and are used to get at more general categories of human difference. That project will be taken up over the course of the next two chapters.

! 161! !

Chapter 4 - Of elves, warriors and autistics Leveraging abnormality in genetics research and advocacy

In our post-genomic present, with the ‘gene-for’ model having proven chimerical for most salient traits and medical conditions (Hayden 2008; Lock 2005; Wade 2009), it has proven unexpectedly difficult to turn observations of the genome into meaningful and potentially practicable information about human health, illness and difference. Biomedical researchers have therefore adopted myriad approaches to overcome this unforeseen impasse. They have turned to the study of gene-environment interaction, epigenetics, bioinformatics and many other frameworks for turning seemingly opaque knowledge about the human genome into meaningful information that can speak to questions we care about.

This chapter discusses a more longstanding, but mostly overlooked strategy for making sense of our genomes: the use of people diagnosed with genomically designated conditions to pursue research on more general, population-level variation. They are what one geneticist described to me at a 22q conference as ‘a low hanging fruit’. When a disorder is fixed to a specific genomic abnormality there is a particularly cogent rationale for studying it because, as we will see, it allows biomedical researchers to ‘control’ for a population with a genetic mutation and then use it as a kind of a natural experiment. This chapter therefore addresses the way that genomically designated conditions can become

! 162! ! privileged sites of biomedical knowledge production despite their generally rare incidence.

I am going to refer to this research strategy as ‘leveraging genomic abnormality’, or LGA.

Examining LGA sheds light not just on a fascinating research strategy, but one with very real consequences, especially for people with genetic disorders. What’s more, it is a research strategy that inherits a long legacy of using the abnormal as a vista on more general questions of human illness, health and difference. Through LGA, contemporary genetics research has raised a series of novel questions about human difference. Some pertain to the pathways that lead from genomic variance to physiological difference and eventually disease states, while some speak to core human faculties and population-level variation. Others are enigmatic, even bizarre: Are there genetically predisposed ‘warriors’ among us? Did the elves and pixies of yore have a microdeletion on the long arm of the seventh chromosome? Taken together, an analysis of the way LGA is pursued helps us understand how abnormality is mobilized in contemporary biomedical research.

LGA is a strategy adopted by many researchers that examines rare disorders caused by discrete genomic etiologies in the hope of using them as a kind of biomedical model for population-level variation. By turning to categories of abnormality that, although rare, have a known genetic etiology, biomedical researchers hope that their findings will speak to those disorders’ symptoms or traits whose prevalence is far higher. Conversely, patient advocacy organizations sometimes consciously leverage the biomedical value conferred by their genomic specificity in order to shape the direction and organization of research. In so doing they hope to attract the huge volumes of funding and research that it takes to garner care and services and eventually develop pharmaceutical interventions despite their small-n status as a population. In other words, this chapter will examine the different ways in

! 163! ! which specific genetic mutations can serve as ‘boundary objects’ (Star and Griesemer

1989) that unite diverse biomedical disciplines, funding agencies, commercial actors and patient/parent activists with divergent interests and frameworks for pursuing knowledge production and social action. Rather than just note that genetic mutations can come to serve as boundary objects, however, this chapter asks ‘boundary objects for whom?’ and examines the particular, divergent kinds of biosocial networks that can be assembled around a genetic disorder.

Drawing on published biomedical research, media and other publically available materials, this chapter will present LGA as a strategy in human genetics and health-related advocacy. I will examine three rare genetic disorders for which developmental delay is the most clear and consistent associated symptom, the further conditions and traits they are leveraged to provide insight to and their contrasting biosocial outcomes. First, I discuss the extremely rare case of Brunner Syndrome, caused by a mutation in the MAOA gene at p11.3 on the X chromosome, and the way it was used to identify a low expression MAOA gene variant that has been widely publicized and commercialized as a ‘Warrior gene’ that can speak to variable proclivities for aggression. Second, I turn to Williams Syndrome and research into the role of the 7q11.23 genetic mircrodeletion in its key traits: sociability, strong language skills and musicality alongside intellectual disability, physical dysmorphism and ‘elfin’ facial structure. Finally, I discuss the way that Fragile X

Syndrome attracts significant levels of biomedical interest because of the hope that it can provide a genetic model for autistic spectrum disorders. I then discuss the contrasting

‘biosocial’ (Rabinow 1992) outcomes of LGA in those three cases, the way they in turn can shape future research enterprises, and situate LGA within a broader history of biology

! 164! ! and abnormality. As we saw in the previous chapter, genomically designated syndromes have to be mobilized within a much broader social circuitry in order to really matter outside of the field of human genetics. Here, we see how not all social circuits are created equal and the emergence of a model for mobilizing genomically designated syndromes as privileged sites of biomedical research and health advocacy that has become a beacon in the field.

Literature review

How can we make sense of LGA in the context of existing social scientific work on biomedicine? The identification of human subjects for biomedical research was recently examined in Steven Epstein’s book, Inclusion (2007), demonstrating how agitation by marginalized constituencies has made the recruitment of representative subject populations a priority for the NIH and other major research organizations. Epstein situates this shift, in part, as the result of contestation about generalizability (ibid: 23): “Whenever one conducts a clinical trial on a set of individuals, the assumption is that knowledge gained from a few can be extended to the many. But what sorts of extrapolations are appropriate?” Epstein’s fascinating research question concerns a series of debates, struggles and changes in accepted practices of knowledge production about what makes a group of human subjects representative of a larger population and the findings that result generalizable. This chapter, by contrast, examines the inverse, when an acutely unrepresentative population’s abnormality is extrapolated or leveraged to better understand far larger categories of people, and perhaps even the population as a whole. Furthermore, it attends to the way contemporary ‘politics of difference in medical research’ (Epstein’s subtitle) are

! 165! ! reconfigured when the goal is to leverage abnormality rather than assemble representative subject populations.

Although a number of studies have touched on what I am calling LGA, most do so only in passing. Rabinow’s seminal statement on the emergence of ‘biosociality’ discussed in Chapter 1, for example, noted (1992:185): “Given the way genes are currently located on chromosomes, i.e., linkage maps, the easiest genome to map and sequence would necessarily be composed of the largest number of abnormal genes. In other words, the pathological would be the path to the norm.” Similarly, Cambrosio et al. (Cambrosio et al.

2009) discussed the ‘bio-clinical hybrids’ like mice models and biomarkers that increasingly circulate between the biologist’s lab and the clinic, blurring the line between the normal and the pathological in fields like cancer genetics (see also P. Keating and

Cambrosio 2006). Perhaps the most direct engagement is found in Rabeharisoa et al.’s

(2012) unpublished discussion of the ‘hybrid collective model’ that sees communities of families, clinicians and researchers working together as a community, especially with respect to knowledge production about a rare disease. In their discussion of the dynamic of

‘singularization-generalization’ they suggest that hybrid collectives make reflexive decisions about either framing their situation as a singular, unique set of challenges presented by a rare disease or instead focusing on the commonalities shared with other conditions so as to emphasize the more general nature of their situation. Rabeharisoa et al. even note the dilemma faced by organizations when researchers seek to use their rare condition as a model for more general biological mechanisms or more prevalent forms of disease (ibid: pp. 27-8). However, none of these studies develop an account of the biomedical or social projects and alliances that see people with genetic disorders, or

! 166! ! abnormal populations more generally, serve as a path to the normal (or to common pathologies) in contemporary biomedicine.

The better place to begin a discussion of LGA then is to look to both the longstanding literature on the use of abnormal people as models in biological research and the extensive literature on contemporary genetics, biotechnology markets and health advocacy. Only then will we be able to grasp how LGA represents a continuation of a longstanding strategy in biological research, but a) does so in the context of prevailing structures for producing, disseminating, interpreting and acting on knowledge about our bodies and especially our genomes, and b) takes different forms that impact the kinds of biomedical research and social action that a genetic mutation can give rise to.

Abnormality in biological research

Leveraging pathological states has a long lineage in biological research, though this history has not been the subject of a systematic analysis. Reading the collection of Georges

Canguilhem’s essays, Knowledge of Life (2008), one cannot help but be struck by the pairing of the final two chapters – ‘The Normal and the Pathological’ and ‘Monstrosity and the Monstrous’. In the first and more widely cited essay, Canguilhem discusses the subjective, environmentally and contextually embedded, dual nature of the distinction between the normal and the pathological and relationship of pathology to both abnormality and anomaly. Beginning in that essay and developed in the next, Canguilhem discusses the role of what he calls ‘monsters’ or ‘monstrosity’ as a tool for biological research.

Canguilhem argues that, in their stark diversions from the normal development of living forms and functions, “the monster bestows upon the repetition of species, upon

! 167! ! morphological regularity, and upon successful structuration a value all the more eminent in that we can now grasp their contingency.” (2008: 135) Furthermore, its divergence from the usual course of development represents a window on to the contingency and the mechanisms underlying the normal, such that (142): “whether in embryology, classification, or physiology, the eighteenth century made the monster not only an object but an instrument of science.” In particular, the rise of teratology and teratogeny in the nineteenth century is when “the scientific explanation of monstrosity and the correlative reduction of the monstrous are elaborated.” Canguilhem explains:

From then on, monstrosity appears to have revealed the secret of its causes and laws, while anomaly appears called upon to explicate the formation of the normal, not because the normal is an attenuated form of the pathological, but because the pathological is the normal impeded or deviated. Remove the impediment and you obtain the norm. (143)

This ‘transparency’, as he puts it (ibid), “cuts monstrosity off from any relation to the monstrous,” and makes it a tool of biological research.

A similar point is made by Nobel Prize-winning biologist François Jacob in his magisterial 1973 work, The Logic of Life (Jacob 1993), particularly in his discussion of late nineteenth century embryology (122-9):

The study of developmental anomalies then acquired a new importance… Experimentation in physiology had usually involved modifying the natural state of an organism in order to disturb one function or another. However, the same result can be obtained by observing certain pathological conditions… In many cases no experiment can reproduce a departure from normality so precisely and selectively. If knowledge of the physiological state was obviously necessary for the interpretation of pathological conditions, the study of pathological conditions also provided a precious instrument to study biological functions. (1973: 123, my emphasis)

Much later in the book, when Jacob is discussing the twentieth century turn towards the molecule as primary site of biological study, he notes how monsters can be created almost at will by altering the ‘programme’ (genome) inside the organism rather than disrupting its

! 168! ! structure or function from without. Jacob explains, “Molecular biology provokes lesions from inside the organism as a result of mutations,” and manipulates bacterial populations such that:

… it is possible to obtain almost at will monsters in which a chosen function is damaged by mutation. By means of such monsters, abnormalities are analysed by physiologists and morphologists in the intact organism, by biochemists and physicists in cell-free extracts. There are no longer two independent fields, but two aspects of the same investigation. What changes had been wrought in the attitude of biology! (265)

Genetics has always required difference and abnormality to drive knowledge production. When it comes to the relationship between chromosomal material and phenotypes, geneticists have long made recourse to model organisms: Drosophila was the object of study in Thomas Hunt Morgan’s seminal work that identified the chromosomes as the main site of inheritance in the early twentieth century (1915), while mice emerged as the primary model system once molecular biology was able to knock out genes and chromosomal loci to observe their phenotypic effects (see Rader 2004; Davies 2010).1

However, for obvious ethical reasons these kinds of mutations cannot be induced in humans, and so the laboratory analysis of ‘monsters’ focuses primarily on non-human animals, leaving researchers to draw connections between human genotype-phenotype correlations and their mostly murine homologs. In this context populations delineated according to genetic mutations, i.e. people with genomically designated syndromes, serve as a privileged site of biomedical research. This dissertation therefore adds a new chapter

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 Throughout, animal models have played a key role in research that aims to understand the relationships between genetics and biological function, both through the breeding of particular variants and the molecular ‘knock-out’ of genes and genomic regions. By comparing ‘wild typles’ (normal), heterozygous mutants (one mutation) and homozygous mutants (mutations on both chromosomes in a pairs), connections between genes, physiology and function can be inferred, and this kind of strategy plays a key role in a number of biomedical fields.

! 169! ! to the history of using ‘monsters’ to further biological research by examining the ways in which medical genetics adopts a homologous strategy. Through what I am calling LGA, this longstanding use of the abnormal as a privileged site of biological knowledge production is emerging as an important strategy in post-genomic research.

Indeed LGA can be traced back to the earliest years of genomic designation. When

Patau et al. (1960) reported their discovery of chromosome 13 trisomy in the foundational paper on what is now often called Patau Syndrome, their discussion noted:

…we may expect that an aetiologically unique group of a limited number of "autosomal trisomy syndromes" will become established. To the human geneticist these will be of continuing interest; he will in particular look forward to cases in which the presence of known genes in the parents can be related to peculiarities in the trisomic child. It seems likely that every component anomaly of a trisomy syndrome reflects, by way of a dose effect of gene action, the presence in the respective chromosomes of at least one gene locus that plays a prominent role in the normal development of the afflicted organ; (792-3)

And beyond the identification of genes in normal development, such a strategy could also provide a research platform for identifying the basis of more common and distinct pathologies:

… indeed, we suspect that many of the trisomy anomalies can also be produced individually in euploid persons by the heterozygous or homozygous presence of a suitable mutated allele of the responsible gene. Polydactyly and hare lip with cleft palate in the present patient may be cases in point. (793)

Patau et al. were advancing an idea that would prove to be highly prescient: “trisomy may become instrumental in establishing for the first time autosomal linkage groups in man.”

(ibid) In short, chromosomal abnormalities were immediately recognized as a potentially vital resource in the broader project of understanding the role of the genome in both normal human functioning and common disease categories.

! 170! !

Similarly, the first editions of Victor McKusick’s Mendelian Inheritance in Man from 1966, which to this day remains the authoritative repository on human genes and genetic disorders (now in online form as OMIM.org), discussed the ‘Usefulness of the catalogs’ thus:

2. Genetic disorders give us insight into the normal. These catalogs of hereditary traits are like photographic negatives from which a positive picture of man’s genetic constitution can be made… As complete knowledge as possible of the normal genetic constitution of man is bound to be useful in the long run. Physicians have a unique opportunity to contribute to the knowledge of what Richard Lewontin referred to as “man’s mutational repertoire.” (1966: xiii)

As we saw in Chapter 3, McKusick and his catalogs were early advocates for what I am calling genomic designation. Investigations of the pathological and the normal, in short, have been intertwined in throughout the history of medical genetics.

Contemporary genetics, advocacy and commerce

How then can we reconcile Epstein’s important point about the politicization of biomedical research design and the mobilization of subject populations with the longstanding use of abnormality or ‘monsters’ as a platform for studying the normal in biology? As we saw in previous chapters, in recent years there has been an abundance of interest in the processes of biomedical research, capital investment, clinical care, advocacy and identity formation associated with human genetics (Freese and Shostak 2009). A few studies are of particular relevance here. Nikolas Rose outlined an ‘emergent form of life’ that is taking shape as contemporary medicine moves from a clinical to a ‘molecular gaze’, including the new forms of ‘biological citizenship’, expertise and capitalization intertwined with the shift to understand life, health and illness at the molecular level (2007: 5–7).

! 171! !

Scholars like Sunder Rajan (2006) have explored the ways in which the contemporary life sciences, and especially genomics, are ‘overdetermined’ (p. 6) by the circuits of capital speculation and accumulation in which they operate. Crucially for the present topic,

Sunder Rajan focuses on genomics and pharmaceutical concerns (see also Adam Hedgecoe

2008) while others have looked at the market for direct-to-consumer (DTC) genetic testing

(Hogarth, Javitt, and Melzer 2008; N. Rose 2008). This line of work, much of it dealt with in greater detail in other chapters, draws our attention to the forms of lay expertise, advocacy, commercial and state investment and other domains that shape the experience of genetic risk and illness, the provision of care and the organization of research programs

(Nelkin 1996; A. Hedgecoe 1998; Kerr, Cunningham-Burley, and Amos 1998; Vololona

Rabeharisoa and Michel Callon 2002; Skinner and Schaffer 2006; e.g. Hacking 2006;

Whitmarsh et al. 2007; S. F. Terry et al. 2007; Schaffer, Kuczynski, and Skinner 2008).

Following Heath, Rapp and Taussig (2004) and Rabehariso and Callon’s lead

(2002), particular interest has been paid to the role of patient and parent mobilization and especially advocacy organizations, with Heath et al.’s specific term ‘genetic citizenship’ serving to indicate the new formations of actors, practices and norms associated with contemporary genetic conditions which “challenge conventional notions of a divide between lay people and experts… [and] have also given rise to new forms of democratic participation, blurring the boundary between state and society, and between private and public interests.” In this vein of research, Panofsky (2011) conducted a comparative analysis of patient advocacy organizations for people with rare genetic disorders and the way they mobilized research programs towards their own ends. Panofsky added the generation and strategic manipulation of ‘sociability’ – the social relationships established

! 172! ! between members of advocacy organizations for genetic disorders and biomedical researchers – as a key factor alongside the five mechanisms noted in the literature: economic resources, social movement-style mobilization, moving early, lay expertise, and organizational controls. As we will see most powerfully in Chapter 6, my research confirms Panofsky’s finding that sociability is indeed a key variable that can impact the success of groups seeking to advance medically relevant research on rare genetic disorders and indicates that future work should extend this line of research to look at the role of expert-activists alongside the role of lay experts. However, this chapter makes an addition to Panofsky’s list of mechanisms through which groups associated with rare genetic disorders can attract resource-intensive biomedical research: they can strategically and sometimes successfully leverage their rare genetic abnormality to attract researchers interested in categories of human difference that are far more prevalent, even approaching universal.

How then do these different concerns – advocacy for genetic disorders, biotechnology and pharmaceutical interests and the use of genetic mutations as models from biological research – come together in what I am calling LGA? In order to get a sense of how LGA can function and the different ends towards which it can be directed, we now turn to our three cases.

A Warrior Gene

In 1993 a pair of articles in The American Journal of Human Genetics and Science by Brunner et al. (H G Brunner et al. 1993; H. Brunner et al. 1993; see also Morell 1993) reported a multi-generation Dutch family with an x-linked form of mental retardation.

! 173! !

Brunner was approached by a member of the family, for whom the existence of a distinct condition affecting some of the males in the family had been apparent for years (Han G

Brunner 1996). While the severity of what we now call intellectual disability was mild and the identification of x-linked mental retardation kindreds was not an uncommon finding,

Brunner et al. reported an unusual behavioral phenotype characterized by aggression and violence. Behaviors included arson, attempted rape, exhibitionism and an attempted suicide. The urine of three of the affected males was found to have disrupted monamine metabolism, genetic linkage analysis indicated a disruption in the MAOA gene at Xp11,

(Brunner et al. 1993a: abstract) and that same year a point mutation that turned MAOA

‘off’ in the probands was identified (Brunner et al. 1993b). MAOA deficiency resulting in mild intellectual disability and impulsive, aggressive behavior came to be known as

Brunner Syndrome. To date, only the 14 cases in the family identified by Brunner et al., with ten of them actually exhibiting the phenotype of aggressive, impulsive and/or violent behavior, have been identified as having Brunner Syndrome. Unsurprisingly then, there is very little biosocial mobilization around Brunner Syndrome as an illness or identity category – no corollary to the Williams Syndrome Association or the Fragile X Syndrome

Foundation discussed below.

But the association between the MAOA gene and human proclivities for aggression, violence and related behaviors did not stop there. Brunner et al.’s findings led directly to further research on the MAOA gene and the discovery of a less severe variant than the fully deactivated mutated gene found in Brunner Syndrome. Just under a decade later another report appeared in Science (Caspi et al. 2002) entitled, ‘Role of Genotype in the Cycle of Violence in Maltreated Children’. Taking a sample of 1,037 children from an

! 174! ! existing multidisciplinary study and focusing on the males (52%), the report takes as its dependent variable the display of ‘antisocial behavior’: 1) adolescent conduct disorder according to DSM-IV criteria; 2) convictions for violent crimes; 3) “a personality disposition toward violence” during a psychological examination at age 26; or 4) symptoms of ‘antisocial personality disorder’ at age 26 by people nominated by the subject as “someone who knows you well (ibid: 852). They then compared composite antisocial behavior scores according to childhood maltreatment (none, probable and severe) and, crucially, genetic data about levels of MAOA activity (low and high). In short, using the example of a polymorphism in the promoter of the MAOA gene that had already been established to significantly reduce monoamine oxidase expression (Sabol, Hu, and Hamer

1998), Caspi et al. reported the impact of a genetic variant on the display of antisocial behavior.

They found that maltreatment had a strong and statistically significant effect on the likelihood of engaging in antisocial behavior, while the low MAOA variant did not.

However, when they examined the interaction effect between the MAOA variant and childhood maltreatment – “the gene-environment interaction” – a strong and significant finding emerged: if a person was maltreated as a boy, then low MAOA activity as measured by the DNA variant in its promoter region becomes a strong predictor of antisocial behavior. What’s more, the effect of maltreatment was much weaker among the majority with high MAOA expression than among those with the low variant. The provocative finding, which the authors called on others to replicate, was that the MAOA polymorphism mediated the impact of childhood abuse on antisocial behavior. In other words, the MAOA findings suggested that it is traumatic experience, mediated by a genetic

! 175! ! endowment, which confers the greatest risk of pathological behavior. The paper has since been cited over 2700 times. Despite mixed attempts to replicate Caspi et al.’s findings (G.

Frazzetto et al. 2007), the field of research on the link between variants in the MAOA gene and aggression had come of age: from an extremely rare x-linked syndrome to an allele associated with anti-social behavior found in around one third of the male population, the

MAOA gene was being turned towards ever-broader questions about human difference.

Ever since Caspi et al.’s 2002 study the MAOA gene and the common variant in the promoter region that decreases its expression has served as a platform for discussions of genetics and aggression in both popular media and biomedical venues. To take one of many examples, the UK’s Guardian reported on Caspi et al.’s study:

Scientists have identified a gene that plays a role in the cycle of violence in men abused in childhood. The discovery could explain why some survive unhappy childhoods, and go on to normal lives, while others turn to violence, crime or antisocial behaviour… The discovery might help explain why violence tends to be male rather than female - the gene is found on the X chromosome. Men have only one copy of the X chromosome, women have two. The gene might also be a predictor for the ability to tolerate mental stress or trauma. The military, the police or firefighters might screen recruits to see if they have the more active form. But the discovery also raises the spectre of biology as destiny, and the argument that people with the less active form of the gene could be social risks, to be treated with drugs. (Radford, 2002)

These reports tended to emphasize the significance of the MAOA finding for our understanding of human behavior while also noting Caspi et al.’s focus on gene x environment interaction rather than genetic determinism. As the Los Angeles Times put it

(Singer 2002), Caspi et al.’s MAOA findings “underscore the interaction between genes and environment in influencing human behavior. ‘People shouldn't think they have no choice of whether to be violent or not,’ [one of the study’s authors] said.”

! 176! !

Further research examined the MAOA variant in new fields. A report on the 2004 meeting of the American Association of Physical Anthropologists in Science discussed work ‘Tracking the Evolutionary History of the “Warrior” Gene’ (Gibbons 2004: 818).

Noting that, “For males, a bit of aggression and risk-taking can earn rewards – just ask real-estate mogul Donald Trump,” the report discusses the way research by Brunner et al.,

Caspi et al. and others on the MAOA gene was used to trace the evolutionary ‘balancing act’ between aggression and its risks. After a German psychiatrist found a similar MAOA variant tied to aggression in macaques, a subsequent study took tissue samples from some

600 apes and monkeys and found the MAOA variant in all apes and ‘Old World’ monkeys but not in ‘New World’ ones, suggesting that the aggression-inclining gene variant has been around for at least the 25 million years since apes and monkeys split but not as long as the split between New and Old World primates. Aggression and the MAOA variant is therefore advanced as a rare example of ‘balancing selection’ – a genetic trait with mixed adaptive implications that can be confer an advantage for some as long as it remains confined to a minority of the population.

Until just a few years ago, the MAOA gene’s reduced-expression variant, unlike the extremely rare disrupted variant identified by Brunner et al., could only be associated with aggression and anti-social behavior in populations that had experienced some kind of environmental insult, usually some form of childhood abuse. However, a 2006 paper in

Proceedings of the National Academy of Sciences by Meyer-Lindenberg et al. (2006) examined the neural mechanisms associated with MAOA variation and found differences in brain function between those with the high and low variants, independent of environmental factors, which “point toward potential targets for a biological approach

! 177! ! toward violence.” (p. 6269) Three years later a study in the same journal by McDermott et al. (2009) catapulted MAOA variation to even greater levels of general significance and popular attention. Armed with 78 male subjects who had been tested for MAOA high and low variants, a behavioral economics study design and a plentiful supply of bottled ‘hot

(spicy) sauce’, this group of three political scientists and one ‘transdisciplinary scholar’ with a background in molecular biology set out to examine behavioral differences in subjects’ reactions to provocation. In each of four rounds, after earning money from a vocabulary task, subjects would have either 20% or 80% of their earnings taken from them by an anonymous (and in reality fictional) alter. In each round, they were given the opportunity to either punish the alter through the forcible administration of the unpleasantly spicy hot sauce or to trade the hot sauce in for money. As expected, those who had 80% of their earnings taken were more likely to retaliate and more likely to administer the maximal punishment (i.e. a whole bottle of hot sauce) to the supposed thieves.

Crucially, those with the MAOA-L variant had both higher rates of retaliation overall and a strong and statistically significant higher rate of retaliation when 80% of their earnings were taken. In particular, those with the MAOA-L variant were more than twice as likely to administer the maximal punishment (three dollars’ worth of hot sauce) in response to either a 20% or an 80% theft (McDermott et al. 2009: 2120). The MAOA-L variant and its behavioral implications was no longer mediated by the environment in the form of a developmental factor, as with Caspi et al. (ibid), but a genetic effect “moderated by the environmental stimulus” (my emphasis). What’s more, McDermott et al. argued:

! 178! !

Although spite has been the ‘neglected ugly sister of altruism’, there is good reason to expect it may have played a significant role in the evolution of human behavior. The influence of genotypic variation among individuals also complicates the notion that humans are “altruistic” punishers because it raises questions about whether one behavioral strategy is really common to a majority of people. Models of the evolution of cooperation might usefully be revisited with this in mind, especially because recent game theoretic treatments find that punishment may evolve in some subsections of the population but not others (raising the possibility of frequency dependence or mixed strategies). Our study suggests that there may be genetic bases for such a hypothesis. Indeed, our results beg the question of why the MAOA-L allele has been maintained in the population [typically 1/3 in western populations, although approaching 2/3 are reported in Maori populations] if it promotes aggressive behavior… One possibility for why MAOA-L has not become universal lies in frequency dependent selection; if everyone were MAOA-L, its advantages would disappear. If everyone were MAOA-H, there may be a niche for more aggressive individuals to exploit.

In short, the MAOA-L allele is used to provide a genomic basis for theories about of human cooperation and conflict, suggesting that biologically distinct kinds of behavioral proclivities have evolved at, among other possible sites, a tandem repeat at 11.3 on the short arm of the X chromosome. Around one third of the (European) population is posited to be genetically predisposed to towards aggression and spiteful punishment in the face of perceived provocation. So where would this line of research go from here?

MAOA research ran into serious controversy in New Zealand when a 2005 conference paper (Rod Lea et al. 2005) reported substantially higher rates of MAOA-L variants among Maori populations and was picked up by the popular press as an explanation for high Maori crime rates. Despite initially favorable coverage a backlash ensued, with leaders from the Maori Party and leading New Zealand biologists roundly condemning the finding on both ethical and scientific grounds (Stokes 2006; Chapman

2009; Anon 2009; Crampton and Parkin 2007). A detailed response by Lea and Chambers

(2007) was published in The New Zealand Medical Journal where they explained how the press had drawn conclusions about “non-medical antisocial issues like criminality” that

! 179! ! were not supported by their research, which only sought to examine alcohol and tobacco response traits. However, they also note that “It is well recognised that historically Maori were fearless warriors,” and that the ‘warrior gene hypothesis’ represents “a retrospective, yet scientifically plausible explanation of the evolutionary forces that have shaped the unique MAO-A gene patterns that our empirical data are indicating for the Maori population.” Their finding is taken to be:

evidence of positive (natural) selection acting at the MAO-A gene. It suggests to us that Polynesian males who embark on long, dangerous canoe voyages and engaged in (and survived) war with other islander tribes carried the AGCCG haplotype, coupled with the 3- repeat allele of MAO-A30bp-rpt, to Aotearoa (New Zealand) where they both increased in frequency due to rapid population growth. (online)

While they acknowledge the responsibility of experts to manage the popular dissemination of their findings while also distancing themselves from the more crass interpretations of their research, Lea and Chambers defend the scientific validity of their findings, their historical/evolutionary explanation of them and the validity of environmentally mediated genetic markers that are not evenly distributed across ethnic groups in the study of human behavior. Unsurprisingly, their reply failed to placate either the Maori Party or their scientific critics and research on the topic appears to have stalled.

An even more recent finding by Beaver et al. (2010) using data from the National

Longitudinal Study of Adolescent Health found that males with a low-expression variant of MAOA were nearly twice as likely to join a gang and use a weapon in a fight.

Furthermore, gang members with low MAOA expression were over four times as likely to use a weapon in fight, suggesting that “variation in violence among gang members may be partially circumscribed by genotype.” (133) After reviewing their findings, Beaver et al. situate them in the history of MAOA research and present their conclusion thus:

! 180! !

An impressive amount of empirical research has demonstrated that the low MAOA activity alleles are associated with a range of antisocial phenotypes, including serious physical violence and criminal behavior… Although the study of gangs has largely proceeded as a sociological phenomenon, this investigation shows that gang formation and activity, like most antisocial behaviors, involves gene-environment interplay.

As with the MAOA/Maori research, media coverage followed, though I am unaware of any backlash to Beach et al.’s study. Katherine Kingsbury reported the finding enthusiastically in Time (2009), gesturing towards the research “leading MAO-A to be referred to as the

‘warrior’ gene,” before explaining that Beaver et al.’s “research, however, takes the association one step further.” They quote Beaver, who explains “’For the first time, we were able to establish a direct connection between the MAO-A gene and the choosing of a violent lifestyle.” Kingsbury then quotes another MAOA researcher, Joshua Buckholtz:

“’What all these risk gene studies show us is that genes do an important job in loading the gun… But it’s the environment that pulls the trigger.” The piece concludes with Beaver noting that early intervention rather than gene-based drugs are the best societal recourse to the finding that 1/3-1/2 males are genetically predisposed to antisocial behavior and violence. Similarly, even a critical article on the misuses of the ‘warrior gene’ in New

Scientist (Yong 2010: 37) concludes thus:

Once we move beyond genetic determinism, the nature/nurture dichotomy and simplistic generalisations, the discovery of genes related to mental or behavioural disorders can only improve our knowledge of ourselves. It will also help us make better decisions. For example, adoption agencies might want to place children with the MAOA-L gene with particularly stable families. If a purported link between MAOA and alcoholism holds up, people with certain variants might choose to become teetotal to avoid the risk of addiction. And perhaps we could tailor effective criminal rehabilitation programmes to the variant of MAOA and other crime related genes a prisoner has. For Moffitt [Caspi’s coauthor], the best defence against the future misuse of genetic information is a “more realistic, nuanced understanding of the causes of behaviour, in which some genes’ effects depend on lifestyle choices that are often under human control”. This view is echoed by Beaver. “We could look at how to manipulate the environment to affect genetic predispositions,” he says. “It’s

! 181! !

a way to harness this type of information and use it in a very progressive and humane fashion.” (my emphasis)

What then is to be done with data about our genetically mediated proclivities for antisocial behavior, violence and aggression?

In a few cases MAOA testing has been invoked in criminal trials (Jorde 2012), including a high-profile murder case. However Young quotes several experts who note that the information could be used to justify harsher sentences, leading defense attorneys to reconsider the strategy of invoking MAOA tandem repeat status in trials. Fortunately, prevailing norms about the use of genetic information, including the Genetic Information

Nondiscrimination Act (Louise Slaughter 2008) will severely prescribe the widespread use of MAOA testing by public organizations, even in the ‘progressive and humane’ vein outlined by Young and Beaver above.

The more straightforward, plausible answer to the ‘what is to be done?’ question when it comes to knowledge about MAOA variation seems to be: sell it. Since November

2010 Genealogy by Genetics, Ltd., dba Family Tree DNA, have offered a ‘Warrior Gene’ test for a mere $99 (Genealogy by Genetics, Ltd 2010). The relevant page on their website, featuring pictures of a football player, a medieval knight and a cross-looking middle-aged man in a shirt proclaims:

WHETHER IN SPORTS, BUSINESS OR ANY OTHER ACTIVITY, SCIENTISTS FOUND THAT PEOPLE WITH THE WARRIOR GENE VARIANT WERE MORE COMBATIVE THAN THOSE WITHOUT THIS VARIANT... Order your DNA test now and find if you have the "Warrior Gene"!2

But what form would the public representation of the MAOA-L variant and this new DTC test for it take? !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2 http://www.thewarriorgene.com/ (accessed April 26, 2012)

! 182! !

Around the same time, National Geographic’s show Explorer aired an episode entitled ‘Born to Rage?’ (Day 2010) where the former Black Flag front man and famously angry Henry Rollins travels around the country meeting a coterie of easily aggravated people, as well as some Buddhist monks, exploring their anger and testing them for a low

MAOA variant. The show opens with a montage of violent urban scenes set to a brooding soundtrack and the narration: “Are some people born to be violent? An extraordinary discovery suggests they are. A single gene has been directly associated with violent behavior. This controversial science has ignited debate over genetic screening and eugenics. Now, for the first time, Explorer takes you inside the Warrior gene.” The show also features interviews with McDermott (of hot sauce fame) and discusses how gene- environment interaction is increasingly undermining the consensus that nurture tends to trump nature in the production of human behavior, “and as a society, it’s something we can try to put right. But in recent years a new breed of genetic scientist is challenging that conventional wisdom like never before. The discovery of the warrior gene suggests that nature has a far bigger influence on our behavior than we’d ever imagined.” The show reaches its denouement with the revealing of Rollins’ ‘Warrior Gene’ test result (it is negative) and concludes: “The links between our genes and our behavior makes it clear that the debate is no longer nature vs. nurture. It’s nature and nurture that make us who we are. And whether we like it or not, the journey inside the warrior gene has only just begun.” As the show’s producer put it to USA Today:

“We raise more questions than answers, because we are really at the very beginning of studying what genetics tells us about our race," Day says. "It is possible to have a gene abnormality and yet control your behavior. But it's also safe to say that if you mix the experiences of a difficult childhood with this Warrior Gene, it can be incendiary." … Day

! 183! !

acknowledges that the gene "is really mistitled, because Warrior Gene has a sort of nobility to it when we may not often be talking about a noble sort of violence. But then again, if you called it the Devil's Gene, I'm not sure people would want to see if they had it." (Della Cava 2010)

Both Day and Genealogy by Genetics Ltd. then seem to be in agreement that it would be better if people wanted to learn whether they have a low-expression MAOA variant.

In mid-2011 Dr. Phil aired an episode (Pennington 2011) with the same title, again featuring McDermott, and a tagline: “Scientists believe they may know why some people are quicker to anger than others. A new study suggests that inside a rageaholic's DNA, "a warrior gene" may be pulling the strings. Could today's guests be genetically predisposed to fits of fury?” The show’s web page recounts the highlights:

"We've given our guests the DNA test to determine if they have a gene that scientists say can lead to aggressive behavior," Dr. Phil says. Dr. Phil says Lori, Bryan and Scott do want to control their frustration and rage. "All three are about to find out if they have the warrior gene… We tested each of you for the warrior gene," Dr. Phil tells the three guests. "Do you think you have it?" "I want to hope that I do," Scott says. "I think it would be a good thing," Bryan says. "You wouldn't have to look so hard to see what's troubling me." Lori says she doesn't think she has it. Even though the warrior gene is rarer in women, test results reveal that Lori does have the gene. "Well, it's an answer for me," Lori says. "It helps me understand why I get as angry as I do. It's a relief there's something linked to this anger, and it's not brought on because I want to do it." Bryan and Scott also carry the warrior gene. "I wish it did explain everything," Dr. Phil says. "I want all three of you to know that when you have a gene, and you have the trigger from the environment, that means you're more likely, or more susceptible; it doesn't mean you have to. It doesn't mean you can't control it… This is not an excuse to go out and rage against people."

Then, at the bottom of the page we are given the link to the Family Tree DNA test, prominently featured in the show itself: “Think you could be a warrior? To get your own

! 184! ! warrior gene testing kit, click here and enter promo code Dr. Phil to receive $30 off your order.” 3

In sum, we have seen how Brunner Syndrome and the MAOA disruption identified as its etiology were leveraged towards more general issues about human proclivities for antisocial, violent behavior. In all the major research papers, Brunner et al.’s findings are cited as the origin of the MAOA line of research into this kind of behavior, and even though Brunner himself has renounced much of that work (Han G Brunner 1996) it is clear that his team’s findings are the point of origin for the ‘warrior gene’. From Caspi et al.,

McDermott et al. and Beaver et al. through to National Geographic and Dr. Phil, gene x environment interaction is invoked to both ground statistically significant findings and create a firewall to charges of reductionism and eugenics, though of course one would struggle to find a genetic determinist crass enough to disagree. The path from Brunner

Syndrome to the MAOA variant and then to the ‘warrior gene’ is suggestive of the way biomedical researchers can leverage the discovery of genetic mutations in severely affected patients to serve as platforms for far-reaching studies about human variation, and eventually even to commercialized allelic tests. However, in the face of significant public criticism (see above and Horgan 2011) and an absence of social mobilization, there is no clear path forward for MAOA research or what to do with it.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 3!http://www.drphil.com/slideshows/slideshow/6293/?id=6293&slide=0&showID=1626&preview= &versionID= (three pages; accessed April 26, 2012)

! 185! !

Our genetic elves

With an incidence of around one in 7,500 (Stromme, Bjornstad, and Ramstad 2002)

Williams Syndrome is not unlike dozens of other rare disorders. However its discrete genetic etiology combined with its distinctive cognitive profile, marrying moderate intellectual disability and a characteristic ‘elfin’ facial structure with unusual sociability, verbal skills and (arguably) musicality has made it the subject of unusually intense biomedical and popular attention. Often called the ‘anti-autism’, Williams Syndrome is a particularly useful object of social scientific analysis because it shows how abnormality is leveraged and mobilized to explain, not only disability and impairment, but also core characteristics of normal human development. A survey of the scientific and social action around Williams Syndrome takes us on a dizzying tour, with sites and objects of concern as varied as the study of the human capacities for music, language, empathy and sociability, a 7q11.23 genetic microdeletion, mouse models, neuroscience labs, foundations, a global advocacy network and, last but not least, Alan Alda.

Many even believe that Williams Syndrome is a modern way of categorizing and understanding the elfin ‘storytelling bands of yore’, raising the fascinating question: did the ‘elves’ and ‘pixies’ depicted in medieval European and other sources have a 7q11.23 microdeletion? According to Howard M Lenhoff, a biologist with over 140 publications to his name and also the father of a girl with Williams Syndrome, the answer appears to be

‘probably’. Lenhoff writes (1999: 150): “I would like to add a new wrinkle to some ideas concerning the origin of the forever young characters of folklore: the pixies, elves, and other fairies,” before explaining his method (154):

! 186! !

For the purposes of comparison, I have assembled descriptions of many of the characteristics ascribed to fairies of folk talks to define the “fairy syndrome.” Searching through anthologies and secondary literature sources dealing with fairies, I found a number of recurring physical and behavioral characteristics remarkably similar to those of individuals with Williams syndrome… Aside from the facial features already noted, I found seven key characteristics: stature; kindness; sensitivity; love of music, song and dance; hyperacusis; fascination with circles and spinning objects; and orderliness and concern for the future.

Lenhoff recounts an episode when John Burn noted in the Journal of Medical Genetics that he could not comment on whether Williams Syndrome children had elfin facies because he had “never seen an elf.” This statement was taken up a few years later, in 1990, by the former president of the Williams Syndrome Association at their annual meeting: “‘Of course Dr. Burn has seen elves before, because he has seen our Williams syndrome children.’ [He] went on to suggest that Williams syndrome individuals had existed for ages, but storytellers unable to fathom the cause of their unique features referred to them as elves, pixes, brownies and other such names…” Lenhoff left readers in no doubt that he agreed with this hypothesis:

In reviewing the physical and behavioral characteristics shared by the fairies of legend and individuals with Williams syndrome, I find the similarities too striking to be merely coincidental. There is no doubt that authors of oral and written folk tales in earlier times encountered dwarfs and other individuals whose physical and mental handicaps set them apart from the main population. Their lack of understanding of genetics and human embryology led them to invent magical and mystical explanations to account for these unusual people and their behaviors. In this essay I provide evidence to support the hypothesis that some of those legends evolved from encounters with individuals having Williams syndrome. (ibid: 157-8)

This theory that Williams syndrome is a biomedical category that is at least partly coextensive with the elves and fairies of folklore is widespread among advocates, bloggers and even experts, though Steven Pinker has suggested (2007:52) that “they look more like

Mick Jagger.”

! 187! !

Making a similar though less detailed argument in a Scientific American piece that reviewed Williams syndrome’s distinctive cognitive phenotype, Lenhoff concludes: “In the past, storytellers created folktales about imaginary beings to help explain phenomena that they did not understand… Today researchers turn to Williams people in a quest to understand the unknown, hoping to decipher some of the secrets of how the brain functions.” (Lenhoff et al. 1997:73) This research was favorably reported in the Los

Angeles Times (Boucher 1994) and crops up in the biomedical literature as well, as with this excerpt from a paper in Neurology:

On a philosophical plane, it is worth noting that there are similarities pointed out by Lenhoff et al. in their recent popular article in Scientific American between Williams subjects and folk such as pixies, elves, and leprechauns, and their folk song, glib talk, and merriment; these "mythical" figures populate the legends of many cultures. One wonders if people with Williams syndrome were not the inspiration for some of these folk tales. Understanding the biochemical and genetic basis for this syndrome has a fascination that reaches far beyond the syndrome itself." (Rossen and Sarnat 1998)

So how was Williams Syndrome delineated as a contemporary biomedical category, and what are these other, more forward looking research domains towards which it is leveraged as a model?

Within months of each other in 1961, a team of doctors in New Zealand and another in Germany independently reported a peculiar syndrome characterized by supravalvular aortic stenosis, mental retardation and a distinct facial appearance (Williams,

Barratt-Boyes, and Lowe 1961; Beuren, Apitz, and Harmjanz 1962). What has often been called ‘Williams-Bueren Syndrome’ soon entered into modern nosology, though ‘Williams

Syndrome’ is now the favored term. More recently, beginning in the mid-1990s, Williams

Syndrome has been associated with a microdeletion of genetic material at site 11.23 on the

! 188! ! long arm of the seventh chromosome and particularly the ELN gene, which codes for the protein elastin (see E Nickerson et al. 1995; LOWERY et al. 1995; Osborne et al. 1996; W.

P. Robinson et al. 1996; Wu et al. 1998; Donnai and Karmiloff-Smith 2000). With its association with sociability, unusual language abilities relative to other cognitive domains, and musicality, Williams syndrome has become an object for researchers interested in some of our most cherished human capacities.

However this discovery had very different, less reassuring consequences for others who had been diagnosed with Williams Syndrome. As we saw in Chapter 2, the isolation of a genetic etiology did more than establish the cause of Williams Syndrome: it actually changed its diagnostic criteria and delineation. A landmark study identified a deletion at

7q11.23 in around 90% of Williams Syndrome cases (E Nickerson et al. 1995), while a subsequent paper that examined 195 patients referred from the Williams Syndrome

Association and 40 clinical referrals found the deletion in 96% of 114 ‘classic’ cases but only 8% of 39 ‘uncertain’ cases and 60% of the 42 cases where clinical information was not available (LOWERY et al. 1995). In the case of Cystic Fibrosis similar discordance between genotype and clinical criteria did not impact nosology, with the latter retaining its status as the basis for diagnosis, while in the case of Huntington’s disease clinically typical cases can be ‘ruled out’ and atypical or even non-symptomatic cases ‘ruled in’ by the observation of a genetic mutation (Miller et al. 2005). Williams Syndrome has followed the path of Huntington’s disease (Miller et al. 2005), with the genetic mutation serving as a necessary and sufficient condition for diagnosis. As the Williams Syndrome Association itself puts it, to reiterate from Chapter 2: “It is important to stress that WS is a genetic diagnosis and an individual who does not have the gene deletion does not have Williams

! 189! ! syndrome (i.e. a person who was clinically diagnosed with WS but was later found not [to] have a deletion in fact, does NOT have WS).”4

Since becoming nosologically affixed to the microdeletion at site 11.23 on the long arm of the seventh chromosome, Williams Syndrome has attracted singular levels of attention from both biomedical researchers and popular media. But why? A paper entitled

‘An Experiment of Nature’ (Reiss et al. 2004: 5009) began by pointing out that Williams

Syndrome’s unusual cognitive phenotype and “well-defined genetic etiology… provides a rare opportunity for elucidating linkages among genetic, neurobiological, and neurocognitive levels of investigation.” Similarly, at a symposium on Williams Syndrome held at the annual meeting of the Cognitive Neuroscience Society in 1998 it was noted that

Williams Syndrome:

is the stuff geneticists’ dreams are made of… Dr. Korenberg explained that the reason for the excitement over the map [of 7q11.2] is that if we could precisely define various aspects of the hypersociability and language abilities in WMS, we could differentiate the WMS markers and thus relate a small number of genes to complex phenotypes… if it is indeed the case that this region of chromosome seven is responsible for variability in traits such as hypersociability and language ability, we might in fact be looking at part of what makes us human. To what extent are these genes variably expressed in each of us? Could these genes have something to do with the evolution of language? (St George 1998:203 my emphasis)

A more recent paper by several leading Williams Syndrome researchers proposed to “[use]

WS as an example of an “experiment of nature” (Järvinen-Pasley et al. 2008: 2–3):

…findings derived from genetically relatively homogeneous populations have direct relevance for understanding brain–behavior linkages in individuals in the general population who exhibit similar but more subtle patterns of cognitive, behavioral, and developmental characteristics… Although the information derived from behavioral neurogenetics will have direct benefit and relevance to individuals with WS, it also holds !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 4 http://www.williams-syndrome.org/diagnosing-williams-syndrome/diagnosing-williams- syndrome

! 190! !

wider relevance in illuminating the ways in which genetic and neurobiological pathways contribute to cognition and behavior in typically developing individuals.

There are many similar statements scattered throughout the biomedical literature on

Williams Syndrome (e.g. Meyer-Lindenberg, Mervis, & Berman, 2006, p. 391)

This approach to the study of Williams Syndrome as a biomedical vista on such core categories of human difference has also attracted significant attention from various forms of popular media. Take the 2001 episode of the PBS series Scientific American

Frontiers (Chedd 2001), when host Alan Alda “enters the worlds of children who are

‘Growing up Different’.”5 After a teaser, followed by an ad by Agilent Technologies that promises to turn the Cs, As, Gs, and Ts, of our genetic code into cures for disease, Alda looks into the camera to explain the show’s rationale. Trying to understand the behavior of others, he avers against the backdrop of children playing at a picnic, is one of the most difficult parts of being a kid. Alda goes on:

But for some kids, this struggle to understand the world is even tougher. That's because the difference that they're born with is so profound, that the world is more baffling than usual. In this program, we spend some time with children who are growing up different and with some of the researchers who are trying to understand why they see the world the way they do. We'll see how the insights that these researchers are achieving are not only helping the kids who are different make sense of the rest of us, but they're also helping the rest of us understand what it means to be human.

The show goes on to explore research and treatment for children with Autism, Cerebral

Palsy and the deaf, but Alda’s first stop is the Salk Institute for Biological Studies in San

Diego and a gathering of people with Williams Syndrome. Still at the same picnic, Alda explains: “At first glance the kids here are like most kids -- certainly they're high-spirited

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 5 http://www.pbs.org/saf/1205/resources/transcript.htm (accessed April 26, 2012)

! 191! ! enough. But they are all linked by possessing a rare genetic disorder called Williams

Syndrome.”

ALDA (NARRATION): The cause is literally visible under a microscope. When stained with a fluorescent dye, the chromosomes of a normal cell show a bright band in the middle of both copies of chromosome 7. In the cell of a person with Williams Syndrome, only one copy of the chromosome has this band. The missing chunk contains only about 25 genes. So scientists are hoping to be able to trace not only the disabilities of people with Williams Syndrome but also some of their special strengths directly back to just a handful of genes. ALDA: I mean, are you going to find out there's a gene for compassion? URSULA: Let's call it sociability and…goddamn it, we might. ALDA: You what? URSULA: We might. We might. ALDA: You might? […] That would be amazing. URSULA: Yeah, it would be, wouldn't it? I think that's sort of the hunt we're on. And I think that's a possibility. ALDA: So you're actually by studying carefully what the roots of Williams Syndrome are, you're actually finding out what the roots of qualities that all of us have are, huh? […] I mean, you're beginning to track down how we are who we are. URSULA I think that's put very, very well and that is true and the added fascination that we've got is that we can understand so much more how the brain does it. And how you can get in unusual ways to these strong qualities.

In short, Williams Syndrome and the 7q11.23 microdeletion through which it is now delineated is used as a platform in both biomedical research and popular culture to investigate the relationship between our genetic inheritance and core human faculties.

While this might seem like the kind of utopian rhetoric that was popular in 2001 as the Human Genome Project was nearing completion, Bellugi is the PI on a $5.5 million grant awarded by the NIH in 2010 “to link social behavior to its underlying neurobiological and molecular genetic basis using Williams syndrome as a model.”6

Williams Syndrome continues to attract attention from biomedical researchers and popular media, recently serving as the subject of a major documentary (Kent 2011), and its

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 6 http://www.salk.edu/news/pressrelease_details.php?press_id=488

! 192! ! foundation has grown in members and resources. Oliver Sacks devoted a chapter of his recent book Musicophilia (2007) to this ‘Hypermusical Species’. That same year a piece by

David Dobbs in the New York Times magazine explored Williams Syndrome at length in a piece entitled ‘The Gregarious Brain’, explaining why a neurogeneticist like Julie

Korenberg would tell him that WS “is the most compelling model available for studying the genetic bases of human behavior.” Dobbs explains (2007):

After being ignored for almost three decades, Williams has recently become one of the most energetically researched neurodevelopmental disability [sic] after autism, and it is producing more compelling insights…. [it] arises from a known genetic cause and produces a predictable set of traits and behaviors… perfect for studying not just how genes create intelligence and sociability but also how our powers of thought combine with our desire to bond to create complex social behavior — a huge arena of interaction that largely determines our fates.

Although Dobbs overstates the scale of research on WS (numerous neurodevelopmental disorders attract substantially more funding and publications) Dobbs is correct to point out that it has proven to be an unusually fecund site for biological research.

Crucially, in the Williams Syndrome Association (WSA) researchers find a community of potential subjects self-organized and motivated to participate in research.

The WSA is a registered charity dedicated to raising awareness and providing support for families affected by WS. For our purposes, it is the fourth and final of their mission statements that is most pertinent: “We will uphold our mission by: Encouraging and supporting research into a wide range of issues related to Williams syndrome.”7

On the one hand, the foundation embraces the biomedical leveraging of WS:

Cutting-edge research is leading to new insights about the workings of the brain in language and spatial processing, and may also contribute to greater understanding of common problems like hypertension and anxiety. The Williams Syndrome Association !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 7 http://www.williams-syndrome.org/content/wsa-mission

! 193! !

(WSA) is the most comprehensive resource for people and families living with Williams syndrome as well as doctors, researchers and educators. By supporting the WSA, you will create life-changing opportunities for people with Williams syndrome and help accelerate research that could have far broader applications… Because Williams syndrome is genetic, it’s possible that findings from research will pinpoint genes affecting certain medical conditions (cardiovascular disease, diabetes); developmental challenges (visuo-spatial problems, ADHD); and even personality traits (affinity for music, anxiety).8

On the part of their site aimed at researchers, they note that “Williams syndrome is gaining wider popularity in the research field due to its unique characteristics and to the wealth of information it is providing to the Human Genome project,” before providing a link to a

Williams Syndrome Registry that aims to provide detailed clinical data on WSA members to “enable authorized researchers to find potential candidates for research programs that can offer new hope for our WS family member's [sic] well being.”9

On the other hand, it is not clear how the foundation and the Registry can encourage work that offers “new hope for our WS family member's well being.” They note that “Knowledge about the features, health problems, and genetic changes in people with

WS has grown steadily over the past forty years. However, progress in developing new therapies has lagged behind.” The Williams Syndrome Association faces a challenging conundrum: the condition and the population they advocate for is characterized by serious developmental and physical disability, but it is used by biomedical researchers to pursue projects of LGA aimed primarily at the relative strengths exhibited in Williams Syndrome and their genetic bases. While it is clear that the WSA welcomes research of all kinds and appreciates the awareness raising function of media coverage, it remains to be seen

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 8 http://www.williams-syndrome.org/media/williams-syndrome-and-research

9 http://registry.williams-syndrome.org/

! 194! ! whether they will be able to leverage the status of WS as a privileged site for biological knowledge production in such a way that advances their own ends as advocates for a rare genetic disorder.

They do, however, appear to be gaining ever-increasing attention from the media, the medical community and biomedical researchers. A report following a Williams

Syndrome feature on ABC’s 20/20 program provides some insight into the role the foundation sees for media coverage (Lovett 2012):

With the network attention and awareness has come credibility. “To me, that’s the biggest piece,” said Monkaba [WSA’s executive director]. “If we can say, ‘As recently seen on ’20/20,’ people tend to pay more attention,” Monkaba said. “Science editors have gotten more interested, so general media attention on WS went from slim to 25 touches.” Monkaba reckoned WSA had received nearly as much media attention in the past year as in the previous 29. The most immediate impact has occurred in the medical community itself. “It’s pieces like ‘20/20’ ‘s that make all the difference,” Monkaba said. “Doctors see it, they go to the website, get interested.” Doctors’ and the public’s heightened awareness has caused more people to be diagnosed with Williams Syndrome, here and abroad, and at younger ages, she said.

Williams Syndrome’s role as a platform for other concerns came next:

Last fall the National Institute of Child Health and Human Development awarded a $5.5 million grant to scientists from several institutions and disciplines to study Williams Syndrome to learn how genes govern behavior. The study could produce drugs and therapies for those with Williams Syndrome. It could also help those with more-common disorders like autism (which affects between one in 150 and one in 500 newborns) by illuminating how genetic differences affect behavior, Monkaba said. “Our 15,000 kids may hold the key to helping millions with autism,” Monkaba said. “What a great legacy!”

Again, we see that research on Williams Syndrome is directed towards general human behavior, and now autism (Williams has been described as the ‘anti autism’, though we will see in the next chapter how that has changed dramatically in recent years). As unlikely as a drug therapy for Williams Syndrome may be given the enormous sums involved and

! 195! ! the kinds of research questions it attracts, the Association seems to at least partly embrace its role as a platform for big questions – biomedical and otherwise.

In sum, Williams Syndrome and its genetic specificity is increasingly being used as a kind of biomedical vista on the relationship between our genetic inheritance and core human faculties like language and sociability. Collaboration between diverse researchers and Williams Syndrome advocates has produced a small but vibrant field of biological knowledge production united by the goal of LGA, but it remains to be seen whether the

WS community will in turn be able to leverage that research towards its own ends.

A genetic model for autism

Fragile X Syndrome’s origins as a medical category bear a striking resemblance to those of Brunner Syndrome: a woman came to a hospital in a major European city seeking support for a family member showing signs of ‘mental deficiency’, which was understood by the family to be an inherited condition affecting only the sons. It was the early 1940s and the setting was the National Hospital in London. Like Brunner et al. in Amsterdam half a century later, J. Purdon Martin and Julia Bell worked to establish a pedigree – the pattern of affected and non-affected members of a family that could be used to determine a

Mendelian inheritance mechanism. When Martin and Bell published their paper, ‘A

Pedigree of Mental Defect Showing Sex-Linkage’, on this particular family in 1943 they inaugurated a field of research about what came to be known as ‘X-linked Mental

Retardation’ or XLMR (J. P. Martin and Bell 1943). In the subsequent studies further cases of XLMR were established according to family pedigrees (William Allan and C. Nash

Herndon 1944; e.g. W. Allan et al. 1944; Losowsky 1961; Renpenning et al. 1962; Opitz et

! 196! ! al. 1965), giving rise to a tricky nosological conundrum: were these all cases of one category, XLMR, or clinically specific disorders caused by distinct X-chromosome abnormalities (see Gillian Turner and Opitz 1980)?

Unlike Brunner et al., Martin and Bell did not report a more detailed phenotype than mental deficiency, noting (ibid: 154-5):

No peculiar features, either mental or physical, have been recognized which would serve to distinguish the disease which afflicts this family from other forms of dementia… Physically the affected males are not abnormal. They are in general of sturdy build, but they do not show any distinctiveness in shape or head or face... Their sexual development appears to be normal.

Nor did they have at the their disposal techniques that could identify particular genes or chromosomal abnormalities underlying these observed, x-linked phenotypes – indeed we saw that it would be some fifteen years before the normal number of human chromosomes was established as 46 and a system for numbering them had been agreed upon – and it was not until 1969 that any kind of anomaly on the X chromosome was associated with XLMR

(Lubs 1969). Yet despite the absence of chromosomal analysis or the observation of a phenotype beyond what we would now call intellectual disability, this study is now considered the foundational report on what came to be known as Martin-Bell Syndrome and then Fragile X Syndrome: a condition characterized by an abundance of CGG copy repeats affecting the expression of the FMR1 gene at 27.3 on the long arm of the X chromosome and variable degrees of intellectual disability, high rates of autism, a distinctive facial appearance and macroorchidism (McLennan et al. 2011; Anon 2011).

Fragile X Syndrome is now diagnosed strictly according to the CGG repeat at

Xq27.3, which delineates quite a different population from the XLMR described by Martin and Bell in 1943. It also diverges considerably from the population for whom a ‘fragile

! 197! ! site’ on some of their chromosomes was visible under a microscope as first reported in

Lubs’ 1969 paper and refined over the next twenty-plus years. Furthermore, the tests developed to test for the CGG repeat in the early 1990s allow for the identification of

‘carriers’ – primarily females whose second X chromosome compensates for the Fragile X mutation resulting in mild or absent symptoms, but whose sons have a 50% chance of being affected. While prevalence studies in the 1980s using cytogenetic techniques put

Fragile X at around 1:1000 (T. P. Webb et al. 1986), subsequent studies using molecular techniques put rates at more like 1:4000 (G. Turner et al. 1996; Morton et al. 1997).

Fragile X is now the most well understood and widely recognized of some 100 genetically specific forms of XLMR (H.-H. Ropers and Hamel 2005). We have come a long way since

Martin and Bell’s work of 60 years ago.

However, even as it was found to be a more rare form of XLMR than had been previously thought, the 1990s saw Fragile X Syndrome grow in leaps and bounds as an object of biosocial action and biomedical research. Three related developments were primarily responsible: first, a reliable molecular test was developed that could identify the

CGG repeat at FMR1 (OBERLE et al. 1991; VERKERK et al. 1991; KREMER et al.

1991; PIERETTI et al. 1991; Rousseau et al. 1991); second, a patient advocacy group whose extraordinary success would make it a model for others to follow was taking its first steps; finally, Fragile X was increasingly associated with autism spectrum disorders, which were beginning their steep rise in prevalence and salience during these years.

Beginning in in the early 1980s autism began to be noted in case reports on Fragile

X Syndrome (Brown et al. 1982; GILLBERG 1983; Rj Hagerman and Jackson 1985; A

Levitas et al. 1983; WATSON et al. 1984). In 1986 a first major study by Randi Hagerman

! 198! ! and colleagues that examined fifty males with Fragile X, “which [they] consider synonymous with the Martin-Bell syndrome,” (abstract) for autism and autistic traits was published in The American Journal of Medical Genetics (Randi J Hagerman et al. 1986).

They reported autism rates of up to ~30%, depending on the diagnostic criteria used (a wrinkle taken up in great detail in the next chapter), and autistic traits in almost all the children examined. Along with some parents of affected children, Hagerman had founded the first major Fragile X advocacy organization in 1984, and we will see below that it would blossom into an extraordinarily successful foundation in the decades that followed.

Fast-forward twenty-five years from her first paper on Fragile X and autism, and

Hagerman (Hagerman et al. 2010: 9) is able to report that “Targeted treatments to reverse these problems are currently being studied in patients with FXS. Many of these targeted treatments may also be helpful for ASD without FXS.” How was so much progress made for a rare genetic disorder?

First or all, Hagerman was not reinventing the wheel. Another author on that 1986 paper was Bernard Rimland, a psychologist who played perhaps the key role in the development of autism advocacy and research as we know it and therefore its ‘epidemic’ increase in prevalence (Eyal 2010; Silverman 2011). Rimland’s own Autism Research

Review International reported these and future Fragile X-related findings enthusiastically

(1987, 1989, 1991). In short, the Fragile X community was able to draw quite directly on the repertoires of collective action and collaborative research pioneered by Rimland and autism parents. Second, autism researchers and advocates saw in Fragile X the promise of a genetic model that could be leveraged to better understand their far more prevalent condition, creating an increasingly powerful rationale to invest resources for research into

! 199! ! a rare genetic disorder like Fragile X Syndrome. Thus the rise of autism as a category of childhood abnormality and Fragile X as a privileged site of biomedical research were intertwined from very early on.

Drawing on many of the repertoires pioneered in autism advocacy, the new Fragile

X community set off on the task of assembling an extensive network of biomedical researchers, parent activists and political allies. Crucially for us, they did not simply draw on the example of autism research and advocacy, but also leveraged the hope that their specific genetic disorder would help to unlock the biological mechanisms underlying autism more generally. Furthermore, they happily embraced the interest from biomedical researchers that Fragile X Syndrome attracted on that basis. As autistic spectrum disorders have loomed ever larger as a public health issue and rubric for classifying childhood abnormality, Fragile X Syndrome research has grown accordingly. As a member of the US

Inter-Agency Committee on Autism put it (Koroshetz 2007:103–4):

…there are these autism-associated disorders which are monogenic and can give really interesting insights into how the brain doesn't develop normally and actually causes symptoms that are very similar to autism. Disorders like Fragile X, Angelman's syndrome, and tuberous sclerosis have genes that have been identified.

What makes these genetic disorders so useful for autism research? Koroshetz explained:

Really, one of the most important things there is that identifying these genes gives scientists tools. They now have animal models, mouse models that they can really work with. Sometimes having the tool is what really attracts the really good scientists. People who are really smart are going to stay away from a problem until an animal model shows up that they can work with. That can be a real attraction. So we are hoping that this research will help these kids with these really bad things, but it may also pay off in the general autism research.

As we will see below, this kind of reasoning features prominently in the justification of

Fragile X’s status as a priority area for NIH investment.

! 200! !

At the same meeting, a leading neurobiologist representing the Simons Foundation and non-governmental sources of autism research funding more generally described the situation like this (Fischbach, 2007, pp. 226–8):

But the great majority of autisms we just don't understand. We call them idiopathic. But we think even here there is a strong genetic component… There is no question genetics play a role here. But the genetics are complicated. I think it is fair to say that no matter how good the groups, no matter how large the patient population, the linkage and association studies have so far been disappointing. They are hard to reproduce, and it is not quite clear which way they are leading. As a result, there are over 100 candidate genes… There are various approaches to this heterogeneity. One is to look at the single-gene disorders, these syndromic autisms of Fragile X, Rett syndrome, and others, and say what is it about those diseases that may teach us something about autism. What is it about the biochemistry. Several groups are doing that, and we are funding several groups doing that.”

In other words, when you know a genetic mutation often leads to autism, as is the case for

Fragile X and several other genomically designated conditions, a range of experimental strategies come into view. For geneticists, and especially neurogeneticists, the preferred point of entry is an animal model. By inducing a mutation like the CGG repeat in FMR1 in mice (it is almost always mice), you can try to infer autistic behaviors, dissect and examine brain structure and function and, eventually, begin to experiment with the compounds that can lead to treatments. You can also conduct all kinds of IRB-approvable studies on people with genomically designated conditions, making small populations like the one diagnosed with Fragile X Syndrome some of the most studied people in the world.

Not only do they have a mutation that is thought to hold important clues about an incredibly salient condition, but as a biosocial community they are organized to facilitate their use as human subjects in biomedical studies. The National Fragile X Syndrome

Foundation has blossomed from a meeting at a kitchen table in 1984 to a registered charity with annual expenditures of around 1.5 million dollars (“National Fragile X Foundation,”

! 201! ! n.d.). Their core missions are to fund and facilitate research, provide information and support of affected families, raise awareness, lobby and establish a network of specialist clinics. While they provide postdocs and research fellowships, their lobbying has helped to make Fragile X Syndrome a major topic of biomedical research with diverse funding streams. Taking the NIH alone, FXS is now priority topic with some 32 million dollars allocated last year alone as part of a steady upward trajectory over the last several years.

Fragile X advocates have even successfully lobbied for millions in research funding through the US Department of Defense as well as the CDC and raised their own funds from corporate donors, foundations and individual donations.

And yet as effective as their activism has been, the potential for LGA from FXS to autism more generally remains something of an obligatory passage point (Michel Callon

1986) – the sine qua non of any justification of allocating so many resources to a rare disorder. The most recent NIH research plan for Fragile X (THE TRANS-NIH FRAGILE

X RESEARCH COORDINATING GROUP AND SCIENTIFIC WORKING GROUPS

2008) noted:

As many as 30 to 50 percent of individuals with FXS meet the diagnostic criteria for autism or autism spectrum disorders (ASDs). FXS is considered a portal for understanding a variety of neurobehavioral disorders, including autism, ADHD, and anxiety disorders. (4)

There are currently multiple research efforts aimed at developing treatments and novel interventions. One such effort is related to clinical trials of pharmaceuticals for FXS includes an ongoing cooperative agreement led by the NIMH in partnership with the NICHD, NINDS, FRAXA, and Autism Speaks to develop therapeutics related to metabotropic glutamate receptor (mGluR) antagonists to treat FXS and autism… The NIH’s focus on efforts to understand the relationships between FXS and autism continues through the Program Announcement (PA) soliciting research to study the Shared Neurobiology of FXS and Autism. (8)

Later in the report, this goal was spelled out even more clearly (18 and 33, my emphasis):

! 202! !

“Objective 4.5. Leverage knowledge about biological pathways in FXS to design treatment studies for individuals with other developmental disabilities that share common pathophysiological mechanisms.” This confluence of concerted activism by FXS advocates and the hope that research on a rare syndrome can provide novel insights about autism has made the National Fragile X Foundation perhaps the single greatest success story among foundations for rare genetic disorders and a model for others to follow.

Their work may be about to bear fruit. At an autism research meeting in 2005, MIT geneticist Mark Bear and colleagues presented evidence that autistic symptoms could be reversed in ‘Fragile X mice’, and today, no less than three pharmaceutical companies have

Fragile X products in late stage trials. The company Bear founded, Seaside Therapeutics, actually has two, one of which is expected to hit the market next year. As Bear put it in a paper last year entitled ‘Toward Fulfilling the Promise of Molecular Medicine in Fragile X

Syndrome’ (Krueger and Bear 2011: 411):

FXS is therefore poised to be the first neurobehavioral disorder in which corrective treatments have been developed from the bottom up: from gene identification to pathophysiology in animals to novel therapeutics in humans. The insights gained from FXS and other autism-related single-gene disorders may also assist in identifying molecular mechanisms and potential treatment approaches for idiopathic autism.

What’s more, trials of the Fragile X compound for the general autism population are now underway at Seaside.

In sum, Fragile X Syndrome is leveraged by biomedical researchers for whom prevalent conditions, primarily autism but also ADHD and intellectual disability, have proven recalcitrant to penetration by the molecular gaze. The Fragile X community has mostly embraced this form of LGA. Indeed they engage in LGA from the opposite direction, seeking to attract research and pharmaceutical investment that can only be

! 203! ! rationalized on the basis of FXS’s capacity to speak to much larger populations of patients.

Where this nexus will lead is an open question. A number of pharmaceutical treatments are entering late stage trials, though whether they will be profitable or help Fragile X patients, never mind the larger autistic population, remains to be seen. Meanwhile, the CGG repeat at FMR1 is leading to the identification of new medical conditions based on smaller repeat permutations that fall short of the threshold for a full Fragile X diagnosis: Fragile X- associated tremor/ataxia Syndrome (FXTAS) causes balance and memory problems in older men, while Fragile X-associated primary ovarian insufficiency (FXPOI) decreases ovary production and can lead to infertility and early menopause in women with an FMR1 permuation. It is extremely unlikely that these conditions would have been recognized without extensive community organizing around full-blown Fragile X Syndrome (Krueger and Bear 2011: 411). Where this biosocial juggernaut will go next will be fascinating for the social studies of science and medicine, but it is clear that mutually productive projects of LGA on the part of those interested in FXS and autism were both crucial to its rise and likely to shape its trajectory moving forward. In any case, if Fragile X represents an important landmark in the surprisingly intractable project of ‘fulfilling the promise of genetic medicine’ (ibid), it was accomplished by an admixture of social mobilization and biomedical work.

Discussion

In all of our cases a genetic disorder was leveraged to open up new and far- reaching lines of research. We saw how the processes of affixing a population to a specific genomic abnormality can require extensive sociotechnical work. In Williams Syndrome

! 204! ! this meant excluding people who had been clinically diagnosed but did not have a 7q11.23 microdeletion; in Fragile X Syndrome key researchers made claims to longstanding categories of XLMR, despite very different clinical profiles, and went to extraordinary lengths to enhance awareness and increase the scope of FXS testing, especially in the mushrooming ASD population; in the case of Brunner Syndrome, the mutation identified in a mere 14 people formed the basis for a research program on a related gene variant found in perhaps two billion.

However, once fixed to specific genomic abnormalities, all the disorders discussed in this chapter were leveraged as what Star and Griesemer called ‘boundary objects’:

“objects which are both plastic enough to adapt to local needs and the constraints of the several parties employing them, yet robust enough to maintain identity across sites” and key to “developing and maintaining coherence across intersecting social worlds.” (1989:

393; see also Star 2010) While in Chapter 2 we saw how genetic mutations can unite otherwise disjunct clinical subfields, here we have seen how multiple biomedical disciplines and a range of actors with divergent interests converged on genetic mutations as objects of knowledge production and mobilization, leveraging them towards different ends.

Not only are the goals of the various groups different – so are their frameworks for understanding genomic abnormality. A genetic mutation can mean many different things: for a parent it is an explanation of their child’s developmental difference or illness, for a molecular biologist it is a deviation from the normal DNA inheritance that results in different levels of enzyme and protein production and for a neuroscientist it serves as a platform for observing abnormal physiological processes in the brain that can be linked to the disruption of normal cognitive functioning. Nevertheless, in each of our cases a

! 205! ! genomically designated syndrome served as a boundary object around which a range of biomedical disciplines and stakeholders were able to coordinate action and pursue overlapping projects.

Simply noting that genetic mutations can serve as boundary objects that enable various projects of leveraging, however, begs the question: boundary objects for whom? In other words, we have seen how the kinds of actors and interests brought together by genetic disorders result in contrasting biosocial networks that impact both biological research and its clinical, commercial and social ramifications. LGA may require that genetic abnormalities get taken up as boundary objects, but we have seen how leveraging can be pursued along very different paths of knowledge production and social mobilization.

Our three cases indicate that LGA is likely to be most effective when the genetic disorder in question is mobilized along multiple fronts and when alliances can be formed between actors with complementary goals. Table 1 (below) summarizes some of the findings from our three cases. We see that FXS is alone both in not being used as a platform for LGA with respect to general traits and in being used for one with respect to medical issues. Furthermore, FXS researchers and advocates moved early to forge alliances with their counterparts concerned with autism. MAOA-L and Williams Syndrome received far more popular attention, at least until recently, due to their role as a platform for LGA about general human traits and capacities, and the Williams Syndrome

Foundation and Family Tree DNA have worked to facilitate media coverage and reaped the rewards. Finally, FXS has been able to become an important object of biomedical investigation and investment, with funding for current NIH projects that list it in their title

! 206! !

or abstracts totaling some $90.3 million compared to around 10.5 for Williams Syndrome

and 8.2 for projects related to MAOA.

Table 1. Leveraging, alliances and resources in MAOA, Williams and Fragile X Syndrome research

MAOA-L WS FXS

LGA towards general traits Y Y N

LGA towards medical issues N Minimal Y

Advocacy for the disorder N Y Y

Alliances with other N N Y advocates Commercial uptake DTC genetic N Pharmaceutical, testing genetic testing Mass media interest Y Y Y (recent)

NIH current funding10 $8,235,993 $10,556,535 $90,317,274

Table 1 only tells part of the story. In order to really grasp the contrasting ways in

which these disorders have been used and transformed by LGA projects we need to

consider the sequences in which different actors came to them as boundary objects. The

potential for LGA with respect to general traits drove media interest in MAOA-L and

Williams Syndrome and led to DTC testing for the former and increased recognition and

resources for advocates concerned with the latter. In Fragile X Syndrome, by contrast, it

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 10 NIH Reporter, current FY funding for “Fragile X”, “Williams Syndrome” and “MAOA, MAOA- L, Brunner Syndrome or warrior gene” in project titles or abstracts.

! 207! ! was the potential for LGA with respect to medical conditions and an early alliance with autism researchers and advocates that fueled the success of groups like the National Fragile

X Foundation and eventually interest on the part of pharmaceutical concerns and, recently, mass media.

The differences in current levels of funding from the NIH – the only major agency to fund WS and MAOA research at a significant level (Fragile X receives funding from several others) – reflect Fragile X Syndrome’s embeddedness in the powerful circuits of biomedical research, treatment and advocacy related to psychiatric and childhood developmental disorders as well as what could be called the more general medicalization of biological research. For while Canguilhem and Jacob discussed leveraging in biological research in relation to basic questions about the nature and organization of life that were foundational and remain present in human genetics, medical concerns have become paramount. As McKusick, often dubbed its ‘founding father’, put it (1993: 2351):

“Medical genetics must be almost unique among clinical specialties; it arose out of a basic science [human genetics] rather than beginning as a craft that later sought out a scientific base.” McKusick notes that human genetics has gone from a discipline dominated by PhDs to one where MDs are firmly in the majority and the premium placed on research with medical relevance is unrivalled. In his review of early research on human chromosomes,

Harper (2006:55) is even more blunt, noting that the early cytogenetics research was almost all done in basic research labs, “for whom the human species had previously had no special significance and indeed was unpromising material by comparison with insect and other species with larger and fewer chromosomes.” Furthermore, the field “had no incentive to pursue human cytogenetics [and] no medical links.” (ibid) Finally, the

! 208! ! emerging human and medical genetics departments that did exist took scant interest in chromosomal research, and few had even considered that cytogenetics might one day serve as a valuable diagnostic tool. That all began to change in 1959. However, as we saw in

Chapter 3 that change was slow and halting – the observation of the genome did not substantially reconfigure medical classification and clinical practice. LGA today works within the context of the medicalization of human genetics and biology more generally, and the potential to speak to common conditions rather than normal traits or capacities attracts perhaps the most powerful alliances and resources. This shift, I argue, has facilitated the movement of genomically designated syndrome from esoteric objects of research in human genetics to sometimes-powerful categories of clinical practice and social action.

Whether oriented towards medical concerns or not, none of our cases are as sociologically straightforward as the ‘monsters’ discussed by Canguilhem and Jacob. In each case, alliances have had to be forged and esoteric questions about human morphology and development have been sidelined in favor of topics that can attract the interest of activists, commercial enterprises, consumers and/or mass media. While Canghuilhem gestured towards the ‘domestication’ of monsters (ibid: 142-3), the mobilization of advocacy by and for people with developmental disorders and other disabilities has transformed the field of research in which LGA operates. The populations of people diagnosed with genomically designated conditions are often organized into groups resembling Heath et al.’s description of ‘genetic citzenship’ (ibid), thereby driving ascertainment and participation in biomedical research. In other words, biomedical researchers often find that people with genetic disorders and their families have done much

! 209! ! of the laborious research of finding caseloads and facilitating the provision of research subjects, clinical data and the precious tissue samples that geneticists need to purse their research programs. Furthermore, the prioritization of ‘translation’ – work that can bridge the surprisingly intractable gap between information about our genomes and medically relevant knowledge – in NIH program announcements is in part the result of patient advocacy, but also a way of justifying biological research expenditures in the face of substantial opposition. Alongside patient advocacy organizations, the commercial and state organizations with access to capital therefore further direct biological research towards work that has translational (i.e. clinical) potential.

While Brunner Syndrome gave rise to a line of research on MAOA variants that has probably garnered the most publicity and (for now) commercial profit of our three cases, the fact that neither the foundational syndrome nor the ends towards which it was leveraged are the subject of any significant organizational activity (advocacy, clinical, political or whatever) leave it with no clear path forward. Williams Syndrome has also been the topic of considerable media coverage for its remarkable phenotypic profile and capacity to speak to quintessentially human abilities, and the Williams Syndrome

Association appears primed to reap significant gains from increased attention and awareness. However, they face the challenge of leveraging that attention and increased interest from biomedical experts towards their own interests. The challenges faced by people with Williams Syndrome and their families are simply not the features that attract biomedical and popular interest. Fragile X Syndrome, by contrast, may not receive the kind of enthralled media coverage as the ‘Warrior Gene’ and Williams Syndrome, but a robust network of researchers and advocates has been assembled around it that operates in

! 210! ! a close, sustained alliance with the field of autism research and advocacy. Through this dual process of leveraging, the interests of Fragile X activists and researchers and their counterparts concerned with autism are coupled and advanced in unison. In so doing, they have garnered hundreds of millions of dollars in funding, a huge increase in caseloads, an extensive network of clinics and advanced-stage pharmaceutical trials. In short, LGA is likely to be at its most powerful when the researchers and advocates dedicated to a genetic disorder are able to align or couple their interests with other categories of social and scientific mobilization.

For a time, MAOA researchers did maintain such a coupling. In some ways, the scope and scale of the biological knowledge produced by LGA from Brunner Syndrome into the research program on the MAOA gene and aggression far outstrips that of our other two cases: Amidst significant media attention the MAOA-L variant has been associated with the variable behavioral proclivities towards aggression of one third of the population, especially the male part; a widely publicized direct-to-consumer product that tested for that gene variant was introduced to the market; a rare case of genomically-grounded differential evolution has supposedly been identified, suggesting that an increased proclivity for violence, aggression and risk-taking are adaptively advantageous as long as it remains a minority of the population; and so on. However, there is no clear path forward for the network of actors interested in the MAOA gene and aggression and, despite interest from media and commercial concerns, the only significant political agitation associated with the line of research leveraged out of Brunner Syndrome has been directed against the use of

MAOA gene variant as a means to understand human difference.

! 211! !

It is certainly tempting to simply write the ‘warrior gene’ off as ‘bad science’ with disconcerting eugenicist undertones (Horgan 2011). That said, it is well grounded in existing research paradigms and has been the subject of dozens of articles in leading, peer- reviewed bioscientific journals. In the absence of a medical population beyond one extended family or a biosocial community to mobilize research, the LGA of MAOA deactivation to isolate the MAOA-L variant, link it to aggression and then commercialize and popularize it was in some sense a remarkable sociotechnical achievement. Indeed by emphasizing gene-environment interaction and direct-to-consumer self-practice, as opposed to a kind of unreconstructed genetic determinism and institutionalized utilization,

MAOA researchers and popularizers have erected a fairly effective firewall to charges of eugenicism. Nevertheless, the MAOA variant appears to be largely stalled as an object of knowledge production and biosocial identity formation. If it is an example of ‘bad science’, from an STS perspective with its commitment to symmetry its failure has more to do with the failure to build a robust network of allies a la Latour (e.g. 1987) and to conform with Fleck (1981:113; above) called the ‘exoteric’ scientific and public domains that any developing scientific field must court rather than with the methods that were employed, the questions that were pursued or the MAOA gene’s relevance for the way we understand and act on human difference. Like XYY Syndrome a few decades prior,

MAOA variants attracted phenomenal scientific and popular attention as an attempt to link mostly male aggression and anti-social behavior with genetic abnormality but, also like

XYY, attempts to mobilize it in institutional settings are unlikely to succeed under prevailing sociocultural conditions, especially in the face of a continuing, acute wariness about eugenics.

! 212! !

In sum, we have seen how the study of rare genetic disorders can give rise to far- reaching claims about human difference. However, the findings that emerge from LGA vary dramatically across cases, as do the implications for the populations whose genomic abnormality becomes an object of biomedical fascination. Despite garnering less popular attention, it is around Fragile X that a diverse network of actors and alliances has been forged, creating a genuine biomedical subfield, reshaping clinical practice, generating significant and sustained capital investment and reorienting identity and daily practice for thousands of people. Crucially, we will see how it has served as a model for others to follow. Using the framework of reiterated facticity to analyze the divergent ways in which these mutations were mobilized as facts in the world therefore helps us to see both how prior network formations enabled and constrained future outcomes and the way paths taken by one genomically designated condition create new pathways for the mobilization of others. Of course, the mutations themselves set parameters for leveraging. However, those parameters underdetermine the paths taken and, what’s more, at least since Latour and

Callon, science and technology studies has been committed to taking the role of ‘actants’

(non-human actors) seriously. In the case of contemporary LGA, we have seen how divergent biosocial networks actually shape research programs as well as clinical practice and social mobilization. This challenges us to consider the ways in which Rabinow’s

(1992) original formulation of the concept of ‘biosociality’ did not make sufficient room for the recursive processes through which genetics research would not only affect social organization and identity formation but also in turn be shaped by it. By carefully answering the question ‘boundary objects for whom?’, science and technology scholars can analyze the ways in which rare forms of abnormality – genetic disorders in our cases – are enlisted

! 213! ! in contrasting networks of research, capital accumulation and social mobilization. In so doing, we can better understand both how knowledge about the many is leveraged out of rare forms of abnormality in contemporary biomedical research and how knowledge about genomic difference is produced and mobilized.

Conclusion

In this chapter we have seen three very different paths from genomically designated conditions to broader understandings of the relationship between genes and certain forms of human difference. The delineation of Brunner Syndrome led to the identification of a gene variant that has received considerable popular attention for its supposed correlation with aggression, violence and antisocial behavior and it has given rise to a direct-to- consumer genetic test for a ‘warrior gene’. The designation of Williams Syndrome to the

7q11.23 microdeletion has served as a platform for genetics research and popular speculation about the relationship between the genome and the human capacities for language, sociability and even musicality, bringing unprecedented attention to the condition and its foundation. Finally, Fragile X Syndrome was carved out of the broader category of X-linked mental retardation and went on to serve as a model for autism genetics research, while the National Fragile X Syndrome Foundation worked alongside autism advocates to make their condition a major object of biomedical research, pharmaceutical investment, clinical care and advocacy. In short, we have seen how genomic abnormality is leveraged to produce knowledge about genetics and human difference whose generalizability goes far beyond the rare disorders from which it originates.

! 214! !

We have also seen how LGA can take very different forms, and how it operates within networks of biomedical research and social mobilization that are both historically specific and variable between cases. Unlike the ‘monsters’ described by Canghuilem and

Jacob, genomically designated conditions and the mutations by which they are delineated can serve as boundary objects not only for biological experts, but also for a panoply of actors and organizations ranging from patient advocates, genetic counselors and healthcare providers to state agencies, charitable foundations and highly capitalized pharmaceutical and biotechnology concerns. What’s more, the abnormal and their families are increasingly organized into communities that, on the one hand, facilitate the recruitment of subjects while seeking to direct the aims and outcomes of biomedical research on the other.

Alongside the medicalization of biological research discussed above, LGA directed towards questions of health and illness therefore has perhaps the most potential to lead to the formation of powerful alliances.

In sum, actors with divergent interests and frameworks for understanding genetic disorders must therefore pursue LGA along dual tracks in order to create robust programs of biomedical research and social action. The divergent ways in which they forge those heterogeneous networks and their varying levels of success profoundly shape the biomedical and social meaning of genetic mutations. In other words, ‘biosocial’ group formation may indeed be initiated by knowledge about our genomes, as Rabinow (1992) famously suggested, but it can also in turn work to direct the way that knowledge is interpreted, acted upon and indeed developed in further programs of research. This is most certainly the case when it comes to genomically designated conditions. We have seen how even the same mutations can take on very different meanings and give rise to very different

! 215! ! forms of practice depending on extant historical conditions and the networks of actors in which they are enrolled. However, as we will also see over the next two chapters, the model pioneered by the Fragile X Syndrome network has proven to be particularly powerful, and actors organizing networks around other genomically designated conditions are increasingly emulating it. Stated in terms of reiterated facticity, we will see how this kind of network formation and collective action has emerged, under current conditions, as the most powerful and widespread way of turning genetic mutations associated with developmental difference into the kind of facts that can really matter across multiple fields of practice. Nevertheless, as the larger project of finding meaning in the human genome continues its inexorable, if unexpectedly halting march forward, we should take stock of the varied ways in which genomic abnormality can be leveraged to speak to far-reaching questions about human difference.

! 216!

Chapter 5 - The trading zone of autism genetics Looping and the intersection of genomic and psychiatric classification

What is more powerful than this common independent history of the science and the families is the fact that neither expected to be brought together by the worldwide efforts to understand and treat autism…Our paths cross today and tomorrow as we the stakeholders – families, basic scientists, clinicians and clinical researchers, trainees and representatives from the autism community – come together with a common goal: to share, learn and plot a course to treatment… The real goal and the reason for this meeting is translation…

- Sue Lomas, President of the Phelan-McDermid Syndrome Foundation, New York Academy of Medicine, March 2011

This chapter takes up the fecund intersection of salient, common disorders like autism and genomically designated conditions like Fragile X Syndrome seen in the previous chapter, extending the story both backwards and forwards. Looking backwards to the conditions for their intersection, I aim to show how social processes played a key role in creating the population-level overlap between ASDs and genetic disorders. Looking forward, I set up a framework for understanding the outcomes of those intersections for research, treatment and advocacy and present evidence regarding their effects thus far.

As with the previous chapter, our starting point is the two-fold complexity of human genetics: most medical conditions are genetically complex, and most genomic anomalies are characterized by phenotypic variability. In response to this unforeseen complexity, a debate has ensued about the proper role of genetics in medical and especially psychiatric classification. Some argue that medical and psychiatric systems of

217

classification, especially with respect to developmental disorders, should be radically revised to be correspond to ‘objective’ molecular findings rather than subjectively defined symptoms (Feinstein 2009; Ledbetter 2009; Loscalzo, Kohane, and Barabasi 2007; Miller et al. 2010; see Insel 2013; Belluck and Carey 2013; Conclusion, on a recent move by

NIMH in this direction), while others have called for more effort directed at “translational research” that would bridge the divide between genetic and behavioral systems of classification (Cicchetti and Toth 2006; Meyer-Lindenberg and Weinberger 2006; Tecott

2001; Woolf 2008). In other words, there are those who would see genomic designation feature much more powerfully in contemporary medical classification and those who see in genetics the promise of improved treatment, and perhaps even cure, for more longstanding clinical disease categories. The different sides in this debate depict it as likely to be resolved by the progress of scientific research and by pragmatic considerations (for example, which approach is better suited to facilitate treatment, surveillance and early detection). Here, I examine the social conditions for, and processes at play in, postgenomic psychiatric classification.

As the main thrust of this dissertation demonstrates, genetic abnormalities can serve as the basis for new categories of human difference and disease despite their wide variation in symptomatology. What’s more, we saw in the last chapter how those genomically designated conditions can be taken up as valuable biological models for the traits and diseases with which they overlap, and that this intersection had proven especially fecund in the case of autism spectrum disorders and Fragile X Syndrome. This chapter shows how social processes made the autism-Fragile X alliance biologically possible and how that alliance has gone on to serve as a model for others. In order to make the former

218

point I will follow the lead of the biomedical experts I study and try to leverage genomically designated conditions as strategic research sites to get at topics of broader significance. I do not just mean the inevitable discussion of what genomic designation can tell us about social theory, the sociology of science and medicine, the implications of emerging genomic technologies – there’s plenty of space for that elsewhere. Rather, what I want to do in this chapter is to say something very specific about the relationship between biological etiology and the categories of human difference that we take them to be causes of.

The point of departure for this chapter is three familiar claims: first, autistic spectrum disorders, or ASDs, are now diagnosable in something like 1 in 100 people (Baio

2012; Kogan et al. 2009); second, autism is highly heritable, indeed: “Autism has the highest estimated heritability (>90%) among behaviorally defined neuropsychiatric disorders.” (Brkanac, Raskind, and King 2008); third, ASD is highly genetically heterogeneous, having already been associated with dozens upon dozens of mutations in the human genome (see Betancur 2011; below). While these three points are commonplace, we have yet to fully grapple with the relationship between them. As we will see, social scientists have made great strides in showing how diagnostic expansion, or more generally what Ian Hacking calls ‘looping’, accounts for much if not most of the rise in ASD prevalence from 1 in 10,000 in 1970 to around 1 in 100 today (Eyal et al. 2010;

King and Bearman 2009; Silverman 2011). But what is the relationship between

‘geneticization’, looping and autism genetics, and can sociology help us understand the genetic makeup of a population? I answer these questions in three parts.

219

First, I present a historical overview of the role that geneticization – put simply, the ascription of a genetic explanation – played in autism’s diagnostic expansion. I summarize the existing work that deals with the discursive role of geneticization in the destigmatization of autism and the enabling of the contemporary field of autism advocacy.

However, I also show how efforts to find a genetic basis for autism provided evidence in support of the idea that autism was too narrowly delineated, and pointed researchers towards a broader autistic phenotype. Whether it was heritability studies or the psychiatric evaluations of cohorts of children with genetic disorders, the search for a genetic basis of autism pointed towards more inclusive diagnostic criteria.

Second, I examine the history of genomic etiology – i.e. characteristics of the human genome, however they are observed, that are taken to be causative – for the psychiatric diagnosis of autism. Of course, it is impossible to observe the genetic heterogeneity of autism over time by looking at genetic studies of autism patients: genetic testing techniques have changed too much and, with few mutations accounting for more than 1% of caseloads, even the largest research cohorts are too small to gain any statistical power on their incidence in the autism population. But what if we go in the other direction

– from research cohorts selected on the basis of specific genetic mutations, i.e. genomically designated conditions, to rates of autism diagnosis? That way, we can effectively control for the genetic etiology and see changing rates of what it is an etiology for. If we see that autism rates have consistently risen at a much higher rate than for the population as a whole, then we can show how autism has systematically absorbed new genetic mutations (or at least higher rates of their bearers) into its ranks. In so doing, I show how changes in classificatory or diagnostic practice can transform the genetic

220

makeup of a population. This is not, however, the random accrual of variance – after all, these genetic anomalies are considered to be the causes of autism. In the context of the dissertation as a whole, this analysis will helps us see how changing diagnostic practices constitute conditions of possibility for genomic designation to gain the kind of traction it has in recent years – after all, it requires a particular conception of ‘autism’ to make the

Fragile X model of activism introduced in Chapter 4 practicable. In short, geneticization had unintended consequences, helping to make autism loop into a category that now has literally hundreds of genomic associations and therefore transforming the way we understand both autism and the genetic mutations that have come to be associated with it.

Finally, I examine in detail the case of one genomically designated syndrome being brought into dialogue with the psychiatric category of autism. Specifically, I look at the field of autism research and advocacy at its interface with the vastly smaller biosocial community concerned with the rare genetic disorder Phelan-McDermid or 22q13 Deletion

Syndrome.

Throughout, by treating diagnostic classifications not just as codifications of scientific knowledge, but as categories with a threefold social character as 1) coordinating devices; 2) identities; 3) sites of looping processes, my analysis demonstrates the extent to which social processes can shape the interchange between genetic research and medical diagnosis. I show how each side has much to gain from their alliance, but that in order to do so they need to establish what Peter Galison (1997) described as a ‘trading zone’ – a framework for communication and cooperation despite divergent expert discourses, understandings of the objects of knowledge in question and goals for the collaboration.

Crucially, I extend the idea of the trading zone in order to include a range of actors –

221

advocates, commercial concerns, regulators and so on – who are not fully-fledged members of a scientific discipline.

The three-fold social role of diagnostic categories

The first point is that diagnostic classifications derive their force, not from their fidelity to reality or their clarity per se, but from the extent to which they facilitate collaboration between diverse stakeholders even when the latter attribute radically different meanings to the same classification (Bowker and Star 2000). Specifically, I will suggest that autism genetics constitutes a complex interface or “trading zone” between biomedical research, clinical practice, surveillance medicine, patients’ and parents’ advocacy groups and other actors. Galison (1997, 483) uses the concept of the trading zone to express the insight that “two groups can agree on rules of exchange even if they ascribe utterly different significance to the objects being exchanged; they may even disagree on the meaning of the exchange process itself. Nonetheless, the trading partners can hammer out a local coordination, despite vast global differences.” The trading zone metaphor is particularly apt because it highlights the fact that geneticists, psychiatrists and advocacy organizations approach autism from different perspectives, and with different commitments and understandings, and yet the diagnosis has been able to facilitate complex and mutually beneficial exchanges between them such that it is unlikely to be abandoned in favor of a genome-based classification scheme.

Galison’s concept of “trading zone”, however, has largely been limited to the analysis of scientific collaboration (see Collins, Evans, and Gorman 2007 for a review).

Adapting it to examine the cases of autism on the one hand and genetic disorders on the

222

other requires that we include not only various kinds of researchers, but also diverse professionals, advocates, indeed even patients and parents who are particularly active and influential in organizing and funding research. The second point, therefore, is that diagnostic classifications are also identities. They stigmatize people or provide them with a way of making sense of who they are, but either way they also serve as focal points around which individuals can meet, recognize each other as similar, and organize. In the case of autism, as I discuss below, geneticization was sought by advocates because it served to destigmatize parents. When strong evidence of a genetic link emerged from twin studies and populations with genetic disorders, parents’ organizations were quick to embrace it and have since maintained a formidable vested interest in the research program into the genetic bases of autism, despite the high degree of genetic heterogeneity and complexity

(Silverman 2011).

Finally, the dynamics of psychiatric classification are shaped by the fact that the kinds classified are interactive. In keeping with the rest of this dissertation I therefore draw on Ian Hacking’s framework of ‘dynamic nominalism’ (1998, 2007), which suggests that the classifications delineated by experts are ‘moving targets’ that interact or ‘loop’ with the people who are so diagnosed in a way that dynamically transforms the classifications, expert practices and the people themselves. However, in this chapter I argue that this research program should be extended beyond the dynamics of identification and labelling to include the way classifications interact with the genetic makeup of the populations who are so classified. Again, I argue the geneticization of autism has likely contributed to its increasing genetic heterogeneity by de-stigmatizing the condition, increasing its diagnostic scope and thereby causing it to overlap with previously unrelated genetic mutations.

223

As we will see below, 22q13 Deletion Syndrome or Phelan-McDermid Syndrome

(PMS) was initially unconnected to the fields of autism research, treatment and advocacy.

In recent years, however, PMS and its researchers, advocates and patients, as well as the

PMS Foundation, have become strongly attached to it. Autistic symptoms are increasingly identified as prevalent among individuals with the deletion and, as a result, resources and attention flow to PMS because it is understood to provide a possible genetic model for autism. PMS provides yet-further legitimacy to the idea that autism is a genetic disorder and therefore the strategy of geneticization adopted by autism activists and researchers.

Yet PMS is far from “swallowed” by autism as just one more susceptibility locus, and nor is the reality of the autism spectrum challenged by the genetic specificity of PMS. Rather, I will show how the two are able to coexist peaceably and profitably in a common trading zone that is loosely coordinated by the idea of a “final common pathway”. This constitutes strong evidence that any radical revision of psychiatric diagnostics on a genomic basis – seemingly epitomized by PMS – will be mediated by the threefold social dynamic described above, making it highly unlikely that autism or similar psychiatric classifications will be abandoned as categories of research, care and identity formation.

Geneticization, genetic evidence and looping

For the first four decades after Leo Kanner first described it in 1943, autism was considered an extremely rare disorder characterized by ‘profound aloneness’, repetitive behaviors, and “islets of ability.” (Frith 2003; Kanner 1943) Moreover, while Kanner

224

himself prevaricated on this matter1, a prevailing psychodynamic interpretation of autism depicted it as caused by cold parenting, especially on the part of overly-intellectual

“refrigerator mothers”. (Kanner 1949; Bettelheim 1967; Silverman 2011) How is it then that the estimated prevalence of autism has increased precipitously over the last three decades to reach the level of around 1 in 100? Amidst considerable debate, social scientists have made important contributions to this puzzle by focusing largely on the complex looping processes, and especially the role of parent advocates, that contributed to diagnostic expansion, awareness and the incentivization of ASD diagnoses. (Eyal et al.

2010; Frith 2003; King and Bearman 2009; Silverman 2011).

The main point I want to make here is that the ‘geneticization’ (Hedgecoe 2001;

Lippman 1991) of autism has played a crucial role in this looping process, enabling the kind of autism advocacy and lay-expert collaboration that is so characteristic of the field today (Bumiller 2009; Silverman 2011). On the one hand, the stigma of “refrigerator mothers” no doubt contributed to the rarity of autism by making the diagnosis unpalatable to parents. On the other hand, among the small number of parents of autistic children, the stigma and the struggle against it entailed the formation of a strong identity as “autism parent”. It was indeed the father of an autistic boy – Bernard Rimland - who wrote the first work arguing that autism was a genetic disorder (Rimland 1964) and his book had a huge

1 Indeed Kanner’s foundational 1943 report concludes by noting the cold relationships and tendencies towards abstraction among the families of the children in question, before noting that the fact that the “aloneness from the beginning of life” means “We must assume, then, that these children have come into the world with an innate inability to form the usual, biologically provided affective contact with people, just as other children come into the world with innate physical or mental handicaps.” He therefore concludes, in the final sentence, with a qualified statement of the paper’s title: “For here we seem to have pure-culture examples of inborn autistic disturbances of affective contact.” (Kanner 1943, p. 250; emphasis in original)

225

impact in setting the research agenda for the decades that followed. Rimland wrote the book, in no small part, in order to dispatch the psychogenic, stigmatizing explanation of autism, and instead mobilized a great deal of evidence to argue that autism was a genetic neurological disorder (Eyal et al. 2010; Silverman 2011). Thus, the geneticization of autism was invested, from the very start, with the interest of autism parents in destigmatization.

However, geneticization also initiated a spiral of looping with unforeseen consequences. In his book, Rimland (1964: 52, 59–60) insisted that autism was rare because he thought rarity constituted evidence that it was genetically determined. If cold parenting caused autism, he reasoned, the diagnosis ought to be much more widely spread, and should come in gradations, as in a spectrum. But autism was rare, and “there is an absence of gradations of infantile autism which would create ‘blends’ from normal to severely afflicted.” (ibid: 52) Yet, by de-stigmatizing autism Rimland helped to set in motion processes that essentially guaranteed that it would become less rare and more graduated. Again, because they function as identities for the affected and their advocates, psychiatric classifications are interactive – they ‘loop’, as Hacking would put it, in such a way that actually changes the population in question. The year after Rimland’s book was published, he founded the National Society for Autistic Children (NSAC). As the society became more assertive and formed ties to therapists and researchers, the numbers of parents of children with autism affiliated with it began to increase (Eyal et al 2010). The following year, indeed, the first large scale survey to estimate autism prevalence was conducted in the UK, and put it at 4/10,000 (Lotter 1966). In the decades since, the diagnostic criteria for autism have been successively revised such that its prevalence has

226

grown by orders of magnitude in recent decades, reaching the ‘epidemic’ rate of around 1 in 100 noted above. The parents of autistic children have been key drivers of this shift in diagnostic practice, raising awareness, resources and decisively intervening in the production of knowledge about autism in ways that have impacted its desirability as a diagnosis (Eyal et al. 2010; Frith 2003; Silverman 2011). Again, at the very origins of this kind of activism lay the task of de-stigmatizing autism parenting by emphasizing that the condition was caused by faulty genes, not faulty parenting.

However, I want to suggest that it was not just the idea of autism as a genetic condition that contributed to diagnostic expansion, but also the results of biomedical research that sought an evidentiary basis for the claim that autism was a genetic condition.

As one of the leading figures in autism research and especially autism genetics, Michael

Rutter, put it in a 2000 review, genetic findings may have ‘seven main benefits’ for autism families: “First, the genetic findings have already had an effect on the prevailing concepts of the nature of autism. We now focus on genetically influenced neurodevelopmental deficits rather than on maladaptive patterns of upbringing.” (Rutter 2000:11) However, the evidence from genetic studies did more than just bolster the move to geneticize autism: it also pointed towards a broader a broader delineation of ‘autism’ as a disease category.

Heritability studies

A decade after Rimland’s book came Folstein and Rutter’s (1977a, 1977b) seminal study comparing monozygotic and dizygotic twin concordance for autism to estimate heritability. Folstein and Rutter found strong evidence of autism’s heritability, received great attention aided by a summary version being published in Nature (1977a), and are still

227

widely cited in support of the now well established biomedical fact that autism’s heritability is in the region of 80-90%. Their findings were seen to vindicate Rimland and no doubt increased the numbers of parents willing to tie the fortunes of their children to the autism diagnosis and movement.

There is a twist to this story, however, since Folstein and Rutter (1977a) found the heritability of classical, Kanner-type autism to be much lower than the heritability of a broader phenotype inclusive of various related “cognitive abnormalities” (p. 308). While the study is taken to be the foundational reference for the 80-90% claim, what Folstein and

Rutter actually found was a monozygotic concordance rate of 36%. What is going on here?

It was a secondary finding that put the MZ concordance of a ‘broader phenotype’ at 82%, and therefore heritability at 80% (DZ concordance was 10%)2 – a finding that was replicated in seminal studies in the 1990s (e.g. Bailey et al. 1995) and onwards for autism simpliciter. As Rutter put it in the 2000 review cited above:

"The replicated evidence from both twin and family studies undertaken in the 1970s and 1980s indicated both strong genetic influences and the likelihood that they applied to a phenotype that was much broader than the traditional diagnostic category of autism… concordance within MZ pairs included a range of cognitive and social deficits and not just the seriously handicapping condition of autism itself. This implied the genetic liability extended beyond “autism proper”. It also raised questions about the diagnostic boundaries of autism and led to an appreciation of the need to consider the likelihood of a broader phenotype of autism, or of lesser variants of the same condition.” (pp.3-4; my emphasis)

That is, they interpreted their results as indicating the need for revising and broadening autism’s diagnostic criteria (see also Rutter 2000; Folstein 1996).

Indeed, Rutter and his allies were key actors in revising autism’s diagnostic criteria from DSM-III onwards (Bishop 2008). Rutter, who was in close contact with the British

2 That is, heritability of 80%, ((.9-.18)/.9).!

228

Parents’ society (Feinstein 2010), followed up by publishing his own version of wider, spectrum-type diagnostic criteria (Rutter 1978). NSAC did the same (Ritvo and Freeman

1977), and two similarly wide and spectrum-like versions were published around the same time by Gould and Wing (1979) and Schopler et al (1980), both in close coordination with either the British or the American parents’ associations, and both similarly influenced by the evidence about genetic heritability of a wider phenotype. All four of the main actors –

Ritvo, Rutter, Wing and Schopler – sat on the DSM-III-R committee rewriting autism diagnostic criteria which essentially adopted Wing’s version. (Waterhouse et al 1992)

Thus, the movement begun with Rimland’s geneticization – which relied on autism being rare and “without gradations” – looped in accordance with genetic evidence to render autism a broad spectrum. In so doing, it may have actually made it a more heritable condition.

Autism and genetic disorders

The second pillar of support for the idea that autism is a genetic disorder came from rates of autism in people with known genetic abnormalities. The foundational case was

Fragile X Syndrome, discussed in the previous chapter – a condition caused by a mutation in the FMR1 gene on the X chromosome and characterized by varying degrees of developmental delay, macroorchidism and mild facial dysmorphism. It began to be associated with autism in case studies in the early 1980s ((Bishop 2008; Brown et al. 1982, e.g. 1982; Gillberg 1983; Watson et al. 1984). As we saw in Chapter 4, the first studies, however, to claim that there was a high autism rate in the Fragile X population were the product of collaboration between Randi Hagerman, who two years earlier founded the first

229

Fragile X advocacy organization, and the aforementioned autism parent/researcher/activist

Bernard Rimland (Hagerman et al. 1986; Levitas et al. 1983). In 1986 a study by Randi

Hagerman and colleagues, including Rimland, examining fifty males with Fragile X for autism and autistic traits was published in The American Journal of Medical Genetics.

They found (Hagerman et al. 1986: abstract):

Sixteen percent of patients fulfilled all of the DSM III criteria for Infantile Autism and an additional 30% fulfilled criteria for Infantile Autism Residual State. Thirty-one percent of patients had autism using the ABC checklist but none of the patients fit the classical Kanner syndrome as described by the E2 questionnaire. Some autistic traits were seen in almost all of the 50 fra(X) patients, including eye avoidance in 90%, handflapping, handbiting or handstereotypies in 88%, and language delays with language peculiarities, usually echolalic speech, in 96%."

At the time, these results were hotly contested by researchers who rejected comorbid autism/mental retardation diagnoses or the notion that the connection held any significance in the absence of significantly higher rates of autism in FXS probands than in IQ-matched controls (e.g. Einfeld, Molony, and Hall 1989; see Cohen et al. 1991 for a review of this debate). Today, it is quite widely accepted that FXS is “the most common known single gene cause of autism” while, conversely, “Autism is a common problem in those with FXS and occurs in approximately 30% of males with FXS. An additional 30% of males who did not meet criteria for autism are diagnosed with ASD.” (McLennan et al. 2011:216; 220)

As with the heritability research inaugurated by Folstein and Rutter, the case of

Fragile X Syndrome therefore suggested that broadening diagnostic criteria brought the category of ‘autism’ into a closer alignment with the genetic evidence. Both heritability studies and research on ASD in genetic disorders therefore pointed towards a broader autism phenotype. In both cases, lead researchers had close ties to the same parent- advocates that were working to transform autism as both an identity and a field of research.

230

In short, geneticization was crucial to the processes of diagnostic expansion that led to autism being diagnosed in one in every hundred American children: it didn’t just change how we understand ASD etiology; it helped change the category of autism itself.

Genetic heterogeneity

There is a final chapter in the looping dynamic of autism genetics: while Folstein and Rutter (1977b, 309) were agnostic about the mode of inheritance – apart from excluding simple Mendelian inheritance – they still hoped to show that autism is determined by a small number of genes (Bailey et al. 1995: 73). However, the evidence has consistently pointed in the other direction, namely that autism is extremely genetically heterogeneous. A recent review of the genetic abnormalities associated with autism

(Betancur 2011) noted that no less than 103 gene mutations and 44 chromosomal anomalies have been associated with autism spectrum disorders (see Tables 1 and 2 below). Furthermore, in many of them ASDs are only observed in a small proportion of cases, while full-blown autism is never diagnosable in 100% of cases. Cumulatively, these observed genetic associations account for only 15-20% of autism caseloads. Autism genetics, it turns out, cannot uncover a small handful of genes that cause autism. Instead, it reveals an ever-growing laundry list of genetic abnormalities that overlap with autism to varying degrees.

I want to argue that there were two sources of increased heterogeneity, both of them traceable to a process of looping that has Rimland’s initial claim that autism is a genetic condition as well as Folstein and Rutter’s findings and evidence from disorders like

Fragile X and so on as dynamic elements.

231

232

The first source of increased observed heterogeneity is simply the heightened scrutiny of this rapidly growing population by geneticists, who are of course also equipped with increasingly sensitive techniques for observing genomic abnormalities that in recent years do not have to be targeted according to clinical suspicion (Ledbetter 2008). As they identified dozens upon dozens of new genetic abnormalities and associated them with autism, the observed genetic heterogeneity of autism increased. Yet, what fuelled this increased scrutiny was not simply scientific interest, but the investment of autism parents and activists in geneticization. Autism organizations raised funds, established research centers and lobbied congress, the NIH and other bodies to direct ever greater sums towards autism genetics research. Between 1997 and 2006 NIH funding for autism increased from

22 to 108 million dollars and the 2006 Combating Autism Act mandated an increase to 210 million by 2011, with basic research representing the majority of projects and genetics the fastest growing field (Singh et al. 2009). Furthermore, parent-led advocacy organizations also provide significant sources of funding for research into the genetics of autism, and they were also instrumental in the creation of the most widely used bio- and databank for studying autism genetics (Tabor and Lappé 2011). It is no wonder then that autism genetics has witnessed secular growth as a field, and hardly surprising that a genetically complex condition like autism would accumulate associated mutations along the way.

However, there is also a second, complementary possibility, viz. that the looping processes that led to autism’s enormous growth in prevalence actually increased not only its observed genetic heterogeneity, but actually changed the population in such a way so as to increase the number of genetic anomalies that have significant associations with autism.

233

In other words, ‘geneticization’, both as an ideological shift in the way people think about autism causation and in the biomedical search for a genetic basis for autism, looped back to contribute to the genetic heterogeneity of the autism population. The rest of this chapter attempts to prove this claim and outline its significance for the way we think about social processes and genetics.

Heritability and heterogeneity

How can we make sociological sense of the idea that autism has become at once more heritable and more genetically heterogeneous? Other sociologists have discussed a similar trend: in their recent paper, Liu, Zerubavel and Bearman (2010) show that autism’s heritability is both on the rise and much lower than previously estimated. The latter point mostly concerns the problems of conducting twin concordance studies on clinically referred, small-N samples – a methodological issue that I cannot speak to, other than to say that the point applies to heritability estimates more generally and does not affect autism’s status as one of the most heritable common conditions in contemporary psychiatry. The former point – that their large dataset of California autism caseloads indicates an increase in autism’s heritability – is what concerns us here. Liu et al. ascribe this extraordinary population change to an increase in de novo genetic mutations that can cause autism in the

ASD population, which predicts their observation that concordance for same sex twins has increased as concordance for opposite sex twins has decreased.3 To foreshadow, I think this explanation is mostly correct. Where I depart from Liu et al., however, is in their

3 Because 0% of opposite sex twins are monozygotic, compared to just over half of same sex twins, population level data about births and sex can be used to generate reliable estimates of the number of monozygotic vs. dizygotic twins.

234

explanation of their explanation, viz. that the increase in average parental age is primarily responsible for this increase in the rate of de novo mutations and therefore in same-sex autism concordance. Liu et al. specifically rule out changes in diagnostic practice as a mechanism for this trend, noting (2010: 340): “The temporal concordance trend reported in this article is not predicted by a diagnostic expansion theory.”

A complementary explanation of autism’s increased heterogeneity does include diagnostic expansion. As I have argued in this chapter, geneticization had looping effects for autism: the genetic evidence both played a key discursive role in the social processes underpinning diagnostic expansion, while also channelling that diagnostic expansion in such a way so as to increase autism’s heritability by including the more heritable, broader phenotype first identified by Folstein and Rutter (above) and the genetic disorders like

Fragile X that attracted such interest from key expert/advocates like Bernard Rimland (see

Chapter 4). In other words, diagnostic expansion followed the genetic evidence. What if, rather than there just being higher rates of de novo mutations in the population as a whole, there are also a greater variety of mutations4 in the autism population? In other words, what if being the bearer of certain genetic mutations confers a much higher risk of an ASD diagnosis now than ten, twenty or thirty years ago because ‘autism’ has changed? Were that the case, we would have an explanation of autism’s rising heritability and enormous genetic heterogeneity that is based, in part, on the very processes of looping towards which geneticization and genetic evidence made key contributions.

4 Most of these will in fact be de novo, but for my argument the minority that are inherited from a parent are equally pertinent because they will be similarly monozygotic-concordant and therefore have the same impact on heritability.

235

Looping and the human genome

So again, I want to explore the idea that autism’s dizzying genetic heterogeneity is in part an artifact of diagnostic expansion, i.e. the looping process of which we saw that geneticization actually played a key role. Can we possibly prove this hypothesis? Can we hope to show that the process of diagnostic shift or looping just outlined did in fact transform the genetic makeup of the autism population? Studying findings from cohorts of autism patients won’t do the trick: for one thing, changes in genetic testing techniques make longitudinal comparisons impossible and, for another, most mutations are so rare that we would need studies of autism cohorts of many thousands of people going back decades.

But what if we approach it from the other direction and take advantage of a strategic research site: cohorts of people with longstanding, genetically specific disorders? By analyzing studies of people with genomically designated conditions over time, we are able to determine whether the diagnostic category of autism has come to overlap with previously unrelated genomic anomalies. In other words, I propose to follow the lead of the biomedical researchers discussed in Chapter 4 by leveraging genomically designated conditions as a strategic research site to say something more general about social processes and the genetic makeup of populations.

Take Hagerman et al.’s (1986) seminal study, discussed above, of autism rates in a

Fragile X Syndrome cohort and their finding that almost all had ‘autistic traits’, 31% met the ABC checklist criteria for autism, 16% met the DSM-III criteria and none fit the

‘classical Kanner syndrome’. By 2009 a study led by Hagerman put the rate at ~30% according to DSM-IV (Hagerman et al. 2009) while a 2011 review of FXS for which she is

236

the last author adds that an additional 30% meet the criteria for ASD (McLennan et al.

2011). Similarly, just this January a paper in the American Journal on Intellectual and

Developmental Disabilities evaluating 182 FXS cases put ASD and autism rates at 83.6% and 48.6% respectively according to the Social Communication Questionnaire (Moss et al.

2013). What is going on here? Clearly, the diagnostic expansion of autism, inspired in part by evidence from heritability studies and cohorts of people with genetic disorders, has only greatly increased the proportion of Fragile X probands diagnosable with autism. But that is not all. I will argue that this same process of geneticization and diagnostic expansion has also greatly increased the number of genetic mutations that are strongly associated with autism. In the case of genomically designated conditions especially, observing ASD rates rising much faster than in the population as a whole therefore allows us to argue that diagnostic expansion has greatly expanded autism’s genetic heterogeneity while therefore increasing its heritability. In this chapter I present some preliminary results from a larger project that will include a quantitative analysis of ASD rates over time in all of the genetic mutations listed by Betancur in Tables 1 and 2 above. First, I present descriptive statistics on the scale of autism and Fragile X work in the chromosomal disorders listed by

Betancur; second, I present a preliminary snapshot of the histories of ten of those conditions with longstanding biomedical literatures and their association with autism over time; finally, I present qualitative results from three illustrative cases: Williams Syndrome,

Phelan-McDermid Syndrome and XYY Syndrome.

237

Impact on the field

In my continuing research on autism and genetic disorders I hope to be able to present a more systematic analysis of changing ASD rates over time. But for now, what has been the impact on the field, and was Fragile X really the pioneer I am claiming it was?

Consider the following charts, which report keyword and title-word rates in a population of papers generated by a Web of Science search string5 that seeks to capture all the papers with title-terms related to the chromosomal disorders listed by Betancur (See

Table 2, above) as associated with autism. It should be noted that this string captures many papers – e.g. gene-mapping at 15q24 rather than studies of the 15q24 deletion – that we would not expect to have any relation to the genetic disorders associated with ASDs, and therefore rates should be read in relative terms and as very conservative estimates of autism’s status in the pertinent fields. Nevertheless, they provide some perspective on the current state of affairs.

Figure 1 (below) reports the total number of papers captured by the search string

(bars, right-hand y axis) since 1991 and rates therein of keywords related to intellectual disability (ID)/mental retardation/developmental delay on the one hand and autism on the

5 TI=(“FRAGILE X” OR XQ27* OR1p36* OR 1q21* OR 2p15–p16* OR 2q23* OR 2q338 OR 2q32q33* OR 2q37* OR 3q29* OR “Wolf–Hirschhorn” 4p16.3* OR 4q21* OR “Cri du Chat” OR 5p-* OR “5P minus” OR 5q14* OR “Sotos syndrome” OR 5q35* OR 5q35.2q35* OR “Williams syndrome” OR “Williams-Beuren syndrome” 7q11* OR 8p23* OR“9q subtelomeric Deletion” OR “Kleefstra Syndrome” OR 10p14p15* OR “10q22–q23 deletion” OR “Distal 10q deletion” OR 11p15* OR “Beckwith– Wiedemann” OR “Silver–Russell syndrome” OR “WAGR syndrome” OR 11p13* OR “Potocki–Shaffer” OR 11p11* OR “Jacobsen syndrome” OR “11q* deletion” OR “Angelman* syndrome” OR “Prader–Willi” OR “15q11–q13” OR 15q13* OR 15q24* OR 15q26* OR “Rubinstein–Taybi” OR 16p13* OR “16p11.2–p12.2” OR 16p11* OR 17p13* OR “Miller–Dieker” OR “isolated lissencephaly” OR 17p13* OR “Smith–Magenis” OR “Potocki–Lupski” OR 17p11* OR NF1* OR 17q11.2* OR 17q12* OR17q21* OR “Down* syndrome” OR “trisomy 21” OR “velocardiofacial” OR VCFS OR “VELO-CARDIO-FACIAL” OR “DiGeorge Syndrome” OR 22q11* OR Phelan–McDermid*, 22q13* OR Xq28* OR MECP2 OR “Turner* syndrome” “monosomy X” OR “Klinefelter” OR XXY OR XYY OR XXYY OR “45,X/46,XY Mosaicism”)

238

other (grey and black lines respectively). We see that the literature as a whole has more than doubled over the last two decades, while the proportion of the literature that has an

ID-related keyword was .2-.25 in 1991 and 1992, .15-.2 until around 2001, and has since grown to more like 30%. ASD keyword rates, by contrast, fell from an early peak of around .05 to around .025 for most of the 1990s before steadily rising to around 13-14% of the literature in recent years, i.e. a much higher rate of increase than ID.

But what underlies these patterns? Consider Figure 2 (below). Here we see the same chart, but with MR/ID keyword rates removed and Fragile X rates included. We see how the early literature on genetic disorders and ASDs is overwhelmingly driven by the Fragile

X/autism association, but also how the former has become largely independent of the latter in recent years as more and more conditions have been associated with autism.

239

Given the relatively high rate of autism keywords in 1991, it is worth tracing the literature back further. However, given the very different and more limited Web of Science keyword capture prior to 1991, such a task much resort to rates of terms in paper titles in order to remain valid over time. As such, Figure 3 (below) reports the same search string for papers with title terms related to Betancur’s chromosomal disorders associated with autism (bars, right-hand y axis), and the co-occurrence of title terms related to autism and

Fragile X (black and grey lines respectively). Here we see even more clearly that Fragile X research, which emerged as and remains a form of x-linked mental retardation, almost completely drove the early research on autism and genetically specific developmental disorders. Indeed the few exceptions are partial at best: for example, four papers about the contrast between autism and Down syndrome (which is now associated with ASDs) appear in the 1970s prior to the first Fragile X-autism association (see Holroyd and Mcarthur 1976 for the most cited paper on autism and a genetic disorder other than Fragile X prior to

240

1999), with similar papers published in the years that followed. These partial exceptions aside, we see in Figure 3 how comprehensively Fragile X dominated the early field of research on autism and genetic disorders.

These are very much preliminary results, and I present them more to indicate the need for further research than as findings in their own right. More refined search strings and citation analysis that can capture the dynamic of the literature over time will be necessary in order to advance sound conclusions. That said, they do bolster one of the key arguments of this chapter, viz. that Fragile X research cleared the path for other genomically designated conditions to become similarly associated with autism. After all, most of the literature is accounted for by genomically designated syndromes with longstanding histories of research, including associations with other developmental

241

disorders. As our case studies below show, it is only in recent years that researchers have started to see ASDs in these genetic disorders, and now that they have they do so at rates that are in many cases similar to that of Fragile X. In sum, the literature on genomically designated conditions suggests that the observed genetic heterogeneity of autism is partly the result of a particular and intimately related set of historical process involving geneticization, the search for a genetic basis for autism and diagnostic expansion, rather than a timeless biomedical fact that was only recently uncovered. What then, do we see when we look at the association between autism and genomically designated conditions over time?

Snapshot of ten syndromes

Initial research on autism and genomically designated disorders about which there is considerable record of biomedical research is presented below. Table 3 (below) reports the histories of 10 genomically designated syndromes’ association with autism. It includes syndromes that have at least a fifteen-year history in the literature, and some with over fifty years. Thus it excludes conditions like the MECP2 Duplication Syndrome which, discovered in 2004 (Ariani et al. 2004) and associated with mental retardation in 2005

(Van Esch et al. 2005), was first associated with autism in 2009 and now has ASD rates of over 90% according to multiple studies (e.g. Ramocki et al. 2009). It provides, in order, the year the mutation was first reported in the literature, the first report that included a cognitive phenotype, the first association with autism and the mean ASD rates and number of papers reporting ASD rates in cohorts of people with the syndromes in question prior to

1990, 1991-2000 and 2001-2012 respectively.

242

Table 3. Histories of autism association in 10 genomically designated conditions

First First Mean ASD Mean ASD Mean ASD Syndrome Mutation cognitive autism rate, pre-1990 rate, 1991- rate, 2001- name/locus reported phenotype report (n studies) 2000 (n) 2012 (n) Fragile X Syndrome/ 1969 1969 1980 31.8% (5) 45.2% (4) 38.1 % (14) CGG repeat, Xq27.3 22q13 Deletion 1988 1992 2000 NA (0) NA (0) 48.7% (3) Syndrome 22q11.2 Deletion 1981 1992 1998 NA (0) NA (0) 32.1% (7) Syndrome Cri du Chat/ 1963 1963 1997 NA (0) NA (0) 23.6% (2) 5p- Syndrome Williams Syndrome/ 6 1995 1995 2006 NA (0) 14% (2) 23.4% (3) 7q11.2 deletion Smith-Magenis/ 1982 1984 1989 NA (0) NA (0) 78.3% (3) 17p.11.2 deletion Potocki-Lupski/ 1996 1996 2000 NA (0) 14% (1) 76% (2) 17p.11.2 duplication 2q37 Deletion 1989 1989 1999 NA (0) 24% (1) 62.5% (1) Syndrome

XYY Syndrome 1961 1962 1971 NA (0) NA (0) 36% (4)

XXYY Syndrome 1961 1961 1977 NA (0) NA (0) 38.2% (3)

Total studies ------5 9 48

We can see how what I am calling a cognitive phenotype – i.e. some kind of psychological or psychiatric observation or diagnosis, most often mental retardation or developmental delay – tends to be reported in the same paper as the one reporting the mutation or very shortly thereafter. (The one exception in the table, 22q11.2DS, is not

6 In the cases of Williams and 22q11.2 I am only considering cases of molecular diagnosis. Thus there were, as we will see below, other reports of ASD in Williams Syndrome but the first to be molecularly confirmed (as most WS cases were after 1995) was a 2006 case study (Herguner and Motavalli Mukaddes 2006).

243

really an exception: initial reports about the 22q11 deletion were with respect to DiGeorge

Syndrome, which was known to cause mild-to-moderate mental retardation, and most of the genetic studies were conducted on infants, many of whom died from heart malformations.) First autism reports, usually in the form of a case study, demonstrate that finding autism in genomically designated probands was considered a significant finding. In other words, we have strong reason to believe that research subjects with genomically designated conditions were being evaluated for and having existing psychiatric diagnoses noted, and that autism would likely have been reported when it was encountered. What then, do we see when it comes to rates of autism in the literature on genomically designated conditions?

In contrast to the typically very short or non-existent gap between discovering the mutation and associating it with a cognitive phenotype, there is generally a long lag before autism is reported in probands. In Potocki-Lupski Syndrome it only took four years, and

Smith-Magenis Syndrome only seven, but those were among the most recently discovered conditions. For XYY it only took ten years (1961-1971), but as we will see below it was another 13 years before a report of autism in XYY was ascribed anything more than incidental significance. In most cases it took well over a decade, and for 5p- Syndrome it took over forty years. The crucial points, however, are these: 1) with the exception of

Fragile X Syndrome, there were no reports of autism rates in research cohorts prior to

1990; 2) despite a handful of rates reported for other syndromes in the 1991-2000 period, it is only in the last twelve years that ASD diagnoses have been reported in significant proportions of people with genomically designated conditions other than Fragile X. In six cases, we see that syndromes with no reports of ASD rates prior to 2001 have since been

244

reported with average rates of 23.6%-78.3% across 22 studies. In the three other cases

(excluding Fragile X) we go from mean ASD rates of 14%-24% across five studies prior to

2001, to 23.4%-76% across six studies since.

As with the rates of keywords and papers titles, studies of cohorts of people with genomically designated conditions indicate that the field has followed Fragile X in seeing high ASD rates in genomically designated populations. In order to examine this process in further detail, I present detailed analyses of ASD associations in three different genomically designated syndromes – Phelan-McDermid/22q13 Deletion Syndrome,

Williams Syndrome and XYY Syndrome – in order to show the multiple trajectories of diagnostic expansion that have contributed to autism’s genetic heterogeneity.

Phelan-McDermid Syndrome

As we saw in Chapter 1, when someone has a deletion at site q13.3 on the long arm of the twenty-second chromosome, they are diagnosed with 22q13 Deletion Syndrome, now known as Phelan-McDermid Syndrome. While there are no clinical diagnostic criteria, most people with Phelan-McDermid Syndrome have moderate-to-severe intellectual disability and severe language delays, and are also likely to suffer from a subset of a long list of associated symptoms. As with Fragile X, autistic behaviors are common. However, even though mental retardation was considered a core feature of the

22q13 deletion from its discovery in the late 1980s (see Phelan, Rogers, and Stevenson

1988; Phelan et al. 1992; Hinkel et al. 1997; Wong et al. 1997), it was another 12 years before an association with autism was even noted in the literature (Goizet et al. 2000). In

2001, the medical geneticist most associated with Phelan-McDermid Syndrome, Katy

245

Phelan, noted that most cases display significant ‘autistic features’ but specifically ruled out a comorbid diagnosis:

… it is somewhat difficult to make an additional diagnosis of autism for children with severe to profound mental retardation… to be diagnosed with autism there must be qualitative differences in language and socialization when compared to nonautistic children with retardation of similar degree. With a cognitive age equivalent of 9.3 months, these children are expected to show some autistic- like features. All children in this sample appear to have language and socialization skills consistent with their general mental ability. (Phelan et al. 2001: 95)

Phelan made a nearly identical assertion two years later (Phelan 2003:2).

And yet by 2008, Phelan had conditionally abandoned her reluctance to countenance ASD diagnoses in 22q13DS patients, as long as they were cordoned off as

“syndromic autism”:

Behavioral features of Phelan-McDermid syndrome include poor eye contact, stereotypic movements, decreased socialization, and language impairment consistent with autism spectrum disorders… Deletion 22q13 has been shown to be one of the common chromosome defects associated with autism. The term "syndromic autism" has been suggested for autism accompanied by dysmorphic features and the 22q13 deletion syndrome was cited as an example of a genetic disorder characterized by autistic behaviour. (Phelan 2008: 14 our emphasis)

Reported rates of autism in PMS are now consistently in the 40-80% range (see Sarasua et al. 2011), while 22q13.3 deletions are estimated to account for around 1% of all ASD caseloads with mutations in the key gene at 22q13.3, SHANK3, accounting for perhaps another 1% (Abrahams and Geschwind 2008: 344). As with Fragile X, the 22q13 deletion has become a cause of ASDs or a ‘syndromic’ form of autism only because the diagnostic criteria have expanded to readily countenance comorbid diagnoses of intellectual disability and ASD. Today, as we will see below, PMS is following the path of Fragile X Syndrome as a model for autism, as SHANK3 has become what Autism Speaks called “the new ‘it’ gene for autism.”

246

Williams Syndrome

Although originally a clinical diagnosis, we saw in Chapter 4 how Williams

Syndrome is now delineated according to a deletion at site 11.23 on the long arm of the seventh chromosome (Donnai and Karmiloff-Smith 2000; Nickerson et al. 1995).7 It is associated with moderate intellectual disability, cardiac problems and distinctive facial phenotype alongside ‘hypersociality’ and strong relative strengths in language and, arguably, musicality. Indeed sociability and strong language skills led some to describe

Williams Syndrome as the ‘anti-autism’ – a developmental disability with a strikingly different behavioral and cognitive profile from classic autism. As Jones et al. put it in the

Journal of Cognitive Neuroscience, autism and Williams syndrome are “polar opposite groups when it comes to social behaviour” (2000:41). They summarize their findings thus:

… social behavioral contrasts between WMS and autism are striking (Courchesne, Bellugi & Singer, 1995). WMS seek out social interaction and eye contact and, generally, do it in a polite and friendly manner... In contrast, the cardinal feature of autism is a profound deficiency in social knowledge, affective expression, and communication. The autistic child avoids eye contact and is poor at discriminating facial expressions. In an early description of an autistic child, Kanner (1943) wrote, "He paid no attention to persons around him... he completely disregarded the people (in a room) and instantly went for an object... he was happiest when left alone.” (ibid: 44)

Their recourse to Kanner, however, already points towards the opening for finding links between Williams and autism.

7 As we saw in Chapter 2, this entailed both excluding the minority of people who had been clinically diagnosed with Williams Syndrome but lacked the deletion and including others who have the deletion but who would have been unlikely to receive the prior clinical diagnosis. In the typology developed in Chapter 2 it is a case of genomic designation that recalibrates an existing clinical category.

247

Indeed already by 1985, Reiss et al. reported two children with behaviorally atypical cases of Williams Syndrome and autism (Reiss et al. 1985), and in 1994 another paper reported four similarly comorbid cases in a population of 60 children with Williams

Syndrome (Gillberg and Rasmussen 1994). One of those authors, Christopher Gillberg, has authored numerous papers making previously unreported associations between autism and genetic disorders (see Gillberg 1998 for a review). Here’s how he put it in this case (ibid:

382):

Although the slightly larger scale studies of the behavioral phenotype in Williams syndrome have not included specific reference to autism -- and most have found overfriendliness, a feature which, at first glance, would seem to be the exact opposite of autism - it is clear from the results of these preliminary investigations that this patient group does have several problems seen in autism. Included in this catalogue are hyperacusis, social isolation, and other types of social impairment (such as indiscriminantly approaching total strangers), distractability, inflexibility, ritualism, obsessiveness, and pragmatic deficits (in spite of relatively excellent superficial expressive speech and language skills), all of which can be hallmarks of autistic syndrome."

Still, they noted that the two boys with Williams Syndrome and autism did not exhibit the typical behavioral features of Williams Syndrome, while the two such girls only did to a small degree (392). In short, it was only in a small handful of atypical cases of Williams

Syndrome that autism could be diagnosed, though overlapping problems like sensitivity to sound and repetitive behaviors had begun to be identified. Nevertheless, the prevailing view was that of Jones et al.’s 2000 paper: autism and Williams Syndrome were strikingly divergent neurodevelopmental disorders, so much so that, “Future studies examining the neuroanatomical differences between WMS and Autism may reveal clues to aspects of the neural and genetic bases of social behavior." (Jones et al. 2000: 44) In sum, despite a couple of papers reporting atypical cases, studies of behavior in Williams Syndrome did

248

not identify ASDs (e.g. Einfeld, Tonge, and Rees 2001:20); rather, they saw their behavioral opposite.

But while early discussions emphasized the differences between autism and

Williams Syndrome, today the two are increasingly reported as comorbid diagnoses. But how is that possible? The most detailed engagement with this question is that of Laws and

Bishop (2004), who note at the outset that: “Individuals with WS have a distinctive cognitive profile; some aspects of non-verbal functioning are impaired, particularly visuospatial cognition, while language and face-processing are areas of relative strength.”(46) Nevertheless, they are able to conclude:

Despite earlier reports that emphasize a strong social interest and empathy, this study suggests that individuals with Williams syndrome have pragmatic language impairments, poor social relationships and restricted interests. Far from representing the polar opposite of autism, as suggested by some researchers, Williams syndrome would seem to share many of the characteristics of autistic disorder. (Laws & Bishop 2004: 45)

How do they square this circle? The answer, it seems, is that the circle has changed. Laws and Bishop appeal to Baron-Cohen’s famous ‘theory of mind’ hypothesis of autism:

If WS represents the opposite of autism, then people with WS could be expected to succeed on the various tasks that have been developed to test theory of mind. There is some evidence that this is the case (Karmiloff-Smith et al. 1995). However, Tager-Flusberg and Sullivan (2000) argued that there are two components to a theory of mind. The perceptual component depends on immediate available perceptual information, such as facial or vocal expressions. The cognitive component is more important for reasoning about the mental states and actions of others. Tager-Flusberg and Sullivan suggest that these components are dissociated in WS. Their experiments showed that social perception was spared in comparison with other neurodevelopmental disorders, but social cognition was not…

By disaggregating social perception and cognition in Williams Syndrome, with the former an area of strength that creates the appearance of strong social skills that contrast with

249

ASDs as typically understood and the latter a relative weakness, researchers were able to point to behavioral characteristics that do resemble autism:

Tager-Flusberg and Sullivan (2000) made the point that differential performance on tasks that tap dissociable perceptual and cognitive domains could explain the paradox found in WS where, despite the social skills and empathic qualities apparent in experiments, clinical reports suggest they have social difficulties. For example, Davies et al. (1998) reported that 96% of parents and caregivers of adults with WS described problems with establishing friendships, and a large proportion also described other social difficulties, such as disinhibition and social isolation. Davies et al. (1998) relied on parent reports based on interviews that included elements from the VABS and the Autism Diagnostic Interview (Lord et al. 1994). This research also indicated evidence for pragmatic language impairment (PLI), including excessive chatter, the propensity for socially inappropriate statements and questions, and for talking to themselves. Further, Udwin and Yule (1991) noted that the conversation of people with WS was not well tuned to the conversational partner. Stojanovik et al. (2001) also noted a high proportion of inappropriate utterances in the conversational speech samples of four children with WS. (ibid)

So autistic behavior in Williams Syndrome is not about aloneness, but difficulty maintaining relationships; not about a lack of speech but talking too much chatter or the inappropriate choice of words and topics; not about an aversion to others but a failure to read the subtle clues of a social interaction. How has this impacted ASD rates in Williams

Syndrome?

From a handful of isolated case reports where a child with atypical Williams

Syndrome was deemed diagnosable with autism, 2007 saw a study put ASD at 2/20 or

10% (Lincoln et al.) and another put it at 14 of 29 or 48% (Klein-Tasman et al.). A more detailed version of the latter study in 2009 that also added one additional subject found autism in 3 of 30 children with ‘geneticially-confirmed Williams Syndrome’ and ASDs in a full 50% of the 30 cases (Klein-Tasman et al.. p. 3). Thus in Betancur’s summary (ibid:

55), “50% of patients with Williams Syndrome meet the diagnostic criteria for ASD.”

250

XYY Syndrome

Autism’s diagnostic expansion at the ‘high-functioning’ end of the spectrum has similarly expanded the scope for associating genomically designated syndromes and

ASDs. Take the case of XYY syndrome. We saw in Chapter 3 that, despite its infamous beginnings as a biomedical category in the 1960s and 70s when it came to be associated with aggression and anti-social behavior in the guise of a ‘Supermale’ Syndrome, today an extra Y chromosome is associated with moderately increased stature, acne, ADHD, around ten IQ points lower than an unaffected sibling and increasingly, ASDs. But again, this was not always the case. XYY is found in around 1 in 1000 male births, so it is hardly surprising that there were a handful of case reports of autism in people with an XYY karyotype (Abrams and Pergament 1971; Nielsen et al. 1973; Gillberg, Winnergard, and

Wahlström 1984). As the second such study noted in 1973, the association was “most probably coincidental” (Nielsen et al. 1973: p. 22).

It was only the team led by Christopher Gillberg (et al. 1984), featured above with respect to Williams and Fragile X Syndrome, who ascribed a broader significance to his autism/XYY case study when he reported it in 1984. He begins and ends his paper with a discussion of Lorna Wing’s ‘maleness’ hypothesis of autism and the significance of the

Fragile X finding, before noting that his is a case of XYY in a proband “fulfilling Rutter’s

(1978) and DSM-III criteria (1980)” for infantile autism. While acknowledging that only a minority of people with XYY also have autism, Gillberg notes the autistic-like behaviors that are more commonly found in XYY cases. By way of ‘summing up’ the existing research on XYY (358):

251

… it is tempting to suggest that the XYY karyotype may predispose the child to speech-language delay, difficulties in establishing social relationships, and overall immaturity of brain development. All these features (language, social and behavior impairment along with delayed development) are typical of autism, but in autism there is another dimension to the problems, regarding severity and quality. The XYY constitution per se does not cause autism, but rather might predispose the boy to milder disturbances of the kind seen in “the triad of language and social impairment” described by Wing and Gould (1979).

For XYY, as with both Fragile X Syndrome and heritability studies, reference to a broader autistic phenotype like the one advanced by Lorna Wing made it easier to associate autism with a genetic underpinning.

Beginning in 2006, we see a series of XYY cohort studies on autism. Tartaglia et al. (2006), on a paper for which the leading Fragile X/autism expert Randi Hagerman

(above) was last author, found 36% or 8/22 cases diagnosable with an ASD – one with autism and seven with PDD-NOS. A 2011 study found that 11 of 58 cases of XYY

Syndrome or 19% had an ASD, while “communicative profiles indicative of mild autistic features were common” among the remainder. Finally, in 2012 a pair of studies by the same team was published in Pediatrics (Ross et al.) and Research in Developmental

Disabilities (Cordeiro et al.) using the SRS diagnostic to examine autistic behavior in children with XYY Syndrome. They found that half of a sample of 40 already had an ASD diagnosis, and indeed that half were in the severe range on the SRS diagnostic while nine of the remaining twenty were in the mild-to-moderate range.

In sum, we have seen three trajectories of diagnostic expansion that brought genomically designated conditions, which is to say genetic mutations, firmly into the autism population: from mental retardation to a comorbid diagnosis of autism in the case of PMS/22q13 Deletion Syndrome; from hypersociality and remarkable relative strengths

252

in language to autistic difficulty in maintaining relationships and using language appropriately in social interaction in the case of Williams Syndrome; from mild behavioral challenges and slightly depressed IQ scores to high ASD rates in XYY Syndrome. In each case, there can be little doubt that the high ASD rates now seen in these disorders, easily qualifying them for Betancur’s list of “genomic disorders and chromosomal abnormalities reported in individuals with ASD/autistic traits” (ibid: 53 (table title)), would have been unthinkable twenty years ago, never mind to Kanner. The same is most likely true of the other syndromes whose histories of association with autism we saw summarized in Table 3

(above). As we saw in our three case studies, changing diagnostic criteria and especially an acceptance of comorbid ASD/intellectual disability diagnoses in the case of Phelan-

McDermid Syndrome has brought these disorders into the autism population. For many years the very same conditions – again, delineated by the very same mutations – were not significantly associated with autism. That these mutations strongly predispose their bearers to ASDs and autistic behavior is therefore dependent upon an understanding of autism that has been transformed, in part by the drive to ‘geneticize’ it. Finally, we saw how autism has grown from a tiny, marginal object of interest in the study of chromosomal abnormalities, and one that was almost entirely confined to the study of Fragile X

Syndrome, to a major object of interest across dozens of different genetic disorders. What then are the implications of these classificatory intersections for our sociological understanding of genetics, looping and disease classification?

253

The trading zone of autism genetics and the ‘final common pathway’

The immense genetic heterogeneity of ASDs constitutes a problem for researchers, forcing them to abandon the chimerical goal of finding a ‘gene-for’ autism and adopt a variety of strategies in order to move forward. On the one hand, we can treat genetic mutations simply as susceptibility loci for ASDs rather than as independent disorders, and indeed this is often the approach implicitly adopted by autism researchers. However, as we have seen throughout this dissertation there is substantial biomedical and stakeholder investment in genomically designated conditions and they are often organized into communities that substantially aid in the organization of research and provision of human subjects. What’s more, even the ones with the highest ASD rates do not approach 100% and they tend to have numerous other clinical associations besides. In short, genetic disorders cannot simply be subsumed by the immense field of autism research and advocacy.

On the other hand, some have suggested that autism might be more usefully split into a handful of more specific conditions that will become evident once the underlying genetic etiologies are understood. As Christopher Walsh, a professor of pediatrics at

Harvard, head of genetics at Children’s Hospital Boston and chair of the Scientific

Advisory Board for the Autism Consortium, put it: ‘I would like every kid on the spectrum to have not “autism”, but a more specific disorder’ based on specific genetic observations

(cited in Pettus, 2008: 29). From this point of view, the genetic abnormalities constitute the

“real” disorder, while autism is merely a symptomatic diagnosis. Indeed, the Betancur piece cited above is described (Ibid: 43) as a “review [of] the different genetic and

254

genomic disorders in which ASDs have been described as one of the possible manifestations.”

Put differently, these researchers attempt to resolve the problem of heterogeneity by subordinating clinical to genomic terminology (see also Collins, Evans and Gorman

2010, 10-11). The main thrust of developments in the field of autism genetics, however, has not followed this ‘splitting’ strategy. Instead, even the researchers and advocates who are primarily interested and invested in genomically designated categories of illness embrace the idea of autism as a common outcome of multiple genetic disorders. We need not waste too much time unpacking this reluctance to relegate ASDs to the status of mere symptom (though we will see that some do hold that view): autism advocacy organizations wield too much symbolic and material power in this trading zone to be so subordinated.

Not only do they control a great deal of funding for genetics research, but unlike most other fields advocacy organizations control the very material infrastructure for genetics research, namely DNA samples as exemplified by the Autism Genetic Resource Exchange

(AGRE).8

8 AGRE was founded by an alliance of parent-activists and geneticists who challenged the assumption that a biobank for autism tissue samples and phenotype data would have to be centralized, closed to unaffiliated researchers and carefully managed by a group of experts. In contrast to the more conventional, centralized and expert-run Autism Genome Project (AGP) and its limited output of published research, AGRE has assembled a far larger database of samples and phenotypic data, has been used by more researchers and has led to hundreds of publications. It has drawn new geneticists to the study of autism and allowed researchers without the means to collect their own samples to participate in the field. It had also been utilized by commercial genomics companies seeking to bring diagnostic tests for autism to the market (Singh 2010; Tabor and Lappé 2011). Crucially for us, it has served as a model for genomically designated conditions’ advocacy organizations in the establishment of their own biobanks and clinical registries.

255

However, the inability of either classificatory strategy to subordinate the other does not mean that genomically designated syndromes cannot reap huge rewards from their association with ASDs, or that molecularly defined disorders cannot play a role in autism research and, perhaps, treatment. As Rutter put it in the review cited above (2000, p. 11)

[T]here is the potential of molecular genetic findings for leads on biological research that will identify the causal neural processes that underlie the development of autism. These will allow a focus on the specific neurobiological effects of susceptibility genes and on the routes by which such effects predispose to autism. As I have indicated, the identification of causal processes could lead to effective means of prevention or intervention… there is the potential of molecular genetic findings for leads on effective drug treatments.

Thus while Rutter is clear that “the research route is likely to prove long and arduous,”

(ibid) the finding of many genomic anomalies associated with autism rather than one or a small handful did not obviate the potential for genetic disorders to advance research and, potentially, treatment for ASDs. As we saw with Fragile X Syndrome, Rutter’s long and arduous research route has already led to several pharmaceutical products reaching late stage trials, and we will see below how an alliance of autism and 22q13 researchers and activists have made substantial progress in following them down that path. How then, sociologically, can we make sense of the way these qualitatively different forms of classification – genomic and psychiatric – are brought together in collaborative projects of research, advocacy and identity formation, and what are the implications of the two-fold ways of understanding human difference that result?

It is clear that genomic designation need not replace or stand in zero-sum opposition to clinically designated categories. First, the two can exist side by side as categories of clinical practice (see Rabeharisoa and Bourret 2009). Second, a recent review of autism genetics by Abrahams and Geschwind (2008) in Nature Reviews Genetics

256

discusses how they coexist as intersecting and overlapping levels of analysis in biomedical research. The study of autism/ASD susceptibility and genetic variation, the authors tell us

(342), is dominated by “two contrasting but valid and potentially compatible paradigms.”

(my emphasis) On the one hand, Abrahams and Geschwind argue that the independent heritability of ‘core domains’ suggests that research should focus on the ‘common pathway’ that leads from genetic endowment to the presentation of the autistic phenotype.

On the other hand, they note that “the fact that functional disruption of single molecules seems to be sufficient to cause disease suggests that the identification of rare variants is also important" and go on to discuss the intersection of ‘molecularly defined syndromes’

(344, 352) with autism and other conditions, especially mental retardation (352-3). We are told (353) “These concepts reinforce the idea that current clinical notions of boundaries between neuropsychiatric disorders need not be representative of the underlying genetic or biological etiologies,” but the way forward in reconciling these divergences is left open, as it clearly is in practice.

Thus the trading zone remains fractionalized, to use Collins, Evans and Gorman’s

(2010, 11-14) terms, organized around the boundary objects (Star and Greisemer 1989) represented by autism and the specific genetic mutations. It is in fact not necessary for the different stakeholders to agree on what autism is, whether it is really a distinct illness, an amalgam of different disorders, or something else. Nor do they need to agree whether a given genetic abnormality is really a susceptibility locus for autism or a genomically designated disorder. Instead, the reigning interpretation of the genetic heterogeneity of autism is that the multiple genetic abnormalities and the changes in protein production that they cause converge on a shared physiological mechanism or ‘final common pathway’.

257

(Betancur 2011:62; Abrahams and Geschwind 2008). We will see how the idea of a

‘common pathway’ from multiple genomic anomalies to the behavioural manifestation of autistic features that Abrahams and Geschwind suggest researchers focus on, while wholly unproven, can be remarkably successful in aligning interests.

Indeed, the phrase “final common pathway” has a kind of double meaning. The first is the one noted above, namely the convergence of many biological lesions onto a single physiological or behavioural phenotype; the second expresses the implicit coordination device operative in the trading zone: the different stakeholders all work together, despite their different understandings and commitments, because they all share the assumption that in the long-term (“final”) it will be shown that their efforts all converge and contribute to the same outcome (“common pathway”). This device allows parents to work with genetics researchers towards a biological understanding of autism that they hope may bring with it therapeutic and even pharmaceutical benefit while also erecting a bioscientific firewall to potential stigma. For researchers, autism genetics brings with it a wealth of genomic and clinical data, funding and the potential for high-impact findings. Furthermore, autism genetics has attracted researchers from fields as varied as genetic epidemiology, neurogenetics, genomics, bioinformatics and clinical genetics and others for whom both ‘autism’ and ‘genetic’ mean very different things. Throw in the other actors with a stake in autism and its status as a genetic condition – state officials, clinicians, special education professionals and so on – and you begin to see the variety of actors engaged in cooperation and exchange in the trading zone of autism genetics. They have been able to cooperate, coordinate action around objects of knowledge production

258

and to trade and share those objects despite ascribing very different meanings and values to them.

The best evidence for how the trading zone of autism genetics functions in the face of seemingly clashing terminologies and classificatory systems comes from cases where a genetic mutation that was not initially associated with autism becomes attached to it as if pulled in by a magnetic force field. While there are significant implications for the scientific and social networks built up around the mutation, they do not get swallowed up by autism, nor is autism abandoned as a category of research, care and advocacy in favor of a ‘more specific’ genetic disorder. Rather, advocates and biomedical experts concerned with genetic disorders tend to embrace their association with autism and the many resources it brings with it, even as they navigate the challenge of pursuing their own interests through alliances with advocates and experts focused on autism. After revisiting the case of Fragile X Syndrome, the first genomically designated disorder to significantly overlap with autism, we turn to the more recent case of 22q13 Deletion Syndrome.

The pioneering case of Fragile X

As we saw above and in the previous chapter, one of the first genetic disorders to claim a significant overlap with autism was Fragile X Syndrome. Indeed, in many respects the Fragile X/autism association has become paradigmatic – an example for others to follow because of its evident success. However, we also saw above how a reconfiguration of the diagnostic criteria for autism was required in order to effect significant overlap with

FXS. In other words, we begin to see how contingent the overlap between genomically designated conditions like Fragile X Syndrome and clinically delineated categories of

259

autism are on the shifting diagnostic criteria of the latter. As we saw above, the seminal paper by Hagerman and Rimland (Hagerman et al. 1986:359) made this contingency quite clear: none of the 50 FXS children in their study was diagnosable with autism according to

Kanner’s original delineation, 16% according to the DSM III and double that figure according to the ABC checklist, while ‘autistic traits’ were observed in over 90%. Today, following the diagnostic expansion of ‘autism’ in DSM III-R and IV, the same lead author

(R. Hagerman, Hoem, & Hagerman, 2010) reports autism in approximately 30% of Fragile

X cases (male and female), with a further 30% diagnosable with PDD-NOS (another autism spectrum disorder) and a huge majority meeting some of the diagnostic criteria for autism. As we saw, some estimates put autism and ASD rates in FXS higher still. This biological fact of FXS-ASD overlap is therefore also, in part, a social fact, and for our purposes it must count among the key conditions of possibility for genomically designated syndromes to gain traction as object of research, treatment, advocacy and capital investment.

By developing and embracing a strong association with autism, as well as alliances with autism experts and activists, and by drawing on the repertoires pioneered by Rimland and autism parents, the Fragile X activists have managed to attract enormous attention and resources to their syndrome, despite its relatively rare incidence of ~1:4000 (Morton et al.

1997; Turner et al. 1996). Otherwise unimaginable numbers of biomedical researchers have turned their attention to Fragile X in the hope of leveraging its specificity to get a better handle on autism spectrum disorders. This dual process of leveraging with respect to

ASDs and FXS now includes a wide range of organizational actors – not just the leading advocacy groups Autism Speaks and the National Fragile X Foundation, but also major

260

research universities and organizations like the Simons Foundation Autism Research

Initiative, federal agencies, pharmaceutical companies and so on. As ASD has loomed ever larger as a public health issue, Fragile X Syndrome research has grown accordingly. Not only do they have a mutation that is thought to hold important clues about an incredibly salient condition, but as a biosocial community represented most powerfully by the

National Fragile X Foundation (co-founded by Hagerman in 1984) and the FRAXA

Research Foundation it is organized to facilitate their use as human subjects in biomedical studies. This confluence of concerted activism by FXS advocates and the hope that research on a rare syndrome can provide novel insights about autism has made the National

Fragile X Foundation perhaps the single greatest success story among advocacy organizations for rare genetic disorders and a model for others to follow. As we saw in

Chapter 4, this dynamic has seen Fragile X become a major object of research and advocacy, and today several pharmaceutical treatments for Fragile X are on their way to the market.

Regarding the argument of this chapter, there are two important lessons to be drawn from the case of Fragile X: first, the population overlap between FXS and ASDs was contingent upon the continual revision of the latter’s diagnostic criteria; second,

Fragile X thrives precisely because it is able to be at one and the same time a genomically designated disorder and a genetic model or susceptibility locus for autism. If it was merely a rare genetic disorder – however well understood and specified – it would not have been able to attract the resources and attention it does. By the same token, if it was merely one more susceptibility locus for autism – however promising – there would be no Fragile X organization, no Fragile X identity or biosocial community. It is precisely the ambiguity of

261

the relations between Fragile X and autism which is so productive for both sides, and which justifies the conceptualization of autism genetics as a “trading zone”.

The making of 22q13 deletion syndrome

In this section, I examine the scientific and social career of one genetic abnormality and the way it has intersected with autism: the microdeletion at site 13.3 on the long arm of the 22nd chromosome. As we saw in Chapter 1, 22q13 Deletion Syndrome/Phelan-

McDermid Syndrome (PMS) is a paradigmatic case of genomic designation and, as we learned above, it has gone from little or no association with autism to a condition characterized by high rates of ASDs. Indeed, there is now considerable biomedical interest in PMS as well as SHANK3 – the gene whose haploinsufficiency is considered responsible for many of its effects – on the part of biomedical researchers and advocates interested in autism. As we will see, SHANK3 has been called the new ‘it’ gene for autism, and there are now preliminary pharmaceutical trials being conducted on 22q13DS patients. By considering the history of the 22q13 deletion I hope to demonstrate, first, that other genetic abnormalities have recapitulated the trajectory of Fragile X Syndrome and become increasingly attached –at the levels of population overlap, research projects and health advocacy – to autism. Second, I will argue that, just like Fragile X, they thrive when they maintain an ambiguity between being a genomically designated disorder and a genetic subtype of autism – an ambiguity supported and made productive by the device of “final common pathway”. To examine how the two sides frame and negotiate their goals so as to create a vibrant biosocial trading zone, I pay particular attention to a symposium held at

262

New York Academy of Medicine in 2011 that brought together leading experts and advocates for both 22q13 and autism.

As we saw in Chapter 1, the initial reports on the 22q13 deletion in 1988 focused on a series of severe physical malformations in a single infant (Phelan et al. 1988), and a decade later a parent’s advocacy group was formed that went on to become a foundation in

2002. Autism did not figure at all in the symptoms associated with the deletion. Further reports over the following years expanded the phenotypic profile considerably, but it was a another twelve years before an association with autism was even noted (Goizet et al. 2000).

As recently as 2011 a study comparing 22q13DS and autism was published (Glaser and

Shaw 2011) whose “findings differentiate the phenotypes of the two disorders.” Indeed we saw above how this association was initially rejected by leading 22q13 researchers on the grounds that it is only possible to diagnose autism in children with moderate-to-severe ID when their autistic behaviors are more severe than IQ matched controls – the same objection originally faced by the FXS-autism association. And yet those same researchers now readily embrace the idea of 22q13 as a form of “syndromic autism” and, in fact, “one of the common chromosome defects associated with autism.” (ibid)

This connection with autism has loomed ever larger in terms of research, care strategies, organizational overlap and indeed the proportion of 22q13.3 deletion cases diagnosable with autism. As we saw above, 40-80% of PMS patients examined in research studies are now diagnosable with autism (see Sarasua et al. 2011), while 22q13.3 deletions or SHANK3 disruptions are thought to account for up to 2% of all autism caseloads

(Abrahams and Geschwind 2008: 344). Interest from autism researchers and organizations has grown dramatically as a result. (Betancur, Sakurai, and Buxbaum 2009; Caglayan

263

2010; Durand et al. 2007; Garber 2007; Manning et al. 2004; Moessner et al. 2007; Peca et al. 2011; Vorstman et al. 2005).

At the expert-advocate symposium the reported rate from a new study in progress put it at 78%, which led Phelan to exclaim approvingly:

That’s amazing… It makes 22q13 very appealing because it umm so much interest and money available (I shouldn’t say money); but autism is so hot right now, to have a community where you have a defined, identifiable cause of their autism is [a] luxury… you have an identified genetic cause of autism and you have a group – a defined group of individuals that are eager to participate in a study to learn more about their condition… is I think an autism researcher’s dream come true. So it is definitely beneficial to the researchers and it’s beneficial to the families. (Interview with Phelan, NYC, March 2011)

Because it is a specific genetic disorder delineated by a particular mutation that confers such a high risk of ASD diagnosis, researchers have begun to turn to PMS as a genetic model for autism.

The PMSF is well aware of this rationale for an explosion of research on their population, and they have welcomed and indeed encouraged this interest. Geraldine Bliss, the chair of the PMSF Research Support Committee, explained the situation in an Autism

Speaks Official Blog post. Note how the notion of “final common pathway” plays a double role in her explanation, both as a justification for why research on a condition that is common to only 1% of autism patients is nonetheless significant for the whole spectrum, and as an assertion that there would be a payoff in the final and distant term:

There has been growing scientific interest in SHANK3, a gene on chromosome 22 along with several other autism-related genes, which portend a new era of understanding and medical treatment. While only about 1% of people with autism spectrum disorders (ASD) have SHANK3 mutations, Shank3 research has broad implications for many people with ASD. It plays an important and central role in synaptic structure, learning, and memory in autism. It interacts with many other proteins critical to neurological functioning, and some of these proteins are already implicated in other genetic forms of autism. A number of researchers have developed mouse knockout models that turn off different parts of the Shank3 protein. These models have led to behavioral, chemical and physiological assays

264

to study the underlying molecular problems in ASD and to rapidly test candidate drugs for future clinical studies. Unlocking the mystery of Shank3 will open the door to understanding its partner proteins, providing a research path towards effective drug treatments for many ASDs. (Bliss 2011)9

Just as autism researchers and activists envision a potential payoff in using 22q13.3 as a genetic model for autism, so does the PMS community perceive distinct advantages in being understood as a genetic lens on autism. First, the estimated rate of 22q13.3 deletions in autism caseloads gives PMS advocates reason to believe that the true prevalence of PMS is at least 1:10,000-15,000 (Rogers presentation 2011) rather than the existing estimate of around 1:30,000. The more prevalent a condition, the more resources and attention it should merit; the more potential allies it can secure. Second, and more importantly, the idea that PMS represents a genetic etiology for or subtype of autism attracts biomedical research and funding on a scale that would be otherwise unimaginable for such a rare disorder.

For these reasons, there is an alliance between PMS advocates and researchers – with their control over this rare, genetically specific, but therefore very valuable population

– and autism advocates and researchers who are primarily concerned with the much larger, but also biologically fuzzier, autism spectrum. While there are considerable differences in the way many in these two groups understand the situation – is autism a disorder that can be caused by a 22q13.3 deletion or is the 22q13.3 Deletion Syndrome a disorder for which autism is a merely a symptom? – these differences do not pose an obstacle to their alliance.

What’s more, the two sides can also draw on the model of Fragile X Syndrome, and its established terms of exchange with the much larger fields of autism research, treatment and

9 http://blog.autismspeaks.org/2011/01/11/the-phelan-mcdermid-syndrome-foundation/

265

advocacy. Indeed, PMS activists and researchers quite consciously emulate what Fragile X activists have done, evidenced most clearly by a talk at the symposium subtitled “lessons from Fragile X”. Thus, we can speak of a burgeoning trading zone of autism genetics, where relatively rare genetic mutations function as boundary objects which coordinate the work of parties with divergent interests and understandings, as long as the mutations remain at one and the same time distinct syndromes and nested within the autism spectrum.

Evidence of how this trading zone functions is furnished by the 2011 International

Phelan-McDermid Syndrome Symposium, co-organized by the PMS Foundation and

Joseph Buxbaum, a leading expert on autism genetics from the Seaver Autism Center at the Mt. Sinai School of Medicine. Geraldine Bliss – the leading organizer of the symposium from PMSF – succinctly stated the purpose of the symposium: “to bring together our stakeholders to develop a plan to maximize scientific resources through coordinated efforts and to find the fastest pathways from bench to bedside.” The 60 researchers participating in the symposium came from a range of fields: neuroscience, psychiatry, pediatric medicine, biology, medical genetics, genetic counselling, basic genetics, stem cell research, physiopathology and pharmacological development. Katy

Phelan, the medical geneticist responsible for delineating 22q13DS, and Curtis Rogers, a genetic counsellor who has been heavily involved in 22q13 research and organization from its early days at Greenwood Genetics SC in the early 1990s, represented the longstanding

22q13.3 experts. They were joined by around 40 advocates and leaders of organizations from Autism Speaks, PMSF and the Simons Foundation Autism Research Initiative as well as researchers from pharmaceutical concerns.

266

Crucially, the basis for the joint endeavour was not an agreement about the most fundamental terms of interest, but the perception of potentially profitable trade. For some biomedical researchers genetic mutations represented the better way to carve up disease categories, with autism recast as more of a common symptom. Thus Catalina Betancur presented tables of genetic mutations associated with ASDs akin to Tables 1 and 2 above, and explained that “Autism, like intellectual disability or epilepsy, is not a single disorder but a behavioral manifestation of tens and probably hundreds of genetic and genomic disorders,” (our emphasis) most of which can also cause intellectual disability with or without autism. (Betancur 2011) For John E. Spiro from The Simons Foundation Autism

Research Initiative (SFARI), by contrast, genetic mutations like the 22q13.3 microdeletion or a SHANK3 mutation were ‘causes of Autism Spectrum Disorder’ (Spiro 2011, my emphasis). Similar statements were made on both sides throughout the symposium, with some situating PMS as the real condition of concern and autism a symptom, while others discussed ways to understand and treat autism on the basis of SHANK3 disruption as one of its many causes.

Yet this discrepancy at the most fundamental level of ontology and nosology – which entities are cast as cause, condition and symptom – did not give pause to participants, nor did it dampen the sense of excitement among PMSF activists. As the president of the PMSF told me in an interview following the symposium (Interview, New

York March 2011): “You know, this autism connection has just changed our lives… totally changed our lives.” At the same time, she was keenly aware that PMS, as a highly organized community with a specific genetic condition, is an autism geneticist’s “dream come true.”

267

This very same PMSF leader also told me that, although she was certain her son was diagnosable with an ASD, she had never sought such a diagnosis because to her autism was merely a ‘subjective’ classification. PMS, by contrast, is the “real, scientific” condition, grounded in causal biological mechanisms and offering the better basis for community organization, research and treatment strategies:

And you know, people ask me, “Does he have autism or does he have Phelan-McDermid first? What does he have first?” And I say, “He has Phelan-McDermid, because that is his scientific answer, right? And possibly he has autism as a symptomatic thing.” […] I mean, I [think] you have to say the science first, and that is we know that at least, let's say, [my son] has a deletion in chromosome-22. So OK, that is his diagnosis. Does he also have the symptoms, because autism is a subjective diagn -- it is not a scientific diagnosis, it is subjective, right? […] if I had a child with autism, and now I knew that you could -- you could um, sort of narrow it down to another um, more accurate, um, diagnosis, and then you could be part of something, I would do that… I mean it is like you could say you have cancer. Well, what kind of cancer do you have? You know? You want to know specifically because not that the general cancer treatments couldn't work for you, but if you know what it is then you have more specific treatments, right? That is how I think about it. (PMSF President, Interview, March 2011)

In short, groups like PMSF are consciously and pragmatically drawing on the model pioneered by Fragile X researchers and advocates: leverage the overlap between your specific genetic disorder and autism to attract research interest, funding and recognition and organize your community in such a way as to facilitate that research. However, this does not require that they allow the genomically designated conditions that are their main concern become subsumed by autism; on the contrary, their association with the massive field of autism research and advocacy it helps them to advance knowledge production and awareness for their rare, ‘more specific’ and ‘scientific’ forms of illness.

With this in mind, PMSF is seeking to establish a registry for researchers modelled on AGRE, develop emotional ties to researchers and to begin funding postdocs as they seek to cultivate a small field of more dedicated researchers for whom PMS is more than

268

just a biomedical lens on ASDs (PMSF leader Interview, March 2011). The chair of the

PMSF Research Committee and symposium organizer noted the importance of showcasing

PMS as an object of research and fostering collaboration, as well as the projects that were initiated as a result of the 2011 Symposium. However, she too indicated an awareness of the need for PMSF to not only facilitate, but also direct biomedical research:

3.) We can help shape the direction that researchers are going. We can use our “moral authority” to get researchers to share information and resources (like animal models) that they would otherwise not be willing to share with competing researchers, and we can use our influence to bring attention to issues that need further research. 4.) Funding for research is extremely competitive. Symposia help to “showcase” some of the work that is highly relevant to PMS, and having decision makers from the major funding agencies there helps to improve their awareness of PMS and willingness to fund PMS –related research.” (Bliss 2011)

PMSF advocates embrace the status of PMS as a model for autism, but they do so critically and with an awareness that they possess a very valuable biomedical commodity – people with a 22q13.3 microdeletion. In this way, the attempt to use genetic disorders like PMS as a biomedical lens on to the much larger population with ASDs can serve to strengthen

PMS as a category of research, treatment and identity formation. The genomic classification thrives on its strong relationship with a psychiatric one.

At the same time, the coordinating device of the ‘final common pathway’ facilitates a fairly seamless trade between the two sides and permits autism researchers and activists to maintain research momentum without needing to broach questions of ontology and nosology; without needing therefore to abandon the category of autism or carve it up. Even though it is only estimated to effect 0.5-2% of autism diagnoses, SHANK3 is now “the new “it” gene for autism” (Kouser 2011) because it holds the promise of uncovering a pathway that could explain ASDs more generally. Perhaps the most enthusiasm at the symposium was generated by discussion between geneticists from several different labs about mice models of autism where the SHANK3 gene would be turned off, and indeed

269

one of the papers presented was published in Nature just days later (Peca et al. 2011).

Autism researchers and activists were enthusiastic about a possible glimpse into the causal mechanisms generating autism, while PMS activists were able to see in such research the potential for developing a pharmaceutical treatment for PMS on the model currently pioneered in Fragile X Syndrome, for which several companies now have drug treatments in late stage trials. No less than three speakers invoked Krueger and Bear’s (2011) discussion of Fragile X Syndrome research as a model for finally realizing the ‘promise of genetic medicine’ (ibid). Mark Bear himself, probably the leading figure in Fragile X pharmaceutical research and production, gave a presentation explaining how the discovery of a genetic disorder’s molecular basis could lead to the development of animal models that would facilitate developing pharmaceutical treatments. He then provided a succinct formulation of the coordinating device of a final common pathway suspended into the future: “ASD is one common consequence of altered synaptic regulation of protein synthesis. …Do many genetic lesions converge onto a few pathophysiological mechanisms? Will treatments developed for one cause of ASD be effective for others?”

With the model pioneered by experts and advocates interested in Fragile X

Syndrome and autism over the last generation or so, things have moved much more quickly for PMS. This hybrid research/advocacy strategy, and SHANK3’s status as the

“new ‘it’ gene for autism,” has already led to the search for pharmaceutical compounds. A

Phase 1 clinical trial is now underway on PMS patients at the Mount Sinai School of

Medicine’s Seaver Autism Center for a compound, Insulin-Like Growth Factor-1 (IGF-1), that has been shown to reverse abnormalities in brain slices taken from SHANK3-deleted mice (Kolevzon 2012). In addition to physical and neurological exams and medical and

270

psychiatric histories, subjects will be assessed using: Autism Diagnostic Interview (ADI),

Autism Diagnostic Observation Schedule (ADOS), the Mullen Scales of Early Learning

Scale or Leiter International Performance Scale-Revised, the Vineland Adaptive Behavior

Scale (VABS), the Repetitive Behaviors Scale (RBS) and the Aberrant Behavior Checklist, as well as other protocols that pertain to ASD diagnoses. Only children with a score of 12 or higher on the ABC Social Withdrawal subscale were included. Recruitment for another such trial is underway at Harvard and Children’s Hospital Boston, just as pharmaceutical products for Fragile X Syndrome like the ones pioneered by Bear et al. (above) are poised to hit the market.

Things are moving fast for PMS and its foundation. As their February 2013

Research Update states:

PMSF’s strategic plan for science has a singular purpose: To improve the quality of life of individuals with Phelan-McDermid Syndrome. When we began executing our ambitious strategic plan for science just two years ago, I had no idea how quickly we would progress. With limited funds, PMSF had to focus on supporting those initiatives where we expected the most impactful returns. We’ve been fortunate to have the expertise of world-renowned geneticists, neuroscientists, neurolgists [sic], and other subject matter experts to help in our decision-making. Wherever possible, we’ve leveraged existing resources to minimize capital investments by PMSF. / Did you know that 80% of individuals with Phelan- McDermid Syndrome have autism? Scientific discoveries related to the gene, SHANK3, implicated in Phelan-McDermid Syndrome, have been among the most promising in the field of autism research. Not only are these discoveries hastening translational efforts related to Phelan- McDermid Syndrome, but they are yielding new insights into the causes of autism and possible treatments. (Phelan-McDermid Syndrome Foundation 2013:2 my emphasis)

Thus in one unbroken passage we see how the PMSF’s ‘singular purpose’ has achieved rapid success with assistance from the avid interest of geneticists and other biomedical researchers as well as the ability to leverage existing resources, followed immediately by a

271

discussion of the high rates of autism in PMS and the way SHANK3 holds great promise for both autism research and translational work with respect to PMS.

For better or worse, people with Phelan-McDermid Syndrome will likely soon be receiving pharmaceutical products from researchers and companies who hope that what works for one genetically specific form of autism will work for the ASD population more generally.

Conclusion

The arguments advanced in this chapter clearly require further development and confirmation. First, interviews with key figures in early autism genetics research will help to determine the degree to which ‘geneticization’ influenced deliberation about the expansion of autism’s diagnostic criteria. Second, a broader and more systematic quantitative analysis will allow us to test the claim that ASD rates in genomically designated conditions have risen at a significantly higher rate that in the population as a whole, thereby either refuting or bolstering the claim that autism’s enormous genetic heterogeneity is in large part an artefact of its geneticization and diagnostic expansion.

Third, we should seek to clarify why certain mutations follow the model set by Fragile X, while others do not. Clearly, one of the key causes is the prior formation of an advocacy organization and the consolidation of the mutation as an identity for patients and parents.

With ASD fast becoming the paradigmatic developmental disorder, we saw how the processes we identified here seem able to pull in even mutations like the 7q11.23 deletion, and therefore Williams Syndrome, despite its characteristic hyper-sociability and relative strengths in language. However, other genetic abnormalities like the 16p11.2 deletion are

272

more likely to be discussed as ‘candidate loci’ for autism (Ballif, Coppinger, and Shaffer

2008; Kumar et al. 2008; Weiss et al. 2008), and even used as a test for autism.10 More research is needed to determine why certain genetic mutations become genomically designated disorders, nested or not within a larger classification, while others become merely one more cause of a given psychiatric classification. Finally, it will be necessary to further examine the developing relationships between autism and genetic disorders and especially to track the results of the pharmaceutical products being developed for conditions like FXS and PMS by researchers and companies trying to develop treatments for idiopathic autism.

That said, I believe I have advanced enough evidence to justify further sociological exploration of autism genetics and, more generally, the relationship between classification, looping processes and the genetic makeup of populations. We saw how ‘geneticization’ played both a discursive and, through heritability studies and research on genetic disorders, evidentiary role in autism’s secular diagnostic expansion. Then I presented preliminary evidence to support the claim, based on ASD rates in genomically designated conditions, that this same diagnostic expansion actually contributed to the vast genetic heterogeneity now taken as a biological fact with respect to autism. On the one hand, it is hardly surprising that changes in diagnostic practice can impact the etiological findings associated with a medical condition. On the other, this analysis helps us to begin thinking about the way classificatory and associated social processes can transform the genetic makeup of a population. Furthermore, and conversely, we can see how what our genes are taken to be causes of is in fact largely contingent upon shifting social and diagnostic terrain, their

10 http://www.childrenshospital.org/clinicalservices/Site1925/mainpageS1925P9.html

273

supposedly foundational etiological status notwithstanding. In this case, the idea of autism as a genetic disorder turns out to be, in part, a self-fulfilling prophecy as genetic evidence served to influence diagnostic practice, but also one with unintended consequences as genetic heterogeneity, rather than clarity, was the ultimate outcome. As a result, a raft of

‘autism genes’ was brought into being, creating new possibilities for both genetic mutations’ careers as actants and for the communities of people they delineate.

We saw how that heterogeneity, in turn, has created a series of fecund biosocial intersections as genetic disorders/mutations that previously had little or no association with

ASDs have been taken up as strategic research sites for autism experts and advocates, leading to a range of innovative and complex programs of biomedical research and advocate collaboration. My analysis of the trajectory of PMS demonstrated that the genetic mutation identifying the disorder functions as a “boundary object” (Star and Griesemer

1989) that is flexible enough to accommodate multiple disciplinary and stakeholder understandings as well as different ideas about etiology, disease classification and symptomatology, but firm enough to coordinate cooperation and exchange. Consequently,

PMS, like Fragile X before it, is increasingly becoming attached to and nested within the autism spectrum, with both conditions (and organizations) deriving strength from this linkage rather than being weakened by some apparent contradiction. We therefore see how the discursive and institutional forms associated with autism have served as a powerful

‘surfaces of emergence’ (Foucault 2002:41) for genomic designation in a way that those associated primarily with mental retardation were not. This robust and profitable trade is enabled by the concept of the ‘final common pathway’, allowing actors to coordinate action in the face of divergent goals and frameworks of understanding. ‘Fulfilling the

274

promise of genetic medicine’ (above) by using genomically designated conditions like the

Fragile X and 22q13 Deletion Syndromes to get at psychiatric disorders like autism, in sum, was the outcome of complex processes of looping and social coordination.

As such this chapter has shown that psychiatric diagnoses such as autism are unlikely to be replaced tout court by genomic classification because of their threefold social role as coordinating devices, identities and sites for looping processes. Indeed, we see how genomic designation can thrive, not by working as a hegemonic genome-centered form of classification, but rather as a novel-but-commensurable way of carving up human difference that can translate the interests of diverse actors and interface with existing classificatory systems. In the terms introduced by Collins, Evans and Gorman (2010:9), the trading zone of autism genetics is likely to remain fractionalized, at once heterogeneous and collaborative, organized around boundary objects and interactional expertise. One reason is that the trading zone discussed here, where a rare genetic disorder becomes a major object of biomedical investigation as a potential model of ASDs more generally, contains too many actors and is of such diversity that there is little prospect of any one group becoming ‘hegemonic’ and imposing its language and system of classification on all others, or even of a new “interlanguage” developing as Collins, Evans and Gorman (ibid:

10-11) argue was the case for hybrid disciplines like biochemistry, particle physics and nanoscience. Rather, as much as boundary objects help to coordinate action, interactional expertise is developed by brokers in both advocacy groups and the many biomedical disciplines involved, so that each becomes able to speak to one another’s concerns.

But the deeper reason is that a trading zone that deals in human health, disease, tissue and especially cognitive development and psychiatric illness will almost invariably

275

involve some degree of looping, i.e. where the various forms of expert knowledge and the people under investigation enter into a dynamic relationship where each affects the other in a continuous, recursive fashion. The paths taken by genomically designated kinds of people and psychiatric diagnoses like autism with which they intersect will therefore likely remain dependent on the interaction between diverse stakeholders in biomedical trading zones. A key theoretical agenda for future work on this topic, therefore, is to bring these two different concepts – looping and trading zones – into dialogue with one another.

Analyses of looping have much to learn from an approach that emphasizes the intersection of divergent ways of understanding and acting on human difference and the interaction between different levels of classification; meanwhile, the trading zone metaphor becomes even more useful when we add advocates and patients to the mix of actors trading and being traded. Finally, when it comes to the study of our genomes, this chapter suggests that we need to take stock of the way looping processes can transform the genetic makeup of populations, creating both opportunities and challenges for researchers and advocates seeking understanding and cure.

276 ! !

Chapter 6 – Assembling a new kind of person: Mobilizing mutations and realigning illness

Disney World might seem like an odd venue for a conference about a genetic mutation, but for the 2012 meeting of the myriad actors united by an interest in the

22q11.2 microdeletion it turned out to have distinct advantages. Not least, it afforded some novel ‘biosocial’ intersections. Take the main reception: I was chatting to a cognitive psychologist about therapeutic approaches for children affected by rare genetic disorders when he abruptly burst out laughing, pointed to the other side of the room where a geneticist was having his picture taken with Mickey, Minnie and his daughter, and exclaimed, “that’s an incredible photo – he basically slaughters mice for a living!” A long line had formed as various kinds of biomedical experts, clinicians, activists, parents and children waited for their turn with the murine guests of honor. Only hours earlier the room we were in had been host to a biomedical session entitled ‘Of Mice and Men’ – part of the two-day program of presentations related to biomedical research on 22q11.2 Deletion

Syndrome (DS) – outlining some of the work being done on mutant mice of a very different kind and the way they might shed light on the condition of many of those same children. The next day, the conference would turn to talk of increased awareness and the work of the International 22q11.2 Foundation, treatment options and the challenges patients and their loved ones might face over the coming years. Outside in the parking lot sat the ‘22q Mystery Tour Bus’, representing the Dempster Family Foundation who had

! 277! ! ! entered the field a few years earlier when the daughter of former Chicago Cubs ace and current Boston Red Sox starting pitcher Ryan Dempster was neonatally diagnosed with

22q11.2DS.

Over the five-day conference we heard from patients, parents, activists, genetic counselors, psychologists, representatives from numerous fields in genetics, medicine, psychiatry and special education, a celebrity photographer, and a commercial pharmaceutical researcher. We heard about the DNA breakpoints that define the key protein-producing regions of 22q11.2, the haploinsufficiency that causes heart defects in mice with homologous deletions and the spectrum of physical and developmental challenges faced by people with 22q11.2DS as well as the latest research on treatment that can prevent or ameliorate some of them; we heard about the pressure that had to be brought to bear in Washington DC to get 22q11.2DS added to the US newborn screening regime, the past and future for the foundation, a 22q summer camp, the networks of support and mutual aid developed between families and the many-varied stories that all culminate with someone being diagnosed with 22q11.2DS and their subsequent presence at the conference.

This motley crew, gathered together in Orlando Florida from all over North

America, Europe and beyond1 – 21 countries were represented in all – constitutes one of a growing group of hybrid networks dedicated to genomically designated conditions. This chapter draws on fieldwork at Elwyn Services in Pennsylvania and with 22q11.2DS experts, clinicians, advocates and parents as a case study in order to examine the contemporary field of genomically designated conditions that the rest of the dissertation has built up to. Throughout, I will try to balance an analysis that, in dialog with publicly !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 In 2010 they met in the British city of Coventry and in 2014 they will gather in Mallorca, Spain.

! 278! ! ! available materials about other conditions, attempts to say something about genomic designation more generally while also trying to remain cognizant of 22q’s specificity as a disease, illness and cause. I should note at the outset that I consider further fieldwork to be the primary direction for further empirical research on genomic designation – work I hope to do over the next few years.

The motto of the International 22q11.2 Deletion Syndrome Foundation is

‘Detection, Care, Cure’. While those may seem like self-evident goals for a health advocacy organization, we will see that for conditions like 22q11.2DS they bring with them novel challenges that require innovative strategies of action. I will therefore attempt to situate 22q as a complex network that intersects and overlaps with a diverse array of fields whose exogenous development in recent decades has contributed decisively to

22q11.2DS’s conditions of possibility as a powerful category of understanding and practice. Furthermore, 22q11.2DS must be mobilized in such a way that it can both work through but also transform and realign existing conceptual and institutional structures for understanding and acting on human difference. As vital as indicators like research funding may be (see Best 2012; Bishop 2010), actors working in the name of conditions like

22q11.2DS must advance on a much broader range of fronts in order achieve their goals.

At stake is not simply the garnering of resources for an established, though rare, disorder; on the contrary, a condition like 22q must struggle to gain traction among researchers, clinicians and institutional structures that are geared towards qualitatively different ways of classifying and acting on disease.

This chapter therefore develops the analysis in Chapter 2, which used historical research and citation analysis to show how the microdeletion at 22q11 served to unify the

! 279! ! ! biomedical literatures on several rare clinical syndromes and delineate the qualitatively new, phenotypically varied condition called 22q11.2 Deletion Syndrome. Whereas that chapter dealt with 22q11.2DS solely as an object of knowledge, this one uses fieldwork, advocacy and historical materials to examine the way it functions as an object of clinical practice, social action and identity formation. As we will see, 22q11.2DS remains in a sometimes-uneasy interface with the ‘clinical gaze’, while simultaneously competing for resources both within a medico-political infrastructure to which it is not always well suited and amongst its different factions. This chapter therefore uses 22q as a case to analyze the power of genetics to realign illness on the one hand and the challenges faced by genomically designated conditions on the other.

My analysis of 22q11.2DS aims to help us understand the potentially dynamic relationship between genetics, social mobilization and the classification of human difference. This chapter therefore seeks to further integrate theoretical frameworks for understanding the impact of biomedicine on clinical practice, the work of disease advocacy organizations, ‘hybrid collectives’ and groups united by ‘genetic citizenship’ (Best 2012;

Epstein 2007a; Heath, Rapp, and Taussig 2004; Rabeharisoa and Bourret 2009;

Rabeharisoa et al. 2012) with the ‘reiterated facticity’ approach discussed above that aims at a causal analysis of the historically specific conditions and processes of collective action that shape what it means to have a genetic mutation. I will argue that this theoretical arsenal is required in order to capture different elements of what I call, following Hacking, the process of ‘assembling a new kind of person’: the combination of heterogeneous elements that can make a novel category of human difference like 22q11.2DS really matter in the world.

! 280! ! !

Conceptualizing the rise of genomic designation

How are we to make sense of the increasing power of genomic designation to shape the way we understand and manage abnormality, especially when it comes to children with medical and developmental challenges? Building on the introduction to Part II, over the next few pages I propose a framework for understanding those cases of genomic designation that, today, have far outstripped their one-time status as esoteric categories of human genetics research and gone on to shape clinical judgment, the provision of resources, the formation of identities and the lived experience of patients and their families.

I begin by recapping some of the literature in the sociology of science and medicine pertaining to the capacity of bioscientific and especially genetic observations to realign clinical judgment and practice on the one hand and the formation of hybrid research and advocacy organization around disease categories on the other. Bringing these literatures into dialog with one another in such a way that we can makes sense of a genomically designated condition like 22q11.2DS, I argue, requires recourse to a broader STS literature alongside a comparative-historical approach. Later, in the discussion, I will analyze the gains made by 22q11.2DS to date as a process of assembling a new kind of person and highlight the heterogeneous points of interface that must be established across diverse fields of knowledge production and practice in order for a novel category of human difference to transform judgment, treatment and experience.

Recent work in the social studies of science and medicine has examined the ways in which genetics can reconfigure diagnosis and clinical judgment. As we saw in Chapter 2, a number of studies have examined the role of genetic testing in the confirmation of

! 281! ! ! clinical diagnoses (see esp. Hedgecoe 2003; Kerr 2000; Miller et al. 2005). More pertinently for this chapter, work like that of Bourret, Keating and Cambrosio (2011) has examined how knowledge about genetic mutations can shape clinical judgment, especially in oncology (Anon 2005; see also Bourret et al. 2006). Of particular relevance here is

Rabeharisoa and Bourret’s (2009) analysis of the way genetic mutations shape diagnosis and judgment in two research-intensive psychiatry and oncology clinics in France. They begin with Keating and Cambrosio’s (2006) analysis of the way post-WWII biomedicine is produced through ‘biomedical platforms’ that establish an interface between lab and clinic, before noting: “Empirically speaking however, investigations have primarily concerned the laboratory, or have studied crossovers between the laboratory and the clinic woven in and around objects such as clinical trials or new technologies.” Their approach, by contrast, examines “the clinical work that makes the development and performance of biomedicine possible.” (p. 693) I applaud and aim to build on this insight, and Rabeharisoa and Bourret provide a deft analysis of the way ‘bioclinical collectives’ focused on psychiatric diagnosis and care handle observations of genetic mutations about which little has been written and few cases have been reported.

However, this chapter challenges some of Rabeharisoa and Bourret’s main findings.

They found that, in cancer genetics, knowledge about genomic anomalies was such that they could be considered “clinical entities with enough robustness to be mobilized in medical judgments and decisions,” while in the case of “psychiatric genetics, this work remains (almost) entirely in the future” (p. 707) given the “the absence of a robust corpus of clinical observations on these disorders… most of which are rare, have been described in children, but to date [lack a] description of their natural history.” (p. 700) To be sure,

! 282! ! ! this is often the case. However, I will argue below that by going beyond both the lab and any given clinic to look at much larger networks of research and advocacy we can see that precisely the same kind of genetic mutations that they discuss have given rise to clinical entities that are increasingly shaping judgment and decision-making in psychiatry and many other fields besides.

22q11.2DS, like other genomically designated conditions, is part of a growing field of rare disease research and advocacy that has been extensively reviewed in other social scientific work (Best 2012; Bishop 2010; Epstein 2007b; Heath et al. 2004; Novas 2006;

Panofsky 2011; Rabeharisoa and Callon 2002; Terry et al. 2007). As this literature makes clear, patient advocacy groups are playing an increasingly powerful role in fundraising, enhancing awareness and especially in the organization of research and clinical resources.

There have been support groups for conditions like Tay Sachs and Cystic Fibrosis since the

1950s, various philanthropic projects devoted to conditions like tuberculosis, cancer and alcoholism throughout the twentieth century, and organizations devoted to causes like women’s health that can be traced back to the nineteenth. However, there is something of a consensus in the literature that, partly inspired by the success of HIV/AIDS and breast cancer activism, disease-related advocacy has experienced both an enormous quantitative growth and a qualitative shift over the last few decades (Best 2012:781; Epstein

2007b:501–2; Terry et al. 2007).

Not only are there now far more advocacy organizations and activists devoted to conditions both rare and common, but they have also taken on increasingly important roles organizing and garnering funding for biomedical research. When it comes to genetic disorders, single-disease advocacy groups can intervene in order to facilitate and direct

! 283! ! ! some of the disproportionate levels of research they attract, for reasons outlined in Chapter

4, towards translational research (i.e. research that can lead to treatments for the condition in question) (Terry et al. 2007). Callon and Rabeharisoa in particular have shown how patients and their advocates have become important actors in the scientific work on human disease and illness as part of ‘hybrid collectives’ (2003, 2004, 2008; 2002; Rabeharisoa et al. 2012). More generally STS scholars have turned their attention to the way that biomedical knowledge is ‘coproduced’ (Jasanoff 2013) by scientific researchers and the social organizations formed around disease categories. Scholars like Steven Epstein have also shown how the distinction between researcher and activist has been broken down by the emergence of lay experts (Epstein 1996; see also Eyal et al. 2010), and Panofsky has shown how advocacy groups for genetic disorders have worked to engender social ties with biomedical experts that will impact their research agendas moving forward (2011).

Finally, particular interest has been paid to the way these developments function with respect to genetic disorders, with work like that of Heath, Rapp and Taussig on ‘genetic citizenship’ (2004) attending to the way that families affected by genetic disorders can form ties of mutual aid and obligation, especially with respect to facilitating and participating in biomedical research (see also Novas 2006; Terry et al. 2007).

To be sure, the networks of research and advocacy that are formed around genomically designated conditions are part of this broader shift in disease advocacy both generally and with respect to genetic disorders in particular and will be analyzed accordingly below. However, as useful as the social scientific literature on these new health movements is, more analytic work needs to be done in order to understand the resurgence of genomic designation in recent years. First, we need to understand more than

! 284! ! ! just the way a particular condition is advanced by hybrid groups of researchers and advocates, but also how the conditions that are born of this new form of classification have to work within and simultaneously seek to reconfigure existing frameworks for understanding and treating disease. As we will see, delineating disease categories strictly according to genetic test results brings with it new challenges and opportunities for both the diagnosis of affected persons and the creation of clinical profiles and treatment strategies. Furthermore, while these new models undoubtedly aid advocacy for many kinds of illness, genomically designated conditions like 22q11.2DS benefit from a more specific set of historical conditions and repertoires of collective action that I will outline below.

Finally, I use the case of 22q11.2DS to go beyond the recognition of hybrid research- advocacy groups dedicated to disease categories and the analysis of their internal dynamics to develop a framework that, while building on that research, examines the processes of network formation and the manifold intersections of fields that are required to make genomically designated conditions into powerful categories of treatment and understanding.

But can we discuss ‘genomic designation’ as a coherent topic of sociological analysis? On the one hand, what we are really talking about is just an aggregate of medical conditions born of the same form of classification. Insofar as there is a self-conscious field of actors and organizations, they do not line up perfectly with what I am calling genomic designation: for example, conditions like tuberous sclerosis feature prominently in the thinking of Fragile X and 22q13DS researchers and advocates as a genetic disorder associated with autism, but as it is associated with multiple genetic mutations and delineated according to clinical criteria it does not feature much at all in this dissertation; conversely, the case of XYY in the 1960s and 70s is one that contemporary researchers are

! 285! ! ! more likely to situate themselves in opposition to, while I have taken it to be an important episode in the development of the field. On the other hand, there is a group of genetic disorders and associated advocacy organizations that watch and communicate with each other, work with the same networks of researchers, emulate the model pioneered largely by the Fragile X Syndrome and confront homologous conditions of possibility and challenges as genomically designated conditions. So while the syndromes delineated through genomic designation do not line up with any self-declared field or class of medical conditions, there is nevertheless a strong rationale for discussing the conditions listed at the end of Chapter 1, as well as the associated networks of research, care and advocacy, as a coherent object of analysis. The case of XYY Syndrome in the 1960s and 70s and Williams and Brunner

Syndrome more recently showed that genomic designation can gain traction in various kinds of social circuitry, and even today there is substantial variation on a case-by-case basis. However, this chapter focuses on the model that is most common and most powerful in the contemporary period: the mobilization of genomically designated syndromes under the rubric of rare disease advocacy, collaborative programs of research and a particular focus on childhood developmental difference.

How then to conceptualize the accrual of biomedical certainty, clinical utility and social power on the part of genomically designated conditions? In order to answer these questions I will employ elements from theoretical work in science studies and the study of classification alongside the framework of ‘reiterated facticity’ outlined above. Of course

Latour (1988, 1993) and many others in the STS literature have shown how the knowledge and objects produced in labs must be able to translate the interests of diverse actors in order to attain the status of taken-for-granted scientific truth or technoscientific utility.

! 286! ! ! Furthermore, the capacity of actor-network theory to take ‘actants’ or nonhuman actors seriously as agents in network formation is central to the analysis of the 22q11.2 deletion presented in Chapter 2 and to the dissertation more generally. Finally, for our purposes,

Latour’s distinction between translation and diffusion (see esp. Latour 1988) is key: conditions like 22q11.2DS do not become categories of clinical practice and social action or get taken up in new institutions or locales simply by virtue of their biomedical status, but rather grow as heterogeneous of networks in which science and social mobilization are mutually constitutive. But how, to return to Fleck’s example (1981:113) discussed above, does an esoteric journal finding metamorphize into a diagnosis that a doctor can relay to a parent? Whose interests need to be translated, what kinds of organizations and fields of practice need to be courted and, in the case of genomic designation, how is commensurability with clinical classificatory practice achieved?

In other words, we need to treat the hybrid networks of research, care and advocacy that form around genomically designated conditions as embedded in or intersecting with much larger fields – various areas biomedical research and clinical practice, special education, hospitals, common conditions like autism, umbrella groups like Genetic

Alliance and so on. We can therefore analyze the changes in other fields that transform the conditions of possibility for genomic designation, whether those changes were exogenous or the result of concerted, agentive action on the part of people seeking to advance the cause of a genomically designated syndrome. Furthermore, we need to attend to what scholars as diverse as Tilly (e.g. 1993, 2003) and Swidler (1986, 2001) have termed

‘repertoires’ of contention or collective action and the way that different strategies have been developed and diffused within and between genetic disorders over time. The

! 287! ! ! proposed framework of reiterated facticity helps us conduct just such an analysis, situating the shifting networks that can turn conditions like 22q11.2DS into real kinds of people within broader conditions of possibility for genomic designation.

In sum, I propose to draw on an STS approach with its focus on heterogeneous networks, and especially the subfields that deal with disease advocacy, classification and standardization, alongside a comparative-historical approach that attends to the shifting conditions of possibility and repertoires of collective action for mobilizing genetic mutations. Such are the tools necessary to understand how a genomically designated condition like 22q11.2DS can become a powerful kind of person rather than a novel and uncertain biomedical finding.

The development of the new model

It is within the broader context of contemporary disease advocacy, discussed extensively in the social scientific literature outlined above, that research and advocacy for genomically designated conditions functions. In a sense, their advocacy organizations are much like the many groups discussed in the existing social scientific literature: as we saw with Fragile X, they lobby, fundraise and work with biomedical experts to organize programs of research. In short, they can be thought of as social movements that seek to redirect resources and reorient research on the basis of the constituencies they claim to represent. Furthermore, I have met numerous advocates who have developed extraordinary levels of lay expertise, and also biomedical researchers for whom genomically designated conditions have become a deeply held personal and political cause that has transformed their careers and research programs. The kinds of hybrid projects of advocacy, research

! 288! ! ! and care seen today in genomically designated conditions are therefore part of a much broader historical shift in the relationship between the biosciences, medicine and stakeholder activism that has transformed all three. Indeed, they see themselves as part of that shift in the role of patient-advocates, draw upon the examples of others and cooperate under the remit of umbrella groups like the Genetic Alliance2 and others. However, we will see that the advocates for genomically designated conditions often face novel challenges that, even when they are not qualitatively different from those faced by other groups, throw important issues into stark relief. Below, I examine the case of 22q11.2 Deletion Syndrome in order to bring some of these issues to the fore.

As we began to see in Chapters 3-5, these models of patient advocacy represent new conditions of possibility for genomically designated syndromes. If XYY and the sociocultural traction it attained was the exception that proved the rule in the 1960s and

70s, today it is just one among many genomically designated conditions that bring together researchers and patient/parent advocates hoping to better understand, treat and perhaps one day cure children with a genetically specific form of developmental difference. At the end of Chapter 3 we saw how Patricia Jacobs – one of the key figures in the history of genomic designation and human cytogenetics more generally – never accepted the downfall of the

XYY-criminality nexus. By the time of her 1982 speech to the American Society of

Human Genetics accepting a career achievement award (Jacobs 1982:696), she was working mostly with spontaneously aborted fetuses in Hawaii, and one cannot help but think that she no longer wanted to deal with social complexity of the kind raised by her

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2 See: http://www.geneticalliance.org/ (4.4.2013)

! 289! ! ! XYY research. After all, we saw how activism with respect to XYY had served to delegitimize and halt research rather than fund and facilitate it.

However, we also saw Jacobs hail the emergence of a new test for “X-linked mental retardation associated with a marker at Xq28” as the dawn of a new era in human genetics in that very same speech. In subsequent chapters we saw how that condition, now known as Fragile X Syndrome, became the object of a path-breaking model of expert- advocate collaboration, alliance formation and activism for genetic disorders. It took another decade for a reliable test to enter into circulation. By then, in the early 90s, rates of institutionalization had long-since plummeted, Fragile X was increasingly associated with autism as well as mental retardation, and parents were increasingly the primary caregivers for their children and able to draw on the rich repertoires of health-related activism that were so key to the autism ‘epidemic’ (Eyal et al. 2010). This was the context in which

Fragile X families and researchers, who began working together to organize research, care and advocacy at a ‘kitchen table’ in 1984, were able to go from strength to strength. That group developed into today’s expansive National Fragile X Foundation that campaigns, fundraises and lobbies in Washington DC. As we saw in Chapters 4 and 5, they have made their relatively rare condition a priority for the NIH, and Fragile X is now both well- recognized and a major subject of biomedical research and pharmaceutical investment.

The case of Fragile X Syndrome, in other words, represented a major biosocial breakthrough for genomic designation, and time and again during my fieldwork the Fragile

X Syndrome Foundation was explicitly cited as a beacon for others to follow. Take Brenda

Finucane, a genetic counselor who has seen and participated in the transformation of numerous genomically designated syndromes into objects of clinical practice and social

! 290! ! ! organization over the last few decades. While she was, like Jacobs, excited by the ability to test for a molecular basis for XLMR and therefore to identify carriers, Finucane was equally enamored with the work of Fragile X advocates. In an interview (Finucane 2011), she outlined some of the FXS Foundation’s achievements like having lobbyists in

Washington and priority status at the NIH and explained, “It is like the model that support groups want to be like... They are unbelievable, they are fantastic.”

Unlike Jacobs, who she actually worked with on FXS in the early 1990s, Finucane has been at the forefront of the kind of parent/expert collaboration that represents such fecund biosocial terrain. She repudiates and discursively distances the contemporary field from the work on XYY in the 1960s and 70s, which she told me was “an old story that was wrong” and that “the genetics profession moved on decades ago… it is certainly not part of our thinking now.” She also bemoaned the fact that it remained one of the most recognized genetic disorders in fields like psychology. Today, Finucane occasionally deals with XYY as an incidental finding, but she does so according to the current understanding of the condition with its focus on increased risk of ADHD, autism and a slightly decreased IQ rather than aggression, criminality and ‘mental sub-normality’. Indeed one cannot help but think that she would have been a potentially important ally to the work of people like

Stanley Walzer, whose XYY research at Harvard attracted the fatal episode of oppositional activism from Beckwith and Science for the People. After all, Walzer wanted to conduct just the kind of unbiased study that advocates and experts for genomically designated conditions today avidly seek to fund, approve and organize, and he was a proponent of early intervention – perhaps the watchword of contemporary advocacy for genetic and developmental disorders. As we will see throughout this chapter, as much as Finucane and

! 291! ! ! others who work tirelessly towards improved outcomes for people with genetic disorders might repudiate the XYY episode of the 1960s and 70s, fundamental continuities remain in terms of the institutions, actors and challenges that comprise the field. Thinking in terms of reiterated facticity, we can therefore both integrate the XYY-criminality project into our narrative and see how, in very different conditions – cultural, technological, institutional and so on – and given new repertoires of collective action for mobilizing genomically designated kinds of people, those continuities take on new meaning and confer novel implications.

Genetic counseling, psychology and the illustrative case of Elwyn Services

The contrast between Jacobs and Finucane is an instructive one. On the one hand

Finucane has, in comparison to Jacobs, made a relatively modest contribution to human genetics as measured in journal articles, citations and so on (though she does have 59 articles to her name according to Web of Science). On the other hand, in her work as a genetic counselor at Elwyn Services specializing in a range of genetic disorders, and now as the president of the National Society of Genetic Counselors, she has made a substantial contribution to a complex invisible college (Crane 1969) that has worked to make genomically designated syndromes matter in schools, hospitals, advocacy networks and homes (see e.g. Berg, Potocki, and Bacino 2010; Griffiths and King 2004; Miller et al.

2010; Reilly 2012; Robin 2006). Finucane has been involved with the Fragile X group since just after their first meeting in 1984. Since then she has sat at a number of other

‘kitchen tables’ and helped to start new organizations like Parents and Researchers

Interested in Smith-Magenis Syndrome (PRISMS) and IsoDicentric 15 Exchange Advocacy

! 292! ! ! & Support (IDEAS) committed to advancing the cause of genomically designated syndromes. She has also sat on the boards of several others, including the International

22q11.2 Foundation discussed below. In short, she embodies the repertoires of collective action discussed above and serves to transmit them from group to group and organization to organization. Furthermore, the status this kind of work has attained in the field is perhaps most easily captured by Finucane’s election as the 2012 President of the National

Society of Genetic Counselors.

At the core of the Fragile X group since its ‘kitchen table’ founding in 1984 was another genetic counselor, Randi Hagerman, who Finucane described to me as “the big guru and saint of Fragile X.” Hagerman, now at the autism-focused MIND Institute discussed below, featured heavily in Chapters 4 and 5 due to her prolific output on Fragile

X and remains one of that condition’s leading researchers and advocates. Hagerman and

Finucane are at the forefront of a new kind of genetic counseling that goes beyond advice and advocacy for individual patients and families and seeks to actually advance the cause of genetic disorders themselves as objects of research and community formation. The contrast with decades past is stark: when medical sociologist Charles Bosk’s conducted fieldwork at a pediatric hospital in the late-1970s and 80s, he found (1992: xix): “Genetic counseling as a service is generally a matter of transferring information to individuals who request it, and then leaving those individuals alone to make the tragic choices based on that information.” Bosk’s book, All God’s Mistakes, is a nuanced and closely observed account of genetic counseling in a particular time and place, but what is most striking for this dissertation is the contrast with the kinds of advocacy and community organizing work that

I see genetic counselors engaging in today. Far from imparting information and then

! 293! ! ! leaving, throughout my fieldwork and broader research I have come across many genetic counselors engaging in the socially complex task of mobilizing observations of genetic anomalies as the essential referent of new medical conditions, organizations and personal identities.

Indeed, as a group geared precisely towards translating genetics research for patients and families, genetic counselors are ideally placed to help raise awareness among broader lay audiences. Furthermore, genetic counselors are able to play a key role as mediators between patients and families on the one hand and biomedical researchers and clinicians on the other (see Navon 2012). Katy Phelan of Phelan-McDermid Syndrome fame was very clear in our interview that it is usually genetic counselors, rather than geneticists, who do the painstaking-but-vital work of setting up the support groups that can eventually blossom into powerful foundations. When it comes to the networks that are formed around genomically designated conditions, it is important to keep in mind that they are distributed rather than centralized (Timmermans and Berg 1997 discussed below).

However, they are distributed in such a way that relies on mediating actors, often in the form of genetic counselors, who can translate the expertise and interests of the various parties involved as well as transmit institutional knowledge within and between networks.

Indeed the institution that Finucane worked at until very recently, Elwyn Services, actually serves as an illustrative, if not entirely representative microcosm of the history of genetics and people with what we now call intellectual disabilities. As a large and longstanding state-run institution for people with intellectual impairment it has been, in turn, a major center of early twentieth century eugenics research (Hansen 2005:14–19), a source of biological samples for esoteric chromosome studies in human genetics from the

! 294! ! ! early 1960s on and, more recently, part of the vanguard for the kind of genetic counseling that can direct parents towards support groups and other ways of mobilizing genomic anomalies into meaningful categories of clinical and self-practice. Long before Finucane came along, Elwyn was home to one of the leading XYY researchers in the US during the

1960s and 70s, Mary Telfer. In an interview I conducted there with Eliot Simon (2012) – a behavioral psychologist, executive director of research and health services at Elwyn and their unofficial institutional historian – the contrast was laid out quite clearly. Telfer, he told me:

… was actually drawing blood, doing research, bench-type stuff. She was not a genetic counselor, I mean, she was a genetics person. And she was doing research on XYY. I mean, she was like the person in the country for this, and so that was of large interest… [W]hether it changed how people here were treated, I don't think so. But I think she was kind of doing her thing, you know, and was not part of the support services of Elwyn during that time period.

That is, in keeping with what we saw in Chapter 3, geneticists at Elwyn in the 1960s and

70s took great interest in institutionalized populations and especially children with developmental deficits because they were a population that yielded fruitful chromosomes complements for basic research, not because they saw any great clinical or social potential in working with them.

Even Finucane, who now epitomizes the turn to mobilize genetic anomalies as new kinds of people, was originally hired as a kind of intermediary3 to provide biological samples to human genetics researchers. As she put it to me in our interview (ibid):

My first job, and in fact an internship while I was still in graduate school, was at Elwyn. They didn't have a genetic counselor there. They were just starting to think about genetics !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 3 I mean this in the actor-network theoretic sense of an actor who translates without reshaping the knowledge or object in question, in contradistinction to the above discussion of genetic counselors as mediators who do crucially transform the meaning of mutations (Latour 1999:307).

! 295! ! ! and people with developmental disabilities and Elwyn had, at that time, like 750 residents there with intellectual disabilities plus thousands of other people in our school programs and other programs. And so there was a job opening for a genetic counselor and it was actually lobbied for by a geneticist who was from a local institution that I am just not going to name, that I will not name, that was trying to have a liaison to the institution to try to get blood samples for research on genetics. So that was ostensibly the purpose of the genetic counselor at Elwyn and why I was hired.

This was in the early 1980s, and clearly by this stage geneticists remained fascinated by people with intellectual disabilities mostly in order to find interesting, i.e. abnormal, biological samples rather than to inform clinical practice. In keeping with with the earlier period discussed in Chapter 3, people with intellectual disabilities, and especially ‘funny looking kids’, have remained probably the most fruitful populations within which to find genetic anomalies. As exemplified by Elwyn, this has been one of the great constants of human genetics from eugenics through to post-genomic medicine.

But alongside that consistent strain we see important discontinuities in the way the genetic mutations that are found in people with developmental anomalies are mobilized in treatment and social action. While early twentieth century genetics researchers, armed with their phrenological and heritability observations, did gain substantial traction in the social and institutional structures of their day, the early decades of research on abnormal human chromosome complements did not. Although Telfer did make something of an impact with her XYY research, and infamously so with her statement that Richard Speck was a prototypical XYY proband (as we saw, it turned out that Speck had just one Y chromosome), for the most part Elwyn remained simply a source of biological samples for esoteric genetics research. Few if any attempts were made to go back into the institution in order to work with psychologists and paraprofessionals in order to assemble detailed behavioral phenotypes or turn the genetic anomalies they found into objects of knowledge

! 296! ! ! that could inform treatment or social organization. Mobilizing mutations was not a

‘problem’ they failed to solve, but a problematization (see Foucault 1990:257) that had to be taken up by actors with very different interests and expertise. Stated in terms of reiterated facticity, the chromosomal abnormalities found at Elwyn well into the 1980s remained almost entirely in the domain of esoteric genetics and therefore not the kind of facts that could constitute a new kind of person.

By the mid-1980s, however, things began to change. New genetic testing techniques were becoming available, Elwyn was moving towards support services and education rather than institutionalized populations, and new repertoires for dealing with developmental difference in general and genetic disorders in particular were coming into view. Again, Elwyn is not representative: Finucane was at the forefront of these stirrings, having become involved with the new Fragile X group a year after its 1984 formation, and so for her ascertaining people with genetic anomalies was not just a means to a journal article, but also a tool to be mobilized in clinical care and, later on, the formation of support groups. As Finucane told me, the situation at Elwyn:

… changed because that geneticist was no longer involved within six months after I was there, so I was left with a position without its original role and so what I did for probably ten years was traditional genetic counseling. We did diagnostic testing of lots of individuals in the Elwyn system. We found – within five years we found literally over 150 individuals with Fragile X which was this new thing that we were being able to test for. We found all sorts of different genetic conditions and so it was really a heady time I have to say, where, you know, you had all these people with these symptom diagnoses, with these behavioral and psychiatric diagnoses, and now we were diagnosing all these specific genetic conditions.

Armed with her knowledge and experience of the Fragile X organization, Finucane helped to establish and develop support groups and then foundations for several of those ‘specific genetic conditions’.

! 297! ! ! Today, almost every child who comes through Elwyn Services (it is primarily non- residential now) is given a genetic workup by Finucane’s department, yielding a high volume of genomically designated diagnoses. In contrast to previous decades, those results do not remain confined to the lab and the field of human genetics. Instead, the patient and their family are likely to be put in touch with the relevant support group. What’s more, the diagnosis is integrated into the clinical care the patient receives at Elwyn. As a behavioral psychologist, Eliot Simon had been aware of genetic diagnoses for years, ever since he had worked in a state developmental center in the 1980s and Elaine Zackai had come out from

Penn to diagnose children for research purposes. Zackai is a major figure in 22q research,

Director of Clinical Genetics at the Children’s Hospital of Philadelphia and the Medical

Director of their ‘22q and You’ Center. As Simon put it, “if you're Elaine Zackai back in, you know, 1980, '81, you are interested in places like Emeryville which had, you know,

400 people with intellectual disabilities in them.” The interest, Simon made clear, was not mutual: geneticists may have been fascinated by the institution’s subjects, he explained, but the institution was not particularly interested in the geneticists’ findings. This lack of interest carried over to Simon’s early career at Elwyn. He explained:

I wasn't into that, you know, I wasn't really into that. But I knew what was going on and I knew it was happening. I knew people who had Fragile X Syndrome, I knew people who had Down Syndrome, 5p Minus, I mean all these things, they were diagnosing them, it was in the chart, but that is where it went – nowhere else… I thought of it as the same thing as somebody having Down Syndrome – you know, it was something that they had, that was the reason they had intellectual disabilities, and that was it… It did not have the status of a clinical condition that was of interest to me as a behavioristic psychologist… You know, but this person is throwing crap across the room in his workshop and I have got to deal with that so I am going to have to look at the behavior, you know, it didn't factor into my clinical [practice]...

! 298! ! ! In short, Simon knew about a number of genomically designated conditions, but they remained causal explanations of clinical categories rather than clinical categories in their own right that could inform understanding and/or the clinical treatment of psychological issues. They were etiology, and little more.

Today, by contrast, Simon sees the role of genetic diagnoses completely differently.

As he put it to me when I asked him what a condition like Fragile X Syndrome means to him when he sees it in a child’s chart now, he explained:

Yeah, now it's the first question I ask. Now it drives – if there is a genetic diagnosis, it totally changes my clinical approach to the family and to the person. It can become the focus of a large part of the intervention. And it can give you a much richer setting to design your behavioral intervention.

His description of the contrast between his approach to a patient with ‘Fragile X’ on their chart experiencing behavioral difficulties thirty years ago, when it would be largely ignored, and his approach today where it can profoundly shape psychological understanding, intervention and prognosis, is telling and worth quoting at length:

If I go in there as Doctor Simon from 1983, I would go in, I would do a functional analysis of the behavior, I would look at the setting variables, I would try to determine what's the function of the behavior, and I would, through manipulation, try and come up with variables to include in their behavioral program for the person. And I could come up with things like, well, maybe the work is not interesting, maybe – whatever, whatever, I do a fairly standard variables analysis. Now let's say I know this guy has Fragile X, right. Now I know where to start. [laughter] I mean, I know people with Fragile X Syndrome don't like noisy places to start with. They tend to be very shy. I can see: do they have this person sitting at a table across from somebody who is looking them in the eye all the time? All those things I may have gotten to – it would have taken me longer – um, so it gives me a place to start.

He recounted the process of coming to Elwyn and discussing genetic disorders with

Finucane and others in an early morning journal club, and how:

… it became somewhat of a defining moment, clinically, for me that, you know, the etiology can be a kind of synthesizing variable for these different fields to get involved in

! 299! ! ! around the person where intellectual disabilities was clearly not – and that is really what we were working around, but it wasn't a good variable to develop programs around.

This ‘synthesizing variable’ represented by genetic mutations, he explained, can yield a wealth of information that the presenting sign of intellectual disability cannot:

If you are a physician, the fact that they have intellectual disabilities is not all that interesting to you because it doesn't tell you anything [laughter] aside from the fact that their IQ is below 70 and they have got some adaptive behavior needs. If you are a psychologist it is also not all that interesting. If you are a social worker, it maybe be a little interesting but not hugely interesting… But now if I tell you this person has Smith- Magenis Syndrome and you are a physician, well now, well that's interesting because people with Smith-Magenis are prone to kidney problems and they have peripheral neuropathy and they have other things, so, I am now interested in this person in a different way. The same if you are a psychologist. They have got a type of self-injury that the developmental courses, this guy is beginning to be ascertained and it is of a particular type and, and you know, now we are working as a team around supporting somebody with Smith-Magenis Syndrome and you know, that is how it can synthesize a team.

So although they cannot be diagnosed according to clinical presentation, genomically designated conditions can direct clinical attention, judgment and treatment in more fine- grained, if uncertain ways than surveillance categories like intellectual disability. What’s more, we see how they can bring together a team – a new network of elements and expertise – according to new jurisdictional formations.

It is important to point out that, despite ascribing great utility to genomically designated conditions in his practice, Simon does not profess to being able to independently recognize, never mind diagnosis them. He is clear that there is a division of labor at Elywn and that it is Finucane’s department that is charged with the task of genetic diagnosis according to the results of the tests they run. He described the division of labor at

Elwyn and his own acumen when it comes to recognizing genetic disorders thus:

They are the experts in doing that detective work, not me. And I am actually pretty horrible at it [laughter] myself, I mean, “Oh, that person has got this or this,” then I am like, “Really? I don't know.” Sometimes I see it, sometimes I don't, but you know, I mean, that

! 300! ! ! is their – that is their bailiwick…. I mean, and why would I do it when I've got Brenda on my staff, [laughter] I mean, it would be silly, you know? And that's where you have a team approach, I mean, I don't think psychologists should become experts in recognizing genetic syndromes, I mean, that's pointless. You know, there's already a group of professionals who that's what they do… You want to know what their genetic diag[nosis is] – I mean, and that's informing their treatment. I mean, whether I can pick the people out with Fragile X out of a crowd is immaterial to me.

So even though he could not diagnose a genetic disorder on the basis of a psychological phenotype – that is someone else’s task, he insists – Simon is adamant that they can powerfully and effectively inform clinical psychological treatment. At a jurisdictional level

(see Abbott 1988), this is perhaps surprising: why accommodate and allow practice to be determined by diagnostic categories that cannot, even in principle, be made according to the expertise of one’s profession? Nevertheless, Simon explained:

We have a team here at Elwyn and Brenda and her department is under my administrative reach here, but you know… we are one of the few if only places in the country like this that [every patient is] going to get a look-see by a genetic counselor when they hit the door. You know, so she is able through her department to diagnose about 60%, and it is a given that it is going to happen… if you are asking me, every single person, in my opinion, with intellectual disabilities, should have a look-see by a genetics counselor to determine etiology.

For Simon, psychology should therefore be willing to cede jurisdiction when it comes to diagnosis in order to gain the new task of working with the distinctive cognitive phenotypes associated with genetic disorders that he deals with in practice.

What’s more, Simon explained how he will encourage patients and their families to seek out support groups for genetic disorders:

And the other thing it gives me a place to do is to involve the family, if there is one, or the person themselves depending on their intellectual disability, in the Fragile X world, you know, which in 1982 I wouldn't have done… I supervise Master's levels clinicians [and] one of the things I impart on them is you know, if you have somebody with a genetic disorder, you as a team have an obligation. Let's get to know about it, get to know is there a local support group? Should your person be going? Should you be involved in that? And

! 301! ! ! you know, what are the advantages of work? So when you have that diagnosis, depending on what it is, it can greatly inform your approach and open up areas of free support that wouldn't otherwise be accessed... it is different to construct a support plan around intellectual disability than it is to construct it around Down Syndrome or Fragile X or 22q or whatever.

That is, Simon considers it an important professional obligation, and one that he instills in his trainees, to direct families towards support groups for genetic disorders and perhaps even become involved with them as relevant professionals. He also thinks that it is important to take advantage of genetically specificy support groups: “…you now can go into the Smith-Magenis support group. And yeah, before you could have gone to the Arc4 and you could have gone to other places but it's different because it is more specific.” The support groups help to bring the characteristics of these ‘more specific’ kinds of people to the attention of parents and their families by offering “a forum for education. And for kind of group support, you know, I mean it is like a peer support network, and given the internet and listservs and the conventions and all of these other things, it provides that service.” He also pointed out that they provide the research subjects for clinical trials, meaning that support group members will get the most cutting-edge treatments, a topic of increasing relevance as Fragile X pharmaceuticals make their way to the market.

Again, Elwyn is most definitely not typical of institutions for managing populations with intellectual disabilities, and Eliot Simon is undoubtedly something of an outlier as such an ardent advocate for the relevance of genetic disorders in clinical psychology. That said, he pointed towards the strides he and other were making in the !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 4 i.e. Arc of the United States, previously the Association for Retarded Citizens of the United States, is the major support and lobbying organization for people with intellectual and developmental disabilities in the US.

! 302! ! ! fields of intellectual disability and special education, and to the fact that Elywn has been contracted by various school systems and, among others, the State of California as it moves towards fuller deinstitutionalization. Although he was initially highly skeptical when he came to Elwyn and started working with Finucane, Simon is now convinced that genomically designated conditions are important objects of clinical practice for psychologists. When I asked him if he thought that perspective was widespread he said that, while it certainly is not, the field’s published literature – especially the American

Association of Intellectual and Developmental Disability's journals – was moving very much in that direction, having “made huge gains in the past ten years.” Simon suggested:

“If you've gotten any journal nowadays, a majority of the articles in there are going to be specific to genetic syndromes, whereas 20 years ago that was not the case.” While a cursory look at those journals suggests that may be something of an exaggeration, it is still significant that, as Simon put it, “the field of intellectual disabilities I think is academically moving more toward acknowledging the importance of etiology again.”

A major remaining obstacle, he explained, was the resistance of fields like developmental psychology and special education to integrating categories derived from genetics research, borne partly out of the historical association with eugenics. As such “for a lot of historical reasons the field of intellectual disability had a great aversion to etiology, and there is a large contingent of people in the field of intellectual disabilities who still do.”

When I asked him to elaborate he explained:

I mean historically the field of intellectual disability's first brush with etiology was disastrous. It resulted in sequestering people with intellectual disabilities into institutions. It resulted in sterilization, eugenics got all mixed up into it and those folks were talking about heredity of, you know, a broad concept that they felt was linked to criminality. So because

! 303! ! ! of that for years the field of intellectual disabilities didn't want to have much to do with genetics at all, and the field of genetics didn't want to have much to do with the field of intellectual disabilities either…. it is only within the past really ten years or so that the two cultures are starting to feel each other out. And academically it has moved faster than clinically.

As such, he noted, “practicing clinicians are still not there in my experience,” while “the regulations for services for people with intellectual disabilities are very behaviorally based.

They don't acknowledge genetic etiologies for instance, you know, so whether the person has Down Syndrome or has Fragile X, the process should be the same for behavioral support… DSM doesn't either.” In short, neither behavioral psychology nor mainstream psychiatry’s codified diagnostic systems are very amenable to the inclusion of genomically designated syndromes. DSM, in fact, removes diagnoses when a strong etiological finding is made, leaving Simon convinced that Rett Syndrome was sure to be excluded in DSM-V

(he was correct). Nevertheless, while Simon may not be representative of behavioral psychology as a whole, he is indicative of the way that genomically designated conditions are making inroads into fields that did not previously take them seriously as categories of practice.

Finally, Simon noted that the support groups act back on the professional fields by serving “a function to educate the professionals because as I said before, a lot of the professionals aren't into this yet and the support groups all have great materials on the actual disorder.” In this way, we begin to see the processes whereby genomically designated categories are taken up by geneticists, other biomedical experts, clinical practitioners and advocates, leading to the recursive development of these ‘more specific’, and yet clinically obtuse, kinds of people. Absent the networks forged largely by advocates and genetic counselors, however, it is hard to see how these complex medical and

! 304! ! ! psychological phenotypes could develop to the point where they would be useful to someone like Eliot Simon.

While the field may be most aware of XYY Syndrome from outdated textbooks and the infamy it attained decades ago, genetics researchers, advocates and insiders like Simon are working to turn genomically designated conditions into meaningful categories of psychological knowledge production and practice, and likewise many other cognate fields.

In sum, against the backdrop of contemporary disease-based advocacy, genomic designation is achieving increasing traction in previously disjunct institutions and fields.

While Fragile X Syndrome is the clear frontrunner and trailblazer in this regard, hospitals and research centers, care programs and commercial genetic testing kits are all engaging with genomically designated conditions more and more with every passing year. Specialist clinics for 22q11.2DS, 15q Duplication, sex chromosome aneuploidies like XXX and

XYY, Williams Syndrome and others are being opened all over the US and Europe.

Numbers of diagnosed patients are rising, media outlets are taking more notice and fields like psychology and special education are starting to relent in their reticence to take genetic etiology seriously in the treatment of people with cognitive and developmental impairments. As a systematic review is not possible at this stage, I will instead draw on fieldwork and publicly available materials related mostly to 22q11.2DS in order to examine the hybrid processes of advocacy and research that can make genomic designation really matter, as well as some of the novel challenges such a goal presents.

! 305! ! ! 22q Detection, Care & Cure

At a 22q conference that she organized in 2011, Finucane told me that she thought the International 22q11.2 Deletion Syndrome Foundation was where the Fragile X group was perhaps 15 years ago, despite the fact that their population prevalence is undoubtedly considerably higher and tests for their respective mutations were developed at around the same time. As Donna McDonald-McGinn – perhaps the central 22q11.2DS expert/advocate in the US – told me in a 2011 interview, the International 22q11.2

Foundation talk about emulating the Fragile X group at every planning meeting. If 22q advocates can directly emulate their Fragile X counterparts then, assuming there is a minimal ‘first-mover’ advantage in the field of rare disease advocacy5 (and there is no prima facie reason to think otherwise), one might think it would be quite easy for it to leapfrog the trailblazing-but-less prevalent Fragile X. Studying the efforts of 22q advocates, as well as the challenges that they face, therefore allows us to trace the conditions for genomic designation and the forms of social action that can make it matter to a diverse range of actors. It also helps us see how genomic designation can realign clinical judgment and redirect resources.

The International 22q11.2 Deletion Syndrome Foundation’s motto reads,

‘Detection, Care, Cure’, while their official 501(c)3 mission statement is “improving the quality of life for individuals affected by the 22q11.2 deletion syndrome through family and professional partnerships.” In many ways their goals are no different to that of the hundreds of other organizations that advocate for research and treatment for rare disorders,

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 5 I mean in terms of the individual conditions themselves; at the level of advocacy organizations for the same or related conditions we might expect just such a first-mover advantage.

! 306! ! ! especially ones that often present with serious medical problems in infancy and can continue to challenge patients and their loved ones throughout the life-course.

Furthermore, like other organizations, they have not only fostered ties with biomedical experts but also count those experts among their most dedicated activists. As

Panofsky (2011) has suggested, fostering ‘sociability’ can help patient advocacy organizations attract greater volumes of genetics research. My experience with the

22q11.2DS field, however, suggests that we should not just look at the way activists on the one hand create ties to biomedical experts on the other, thereby impacting the latter’s research agenda. Rather, those ties can substantially erode the distinction altogether, leading biomedical experts to become activists in their own right. Thus for Peter Scambler, the medical geneticist most responsible for establishing the link between Velocardiofacial

Syndrome (VCFS) and 22q11 deletions, what began as a straightforward finding has become a life’s work. He acknowledged in an interview (Scambler 2011) that 22q has now become a cause for him, and that if it were not for the personal relationships built with families and colleagues over the years he would have moved on to topics that were “sexier, like stem cells or something.” Instead he keeps working on 22q, runs marathons to raise money for the foundation, and gives support to their activism when he can. Donna

McDonald-McGinn tells a similar story (2011): "I never went in wanting to do this..." The original idea was simply "what is the diagnosis, keep moving"; but she wound up around a kitchen table in Philadelphia in the mid-1990s and, ever since, she has focused her career on 22q11.2 Deletion Syndrome and worked closely with that group through its metamorphosis into an international foundation. She organizes both research programs and foundation events, agrees that 22q is now a cause for her, and insists that she "can't stop

! 307! ! ! until every clinician is familiar with 22q11.2 Deletion Syndrome."

However, 22q11.2 Deletion Syndrome presents researchers and advocates with a series of unusual challenges. Take McDonald-McGinn’s last statement: why is 22q11.2DS, which is actually not so rare at all, still so slow to achieve recognition in the medical community in general and the pertinent specialties in particular? Why do rates of diagnosis remain so low? How does one communicate such a diagnosis with its hugely variable clinical profile to the relevant professions and broader publics, and how does one work to improve care and support for people with 22q11.2DS and their families? Finally, how does a condition like 22q11.2DS realign clinical judgment and the relationship between the normal and the pathological, and how does it thereby redirect resources and practice?

Following the International 22q11.2DS Foundation’s own motto, the rest of this chapter is organized around the challenges facing 22q advocates in improving detection, procuring care and pursuing the goal of a cure.

Detection

For a condition like 22q11.2DS, the problem of ascertainment looms large. While the detection of people with 22q11.2DS may be a prima facie straightforward goal, the reality is anything but. For one thing, we might ask ‘detection of what?’ After all, it is both unusually clear and at the same time exceptionally uncertain what 22q11.2DS really is. On the one hand, it is diagnosed strictly according to the observation, through genetic testing, of the absence of DNA at site 11.2 on the long arm of the twenty-second chromosome; on the other, despite this genetic specificity, 22q11.2DS is characterized by over 180 associated symptoms ranging from serious congenital heart defects, cleft palate and

! 308! ! ! schizophrenia to small ears, tapered digits and ADHD, with some patients affected by dozens of serious ailments and others by only a few or even none. For another, the more we learn about it the more we realize that it might not be very rare at all: with a relatively low profile and such a broad spectrum of clinical indicators (i.e. a phenotype) there is reason to believe that there are many, many people with 22q11.2 microdeletions who have not be diagnosed, in part because they do not conform to the longstanding-but-biased profile of 22q11.2DS in the medical field. How then, are we to make sense of the challenge of detecting 22q11.2DS?

First, we need to keep in mind that, in its capacity as a social movement, 22q detection is really a question of recruitment. Each newly diagnosed person represents not only a tiny increase in total numbers, but also a client for treatment, care and services, a potential research subject and, along with their families, potential members of the 22q foundation or one of the many smaller support groups scattered around the US and Europe.

As is well recognized in the literature, numbers of affected patients is one of the key bases upon which disease advocates advance claims for resources (see e.g. Best 2012; Bishop

2010) and social movement scholars have noted that it is also a central basis for claim- making more generally (e.g. Tilly 2004). But for a condition like 22q11.2DS the discrepancy between likely population prevalence and numbers of actually diagnosed people is almost certainly vast. Increased detection may therefore hold the key to

22q11.2DS becoming a widely recognized condition, with all the many-varied resources that status would entail for patients, researchers, advocates and their organizations.

! 309! ! ! Awareness

It is therefore also a question of awareness – of having various fields and publics cognizant of and interested in 22q – as leading researchers and advocates noted time and again. As the foundation’s president, Carol Cavana, put it in her opening address at the conference:

I wake up rededicated to the cause, to awareness and early detection… Our goal is simple: it is to educate the public about the 22q community. The foundation is working on changing this by raising awareness and visibility around the world with our slogan, 'detection, care, cure'. We know early detection for all will lead to the best care possible, and I would like to thank all the physicians, the clinicians, scientists, specialists, therapists, educators, who have worked so tirelessly to help make the 22q community a better place for our children.

Indeed as the foundation’s secretary and one of its founding members, Wendy Rose, put it in a talk entitled ‘Detection, Care, Cure: a Common Goal’, “We have three wide, broad focuses as an organization: awareness certainly is the first.” What’s more, they were focused on the challenge of raising awareness among different constituencies: “We struggle and focus every day on informing the medical community that so many of you focus on as well, whether you are a part of the medical community yourselves or are parents trying to educate your own physicians wherever you are.” Towards this end Anne

Bassett, Donna McDonald-McGinn and other leading 22q expert/activists spearheaded the organization and writing of an authoritative overview and clinical guidelines by a ‘22q11.2

Deletion Syndrome Consortium’ that was published in 2011 in the ‘Grand Rounds’ section of the Journal of Pediatrics (Bassett et al. 2011). This was perhaps the flagship achievement of a whole host of projects aimed at raising awareness among medical professionals, and in fact the foundation took out a full-page ad in the same issue of

Pediatrics and was working towards having the guidelines translated into other languages.

! 310! ! ! As well as professionals, Rose also discussed the challenge of raising awareness among “the general public, which the Dempsters work so hard with us on – we've focused really one of our major awareness events for two years in a row and we have already got a date for the third, with our 22q at the Zoo Worldwide Awareness Day.” The first ‘22q at the Zoo’ was originally going to be a local Philadelphia event, but after a dramatic change of plans it wound up being held at 62 zoos around the world with the foundation estimating that, as Rose put it, “10,000 people assembled across the world on that day to celebrate, to meet one another, to support one another, and to tell the world about 22q.” They hope that the third iteration this summer will be bigger still. Alongside myriad other efforts, ‘22q at the Zoo’ sees the foundation engaging in the kind of awareness raising that we would expect from a rare disease advocacy organization.

However, while this dual goal of detection and awareness might all seem quite straightforward at first glance it is really anything but. After all, if a general practitioner, a specialist or a genetic counselor orders a test for 22q, or a parent successfully demands one, a definitive result will most likely come back: they have a 22q11.2 deletion or they do not.

The rub, once again, is this: how do you get the tests ordered in the first place? On the basis of which clinical signs? How do you get 22q higher up on clinicians’ differential diagnostics for children with heart and palate malformations, developmental delays and certain craniofacial features without exasperating ascertainment bias and missing mild or atypical cases? It is not just a question of having people referred to the correct clinical specialist. After all, there are no clinical diagnostic criteria for 22q as such, just differential indicators, and so a clinician may simply treat each presenting symptom or clinical pathology individually without thinking to order a test for 22q. Even those familiar with

! 311! ! ! the typical symptoms are likely to miss all but the most classic cases of 22q11.2DS, in other words those who would have been eligible to receive a clinical DiGeorge Syndrome or VCFS diagnosis. In other words, how can 22q11.2DS be made commensurable with existing nosology such that it can be routinely ordered on the basis of certain, but not too specific, clinical signs? This is not a matter of what Hacking (1998:98–101) calls

‘calibration’ – the way a new measurement refines but also validates and affirms existing expert knowledge and categorization. Instead, it is about making a new way of organizing and prioritizing different kinds of observations capable of interfacing with a qualitatively different modality of classification. We will see below how this is played out with respect to clinical judgment and care for 22q11.2DS patients, but what about the challenge of getting people referred for a test that can detect a genetic mutation like the 22q11.2 deletion in the first place? In other words, how do you get clinicians to think about genetic diagnosis?

Take this speech from Sheila Kambin detailing the story of her son and the

‘diagnostic odyssey’ they experienced before finally receiving a 22q11.2DS diagnosis, which I quote in full:

Our journey began when my son was originally delivered at term for severe growth restriction. He spent several weeks in the NICU at very well-known teaching hospital. He was evaluated to thrombocytopenia, hypoglycemia, poor feeding, and temperature regulation. Multiple diagnostic procedures were performed. He was unable to latch for breast-feeding. He had a sacral pit, low set ears, and thrombocytopenia of unknown origin yet no one recommended genetic testing. He came home at a weight of 4- 11. I kept trying to breast feed him and he lost more weight. I had to supplement with formula and pump into a bottle. We saw four doctors and three lactation consultants and no one ever bothered to examine his palette. At four months of age he underwent a bilateral ingrown hernia repair. Next he was late on every single developmental milestone by at least six months. He finally walked at 18 months. We saw physical therapists and occupational therapists. We saw orthopedics for his ligament laxity and hammertoes at age two. That doctor sent us to a geneticist for

! 312! ! ! an evaluation for Ehlers-Danlos Syndrome. His medical and surgical history were carefully reviewed. He had a full physical exam and a seasoned geneticist looked at my son's bulbous nose, his narrow pelcretal fissures, his malar hypoplasia, his retrognathia, his higher palette, his bifid uvula, his hypernasal speech, his long tapered fingers, his growth and developmental delays, and his thin appearing skin. He told me no testing is indicated, let's just get a cardiac echo and when he turns four to be safe. Time marched on. We were sent to an orthotist for his anomalous toes and feet. He was fitted for countless braces. His teacher started to raise concerns. We saw a speech pathologist who failed to recognize his anatomy, his hypernasal speech, and the fact that he had a sub-mucus cleft palate. She told me that his speech was normal. When he was four, one of his teachers told his father and I that she thought something was wrong with Aiden. She delicately explained that she had been teaching preschool for over 30 years and she felt that Aiden needed to be evaluated. But by who? We brought him to a world-renowned developmental pediatrician who spent four hours evaluating him. That doctor told us that he might be on the autism spectrum and sent us to 14 specialists. Those specialists included PT, OT, speech again, othropedics, audiology, back to genetics, psychology for cognitive testing, behavior therapy, and social skills evaluation. Overwhelmed with dampened spirits, his father and I left the office and he said to me, “Aiden does not have autism, and we are not going to make any of these appointments.” I said, “I know he doesn't have autism but we are going to see every specialist on this list until we find an answer.” And he agreed with me. It took another six months and fourteen visits. When he was five we saw another geneticist. That doctor also looked carefully at Aiden and again said, “We won't find anything, but I will send a micro-array just to be thorough.” His diagnosis came back as an incidental finding. That geneticist said to me, “Sheila, this is not the first time I have been burned my 22q. I really need to keep it higher up on my differential.” Aiden is now eight years old. His palette was not repaired until this year because he missed the chance for intervention. His speech patterns are set and most surgeons did not want to touch him. He requires intensive speech and articulation therapy three times per week. It is possible that he suffered from hypocalcaemia at birth. This may play a role in his psycho-social development and cognitive function but we will never know. What I left out of this story until now is that besides of the 27 clinical specialists that he saw, his two parents are both physicians who practice at major academic institutions in a metropolitan area. His father is an ENT, an ears, nose, and throat surgeon, a very bright and accomplished ENT who did not realize the velopharyngeal deficiency. I am an obstetrician who always thought that her son's ears looked funny, but never put it all together. Furthermore, I am an obstetrician who delivers some of the sickest babies in the world, and to this day I am not certain that I could reliably make this diagnosis in the delivery room. My story is not one of a child who almost died, but my story is a very common one. I believe in my heart that Aiden would be undiagnosed to this day if he had not been evaluated at CHOP or had parents with the tenacity to keep pushing for evaluations. Newborn screening would have given Aiden a chance for early interventions which would

! 313! ! ! have improved his prognosis substantially. Newborn screening would have prevented this type of diagnostic odyssey. Thank you.

Why have I chosen to include this speech, delivered at a 22q conference last year, in full?

First of all, I want to give a sense of how serious a condition like 22q can be even when what is perhaps the cardinal clinical sign – a congenital heart defect – is absent. Second, despite the fact that both parents have extensive and pertinent expertise and worked at perhaps the world’s leading center for research and care for 22q, Aiden was only diagnosed as an incidental finding on microarray. In short, no test for 22q per se was ever ordered despite the relatively strong indicators and the abundance of 22q expertise at

CHOP. We saw how an unsatisfying autism diagnosis was offered as part of their

‘diagnostic odyssey’ and how the 22q diagnosis resolved it, not by obviating the need for further referrals, but by providing an etiological diagnosis with a dedicated network of clinical and social practice. Finally, we see how a counterfactual past is imagined in which a 22q diagnosis had been made earlier, perhaps even at birth, allowing for that most cherished opportunity: early intervention. The fact that her son did not fit the typical

22q11.2DS profile would not have made him a ‘patient in waiting’, to use Timmermans and Buchbinder’s evocative term (2010), but instead would have provided an explanatory anchor to prevent years of diagnostic odyssey and a lens through which to understand his manifold challenges.

Kambin’s story speaks to the barriers to diagnosis that exist even for seriously affected patients with access to leading medical specialists. Conversely, it also speaks to the potential for a new generation of non-targeted genetic testing techniques like microarray to increase ascertainment as vs. fluorescence in situ hybridization or FISH tests

! 314! ! ! which has to be specifically targeted to certain chromosomal loci. It encapsulates the belief in early intervention, the need for screening and the clinical utility of a 22q diagnosis. For even though the clinical guidelines published in Pediatrics emphasize treating symptoms according to standard practice, Kambin articulates the widespread position that a 22q diagnosis helps practitioners to see problems that would otherwise be missed (e.g. the submucous cleft palate, infant hypocalcemia) and informs interventions that might not otherwise be undertaken (speech therapy, ENT procedures etc.). By trying to establish points of interface with numerous clinical specialties, 22q advocates hope to have both more patients referred for 22q11.2DS testing and otherwise non-indicated forms of evaluation and treatment made available to 22q11.2DS patients.

Like almost every parent of a child with 22q that I have met, Kambin’s is a dramatic story about the trying process of receiving a diagnosis, whether it is a prenatal finding that transforms parents’ outlooks for the future or an adult diagnosis that brings with it the poignant question of ‘What if?’. While some parents receive a diagnosis pre- or neonatally, the experience of ‘diagnostic odyssey’ is the norm: a range of medical and/or developmental challenges in early childhood and often beyond – from minor and trivial to life-threatening and profound – amounting to a bewildering and stressful experience for both the patient and their family. A 22q11.2DS diagnosis, it is generally agreed, ends the odyssey by providing a single and encompassing explanation and the opportunity to meet other families dealing with the same condition, if not the same clinical manifestations.

Indeed the capacity of a genetic anomaly of the kind that can lead to genomic designation to put an end to a family’s diagnostic odyssey is considered one of the major benefits of genetic testing when it comes to children with multiple medical and/or developmental

! 315! ! ! challenges (Ledbetter 2009; Bales, Zaleski, and McPherson 2010). As Costain et al. write in a piece in the Journal of Intellectual Disability Research, and whose last author runs a major 22q11.2DS clinic in Toronto and is a central actor in the 22q Foundation:

… diagnostic certainty alone appears closely linked to psychological benefit, even when uncertainty remains about the extent of expression of a particular genetic anomaly in any particular individual. This may be overlooked by clinicians who equate diagnostic benefit solely with improved medical management. Identification of a 22q11.2 deletion after early childhood provided many adults with 22q11.2DS and families the satisfaction of an explanation for their lifelong challenges, sometimes after years of fruitless investigations (the ‘diagnostic odyssey’). As for autism, genetic findings that explain ID, schizophrenia or other stigmatised neurodevelopmental conditions may be particularly valued for their explanatory value. (2012, p. 648; references omitted)

That is, a genomically designated diagnosis provides both psychological and social benefits for patients and their families that are not obviated by their sweeping clinical range and cannot be reduced to their potential medical benefit. An explanation is thought to be beneficial in its own right.

However, Kambin’s speech does not include the next chapter: the bemusing experience of reading voluminous material about 22q and its legion clinical and psychological associations, the challenge of getting their local clinicians to embrace the

22q diagnosis and its attendant risks and referrals, and then the experience of finding other

22q families who have faced kindred challenges, even when they are in fact markedly different. One parent poignantly described the experience of reading about 22q and attending her first conference to me as one of ‘mourning for the future’ as she met seriously affected patients and learned about the various problems that her son may or may not encounter as he grows up and on into adulthood. By the time I met her, however, the

22q11.2DS diagnosis and the community of people united by it provided a rubric for understanding whatever lay ahead and a network of support to help her cope with it.

! 316! ! !

The move towards newborn screening

Kambin’s speech above closes by referencing a bold move that the foundation has made: an appeal to have 22q11.2DS added to the US newborn screening regime. If her son was not tested for years, despite exhibiting a number of the relevant signs in what is probably the leading center for 22q research and treatment in the world, then how can 22q ever hope to even approach complete ascertainment? As Kambin put it herself:

My son, Aiden's diagnostic odyssey incorporated 27 specialists over a 5-year period at major medical centers. Despite having 18 findings associated with 22q, Aiden remained undiagnosed. The cost was over $500,000, but what cannot be measured in dollars is Aiden's lost chance for early interventions, interventions which I believe could have substantially improved his prognosis. What would Aiden's IQ and speech be like today if he had come to attention in infancy? We will never know. I am a parent. I am also an obstetrician/physician who has coped with her son's medical diagnosis by medicalizing every aspect of it. I can recite every anomaly associated with the syndrome. I also work on a special delivery unit which was built to deliver babies with congenital anomalies, specifically babies with congenital heart disease. And I came here to tell you today that I could not reliably make this diagnosis in a delivery room. Newborn screening is the only solution to this complex problem. Please do right by these wonderful children and recognize having newborn screening for 22q.

Kambin was speaking alongside a group of leading 22q11.2DS advocates, clinicians and researchers at the US Department for Health and Human Services’ Secretary’s Advisory

Committee on Heritable Disorders in Newborns and Children (SACHDNC) meeting in

January 2012. While their first bid failed, they were not discouraged: first attempts to add a condition to the screening kit are rarely successful, they were invited to submit further evidence at future meetings, and it was clear from the meetings and the conversations afterwards that allies were won and goals were refined as a result of the presentations.

This is a bold but perhaps brilliant claim on state resources by 22q11.2 activists: no other genomically designated conditions are on the screening kit, and with 22q’s high

! 317! ! ! prevalence, sometimes life-threatening medical associations and broad phenotypic variability, they have both an enormous amount to gain and a real chance of success. At the

International 22q11.2 Foundation’s 2012 conference, their president discussed what a neonatal diagnosis might have done for her severely affected son, before explaining how early detection and awareness are two sides of the same coin. Indeed she opened her presentation, ‘Making Newborn Screening a Reality: Why Parents Care’ by telling the audience, “Immediate detection is the most important, which will [bring] rapid intervention, optimal care, and global awareness.” Donna McDonald-McGinn made the point even more clearly after receiving a kind of lifetime achievement award from the foundation:

So how can newborn screening help us along this journey as we move forward? Because newborn screening will lead to early detection for both deletion and duplication, lead to proper care in the newborn period, lead to early interventions as needed, and provide accurate prevalence figures which will lead to greater awareness by the public, greater understanding by the medical community, more research dollars by government agencies in search of better treatments and outcomes.

With significant rates of cardiac defects and hypocalcemia, for which early intervention can undoubtedly save lives and resources, and a test that can be affordably integrated into the existing heel punch, there is a real case to be made for adding 22q11.2DS to the newborn screening kit.

However, the reasons both to and not to include it bear a striking resemblance to the debate that swirled around the newborn study of XYY, discussed in Chapter 3, undertaken at Harvard in the mid-1970s but then abandoned in the face of bitter opposition.

Take this selection from Bales et al.’s summary in Genetics in Medicine of some of the benefits and risks involved with adding 22q11.2DS to the Wisconsin newborn screening kit (2010:140):

! 318! ! !

Benefits(( Risks( ( ( Consideration!of!adding!screening!provides! Individual!values!and!principles!ignored! impetus!for!development!of!effective! (some!families!may!prefer!not!to!know!of! inexpensive!screening!for!a!potentially! medical!conditions!for!which!urgent! serious!problem!that!now!is!often! treatment!may!not!be!needed)! unrecognized! ! ! Early!interruption!of!parent/child! Societal!benefit!of!delineating!full! relationship!(effects!on!bonding,!stress)! phenotypic!spectrum!and!natural!history!of! ! 22q11DS! May!set!precedent!for!other!syndromes! ! ! Early!detection/treatment!for![medical! Possible!detection!of!defects!that!would! issues]! resolve!without!treatment! ! ! Early!intervention!for!developmental!delay!! Risk!of!labeling!child!who!might!prove! ! mildly!affected! Early,!timely!recognition!of!treatable! ! complications!(i.e.,!mental!illness!and! Concerns!of!creating!“vulnerable!child! learning!disabilities)! syndrome”!for!mild!cases! ! ! Prevent!“diagnostic!odyssey”!for!families!! Phenotype!varies!and!difficulty!to!predict! ! presence!or!absence!of!features,!therefore! Recognition!of!familial!cases,!recurrence! cost!associated!with!interventions!that!may! risk!counseling! not!be!necessary!for!every!patient! ( ! ( Anxiety!with!false!positives!or!mild!cases! ( not!requiring!urgent!treatment! !

I do not want to advance the comparison too forcefully – many factors make XYY Syndrome in the 1970s and 22q11.2DS today disanalagous, not least the frequency and severity of serious medical complications and the wildly divergent social circuits within which they are mobilized. Nevertheless, we see among the benefits the delineation of the full phenotypic range of the disorder (i.e. an unbiased sample) and early intervention, which were the primary objectives of Stanley Walzer’s XYY screen in Boston; among the risks, we see the fact that consent will not be required, the interruption of parent/child relationships, the risk of labeling the mildly affected and the creation of “vulnerable child syndrome”, overtreatment and

! 319! ! ! unnecessary anxiety for mild cases, which all bear strong similarities to the objections to

XYY screening raised by Science for the People and others. In short, screening for a genomically designated condition whose total population may include very mildly affected people has both benefits and drawbacks that transcend the very different profiles and social contexts of XYY in the early-1970s and 22q11.2DS today. This is why reiterated facticity is so useful for our analysis of genomic designation: we see how deep continuities in the objects of knowledge generated by human genetics nevertheless give rise to strikingly different networks, practices, episodes of contention, meanings and indeed kinds of people under new historical conditions and according to contrasting repertoires of mobilization.

However the network of 22q11.2DS research and advocacy has much to gain from the fact that newborn screening is likely to pick up large volumes of mildly affected cases. Not only will it allow for early intervention for both mild medical issues and developmental challenges, but it will also reveal what they believe to be a much higher prevalence than currently estimated and therefore allow them to reap the rewards that come with being a common disorder. What’s more, as the leader of the main 22q group in the UK put it, she was there to support the “International Foundation who are like vanguards who drive things forward, and I can only hope that when we will get Washington to agree, the North Atlantic

Drift will bring it over to Britain.” Finally, by absorbing cases of 22q11.2 duplication into their constituency, the foundation is moving to gain in numbers even if it means absorbing even more clinically variability – duplications of the same region tend to lead to a milder phenotype, though research on the subject is far less developed (Courtens, Schramme, and

Laridon 2008; Ensenauer et al. 2003; Pagon et al. 2009; Portnoī 2009; Wentzel et al. 2008).

As the foundation explains on their website, “Many people with the duplication have no

! 320! ! ! apparent physical or intellectual disabilities. Often times, parents of children with the duplication find out that they also have it only after their child is diagnosed.” The page then goes on to list many of the same clinical associations as the deletion.6 Whereas Timmermans and Buchbinder (2010, 2012) have shown how the ascertainment of mild or near- asymptomatic cases of conditions included in California’s newborn testing regime requires clinicians to engage in ‘bridging’ work between genotype and phenotype (see above), for a genomically designated condition like 22q11.2DS the ascertainment of mild cases who may never have been diagnosed holds huge potential and presents an empirical rather than an ontological challenge.

Overcoming and interfacing with the old nosology

Complicating matters further, consider that 22q11.2DS’s biased profile is partly born of its complex nosological relationship to the older diagnoses DiGeorge Syndrome,

Velocardiofacial Syndrome (VCFS) and others, discussed in Chapter 2. That history, in turn, creates further obstacles to the accrual of 22q11.2DS diagnoses. Even though we saw that

‘22q11.2 Deletion Syndrome’ is very much in the ascendancy when it comes to the nomenclature of biomedical research (see Figure 6 in Chapter 2), DiGeorge and VCFS remain common names for diagnosis upon receiving a positive test for a 22q11.2 deletion.

Meanwhile,VCFS is still used by a number of advocacy organizations. Indeed, the American- based Velo-Cardio-Facial Syndrome Educational Foundation, Inc. (VCFSEF) remains a major player in the field and the relationship between their leadership and that of the International

22q11.2 Foundation is strained, to say to the least. At the center of the VCFSEF is Robert !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 6 See: http://www.22q.org/index.php/what-is-22q/overview

! 321! ! ! Shrintzen, who published the foundational paper delineating VCFS in 1978. The origins of the dispute are contested, but they go back to an attempt to collaborate on an NIH grant proposal in the early 1990s that went awry. In the years that followed the 22q group was established and each side has repeatedly accused the other of deception and attempts to undermine their work. While some experts have been able to maintain good relations with both groups, key people on either side readily offered recriminating tales and opinions about one another in our conversations. Shprintzen has written about ‘The Name Game’ (Shprintzen

1998) and in a speech entitled ‘Science not Politics’ at the VCFSEF annual meeting I attended in New Brunswick in 2011 he excoriated the very idea of calling the condition delineated by the 22q11.2 deletion 22q11.2DS. To him, the fact that 22q lacks a clinically descriptive element and the fact that other conditions that have been associated with a mutation, like

Williams Syndrome, had not been so renamed made the use of 22q11.2DS unnecessary and counterproductive (though he was at pains to point out that he held no attachment to VCFS as the name he introduced to the world, or to the once-frequent use of “Shprintzen Syndrome”).

Others on the 22q side object that VCFS is misleading, insofar as it is descriptive, as people with 22q11.2DS can have neither velopharyngeal insufficiency (V), cardiac defects (C) nor the typical facial phenotype (F), and as we saw in Chapter 2 even Shprintzen now considers a

22q11.2 deletion to be necessary and sufficient for diagnosis with VCFS. In short, even though genomic designation has won the day in terms of nosology, when it comes to the actual name of the condition delineated and diagnosed according to the 22q11.2 microdeletion the old clinical nosology continues to represent an impediment to its development and growth as a kind of person.

! 322! ! ! The International 22q11.2 Foundation considers the division and confusion caused by multiple syndrome names to be a serious impediment to increased awareness and unity among the population, as well as a barrier to parents and clinicians identifying the full range of information and resources available for 22q11.2DS. As a result, the Foundation has launched a ‘Same Name Campaign’, headed by Brenda Finucane, with the goal of getting clinicians, publishers, parents, media and everyone else besides to drop the older clinical terms like

DiGeorge Syndrome and VCFS in favor of 22q11.2DS. They have sent postcards to journals, genetic counselors and others, lobbied professional organizations and convinced groups like the Dempster Foundation and the UK’s main group, Max Appeal, to adopt 22q11.2DS in place of their previous use of DiGeorge Syndrome. As Finucane explained the situation to a group gathered for a conference on 22q11.2DS and education in Indianapolis in 2011:

So I think the name problem that we have is because the history – these were things that were described before we had the DNA technology to understand what was causing them. They were described based on the features that were being seen, and Velocardiofacial Syndrome, when it was described, and DiGeorge Syndrome, people thought these were two completely different syndromes, et cetera for all the other syndromes that we know. So VCFS and DGS and the other names of other syndromes that associate with the deletion, they are older terms that described specific, clinical – what are called phenotypes or patterns of features, prior to the discovery of the 22q11.2 microdeletion. However, we are smarter now, and genetically we know there's no detectable difference in the microdeletions between people with VCFS versus those with DGS or the other syndromes – it's the same thing. There's no difference between the child with the DGS diagnosis and VCF. Now, could they be very different children? Of course they could because there's a lot of variability in the syndrome, but genetically it is all just one big continuum or spectrum, all due to the same underlying microdeletion, and as Donna put it I think very well, the one thing they have in common… is the underlying microdeletion.

So here we have the nosological consensus – again, the VCFS group accepts that two ‘very different children’ can have this condition if and only if they have the deletion – that relegates clinical observation to the level of symptom and carves up disease strictly according to genetics. Initial optimism that variations in the breakpoint of the deletion would explain the

! 323! ! ! contrasting phenotypes proved to be mostly incorrect, while new ventures to examine gene-x- gene interaction effects with other genomic regions are in their early phases. Whatever the outcome, the goal is not to re-partition 22q and it is certainly not to validate VCFS and

DiGeorge Syndrome as independent conditions. The older, lingering clinical categories are therefore treated as a kind of historical efflux.

However, that does not mean that they have no impact on the field today, as Finucane pointed out:

What are the implications of this? Well, the issue is that we have got several different names being used for the same condition – it creates tremendous confusion not just among families but also the professionals out there… it further divides the resources for an already rare syndrome, so you know, even though it is common and we talk about it being common, it's still an unusual syndrome in the general population, and when you have a syndrome where you are trying to band together resources, to divide it into these little pockets of, you know, money for DGS, money for VCFS, et cetera, it kind of divides these efforts, and it hampers unified efforts toward research and support… You really can't move forward and get to a place where other syndromes are in terms of dollars and research until we get the whole gang together and move all forward in the same direction.

As long as this situation persists, Finucane noted, 22q11.2DS would never attain the status and resources it deserves. By contrast, “even though [22q] is so common, and it is even more common than let's say, Fragile X Syndrome, which has much more momentum for targeted pharmaceuticals, they're a multi-million dollar organization for kids with Fragile X. We're nowhere near that… a lot of it has to do with the name.” If the foundation could get families and especially professionals to ‘think 22q’, then eventually children would cease to be diagnosed with the older names and awareness, money and other resources would come more easily to 22q11.2DS. We therefore see how 22q advocates are acutely aware of the challenges represented by the more longstanding clinical nosology discussed in Chapter 2, and in the

Same Name Campaign they have embarked on an explicit mission to ensure that the

! 324! ! ! genomically designated 22q11.2DS reconfigures the field and does away with the older symptom-based terminology. I have spoken to numerous parents at both 22q and VCFSEF conferences and they generally express frustration at what they see as an unnecessary and costly split in their movement. There have also been some moves towards rapprochement with VCFSEF in recent years, with each organization appearing on the other’s website and key figures attending each other’s events. Furthermore, with Shprintzen retiring and with the biomedical literature and important actors like the Dempster Foundation increasingly favoring

‘22q11.2DS’, perhaps we can expect the International 22q11.2 Foundation to establish hegemony in the field. Time will tell.

Whatever the outcome, the uneasy relationship between 22q11.2DS and the older syndromes that it has mostly subsumed and supplanted is indicative of the way genomic designation, as a form of classification, can be impeded by its relationship to older clinically derived categories or terms. In some cases, as with autism and both Fragile X Syndrome and

Phelan-McDermid Syndrome, a productive interface between genomic designation and clinical conditions can be achieved that actually facilitates knowledge production and advocacy on genomically designated conditions. However, each attempt to achieve commensurability or to establish hegemony with respect to clinical classification will likely bring with it its own risks and opportunities.

Indeed we need look no farther than the same speaker at the same conference to see how 22q11.2DS experts and advocates attempt to achieve commensurability and a productive interface with other kinds of phenotypically designated diagnoses. As Finucane explained to the audience in Indianapolis:

! 325! ! ! Psychiatry, psychology, and genetics, really, over the past many decades, have kind of grown up independent of each other and have really developed their own ways of looking at developmental differences… and unfortunately, up until relatively recently, these three disciplines really didn't talk with each other, so they were all doing their own thing and not really doing a lot of cross-fertilization, so as a result we've come up with different diagnostic systems, and you have one child in the middle of that, may get multiple diagnoses depending on who is looking at the child and what the background is.

On the one hand, she told the audience that when three kids all present with a learning disability, obsessive-compulsive disorder and anxiety disorder, “from the school's point of view, these three kids all have the same thing, and yet when you do the math, so to speak, this little boy has Smith-Magenis Syndrome, this child has Williams Syndrome, and this little guy has the 22q11.2 Deletion Syndrome. Three totally unrelated genetic disorders.” She explained that psychiatry, psychology and genetics categorized people according to very different kinds of observations, and that for schools it is primarily the psychiatric diagnoses that determine education plans and eligibility for services. This meant that, for pragmatic reasons, 22q parents should not eschew psychiatric diagnoses:

Educational and behavioral diagnoses – not the genetic diagnosis – are generally the ones that determine eligibility and services within the school setting…. We know that kids with the 22q11.2 Deletion Syndrome, they often meet criteria for one or more behavioral or educational diagnoses, and I would say to use these diagnoses for everything [they’re worth] realizing and keeping in mind and trying to educate people that 22q is the underlying cause and these other things are all symptoms of it. But don't be afraid to use those other diagnoses.

In short, while 22q advocates are essentially waging war on DiGeorge Syndrome and VCFS as diagnostic categories, when it comes to the conditions that shape mainstream practice in schools and hospitals the goal is to establish an interface rather than hegemony.

That said, the same kind of logic that applies to the medical diagnostic odyssey discussed above is also applied to the accrual of behavioral and cognitive diagnoses – a logic that attempts to both accommodate the practical necessity of symptom-based classification

! 326! ! ! and ascribe a kind of ontological primacy to genomic designation. Thus Finucane told the parents and professionals in Indianapolis:

…psychiatric diagnoses are really important, but they are different from etiological diagnoses. So here are some of the psychiatric diagnoses that we often see in school-age children…: oppositional defiant disorder, obsessive-compulsive disorder or OCD, learning disabilities, ADHD, bipolar, intermittent explosive disorder, et cetera. Now, in our experience at Elwyn, and Elwyn is primarily a school system in many respects – [an] organization that works with children and adults to help them with their educational, behavioral, vocational needs. Very often we see children and adults with many, many different diagnoses… practically speaking in the States, this is the kind of thing we see happening, and in the school system, if you pick up the chart on a 17-year-old, you will see six and seven different diagnoses because every time a different symptom constellation comes up, or as the child matures and behavior differences happen, then you are starting to see different diagnoses. The child may be diagnosed with bipolar disorder and then later on may be diagnosed with anxiety disorder – maybe because he has too many diagnoses and now he is really anxious. [audience laughing]

Where then does a diagnosis like 22q11.2DS fit into this picture? Finucane continued:

But this is not unusual, and so then you have got, hovering in this [diagnostic] soup here, also the 22q Deletion Syndrome, and the reaction of a parent when she finds out, yet again, that her child would get another diagnosis [audience laughing] it is pretty understandable I think, you know, after a while you're kind of like, “What is going on? Every year it seems to be a different thing, every year we are told a different diagnosis.” And the bottom line is this: a lot of times parents, and even professionals in education systems will say to me, “This poor kid, he ended up with seven different things. Oh my gosh, how could you get struck by lightning seven different times?” The bottom line is he didn't get struck by lightning seven different times. This child has one underlying condition – the 22q11.2 Deletion Syndrome which results in patterns of behavior and learning that as the psychiatric and educational diagnostic systems we can classify by these other names.

In other words, a genomically designated diagnosis transforms multiple comorbid or shifting conditions into simply a series of symptoms caused by a genetic disorder. It may not yet be widely recognized in schools and hospitals, but a condition like 22q11.2DS allows parents to recast a shifting, stressful and confusing series of medical and behavioral diagnoses both big and small into simply an evolving spectrum or pattern of symptoms caused by a single genetic disorder.

! 327! ! ! Crucially, the point is not simply to realign the understanding of patients and their parents, but for both parents and experts to work towards getting school districts around the country to recognize and appreciate the value of genetic classification:

So it may be on you to educate education systems as well about this… I do a lot of consultations all over the country dealing with school districts, and we often will start out with a little presentation kind of like this. I'm always floored in school districts where it's like light bulbs start going off over people's heads. They never realized this, that these genetic diagnoses that are now starting to come in are actually the underpinnings for many of these other behavioral diagnoses, and that is why it is important: because it is kind of like we've gotten down to the bottom line.

Finucane explained how it was not necessarily a problem that “Schools and teachers are very unlikely to be familiar with the syndrome” because “they don't use this diagnostic system.”

Nevertheless, she explained that this “does not necessarily mean that they can't provide excellent services.” She explained how even schools that had never knowingly dealt with a case of 22q11.2DS before were often able to learn a lot about it and put excellent, tailored programs in place: “An open mind, willingness to learn about 22q, and a creative approach to meeting a child's needs are just as important – if not more important – as experience with this syndrome.”

Visit by visit and child by child, the network of 22q experts, parents and activists are seeking to make 22q11.2DS a diagnosis that matters in institutions organized around different classificatory systems. They work to create commensurability – a framework in which categories born of qualitatively different forms of classification can nevertheless inform and redirect one another – such that 22q can profoundly shape every aspect of a patient’s treatment, care and support, even in fields that are not used to dealing with genetics. In sum, as a closely linked set of actors and organizations the 22q11.2 field is striving towards the greater numbers, awareness, resources and alliances that will make it a more powerful kind of

! 328! ! ! person. Doing so, however, is not simply a question of accruing diverse resources. In order to really matter in the world, genomic designation in general and 22q11.2 in particular must forge commensurability with existing forms of classification and diagnostic categories while also realigning clinical judgment and redirecting practice.

Care – Realigning and redirecting clinical practice

We have seen how 22q11.2DS experts and advocates are working towards increased numbers of diagnosed patients, detailed biomedical knowledge and clinical guidelines, as well as started to glimpse the institutional infrastructures that make it matter to patients and their families. One major aspect of their efforts is the formation of local alliances of parents, genetics experts and clinicians who in turn lobby local hospitals to create multidisciplinary centers of excellence for 22q11.2DS. Given the range of resources and expertise required, these centers tend to be formed at major research hospitals and the foundation is working to promote their creation and on criteria to ensure that these centers meet certain minimal standards. The goal, as Donna McDonal-McGinn explained, is that everyone should be within driving distance of a dedicated 22q center capable of providing counsel to the newly diagnosed and their families and meeting the diagnosis’ many clinical demands. Referencing the hundreds of families who have traveled from far afield to the 22q and You clinic in

Philadelphia to receive such care, she proclaimed: “no one should have to come to CHOP.”

But when patients find themselves in a specialist center upon diagnosis with

22q11.2DS, what are the implications for their clinical evaluation, treatment and prospect horizons? First, what forms of medical evaluation and treatment are recommended upon diagnoses with 22q11.2DS? As the 22q11.2 Foundation’s newly refurbished website (made

! 329! ! ! possible by a grant from the Robert Wood Johnson Foundation) puts it in their ‘FAQs’7 section:

How should individuals with the 22q11.2 deletion be followed medically?

Ideally, children with the 22q11.2 deletion should receive coordinated care from centers that offer multidisciplinary teams of clinicians, often drawn from more than 20 specialties. Centers address each child’s individual health problems, as well as issues such as speech or learning delays, in order to help these children and their families lead the best life possible. Upon initial diagnosis, the standard assessment and work-up for all ages generally includes: • Cardiology • Endocrinology • Immunology • Speech/Language/Developmental Assessments • A Renal Ultrasound (to check the kidneys) • X-rays of the neck (in children old enough to cooperate and where the bones are well ossified – so ~ 3 to 4 years of age)· • Deletion studies in both parents when available

Thereafter, the work-up is individualized, depending on the symptoms, but may include any or all of the following: • Plastic Surgery/ENT/Audiology • Gastroenterology/Feeding Team • Hematology • Urology/Nephrology • Orthopedics • Ophthalmology • General Surgery • Dentistry • Rheumatology • Neurology/Neurosurgery • Psychiatry

Any given 22q11.2DS patient may require treatment from many or all of these clinical specialties, or they may require none, but with 180 associated symptoms there is certainly plenty to worry about. (Conversely, a patient can of course have all of the most commonly associated symptoms and not, in fact, have 22q11.2DS.) Even for mildly affected cases, however, a 22q11.2DS diagnosis realigns clinical attention in such a way that problems are

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 7 See: http://www.22q.org/index.php/what-is-22q/frequenly-asked-questions

! 330! ! ! likely to be found even when they may not have been considered problems at all in the absence of a 22q diagnosis. Indeed as ascertainment broadens and becomes less biased towards patients presenting with cardiac and palatal issues we can expect 22q to realign clinical judgment and the ascription of abnormality in a growing array of people.

After all, when it comes to care for genomically designated conditions like 22q11.2DS, we are confronted with a challenging question: care for what? The tautological answer – care for the symptoms caused by a 22q11.2 microdeletion – does not get us very far. For unlike other diagnoses that are made according to biomedical observations, genetic mutations do not have any immediately apparent physiological specificity. It is not a question of a clinical sign, nor is it even a question of an observation that may predispose someone to a clinically significant sign at a later time. To repeat, as Donna McDonald-McGinn put it to me in an interview (2011): “22q is 22q… it's a spectrum from no symptoms to every malformation under the sun.” So while 22q11.2DS has an enormous number of clinical associations – around 180 and still growing – having 22q11.2DS is simply a question of having a deletion at

22q11.2. You may be at risk of any number of medical or developmental pathologies, but if you are healthy you are not just asymptomatic or a carrier, you are mildly affected.

Indeed conditions like 22q11.2DS challenge the idea of incomplete penetrance: the term for a genetic mutation that does not always produce the associated phenotype or disease entity. In my conversations and interviews with experts I heard of numerous parents who were found to have 22q11.2DS only after their more seriously affected child was diagnosed, though unfortunately I have not been able to formally interview them. What is striking about these cases is that, even though they usually did not have any clinical signs that would have led to a referral for a 22q11.2DS test, they all considered themselves to be affected by the disorder

! 331! ! ! rather than carriers who passed an inheritable condition on to their children.8 One such woman had been fairly successful and healthy throughout her life, with a graduate degree, a good job and no serious medical problems, but when she found out she had the 22q11.2 deletion like her more severely affected son she took it to explain her relative lack of success, stature and light complexion as compared to her high-flying, tall, blond siblings. Another parent had struggled in high school and had nasal speech, and felt a sense of relief and exoneration when she found out about her 22q status. In both cases it helped to explain their pasts and who they are today. I heard about the older sibling of a seriously affected child who was absolutely delighted to learn that he too had 22q11.2DS, immediately embracing the idea that he could be an even more powerful advocate as a mildly affected case than he could as a mere sibling. Their parents, however, did not want to find out which one of them had passed on the deletion (though after seeing a family picture the experts who were there when this story was related had little doubt as to which one it was based on their facial features).

Incidental findings like these9 lead 22q11.2DS advocates to believe that the real prevalence of the disorder is in fact significantly higher than the standard 1 in 2000-4000 estimate, and provide further impetus to pursue newborn screening.

But why treat people with a 22q11.2 deletion who do not even have any of the major clinical associations as anything more than carriers or asymptomatic cases? A major part of the answer has to do with the way that genomically designated conditions serve to realign

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 8 It is thought that around 90% of 22q11.2DS cases are the result of de novo mutations. However, parents with 22q have a 50% chance of passing it on to their child so it is a heritable condition even though it is not usually actually inherited from a parent.

9 Other incidental findings that are thought to support the idea of substantially higher prevalence include identical twins where one is mildly affected and several pairs of cousins with de novo cases of 22q.

! 332! ! ! clinical judgment. At the 22q conference held in Orlando last year, I saw a highly instructive talk by Laird Jackson from Drexel University and Children’s Hospital of Philadelphia on a

NICHHD prenatal study he had participated in. The study had tested 4,400 fetuses using microarray. Speaking shortly after the appeal for newborn screening, Laird explained:

“Prenatal diagnosis in the main is somewhat similar to newborn screening in that the majority of mothers who qualify for a prenatal diagnosis do not have any prior suspicion of a clinical syndrome.” Indeed in this study 3000 had not been referred for testing on the basis of clinical suspicion. What did they find? Of around 25 anomalies observed at 22q11.2, 10 were in the form of the ‘classic’ 22q11.2DS 2.5mb deletion, and:

In the classic deletions, 7/10 were indicated… by ultrasound imaging, usually a tetralogy or PSD, but three of the ten classic deletions were in otherwise normal pregnancies with no fetal indication and nothing to signal a change except what was found by the microarray. So that was about 3000 of the 4400 patients, giving us an incidence, in this study, of 1 in 1000 since three out of three thousand had the classic deletion.

That is, they found an incidence of 1 in 1000 for non-indicated pregnancies and 7 in 1400 for ones with an indication that there may be complications. As Laird explained, “you’ve already heard that there’s a great deal of variability, and this is found throughout the things that we’ve found with CNVs – copy number variations –that you see on the microarrays. And they are obviously found sometimes in apparently normal [people].”

What would this mean for our understanding of the syndrome itself and for genetic disorders more generally? For Jackson, and indeed for most researchers interested in genomically designated conditions, the failure to meet clinical thresholds of significance does not mean that people are unaffected by the mutation in question or simply ‘carriers’ of the disorder. Rather, traits that are within the clinically normal range can still be thought of as symptoms of the disorder:

! 333! ! ! So there [is a term] that geneticists use for this: incomplete penetrance, if you can't really find anything clinically that you would say is affecting the individual. But really, that is just a lower-hand of a variation of expression which goes across a spectrum, and if you are able to look, instead of looking at qualitative dichotomous traits – either you have it or you don't – but look at continuously distributed quantitative trait, you would see that there is a Gaussian distribution of such traits.

Jackson proceeded to discuss the example of intelligence quotients, and explained how the fact that the distribution was shifted to the left in 22q11.2DS meant we could treat the substantial proportion of patients whose scores fell in the normal range as still being affected by the 22q11.2 deletion.

This is the logic that sees ‘an IQ of ~10 points lower than an unaffected sibling’ become a key feature of a condition like XYY Syndrome (see Chapter 3). As Jackson put it:

[I]n developmental delay, there are some tools if you can use to measure IQ, and you could use those to show that in the normal population you have a distribution like this which artificially is centered at an IQ of 100 with some people above that and some people below. And we define, arbitrarily, intellectual disability when it falls two standard deviations below the mean of normal intelligence. And in some conditions such as Down Syndrome or the Fragile X Syndrome, most of the individuals affected by this type of disorder have clearly recognizable intellectual disability although even some of those overlap in distribution of normal intelligence. In other conditions maybe there is a different sort of distribution so that here is a curve illustrating the distribution of IQ in persons with Deletion 22q11.2 Syndrome indicating that the gene is always working there, but it is working variably, and the distribution mean of these individuals is in fact two standard deviations below our assumed mean of normal intelligence.

In other words, the characteristics of people with genetic conditions like 22q11.2DS, even when those characteristics do not meet the threshold for clinical significance, are almost automatically coded as symptoms of their genetic disorder. Laird’s is simply an unusually explicit articulation of the implicit logic that guides clinical judgment for genomically designated conditions. He is calling for a redistribution of clinical judgment that can accommodate symptoms that would not have been considered symptoms in the absence of a

! 334! ! ! genetic anomaly. To take the example above, an IQ of 85 in a person with 22q11.2DS can be considered a symptom of the disorder – as the result of a ‘gene effect’ – even though it is not even on the borderline of the usual threshold for clinical significance. So, Laird concluded, his study suggests “there probably is a higher incidence or prevalence [of 22q11.2DS] than is commonly assumed and the gene deletion effect probably operates all the time, but with variable consequences, which is up to us to understand.” As ascertainment increases the caseloads and observed clinical range of people with genetic disorders, the case of 22q11.2DS is indicative of the capacity of genomic designation to realign clinical judgment so as to incorporate and potentially medicalize hitherto benign forms of difference.

Furthermore, genomic designation can redirect clinical judgment. Consider again the case of IQ in 22q11.2DS. One commonly noted feature of 22q11.2DS’s cognitive phenotype is the frequent divergence of scores on the Wechsler IQ test’s verbal and performance subtests, rendering full-scale IQ scores invalid. As Donna McDonald-McGinn put it to an audience of parents at a conference geared towards development and education in 22q:

[U]sing the appropriate Wechsler IQ test we found out about 18% of kids had an IQ that was in the average range, 20% below average, 32% borderline, and then 30% in the mentally retarded range… and so you know, these are not really too bad in terms of numbers. But we found that the full-scale IQ generally doesn't reflect the individual's cognitive ability. So how so? Well, for instance in this child, he had a full-scale IQ of 87, but his verbal IQ was 111 and his performance IQ was 65. So verbal are things like reading and rote memorization and performance is math and abstract reasoning, things like that. So this is an incredible split which really renders that 87 invalid.

After finding the verbal/performance IQ split in this particular child, what did McDonald-

McGinn and her colleagues do? They conducted further such tests on other children to see whether his case pointed towards an important feature of 22q11.2DS more generally:

So after we saw this child we went back and looked at other kids and we found a greater than 12-point split between the verbal performance IQ in 65% of patients. So therefore, the full-

! 335! ! ! scale IQs don't necessarily represent these patients accurately, and their verbal and performance IQs should be considered separately, and again, this has direct ramifications for cognitive remediation in this population because only 1% of the general population has that kind of split. Everybody else has a verbal learning disability here, not that some kids don't have deficits in that area, but the schools are not generally prepared to deal with this, and so it is important, you know, that you know whether your child has this kind of split or not, and that the remediations are set up specifically for your child and not for, you know, the whole classroom.

Again, we see how parents are encouraged to delve deeper into a child’s IQ score on the basis of the fact that they have a 22q11.2 deletion in such a way that a full scale IQ of 87 – well within the otherwise normal range – can be disaggregated in order to reveal a serious deficit in a more specific phenotype: performance IQ. While the 12-point verbal-performance split is only found in 65% of 22q patients, McDonald-McGinn implores parents to make sure that educational and remedial services are reoriented around this deficit in contradistinction to the usual focus on language deficits. In short, 22q11.2DS is mobilized to redirect clinical judgment so as to find specific deficits that would otherwise have been unlikely to come to expert attention or inform intervention. In this way an observation that would have put a child squarely in the normal range of variation can be recast because of our knowledge of

22q11.2DS, finding a deficit in need of individualized intervention that would otherwise have gone unnoticed.

What’s more, this relative strength in language is observed despite the usual developmental trajectory of 22q11.2DS. As McDonald-McGinn explained in the same talk:

But across the board kids have delays in emergence of language, with a mean age of talking at two-and-a-half which is, you know, most kids are talking pretty well by eighteen months, two years… We have had kids who were seven who didn't say a word and then suddenly were speaking in full sentences. It is very interesting, um, and it is something that deserves a lot more investigation. Why do we have these delays in language and then it becomes a strength because the kids, when they are older, are – are really eloquent speakers, so that is something we have to pay attention to. And then you'll hear that there's a subset of kids with frank autism or autistic-like features, and no question, all benefit from early intervention strategies, and the

! 336! ! ! sooner we can get kids into speech therapy and OT, PT, et cetera, the better things will move along.

So not only is there this specific cognitive phenotype where language becomes a relative strength by school age, but there is also a particular trajectory that goes from delayed speech to relative strengths in reading and language. Nevertheless, despite enormous variability parents are told that all of their 22q11.2DS-diagnosed children will benefit from multiple forms of early intervention. A 22q11.2 deletion therefore constitutes a mandate to realign the normal and the pathological, but also to justify scrutiny and intervention where it would not have otherwise been called for.

While not nearly as extreme as in the case of Williams Syndrome, presented in

Chapter 4, 22q11.2DS’s verbal-performance IQ split is but one of the more concrete ways in which distinctive cognitive phenotypes are identified in genomically designated syndromes: the aversion to eye contact in Fragile X, the anxiety associated with 22q11.2DS, the upper- body spasmodic tic in Smith-Magenis Syndrome to mention just a few. Indeed the ‘upper- body spasmodic tic’ is illustrative: as Finucane explained to me, she published a paper detailing that particular phenotype (Finucane et al. 1994) only after teachers became aware of their Smith-Magenis students and reported what they described (and indeed Finucane still used in conversation) as ‘this adorable self-hugging thing they do when they’re happy’. While their characteristic destructive outbursts would certainly get a Smith-Magenis proband into the psychologist’s practice absent a genetic diagnosis, it is these more subtle traits that are observed in high proportions of genomically designated diagnoses that comprise the specific cognitive and behavioral phenotypes that would certainly never have been clinically delineated. Fixity at the level of the genome allows not only clinicians, but also other

! 337! ! ! caregivers like parents and teachers to identify subtle and probabilistically distributed phenotypes that may otherwise have remained unidentified or, at best, considered an individual particularity rather than part of a specific disease profile. They can then make variegated observations on the basis of genomically designated syndromes that realign the normal and pathological, from low growth that is actually normal for 22q (see below) to an IQ of 87 that masks an important, gene-based deficit.

As Tony Simon, a cognitive neuroscientist at UC Davis’ MIND Institute who now specializes in 22q11.2DS, put it to me in an interview (2012), early work comparing Down and Williams Syndrome showed how:

the early developmental trajectory separates and they are just completely disparate, unrelated [laughter] genetic populations that create two different kinds of subspecies of human being. There's a sort of real developmental emergence of these phenotypes and so I think, you know, that really started though with more of that understanding. And to me, it is extraordinary. You see it in these meetings and other meetings about genetic disorders.

To be clear, Tony Simon is no genetic determinist. Our conversation took place after a talk about the work being done to explain 22q11.2DS’s variable phenotype according to the deletion’s interaction with SNPs and haplotypes, i.e. gene x gene interaction. Simon is close to many of these researchers, but he told me, “I am sitting there going, huh, you know, these people are alive… do they not even realize that these people are alive, you know? There are other things in the world.” In fact it is precisely the detailed attention to the patterns of these living, thinking, environmentally embedded people with genomically designated conditions that makes them potentially so powerful as categories for understanding and acting on difference. Having first worked on 22q11.2DS at CHOP in Philadelphia, it was coming to the

MIND Institute with its cutting-edge research programs on autism and Fragile X Syndrome

! 338! ! ! that really allowed Simon to develop a dual research-clinical environment in which new aspects of the 22q phenotype could be discovered.

Seeing more children, Simon recounted, meant that “suddenly after a little while we began to see that there was a pattern here, and in fact, if you go back and look in the literature, mostly studying slightly older kids, it was already out there.” He explained:

What we noticed is that the kids were predominantly very, very anxious. The literature says 60%, but fewer than 20% were coming through our doors with any recognition of anxiety. What we gradually began to see is, you know, I'm sitting down with the parents and they'll say, “My child won't go to school, my child has stomach aches and headaches and picking at their skin and they're OCD and they can't sleep at night.” It's like, “OK, that's not seven different things, that's all anxiety.”

And that anxiety, recognized in 22q11.2DS children by specialists like Simon even though it had not been previously identified, is caused by the specific cognitive phenotype of 22q with its differential capabilities. The experience of being at an age-appropriate level for reading and other subjects but potentially several years behind in subjects like math may induce the high rates of anxiety seen in 22q11.2DS and, Simon reasons, contribute strongly to the high rates of schizophrenia seen in later development. This stressor can in fact be exacerbated as much by the relative strength as the weakness, such that Simon has found that “we see that IQ is not predictive of adaptive functioning at all in this population. It is almost a 1:1 correlation in regular, typically developing people, and even in other neurodevelopmental disorders it is high. Here you have a zero correlation.” It is therefore vitally important to devote time to

“training the parents a little bit to adjust their expectations and their emotional responses to the kids because the parents are a critical part of this [and] can change what we now call the behavioral ecology of the child. We think we can change the developmental trajectory.”

! 339! ! ! Simon’s work sits on the boundary of basic science and treatment, marrying a realignment of clinical judgment and the development of specific cognitive profiles with the creation of points of commensuration or interface with existing diagnoses and treatment programs. Trained as a basic scientist, Simon rhetorically asks, “Who am I to talk about this stuff? But this is what happens with translation of science, and that is what the MIND did to me.” He explained:

So I gradually began to have a firm team set up… so suddenly we had this interdisciplinary team: a developmental-behavioral pediatrician, a psychiatrist, a clinical psychologist; and I suddenly kind of woke up in the middle of the night one night and said, “Gee, you know, my model is Randi [Hagerman]. Fragile X, it comes back to the thing it always comes back to: Fragile X being the model.” She had a treatment in a research center. That is what I wanted to develop.

This creation of an interdisciplinary team dedicated to studying and treating the behavior and cognition of children with 22q11.2DS was aided by the participation of Nicole Tartaglia, a leading sex chromosome aneuploidy researcher (including XYY) who “was really well- trained in doing this kind of work with Fragile X, working with Randi and the sex chromosome aneuploidies.” This kind of resource can mean a lot for a 22q patient and their families:

… within a very short period of time, there was this life-changing experience because the parents could come and bring their kids to our study and they would get an evaluation by the very first pediatrician they'd ever seen who didn't go, “Huh?” When they told them that their kid had 22q. [laughter] And so that was wonderful for them.

It is no coincidence that this kind of innovative interdisciplinary team was assembled at the MIND Institute. It was originally founded with funding from three wealthy ‘autism dads’ in order to bring innovative research approaches to bear on neurodevelopmental disorders. For some time it has been the leading center for research on Fragile X Syndrome and at the core of the FXS-autism nexus discussed in Chapters 4 and 5. Even though 22q11.2DS is not as

! 340! ! ! strongly associated with autism as other genomically designated conditions and is more likely to be taken up as a model for schizophrenia, as we will see below, Simon’s research and treatment center has thrived at MIND. He explained why he and his group have had so much success in both research and treatment:

So it's the fact that it's an integrative team approach. The MIND was put together, you have a huge multidisciplinary, but more importantly, I think it was interdisciplinary team of completely different people. So we've got people who are epidemiologists and environmental health people and immunologists and blood genomics people and neuro-anatomists and blah blah blah blah blah. So it's all – we're all focused on one thing: neurodevelopmental disorders, so they all have a behavioral manifestation in the end, but there's a variety of biological causes, so it is all about endophenotyping as widely and broadly, as interdisciplinarily, as wide as you can possibly get. So essentially what happened was the MIND created the environment for me to do what the MIND is supposed to do. I think it is an incredible success story that it simply by being there, I was able to literally – you know, it changed my career, research direction because it educated me on how that works, and simply provided the conditions for us to be able to do it… And now we have a real clinic. We call it the 22q Healthy Minds Clinic.

Today, they bring in people who do not qualify for the research studies in order to provide treatment, and they are starting to have success in getting people referred to the MIND for 22q assessments through their insurance plans. They now combine detailed interdisciplinary workups of every child, but also an expertise that allows Simon to say, “I've seen several hundred kids with this disorder. The first 20 seconds I go, ‘Oh, I think I see roughly where this kid fits.’” They then provide contact details for experts from the family’s local area and often conduct follow-ups down the road. Later that evening, as I was talking to Simon during a reception, numerous children with 22q11.2DS would excitedly approach and often embrace

‘Dr. Tony’ while their parents told me how wonderful he was. And yet, were it not for

22q11.2DS he may have never truly worked with a patient in his career. The scientific opportunity represented by a genomically designated condition like 22q11.2DS, combined with the hybrid research/clinical institutional set-up pioneered by Fragile X researchers at

! 341! ! ! MIND, have allowed Simon to move in directions he says he never could have imagined.

From an esoteric interest in the cognitive basis of the human capacity to deal with numbers

(e.g. Simon 1997), he now maintains active research interests in 22q11.2DS and autism and what amounts to a clinical practice dedicated to cognition and behavior for children with

22q11.2DS that he repeatedly noted had completely changed his life. He summarized his new, cherished role with what he calls ‘a sad little joke’: “I am not a doctor, but I play one at work.”

While the MIND Institute may provide singular opportunities, that does not mean it cannot play a broader role within the 22q network. Families visit from far afield, learn about their children and about treatment strategies and take that knowledge back to their local care providers, schools and communities. Simon is found at almost every 22q conference and there is a constant stream of trainees from various fields passing through his clinic and other centers at MIND. Alongside medical centers like the one at CHOP and institutions like Elwyn, these central nodes in the 22q network allow for not only the production and diffusion of knowledge, but also the proliferation of expertise. What’s more, experts who have trained at places like CHOP and MIND are often the same experts who work with families to set up the new interdisciplinary centers of excellence for 22q that are now scattered all over the US and

Europe. To be sure, certain infrastructures and expertise need to already be in place, and yet in each instance they must be reorganized in order to work for 22q11.2DS just as rubrics clinical attention and judgment must be recalibrated according to its particular phenotype. Finally, jurisdiction over diagnosis must be ceded to geneticists as genomic anomalies take precedence over clinical observation as the basis for delineating disease. After all, for Tony and Eliot Simon (no relation) to explore and master the cognitive phenotypes brought into

! 342! ! ! view by the ‘synthesizing variables’ represented by genetic mutations, they both must accept that their expertise can only flesh out and treat a condition, never actually diagnose it.

In this way, even though they are not clinically specific enough for diagnosis, deep phenotypic commonalities are observed even between clinically very distinct patients. One constant refrain is that kids with 22q ‘look like siblings’, even though the craniofacial phenotype is variable as well as mild and unlikely to be picked up by anyone without dedicated 22q11.2DS expertise. This is also the case for at least several other genomically designated conditions, though some craniofacial phenotypes are considered highly specific

(e.g. Williams Syndrome) and others considerably less so. Indeed facial appearance is often the only feature that is thought to indicate a particular genetic disorder and, even when not specific, it is often the basis for referral for a genetics evaluation (Hammond 2007; Morelle

2007; Shaw et al. 2003). Still, only the observation of a genetic abnormality is sufficient for a genomically designated diagnosis. Furthermore, at least in 22q11.2DS, the fact that it is the facial phenotype that often leads to testing results in decreased ascertainment of non-

Caucasian patients as their craniofacial features are less distinct (or so say the mostly-white doctors who fail to recognize them).

The post factum observation of distinct phenotypes extends beyond behavior and cognition. 22q11.2DS is also distinct when it comes to other characteristics. Take the example of growth patterns. Comparing children with 22q11.2DS to the standard Centers for Disease

Control and Prevention (CDC) growth charts is considered to be invalid. For while children with 22q11.2DS tend to fall into the very low end of the standard distributions of height and weight in their first years, they also tend to catch up before they reach adulthood (Habel et al.

2012). This point is made to parents at 22q/VCFS meetings in order to assure them that their

! 343! ! ! children’s low growth in their early years compared to the standard CDC chart is not necessarily a cause for concern, and that they may indeed catch up given the divergent growth trajectory observed in 22q11.2DS. As Robert Shprintzen put it at a VCFSEF meeting for parents and professionals in New Brunswick in 2011:

Now, remember, in growth issues, children with VCFS have a different growth velocity. I have had conversations with a number of parents during the course of this meeting, all of whom are having their children compared to the CDC growth charts. The CDC growth charts are normed against the general population, the normal population. People with Down Syndrome, people with Williams Syndrome, people with achondroplasia [a common form of dwarfism], people with VCFS should not be compared with the CDC growth charts, all right? They're – you are seeing a number of adults with VCFS walking around, you know, during the course of this meeting, all right, I mean, [holding hand flat around waist level] are any of them this tall? OK, now that's anecdotal, I know that, but I know hundreds of adults with VCFS and they are all in the normal range of height and people with VCFS, they get there on a different path than other children. So their group velocity is different but their eventual growth goals are essentially the same as the general population, and we have data that supports that and I reported on that last year in Salt Lake City.

Again, even Shprintzen considers VCFS to be coextensive with the 22q11.2 deletion. He also gestures to the same point about divergent growth patterns in other genetic disorders like

Down and Williams Syndromes (achondroplasia, by contrast, is specifically a growth disorder and a leading cause of dwarfism). So while 22q11.2DS and other genomically designated conditions may not be diagnosable or even indicated by low growth, the post factum observation of different growth patterns in probands can still become part of their distinctive phenotypes. Thus the recasting of expectations and interpretation for people with 22q11.2DS when it comes to IQ and growth charts, which David Armstrong cites as a quintessential tool and image of twentieth century surveillance medicine (Armstrong 1995:396–7), is indicative of the way genomic designation can realign the normal and the abnormal.

But what do parents do with all this information when they leave the annual conference or the specialist clinic, and how does 22q11.2DS redirect clinical practice in the

! 344! ! ! vast majority of institutional settings that are not designed for it and with practitioners who are not familiar with it? While a good geneticist or pediatric cardiologist will likely have some familiarity with 22q11.2DS, most general practitioners, pediatricians and specialists in other fields do not. Yet on the medical front alone, a diagnosis like 22q11.2DS brings with it risk factors for a plethora of ailments ranging from the mild to the debilitating and across a host of bodily systems: scoliosis, cleft palate, endocrine and immunological imbalances, eye and dental problems, constipation and so on. Nevertheless, parents are actively encouraged to follow the clinical guidelines for 22q11.2DS as faithfully as possible, requiring referrals, evaluations and sometimes treatments that are not indicated by the usual clinical standards that inform most medical practice. It is therefore often left to parents to try to redirect clinical resources according to 22q11.2DS, even though they usually have no formal medical expertise. Parents therefore engage their local practitioners with assistance from summary sheets provided by the foundation and the associated biomedical literature, with the Journal of

Pediatrics guidelines serving as a major resource in this regard. In my interviews with parents, this struggle to acquire the evaluations and care they have learned their children need comes up continually. To take one example, in an interview with three mothers of children with

22q11.2DS they all agreed when one explained:

Well, they trusted me because I knew about 22q and I would tell them all about that… I usually could walk in there telling them, “OK, this is what the findings show.” I mean, I would walk in with literature, and I did that just on Friday, he had an eye appointment with an eye doctor he hadn't seen before and I pulled up the website and I said, “These are common eye problems in children with 22q, here is my smart phone, please take a look,”… just like I did for the diagnosis with human growth hormone deficiency, they said, “Oh, well, he is way too young to have that, it is not soon enough,” and I said, “He is not on any growth chart that you have got” …and sure enough, you know, they found two problems. So I think that it helped, having the literature showing, you know, human growth hormone deficiency is present in children with 22q, and I think it also helped having a good collaborative relationship.

! 345! ! ! In some cases, a credentialed 22q11.2DS expert or a genetic counselor is enlisted to negotiate referrals or courses of treatment, but in many others parents armed with the right resources

(pertinent biomedical papers, good insurance, cultural capital and so on) are able to redirect resources according to knowledge about the implications of a 22q11.2 deletion.

Indeed even though most parents lack professional credentials, they nevertheless often gain and employ impressive levels of lay expertise. I asked that same mother whether she felt that her expertise in 22q, combined with resources like the recently published clinical guidelines in Pediatrics, has meant that she could interact with medical professionals in a different way – as a kind of complementary medical expert. She replied:

Definitely. Definitely. And oftentimes – and I have walked in with a book and they've said, “I don't have time to read it,” and I said, “Fine, here's pertinent information for what we are dealing with now. This is the age range, this is the findings.” When I walked in and told my pediatrician, you know, they are now recommending cervical spine X-rays, he looked back in the charts and said, “Well, we don't have that, we'd better get them then.”

Of course, this takes capital on the part of the parents. That same mother had been to 22q conferences all over the country, visited specialist clinics and even traveled from Florida to

Sacramento to have her son evaluated at Tony Simon’s aforementioned 22q center at MIND.

Her son, who hours earlier had excitedly talked to me about all the books he was reading, did indeed fit the Simon’s finding that children with 22q experience anxiety as a result of their relative weakness in math:

… we took him out to Tony Simon, MIND Institute, and they told me, “Yeah, he is a pretty good case for ADHD. He has got anxiety, generalized anxiety,” which I had no clue of because the school didn't think that it was relevant that every time math came up he developed a severe stomach ache and had to go run into the clinic every day because they said, “Oh, well, we knew he was just faking it so we didn't want to worry you about it.” … But thanks to the MIND Institute, we got him, you know, in to see a therapist.

! 346! ! ! Hence this recognition of anxiety, caused by the particular cognitive phenotype of 22q11.2DS, can lead to adjustments in expectations on part of the parents and entrée into standard psychiatric treatment for anxiety disorders. It is clearly an instance of medicalization-via- genetics, but it comes from a non-medical expert (even if he does ‘play doctor’ at work) at an institution founded by ‘autism dads’ and it empowers parents with a new resource as they seek to garner care and support for their children. With sufficient (but by no means abundant) cultural and financial capital, this particular mother has been able to intervene in the clinical and educational evaluation of her son and receive referrals and treatments that would likely not have been sought or obtained in the absence of a 22q11.2DS diagnosis.

Hers is one of countless such stories I have heard, and as 22q11.2DS grows in scope and strength as a network composed of experts, advocates, published papers, tests, centers and so on, less and less capital is required on the part of any given family in order to make the observation of a 22q11.2 deletion transform the way a person is understood, treated and cared for. It is also a micro-example of the network accruing ties at the local level – with schools, various clinical practices, labs etc. – according to the work done by 22q researchers and advocates to make 22q11.2DS commensurable with clinical classification. Only with the various resources built up through hybrid collaboration and collective action can they interface successfully with the massive fields of clinical medicine, education etc. and their numerous constituent fields.

Cure

In contrast to the extensive work being done to detect 22q and create infrastructures for treatment and care, there is very little work aimed at curing 22q11.2DS. As the piece by

! 347! ! ! Bales et al. succinctly states (ibid, p. 138) “Because the missing genetic material cannot be replaced throughout the body, there is clearly no cure for 22q11DS. Once the syndrome is identified, the goal is detection and management of treatable complications.” In other words, cure is neither currently possible nor on the horizon; care is the only practicable goal.

Nevertheless we can see the idea of cure as a coordinating device that helps to justify the many biomedical studies on 22q11.2DS that require tissue samples and other interventions. Indeed, we do see elements of the same kind of interface that have been achieved by networks like the Fragile X and Phelan-McDermid Syndrome groups with autism research and advocacy, though for 22q they are far less powerfully developed. Researchers may be attracted to 22q11.2DS as a model for congenital heart malformations like tetralogy of fallot, psychiatric disorders like schizophrenia, and increasingly with autism, although I have seen parents and experts both question whether the autistic behaviors exhibited by children with 22q11.2DS warrant an ASD diagnosis. As a recent paper entitled ‘The 22q11.2 microdeletion: Fifteen years of insights into the genetic and neural complexity of psychiatric disorders’ (Drew et al. 2011:259; see also, e.g. Karayiorgou, Simon, and Gogos 2010) put it with respect to schizophrenia:

The identification, fifteen years ago, of 22q11.2 microdeletions as a causative factor for schizophrenia was the first demonstration that deletions, or duplications, of chromosomal regions may play an important role in schizophrenia etiology in many cases of the disease. Such mutations, known as copy number variations (CNVs), result in altered gene dosage and a rapidly expanding literature shows a high prevalence of such mutations in schizophrenia patients. By allowing the generation of etiological valid animal models, identification of highly penetrant mutations such as these offer unprecedented opportunities to determine the key neural changes that can lead to psychotic illness. Here we argue that the 22q11.2 deletion can act as prototype for this type of investigation. That is to say, the rigorous definition of the pathway from mutation to disease phenotype in models of this mutation will provide invaluable insights into the etiology and pathophysiology of schizophrenia as a whole. (my emphasis; references omitted)

! 348! ! ! As a former CHOP/22q11.2DS researcher who now works for the same company that has pioneered Fragile X/autism pharmaceuticals told the crowd in Disney World, genetically specific disorders are like a ‘low hanging fruit’. This phrase, which came up in Chapter 4 and repeatedly during my fieldwork, is indicative of the fact that the primary goal is not to cure the specific genetic disorder but to work towards insight and perhaps cure for the more common conditions for which it is supposed to serve as a model. While that may have secondary benefits for a condition like 22q11.2DS, given the range of genes deleted and bodily systems affected as well as the resources required to develop a cure for even most single-gene disruptions, cure should be seen more as a coordinating device to facilitate research or a rhetorical, abstract ambition than a concrete goal.

Discussion

I have argued that genomically designated conditions have the capacity to realign clinical judgment and redirect practice. This chapter therefore confirms Rabeharisoa and

Bourret’s aforementioned finding (2009. p. 697) that, “Contrary to the discourse on the

‘geneticization’ of diseases… the introduction of genetics into oncology and into psychiatry does not by any means amount to reductionism,” but rather “genetics reinforces the complexity of pathological categories.” I also found that conditions like 22q11.2DS compel researchers, clinicians and advocates to work to “invent or reinvent a clinic capable of working on these multiple and complex pathological entities.” However, contrary to

Rabeharisoa and Bourret’s findings, we have seen that mutations outside of oncology can also give rise to “a robust corpus of clinical observations” and therefore “clinical entities with enough robustness to be mobilized in medical judgments and decisions.” (ibid) Rabeharisoa

! 349! ! ! and Bourret cite a presentation wherein “a child psychiatrist reviewed his collaboration with the pediatric geneticists, and reported on 26 ‘diagnoses’… Apart from 11 diagnoses of Fragile

X and some isolated diagnoses of rare syndromes, the ‘diagnoses’ mentioned locations of chromosomal abnormalities.” (707) At most, this group discussed these chromosomal abnormalities as potential ‘binding objects’ between pathological categories, rather categories in their own right. However, we can see how Rabeharisoa and Bourret’s distinction between the apparently unproblematic category of Fragile X and the uncertain “’diagnoses’ [that mention] locations of chromosomal abnormalities” (i.e. most of the syndromes discussed in this dissertation) is primarily a product of collective action on the part of Fragile X advocates rather than its biological status. Rabeharisoa and Bourret were undoubtedly correct with respect to their particular field site in France several years ago. In other settings, however, disease categories named after chromosomal locations have achieved both the status of a diagnosis sans quotation marks as well as a knowledge base that enables them to inform and redirect clinical judgment and attain profound meaning for the patients and families to whom they are delivered.

In short, in order to understand how biomedical objects of knowledge like genetic mutations can become bona fide clinical entities in psychiatry and fields of medical practice other than oncology, we cannot stop at “the clinical work that makes the development and performance of biomedicine possible.” Rather, we have to examine the networks of knowledge production and advocacy that span labs, clinics, repositories, protocols, textbooks, families and foundations as well as far-flung locales. No single ‘bioclinical collective’ could hope to produce the knowledge base for a kind of person like 22q11.2DS. Rabeharisoa and

Bourret do note that debates about the nosological status of mutations create ‘strong tension’,

! 350! ! ! particularly with respect to an episode where “The head of the medical genetics department argued during a staff meeting that he would like ‘to find the molecular marker, to be able to say to the parents: “This is what your child has” and not to haphazardly put labels on her or him’.” (708) For us, this particular medical geneticist might be likened to McKusick’s strident

1966 statement about the nosological role of genetic observations cited in Chapter 3: a strong statement in favor of genomic designation that lacks an understanding of the complex networks that need to be built in order to turn a ‘molecular marker’ with a diffuse phenotype into a robust category of clinical and social practice. As I argued in the previous chapter, diagnostic categories have a threefold social character as identities, coordinating devices and sites of looping. In order to make genomically designated syndromes into all three, a complex network needs to be assembled. This chapter therefore contributes to the literature by examining the way knowledge and practice with respect to genetic mutations like the 22q11.2 deletion are generated in settings that transcend particular biomedical platforms and bioclinical collectives. It shows how genetic mutations can be mobilized, not only in oncology clinics concerned with different kinds of malignant growth, but also in the many-varied settings concerned with different kinds of people.

Furthermore, my analysis of genomic designation also challenges the idea that ‘postgenomic’ medicine precludes a radical reorientation of nosology according to genetic mutations in favor of an emphasis on genetic heterogeneity (Rabeharisoa and Bourret 2009, p.

697). Rather, the observation of genetic anomalies can give rise to new medical conditions with their own clinical profiles and hybrid networks of research and advocacy in a manner that does not obviate the possibility of an epistemic model that focuses on complexity. On the contrary, fixing disease categories to genetic mutations is shown in the case of 22q11.2DS to

! 351! ! ! enable not only the embrace of complexity on the part of multidisciplinary ‘bioclinical collectives’, but also new forms of clinical observation, judgment and treatment that would otherwise not have been possible. In other words, the case of 22q11.2DS shows how genetics can marry a move towards claiming etiological and nosological primacy with a realignment of clinical judgment that in many ways facilitates more complex forms of observation with respect to organic and psychological phenotypes.

To be sure, getting the full 22q experience – receiving the diagnosis, interacting with medical practitioners, visiting a leading center like CHOP, regularly visiting regional centers that may still be far away, attending conferences, accessing information online and so on – often requires significant financial, social and cultural capital on the part of parents. However, that is decreasingly the case as the field itself accrues capital (Bourdieu 1984). For as 22q advocates establish more centers, garner awareness, accrue allies and clinically-oriented journal articles like the aforementioned Pediatrics piece, and as testing becomes cheaper and more available, 22q11.2DS requires fewer resources on the part of individual families in order to be taken up as a powerful category of illness. Furthermore, we have seen how geneticists and others are less likely to recognize ‘funny-looking’ non-Caucasian kids and refer them for genetic testing. Finally, there are no specialist clinics for conditions like 22q11.2DS, as far as

I know, outside of highly-developed countries, though there are research reports on diagnosed patients in countries like India (Gawde et al. 2006; Halder et al. 2010), Thailand

(Ruangdaraganon et al. 1999) and Chile (Repetto et al. 2009; Guzman et al. 2012). Thus, as a kind of person, 22q11.2DS is certainly not (yet) readily available to the majority of people with a 22q11.2 microdeletion. Nevertheless, as a network of research, clinical practice and advocacy 22q11.2DS continues to make great strides, accruing resources and ties to broader

! 352! ! ! fields that are making it an increasingly powerful and practicable category of human difference.

I argued in Chapter 2 that the actor-network theoretic perspective helps us to understand the role of the 22q11.2 microdeletion as an actant that reconfigured clinical nosology, and throughout I examine 22q11.2DS as a heterogeneous network composed of various experts, scripts, activists, technologies and the 22q11.2 deletion itself, to name just a few of the most pertinent actants. However, examining genomic designation simultaneously within the rubric of reiterated facticity allows us to take into account elements that cannot be considered to be part of the network per se, but which nevertheless help comprise the conditions of possibility for a genomically designated syndrome to matter. For one thing, we have seen how a loose-but-distinct ‘invisible college’ (Crane 1969) epitomized by actors like

Finucane is working to advance the recognition and practicability of genetic disorders in general and genomically designated conditions in particular.

Furthermore, the network of actors more or less directly concerned with 22q11.2DS is enabled by exogenous shifts in broader, independent fields, from increased receptivity to genetic diagnosis and commercial investment in new genetic testing techniques to deinstitutionalization and the proliferation of new communications technologies. Indeed we should not underestimate the affordances of new genetic testing kits: not only do they make testing more accurate and affordable, but as we saw with the diagnosis of Sheila Kambin’s son the capacity to scan the entire genome, obviating the need to target particular chromosomal loci according to clinical suspicion, may dramatically increase and also broaden ascertainment of conditions like 22q11.2DS. The accrual of clinical relevance and social power on the part of genomically designated syndromes therefore depends on developments

! 353! ! ! in the many related fields that have taken place over the course of the last fifty years. But it also requires the agentive work done to establish contact, exchange and interface with various fields such that conditions like Fragile X, 22q13DS and 22q11.2DS can really matter in the world. Reiterative facticity allows for comparison within and across cases, and therefore for the explication of the shifting networks of actors organized around genomically designated syndromes as well as the historical conditions and processes of collective action that make those formations possible and transform genetic mutations as facts.

The categories born of genomic designation must establish themselves with respect to the existing, primarily clinical nosology and the social formations that have been built up around it. While the challenges represented by the older clinical nosology are perhaps unusually acute in the case of 22q11.2DS, given its complex relationship and competition over the same constituency of patients with DiGeorge Syndrome, VCFS et al., they are nonetheless widespread. That said, as we saw most clearly in Chapters 4 and 5, when it comes to broader surveillance categories – especially cognitive and behavioral conditions like intellectual disability, autism, ADHD and so on – genomic designation functions most powerfully when conduits of interface and exchange can be established. Despite attracting considerable attention as a potential biological model for schizophrenia, 22q11.2DS has not established an alliance nearly as powerful as the ones seen earlier in this dissertation between autism and Fragile X Syndrome or Phelan-McDermid Syndrome. However, with its high prevalence and cacophony of clinical associations, 22q experts and advocates have been able to forge multiple points of interface spanning several fields of research and clinical practice – cardiology, psychiatry, behavioral psychology, ENT, endocrinology and so on – in the form

! 354! ! ! of journal articles, specialist interdisciplinary centers and parents convincing local clinicians to modify their practice according to knowledge about 22q11.2DS.

I argue, with Timmermans and Berg (1997), that each implementation of a standard requires the creation of ‘local universality’. As they put it, “Local universality emphasizes that universality always rests on real-time work, and emerges from localized processes of negotiations and pre-existing institutional, infrastructural, and material relations.” New standards, they argue, do not negate local specificity but rather require a process of

“transforming and emerging in and through it.” (ibid, p. 275) Crucially for us, this argument applies equally to a new form of classification, the categories it generates and the protocols associated with those novel categories. For example, a condition like 22q11.2DS will only be diagnosed when it has achieved traction and commensurability within certain clinical specialties, been included in genetic testing kits or made its way on to an existing institution like a newborn screening regime. Furthermore, implementing best practice for 22q requires that existing resources in any given locale are enrolled in the larger 22q network: while a small group of 22q11.2DS experts in a specialist clinic may serve as a new point of mediation between a patient and clinical services, and even though they may reconfigure the network of clinical expertise brought to bear on a patient with 22q11.2DS, they nevertheless rely on the resources and expertise available at any given hospital or locale. However, when certain standards or forms of classification have not achieved widespread recognition, as is the case with genomic designation, local universality entails even more work towards translation such that it often requires overt collective action in the form of patient/disease advocacy.

But how does a network like 22q11.2DS grow in size and strength such that it can truly impact practice and lived experience? As Timmermans and Berg argue, the

! 355! ! ! implementation of universal knowledge or practice is dependent on ‘distributed’ activity and therefore a wide range of actors and actants. In our case, a local coalition of perhaps a few parents, a genetic counselor and one or two interested clinical specialists can truly make a condition like 22q11.2DS a powerful category of understanding and practice in a given locale, while a new test may make its detection far more practicable. Indeed I have often found genetic counselors and patient advocates playing vital roles at the center of networks geared towards research and care for genomically designated syndromes, not least because they help erect the social infrastructure for both clinical treatment and further knowledge production. To be sure, their efforts would likely be futile were it not for the biomedical literature available on 22q11.2DS, but at the same time that literature ceases to be merely esoteric only insofar as it is mobilized by actors seeking to remake the world in its image.

In other words, the conditions of possibility for genomic designation are both historically and locally specific. They require both the agentive work of diverse actors seeking to mobilize and reorient existing resources as well as the conditions under which they can access the repertoires and resources that make such a project practicable. A condition may establish scattered bastions in which to develop and flourish – Elwyn Services, Children’s

Hospital Philadelphia, Great Ormond Street Hospital in London, the MIND Institute, to take a few examples for the case of 22q – that in turn enlist new actors like the previously skeptical

Eliot Simon (discussed above) and developing professionals undergoing their training before moving on and bringing their newfound expertise with them. Again, knowledge and practice about a condition like 22q11.2DS does not simply undergo a process of diffusion (even or otherwise); rather, actors work within existing structures must work to forge a larger, stronger network across space as well as within and between institutions, disciplines and so on. One

! 356! ! ! implication is that a 22q11.2 microdeletion has very uneven chances of being detected even within the same highly developed polity with access to the same genetic testing kits and biomedical literatures. Furthermore, even once detected there is enormous local variability in capacity of knowledge about a 22q11.2 microdeletion to reconfigure clinical judgment and practice, though parents with sufficient capital can overcome a local paucity of resources by tapping directly into the larger network. In sum, new kinds of classification and the categories they delineate are just as dependent as standards upon distributed, local action, and perhaps even more so. So, therefore, are new kinds of people.

Bringing these various approaches together allows us to make sense of way the categories born of genomic designation can constitute what I called in the Introduction, drawing on the work of Ian Hacking, a new kind of human kind. On the one hand, we saw in

Chapter 3 how it emerged quite rapidly and unproblematically in the field of human genetics as what Foucault would call a ‘discursive formation’ once it became possible to observe abnormality at the level of our chromosomes. However, while we can perhaps talk about the earliest genomically designated syndromes as would-be kinds of people, we saw how they remained too thin to impact practice or inaugurate significant levels of what Hacking calls looping. In short, very little could be done with genomically designated syndromes outside of human genetics, with the notable exception of terminating pregnancies on their basis. They were facts in the domain of human genetics, but they were not really medical facts, psychological facts or social facts.

On the other hand, I have used the framework of reiterated facticity to outline the hybrid forms of collective action, research and network formation that have seen some genomically designated conditions achieve the status of both fully-fledged kinds of people

! 357! ! ! and what Foucault (1977) would call an apparatus of knowledge/power that can reconfigure the relationship between existing elements, respond to historically given needs and reorient our understanding of the normal and the abnormal. Becoming a practicable fact in diverse fields of research and practice, at least in the case of 22q, involved multiple processes at the intersection of biomedical research, clinical practice and social mobilization. Some of those were mostly exogenous to genomic designation: deinstitutionalization and the emergence of contemporary patient advocacy helped to make available the repertoires of collective action that have served conditions like 22q11.2DS so well; the looping processes at work in other kinds of people like autism created new opportunities for genomically designated conditions like Fragile X to forge productive points of interface with clinical medicine and further contributed to the material and symbolic resources available to them; advances in genetic testing and communications technologies have made it far easier to detect genetic mutations and to assemble networks on their basis; changes in related fields like genetic counseling and clinical psychology, the broad discursive power of genomics and the increasing focus on personalized medicine have all made it easier for genomically designated syndromes to develop as kinds of people. Other processes, as we have seen throughout this chapter, depend crucially on the agentive action on the part of experts and advocates laboring under the banner of particular genomically designated conditions. In sum, we have seen how different assemblages of actors, repertoires of action and conditions of possibility have reiteratively made and remade genetic mutations as facts.

Bringing together the sociology of science and medicine and a comparative-historical approach therefore helps us grasp not only how a new disease entity accrues salience and traction, but also how a new form of classification begins to reshape knowledge and clinical

! 358! ! ! practice. Through all of the diverse forms of action undertaken by 22q researchers and advocates a heterogeneous network is being assembled that can underwrite what Hacking

(1995, 1998, 2007) has called a new kind of person, and one that realigns and redirects existing frameworks for understanding and acting on human difference. Future work should attend to the processes of identity formation and looping at work in 22q11.2DS and other genomically designated conditions as ascertainment expands, biomedical research continues, clinical practice is redirected and patients and their families forge increasingly powerful networks of mutual aid and understanding. This chapter has focused on a prior question, namely how a network has been assembled such that 22q could become a bona fide kind of person, even though doing so necessitated a realignment of dominant modes of classification, judgment and practice. In sum, I have tried to show how knowledge about a genetic mutation can be used to carve out new kind of person capable of informing and also redirecting understanding and practice across a wide range of fields.

Conclusion

This chapter has drawn on a range of social scientific literatures in order to capture the way 22q11.2DS is being assembled as a new kind of person. It has built on the recognition that diseases can become the objects of hybrid movements that bring together biomedical experts and advocates and indeed blur the distinction between the two. It has sketched out the heterogeneous networks of expertise and advocacy that can turn an esoteric ‘discursive formation’ like genomic designation into an apparatus of power/knowledge like 22q11.2DS today (see Eyal 2013 for an analysis of the autism epidemic that similarly draws on Foucault and actor-network theory in tandem). It made recourse to my proposed framework of

! 359! ! ! reiterated facticity to examine the shifting repertoires of collective action and the external fields whose transformation over the last half-century constitute the conditions of possibility for genomic designation’s contemporary resurgence. Finally, I have tried to show how genomic designation, represented here by 22q11.2DS, constitutes a new kind of human kind that can realign clinical judgment and care through a dynamic interface with other forms of classification.

Perhaps this is a case of too many theoretical cooks spoiling the explanatory broth, but removing any one makes it harder to grasp the way 22q11.2DS has come to increasingly shape understanding and practice. Clinical medicine may have achieved a more rapid and decisive victory over its predecessor, at least as described in Foucault’s Birth of the Clinic with its more forthright theoretical argument (see Armstrong 2011 for a more circumspect account of clinical vs. symptom-based medicine). Genomic designation, by contrast, seems destined to grow indefinitely within the womb of the old if it is to thrive at all as a form of classification. It is precisely this exigency of establishing commensuration and points of interface with clinical medicine, as well as psychology, special education etc., which makes a case of genomic designation like 22q11.2DS so irreducibly complex and impossible to capture in one site or with one analytic framework.

We have seen how genomic designation has, in recent years, found a set of institutional structures and repertoires of collective action with respect to disease and childhood abnormality, professional and public orientations towards knowledge about our genomes and advances in genetic testing that have transformed its scope for producing powerful and practicable categories of human difference. Nevertheless, in order to gain traction a genomically designated condition like 22q11.2DS must be taken up by a

! 360! ! ! heterogeneous network of actors working to mobilize it across diverse spheres of practice. As a kind of person, it had to be assembled using multifarious elements that were not always intended to fit together, but instead had to be made to do so. This capacity to interface with divergent fields, while perhaps still tentative and partial, has transformed 22q11.2DS as an object of knowledge, practice and mobilization as well as the experience of illness for thousands of patients and their families. As they continue to make gains, it will be important to continue to monitor the networks of research, treatment and advocacy like the one underwriting 22q11.2DS as a kind of person, for what we can learn about both genomic designation in particular and the relationship between the biosciences, medical classification and collective action more generally.

! 361! !

Conclusion

So far, this dissertation has focused on genomic designation as a form of classification that rigidly designates categories of human difference according to specific abnormalities in the genome, as well as the contrasting kinds of clinical, social and self- practice that it has given rise to across different cases, places and historical moments. We have seen how it challenges and extends existing concepts in the social scientific study of genetics and medicine; the various ways in which it can reconfigure nosology; its early history as a practice confined mostly to the esoteric field of human genetics and the various kinds of human difference towards which genomically designated syndromes can be

‘leveraged’ as privileged sites of biomedical research. Treating diagnostic categories as having a threefold social status as identities, coordinating devices and sites of looping, we saw how genomic designation can achieve traction amidst shifting systems of clinical classification, and specifically how autism has looped to intersect with and thereby provide a fecund terrain for a growing number of genomically designated syndromes to garner resources and gain traction.

Our journey culminated with a discussion of how these novel categories of human difference – carved out at the level of the genome but lacking clinical coherence – have come to really matter to growing numbers of people. We saw how actors have worked to forge the heterogeneous networks that could turn a genetic mutation like the 22q11.2 microdeletion into an increasingly powerful apparatus of knowledge/power and a bona fide

! 362! ! kind of person. However, we also saw that they did so under conditions not of their own choosing or their own making: the contemporary power of conditions like Fragile X and

22q11.2 Deletion Syndrome as they exist today, in other words, supervene upon historically specific conditions of possibility. Using what I called ‘reiterative facticity’ as a comparative-historical approach, we traced the shifting landscapes and the many forms of research, network formation and collective action that turned genetic mutations from esoteric objects of knowledge into rich categories of clinical practice and social mobilization. What then, can we expect of genomic designation moving forward?

On the one hand, there are reasons to think that genomic designation will go from strength to strength as a form of human classification: as we will see below, there is certainly a resurgence of genomically designated conditions in the biomedical literature as well as increased acceptance in the field for ‘genotype-first’ discovery and diagnosis when it comes to children with developmental disabilities and perhaps further afield. The repertoires of collective action, established by HIV/AIDs, breast cancer and especially autism advocates paved the way for a model, pioneered by Fragile X Syndrome advocates, that has transformed the field environment for mobilizing genetic mutations into new kinds of people with expansive and well-funded programs of research, care and activism.

Genomically designated conditions are also more likely than ever to be taken up by biomedical researchers and pharmaceutical companies as models for more common conditions, and they are also increasingly giving rise to clinical protocols and specialist interdisciplinary centers. The genetic intransigence of most common conditions alongside the continued discursive power of genomics and the still-growing emphasis on personalized medicine all point towards the continued development of genomically

! 363! ! designated syndromes as objects of research, care and social action. Finally, continual innovation and penetration of communication technologies and especially cheaper and more accurate genetic testing techniques that do not have to be targeted according to clinical suspicion mean that more people will receive genomically designated diagnoses and be able to join communities on that basis. These developments and others give us reason to think that genomic designation will be a far more widespread and powerful way of understanding and acting on human difference in the years ahead.

On the other hand, genomic designation may remain quite marginal, confined to a handful of cases and could ultimately amount to little more than a biosocial oddity that, while still affecting thousands of people in an interesting case study for the sociology of science and medicine, peters out if and when the genome loses its aura as a privileged site for understanding difference.

However, I want to conclude by discussing a more disquieting possibility that has loomed larger and larger over the last few years: a secular increase in the prenatal ascertainment and termination of fetuses with genetic anomalies large enough to significantly decrease genomically designated population sizes moving forward. Prenatal testing for genetic disorders helps parents to make informed reproductive decisions and, for some, represents an important opportunity for society as a whole. Thus we saw Bentley

Glass speculate in his presidential address to the American Academy of Arts and Sciences in 1970 (1971:28) that humanity might embrace the power of scientific progress and use

“perfected techniques of determining chromosome abnormalities in the developing fetus to rid us of the several percentages of all births that today represent uncontrollable defects such as mongolism (Down's syndrome) and sex deviants such as the XYY type.” While

! 364! ! this did not come to pass in the years following Glass’s address, I will argue that new twenty-first century ‘techniques of determining chromosome abnormalities in the developing fetus’ are once again bringing this prospect into view. How so?

A long hypothesized but only recently proven technique now allows for non- invasive prenatal genetic testing in early pregnancy. The first such tests hit the market last year for Down and Edwards Syndromes, and both commercial uptake and refinement of the technique to include much smaller genomic abnormalities is expected to be swift. As I discuss below, this innovation in prenatal testing technology has the potential to serve as a massive exogenous shock to genomically designated conditions like the ones discussed in this dissertation. By significantly reducing their population prevalence and therefore the recruitment of activists, prenatal genetic testing and the selective termination of fetuses with genetic abnormalities has the potential to dramatically impact the way genomic designation unfolds moving forward. Some of the same historical conditions animating the rise of disorders like 22q11.2DS in recent years – the discursive power of genomics, the validity of genotype-first diagnosis, the development of complex clinical profiles and especially considerable commercial investment and innovation in genetic testing technologies – could combine to decimate populations of genomically designated conditions and undermine all the hard-fought gains that conditions like FXS and

22q11.2DS have made in recent decades. Indeed the very gains made by networks like the one that has formed around 22q11.2DS, for example the development of clinical profiles or their drive for inclusion on the newborn testing regime, may well contribute to the likelihood that a given condition is added to the new prenatal testing kits. A major question will be this: can the existing networks of researchers and advocates achieve an interface

! 365! ! with prenatal testing such that they can mitigate the potential loss of population and potentially even benefit from increased ascertainment and early engagement with the diagnosed and their families? The Down syndrome and sex chromosome aneuploidy organizations have already taken steps in that regard, with the former explicitly making reference to the risk represented by the new wave of prenatal tests.

By way of conclusion then, I will discuss this prospect and then reflect on some of the ethical, historical and theoretical issues raised by genomic designation as well as some of the work that remains to be done on this topic.

Terminating pregnancies – A lonely clinical constant

As we saw in Chapter 3, perhaps the one longstanding form of clinical intervention on the basis of genomic designation has been the prenatal diagnosis and termination of fetuses with genetic anomalies. With the development of techniques for karyotyping fluid extracted through amniocentesis in the mid-1960s (Steele and Breg 1966; see Rapp 2000 for the authoritative social scientific account of amniocentesis), genomically designated syndromes like Edwards Syndrome, Patau Syndrome and sex chromosome aneuploidies like XYY became routine objects of clinical diagnosis and intervention in the early 1970s

(see Harper 2006:162; Reinhold 1968).

Reliable figures for prenatal genetic testing and the termination of pregnancies on the basis of positive results are not available. Figures for the period discussed in Chapter 3 are particularly hard to come by, with the exception of scattered studies. To take a couple of representative examples, one study found that rates of amniocentesis in the state of Ohio for women over the age of 35 increased from 21 or 0.21% of total pregnancies to 2400 or

! 366! !

23.4% between 1972 and 1983, with population density being the main predictor of uptake

(Jewish women were also more likely to get tested) (Naber, Huether, and Goodwin 1987).

Meanwhile, in New York by 1980, 35.3% of women over 35 had prenatal cytogenetic testing, with a rate of 41.2% in New York City and 29.1% in the rest of the state (the equivalent rates for women below 35 were 0.76%, 0.93% and 0.62% respectively). 58 cases of Down syndrome were reported along with 48 cases of other aneuploidies including six XXX and four XYY (Hook and Schreinemachers 1983). The highest rates of prenatal genetic testing in these first decades appear to come from Denmark, where by

1979-80 Mikkelsen et al. (1983) could note that the Down syndrome births stood at 0.86 per 1000 compared to 1.17 when prenatally diagnosed cases were included.

The best American study comes from Verp et al (1988) who combined an analysis of all pregnancies in which a fetal chromosomal abnormality was found between January

1977 and June 1986 at Northwestern Memorial Hospital with a systematic review of existing studies available at that time. As they begin their paper (613-4): “Although when faced with the diagnosis of a fetal abnormality most couples do choose pregnancy termination, we have found that parental behavior is by no means uniform.” Of the 4,684 amniocenteses performed, 95% were conducted as a fetal chromosomal analysis and over

90% of those were for women over 35, with a small number of younger women receiving it “solely for parental anxiety.” (614) In addition, 243 patients received CVS testing between 1984 and 1986. In the group who received amniocentesis, 21 of 24 (87.5%) autosomal aneuploidies led to termination compared to 7 of 17 (41.2%) for sex chromosome aneuploidies. All seven aneuploidies detected by CVS were terminated (two

SCAs and five autosomal trisomies). In their combined meta-analysis of studies from the

! 367! ! late 1970s and 1980s, covering over 100,000 pregnancies, 96% of 554 of autosomal trisomies detected using amniocentesis were terminated compared to 67.3% of 281 SCAs.

In the smaller volume of studies on CVS testing, 109 of 110 (99.1%) autosomal trisomies were terminated as were 15 of 17 (88.2%) SCAs. Verp et al.’s explanation of this discrepancy is openly speculative, but, I think, cogent:

CVS patients appear more likely to terminate a pregnancy with a sex chromosome abnormality. This was true in our small series as well as in several other series (Table III). Possible reasons for this difference include 1) different patient populations choose CVS as opposed to amniocentesis, and 2) less reluctance to undergo pregnancy termination in the first trimester than in the second trimester. First trimester abortion procedures are simpler, safer, less expensive, and presumably emotionally easier in that the couple may be less attached to a first trimester fetus. In addition, family and friends may be unaware of the existence of a first trimester pregnancy, thereby affording the mother privacy regarding her decision. (ibid: 620, my emphasis)

As I will argue below, a similar logic suggests that with the introduction of a non-invasive test that 1) can detect the many genetic abnormalities smaller than full aneuploidy that are not as strongly predicted by maternal age, and 2) can already do so from the ninth week of pregnancy (see below), we should expect both uptake of prenatal genetic testing and terminations based on the diagnosis of genetic disorders, even when often quite mild, to increase dramatically.

More recent studies make it clear that both testing and termination rates have varied considerably according to individual-level characteristics, national and institutional context, time period and the genetic mutation in question (Hamamy and Dahoun 2004; Jeon, Chen, and Goodson 2012; Mansfield, Hopfer, and Marteau 1999; Shaffer, Caughey, and Norton

2006). Throughout, genomically designated conditions have been reported in prenatal testing results and included in screening programs. Edwards Syndrome, a devastating

! 368! ! condition whose probands rarely survive early childhood, has termination rates that are only rivaled by Down syndrome according to most studies. The sex chromosome aneuploidies (SCAs), which tend to be far less debilitating, are more mixed. Klinefelter and Turner Syndrome were well-established clinical entities before their association with

SCAs. XXX and XYY Syndrome, by contrast, were among the first genomically designated conditions and remain rarely diagnosed because of their generally mild or negligible clinical manifestation. Nevertheless, studies show that termination rates upon prenatal detection of XXX and XYY Syndrome are often well over 50%, and their inclusion in screening programs in countries like Denmark as well as their ease of detection on a standard karyotype means that they are routinely detected.

In a meta-review of eight studies that reported rates of termination on the basis of sex chromosome aneuploidies published between 1987 and 2004, Hamamy and Dahoun

(2004:61) found that XYY pregnancies where no anomaly was found on ultrasound were terminated between 20% and 54.5% of the time while XXX fetuses were terminated at a rate of 17-70%. In a larger study conducted at UCSF, Shaffer et al. (2006:669) found that of the over 800 prenatally detected cases of chromosomal aneuploidy following amniocentisis or chorionic villus sampling (CVS), termination rates for conditions discussed in this dissertation were as follows: trisomy 21 (Down syndrome), 87%; trisomy

18 (Edwards Syndrome), 81%; trisomy 13 (Patau Syndrome), 90%; X (Turner Syndrome),

65%; XXY (Klinefelter Syndrome), 70%; XXX, 57%; XYY, 40%. We will see later on how prenatal testing is increasingly picking up a range of genomically designated syndromes delineated according to abnormalities much smaller than whole chromosome aneuploidies. As should be apparent from the above rates of termination for XXX and

! 369! !

XYY syndrome, the dilemmas represented by prenatally discovered cases of genomically designated syndromes remain acute even when we know the phenotype tends to be very mild. Furthermore, Shaffer et al. report that termination rates for SCAs were higher following CVS rather than amniocentesis testing (but not for the more severe autosomal trisomies) – “77% vs 55%, p = 0.015” (ibid: p. 669) – a result they plausibly ascribe, like

Verp et al. nearly 20 years earlier (above), to the fact that CVS tends to be performed in the

10-12th week of pregnancy as vs. the 15-20th week in the case of amniocentesis, as well as the decreased tolerance for genetic abnormality of women opting for CVS testing. What then, if we could non-invasively test for fetal chromosome abnormalities both large and small even earlier into pregnancy?

Extracting fetal DNA

On October 21st 2011, a press release from Seqeunom and articles in major media outlets like the New York Times (Pollack 2011) reported the availability of Sequenom’s new MaterniT21™ test and the validating findings of a Seqeunom-funded study published a few days earlier by Palomaki et al. in Genetics in Medicine. MaterniT21™ uses

Sequenom’s SEQureDX™ technology that utilizes ‘massively parallel sequencing’ (MPS) to prenatally identify, as early as ten weeks into pregnancy, extra material from chromosome 21 in fetal DNA found in a maternal blood sample indicating trisomy 21, i.e.

Down syndrome, in the fetus. The test represents a significant advance in prenatal genetic testing because, unlike amniocentesis or CVS, it is non-invasive, carries no risk to the fetus and can be reliably conducted in the first trimester of pregnancy. As long as such tests can ensure that it will return no false negatives and only a very small rate of false positives, it

! 370! ! can be offered on the assumption that a confirmatory amniocentesis or CVS test will be ordered following the detection of an abnormality.

In fact, media, experts and stakeholders had the chance to wrestle with

MaterniT21™’s implications a couple of years earlier when Sequenom announced that a

SEQureDX™ test for identifying Down syndrome prenatally was about to hit the market.

As it turned out, the studies that purported to demonstrate SEQureDX’s accuracy and reliability were deeply flawed, leading Sequenom to withdraw the test, the collapse of its share price, and the firing of its CEO and Vice President of research and development, among others. In addition, a shareholder lawsuit eventually resulted in a 14 million dollar settlement and an ongoing investigation by the Securities and Exchange Commission into insider trading. The former Vice President of research and development had plead guilty to a charge of conspiring to commit securities fraud by knowingly distorting data before she died in early 2011. The new CEO opened their 2010 annual report by writing, “In 2009

Sequenom opened the year at a euphoric high and in late April experienced the depths of despair as an organization.” Nevertheless, the report was clear that Sequenom’s “Down syndrome test program will remain our largest single investment in 2010. The company remains committed to completing the development, validation and launch of a noninvasive

Down syndrome test.” (Sequenom 2010) As CBS news put it in late March 2011 (Edwards

2011), “If San Diego-based Sequenom could bring a non-invasive prenatal test for Down

Syndrome to market it would likely be a blockbuster: Every pregnant woman would want to take it. The data scandal forced the company to start its research from scratch.”

Only months later Sequonom announced that the new research had been completed and MaterniT21™ would be available in 20 major metropolitan areas in the US. So far,

! 371! ! business has been good. Sequenom’s most recent quarterly report shows that their revenues for diagnostic services for the nine months leading up to September 30th 2012 were

$25,392,000 up from $5,494,000 for the same period the year prior. The report noted (p.

22): “We have continued to see growth in our diagnostic services test volumes, with increases attributable in particular to the MaterniT21 PLUS LDT since its product launch in October 2011.”

The MaterniT21™ test has rightly received extensive attention for its likely impact on prenatal testing and termination rates for Down syndrome. The New York Times

(Pollack 2011) quoted one expert who described this new wave of prenatal tests as ‘a game changer’, noting that two other companies planned to release functionally equivalent tests in 2012. The article further cites concerns that tests like MaterniT21™ will lead to an increase in abortions due to prenatally diagnosed trisomy 21 and a decline in Down syndrome prevalence that is likely to decrease support for that population. Indeed it is widely recognized that prenatal testing1 has already had a substantial impact on Down syndrome populations, and that even as rising maternal age should have led to a huge increase in Down syndrome births, all else being equal, termination rates following a positive test have prevented this from happening. For while a minority of Down syndrome fetuses are diagnosed prenatally, most often because the pregnant woman is over 35, a large meta-analysis covering 5 countries found that ~92% of 5035 cases were aborted

(Mansfield et al. 1999), although a more recent study suggested that rates in the US are somewhat lower and highly variable across sites (Natoli et al. 2012). In many countries

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 1 This refers to both genetic and other forms of prenatal testing, e.g. sonograms, but genetic tests are still generally used to confirm.

! 372! ! increasing maternal age and relatively low prenatal testing rates ensure that a small increase in prevalence is still observed, while in others it is the inverse: for example in

Paris and Taiwan, it is estimated that 85% and 80% of Down syndrome fetuses are aborted due to much higher prenatal testing uptake and similarly high rates of termination upon receiving a Down syndrome diagnosis (Leroi 2006).

A number of scholarly studies have therefore examined the eugenic implications of prenatal testing for Down syndrome (e.g. Dixon 2008; Epstein 2002; Harmon 2007; Kerr,

Cunningham-Burley, and Amos 1998; Lippman 1994; McCabe and McCabe 2011;

Shakespeare 1998; Skotko 2009). Meanwhile commentators on the right have made not entirely dissimilar arguments despite their very different orientation and tone (for perhaps the most thoughtful, remarkably, see Carlson 2012). The National Down Syndrome

Society has not taken a position on prenatal testing or termination per se, choosing instead to focus on shaping the provision of information and access to genetic counselors and other genetic experts for parents who have received a prenatal diagnosis.2 To this end, they have drafted model legislation, which in June 2012 passed in Massachusetts. As their press release explained, the legislation ensures that “medical professionals are required to give parents who receive a prenatal or postnatal diagnosis ‘up-to-date, evidence-based, written information about Down syndrome that has been reviewed by medical experts and national

Down syndrome organizations’” and that parents are “given contact information for the MDSC’s Parent's First Call Program and support services” as well as other Down syndrome advocacy resources. Crucially, the press release noted the significance of this

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 2 see: http://www.ndss.org/About-NDSS/Media-Kit/Position-Papers/Down-Syndrome-Prenatal- Testing/ 3.22.13

! 373! ! legislation given the new wave of non-invasive prenatal genetic test kits, which, they explained:

… signaled a not-far-off future in which expectant parents will routinely receive an accurate prenatal diagnosis for Down syndrome and other chromosomal conditions early in their pregnancy… Today, there are at least two other such non-invasive prenatal tests on the market, all of which are being used in a limited fashion throughout the country and internationally. These tests are not yet routine for all pregnant moms, but we know they will be within the next decade if not sooner.

In short, the implications of non-invasive prenatal genetic testing for the Down syndrome population build on longstanding concerns and were immediately apparent, as is clearly recognized by the major network of Down syndrome advocacy in the US.

It is entirely understandable that Down syndrome dominated these initial discussions. But today, less than a year and a half later, Sequenom has rebranded

Materni21 LDT as Materni21 PLUS LDT and added the other major autosomal trisomies to their test: Edwards and Patau Syndrome (trisomy 18 and 13 respectively). In addition, at least three rival companies have brought competing products to the market: Arisoa

Diagnostics Harmony Prenatal Test, Verinata’s verifi® and Natera’s Panorama. None are limited to Down syndrome or the autosomal trisomies 18 (Edwards Syndrome) and 13

(Patau Syndrome).

Verinata Health introduced their verifi ® prenatal test in the first half of 2012, and in June of that year they presented results at the at the 16th International Conference on

Prenatal Diagnosis and Therapy that demonstrated the ability to detect sex chromosome aneuploidies using massively parallel screening. On November 6, 2012 a press release from Verinata announced that they were “expanding the verifi® prenatal test capabilities to include detection of the most common sex chromosome abnormalities. Clinicians will now

! 374! ! be able to select the sex chromosomes option on the verifi® test to access this addition.”

They explained:

Beginning December 3, 2012, the verifi test will offer the most comprehensive non- invasive prenatal test detection menu available. The verifi test detects the most common chromosomal fetal abnormalities seen in pregnancy, including Down syndrome (trisomy 21 or T21), Edwards syndrome (trisomy 18 or T18) and Patau syndrome (trisomy 13 or T13). The optional test expansion now includes not only detection of Turner syndrome (Monosomy X), but also Trisomy X (XXX), Klinefelter syndrome (XXY) and XYY syndrome, the most common fetal sex chromosome abnormalities…. According to recent scientific publications, sex chromosome aneuploidies represent approximately five percent of all reported fetal aneuploidies. The most common sex chromosome aneuploidies result from a deletion or addition of an X or a Y chromosome to the expected two sex chromosomes (XX or XY). Subtle neurodevelopmental, language and learning difficulties, as well as anatomical changes result from most forms of sex chromosome aneuploidies. Klinefelter syndrome (XXY) is the most common sex chromosome aneuploidy, affecting approximately one in 500 males. XYY syndrome affects approximately one in 1000 males, whereas trisomy X (XXX) affects approximately one in 1000 females. Turner syndrome (Monosomy X) occurs in one in 2000 female births.3

Another press release on the December 3rd release date reiterated the inclusion of the SCAs

(referring to XYY as ‘Jacobs Syndrome’, which is quite bizarre given that term’s extremely rare use), with their Executive Chairman and CEO noting “Overall, the verifi® prenatal test expands the option to obtain more complete prenatal information than previously possible through non-invasive means, and potentially reduces unnecessary invasive testing.” Remarkably, the press release also included a supportive blurb from the

Executive Director of the leading SCA advocacy group KS&A, Jim Moore:

Today, SCAs are severely under-diagnosed. In fact, one in 500 babies born in the United States has an X or Y chromosome variation, with less than a quarter of the population diagnosed today… We are pleased to see Verinata Health expand its verifi® prenatal test to include detection of sex chromosome aneuploidies during pregnancy. Early identification

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 3 http://www.verinata.com/news/verinata-healths-verifi-prenatal-test-expanded-to-include-the- most-common-sex-chromosome-abnormalities/ 3.17.2031

! 375! !

and subsequent intervention could end the ‘diagnostic odyssey’ and lifetime of struggle that so many face as a result of never being diagnosed. KS&A stands ready to serve this community through educational resources and support. (my emphasis)

The press release then invites the reader: “For more information on KS&A, please visit www.genetic.org.”4

Similarly, on March 1st of this year Natera launched their Panorama test that includes the autosomal trisomies, triploidy and announced that it “will be the first to screen for sex chromosome abnormalities in every sample.”5 They explained:

Why are sex chromosomes important? Recent studies have demonstrated early interventions to address physical, developmental and emotional issues related to these aneuploidies can be effective. If you, your patient and her pediatrician know to look for these characteristics, they are likely to be treated much earlier, improving the quality of life for her child.1,2,3,4

Early treatment is recommended for sex chromosome aneuploidies An individual with an extra X and/or Y chromosome can have5: • Neurodevelopmental differences, with increased risk for delays • Language-based learning disabilities • Cognitive impairments • Behavioral and psychological disorders6

Not to be left out, Sequenom announced that they would also include the major SCAs on all test samples received from February 4th 2013. Sequenom's Chief Medical Officer explained the rationale for their inclusion in the press release:

Sex chromosome abnormalities may be recognized at birth, as part of the spectrum of less severe chromosome abnormalities. They can also be found in adults, many being incidentally discovered in the course of evaluating patients for infertility or endocrine !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 4 http://www.verinata.com/news/verinata-healths-non-invasive-verifi-prenatal-test-now-includes- unique-ability-to-detect-additional-fetal-chromosomal-abnormalities/ 3.17.2031

5 http://www.panoramatest.com/panorama-test-launched 3.22.2013

6 http://www.panoramatest.com/importance_detection 3.22.2013

! 376! !

problems. Identifying these conditions through the MaterniT21 PLUS test will allow the health care provider and patient to discuss the medical issues associated with these conditions as well as to develop both short- and long-term care plans.7 (my emphasis)

The history of prenatal genetic testing casts serious doubt on the idea that the development of early intervention and care plans will be the main form of intervention undertaken on the basis of positive prenatal SCA results early in pregnancy. While the KS&A is surely right to try and at least gain some involvement in the information given to parents about

SCAs (as with Down syndrome advocates seen above), there can be little doubt that these new tests have the potential to lead to a significant increase terminations and perhaps seriously impact the SCA population moving forward.

Non-invasive prenatal genetic testing will not, however, stop with the aneuploidy syndromes. Indeed the key patent in the field, ‘Noninvasive Diagnosis of Fetal Aneuploidy by Sequencing’, was issued to Stanford University bioengineer and Verinata founder

Stephen Quake, and his student and Hei-mun Christina Fan. In it, they specifically note

(Fan and Quake 2010:6):

The present method is applicable to large chromosomal deletions, such as 5p-Syndrome (five p minus), also known as Cat Cry Syndrome or Cri du Chat Syndrome. 5p-Syndrome is characterized at birth by a high-pitched cry, low birth weight, poor muscle tone, microcephaly, and potential medical complications. Similarly amenable disorders addressed by the present methods are p-, monosomy 9P, otherwise known as Alfi's Syndrome or 9P-, 22q11.2 deletion syndrome, Emanuel Syndrome, also known in the medical literature as the Supernumerary Der(22) Syndrome, trisomy 22, Unbalanced 11/22 Translocation or partial trisomy 11/22, Microdeletion and Microduplication at 16p11.2, which is associated with autism, and other deletions or imbalances, including those that are presently unknown. (my emphasis)

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 7 http://sequenom.investorroom.com/2013-02-04-Sequenom-CMMs-MaterniT21-PLUS-LDT- Now-Reporting-On-Gender-Specific-Chromosomal-Abnormalities

! 377! !

In other words, the capacity to use massively parallel screening to detect genetic anomalies in the fetal DNA found in maternal blood was seen from the outset as commercially applicable to mutations both large and small, and at least one of the major companies is already moving in that direction. Verinata recently reported findings at the American

Society of Human Genetics that suggest non-invasive prenatal testing for chromosomal abnormalities smaller than full aneuploidy are on their way, leading the their CEO to say:

The verifi® prenatal test offers pregnant women access to a blood test that is safe, early and accurate… Looking beyond our current product offering, our early work presented at ASHG represents an exciting possibility for future non-invasive prenatal tests. While this initial work serves as a proof of principal, it clearly demonstrates the power and potential of massively parallel sequencing to detect alterations within individual chromosomes, similar to chromosomal microarrays.8 (my emphasis)

Indeed, one lab already managed to sequence an entire fetal genome (Kitzman et al. 2012).

Given the long list of disorders already offered on prenatal microarray tests following amniocentesis or CVS, what can we expect these expanded non-invasive tests to include?

Take, for example, Signature Genomics’s current offering. Signature was cofounded by a leading figure in the field of medical genetics and a strong proponent of the ‘genotype-first’ diagnosis and delineation of syndromes, Lisa Shaffer (see e.g. Ballif et al. 2007; Bejjani and Shaffer 2006; Shaffer et al. 2007, 2007, 2011, 2006; Shapira et al. 1997). Their

PrenatalChip®OS “Evaluates over 245 known genetic syndromes and over 980 gene regions of functional significance in human development.”9 Most of those ‘known genetic

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 8 http://www.verinata.com/news/verinata-health-announces-new-findings-at-the-american-society- of-human-genetics/

9 http://www.signaturegenomics.com/prenatalchipOS.html (my emphasis) 3.22.13. For the complete list of ‘syndromes’, see: http://www.signaturegenomics.com/documents/SG_DisordersTestedOS_GC-BR-003v3_v2_04- 2012_nosec.pdf

! 378! ! syndromes’ are actually poorly characterized cases of genomic designation – i.e. it includes all the strong cases that have given rise to detailed clinical profiles and advocacy organizations listed in Table 2 of Chapter 1 as well as dozens upon dozens that have not.

Clearly, the field of non-invasive prenatal genetic testing is moving very quickly, and we can now expect this highly lucrative market to be fought out on the basis of price, accuracy, the range of conditions chosen and the ability to perform the test earlier and earlier in pregnancy. In January, Illumina Inc. bought Verinata for $350 million dollars with a commitment to invest a further $100 million to help them expand their reach

(Reuters 2013). Sequenom is working with insurance companies to ensure a low co-pay, pledging that insured consumers never pay more than $235.10 Ariosa’s test is only $795 and is covered by California’s Medi-Cali – a state program covering prenatal health for low-income women. One report cited an investment analyst who said that Sequenom’s above-cited sales (it is the only publicly traded company) have “substantially exceeded expectations” and that the non-invasive prenatal genetic testing industry could soon be worth over a billion dollars a year, helping to explain why there are several patent lawsuits between the various companies working their way through the courts. The same report also noted that observers expect the advantages of non-invasive testing to increase annual uptake of prenatal genetic screening “from fewer than 100,000 to as many as 3 million.”

(Check Hayden 2012) While a qualitative technological advancement was required to turn the excitement regarding the discovery of fetal DNA in maternal blood in 1997 (Lo et al.

1997) into a biomedically and commercially viable reality, it is now a race to towards the quantitative improvements discussed above: price, specificity, reliability, age at which a !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 10 http://www.sequenom.com/home/media/news/ accessed 2.1.2012

! 379! ! fetus can be tested and so on. Indeed Natera now offers their test just nine weeks into pregnancy.11 Given the progress seen in just the last few months, it is hard not to anticipate rapid advances on each of these fronts.

A trapdoor to eugenics?

The Signature Precision Panel™ | Prenatal kit tests for trisomies 13, 18 and 21 as well as the SCAs, which they note are found in “up to 1 in 300 births,” and “20 common and severe microdeletion/duplication syndromes” like 22q11.2DS, Cri Du Chat, Williams

Syndrome, etc., whose total incidence they list as 1 in 1,126, bringing the total incidence of conditions covered to 1 in 237.12 This is almost certainly a huge underestimate: for example, Signature cite the prevalence of 22q11.2DS as 1 in 5,95013, even though most experts agree that the standard 1 in 2-4000 figure is almost certainly too low and, as we saw in Chapter 6, some expect it to rise to at least one in a thousand. What the total incidence for findings covered by their new PrenatalChip®OS test and its “245 known syndromes” (above) will be, not to mention the uncharacterized “980 gene regions of functional significance,” is impossible to say. One percent of all births would certainly be a conservative estimate. Furthermore, almost all of these conditions’ clinical profiles are ridden by a systemic ascertainment bias: as we have seen throughout this dissertation, genetic testing is almost entirely reserved for patients with congenital abnormalities or developmental and/or intellectual deficits.

!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! 11 http://www.panoramatest.com/SMFM-Validation-Data-Announcement 3.23.13

12 http://www.signaturegenomics.com/precisionpanel.html 3.23.13

13 http://www.signaturegenomics.com/precisionpanel_disorders_tested.html 3.23.13

! 380! !

Given what we know about longstanding termination rates even for typically mild conditions like XXX and XYY, it is not unreasonable to expect that a very large proportion of fetuses bearing these mutations will be aborted. If the new wave of non-invasive tests continues to expand its market penetration and the conditions tested for, we are talking about a technology whose individual-level affordances may eventually have demographic effects in the aggregate. In other words, the current trajectory of non-invasive prenatal genetic tests raises serious issues about the emergence of a new kind of non-centralized eugenics in which genomic designation would feature centrally.

There are of course compelling and, in my opinion, ethically sound reasons to offer prenatal genetic testing to expectant mothers: the right to make reproductive decisions based on the information available and the ability to plan for the care of a child with a genetic disorder. As such, prenatal testing must now be offered to expectant parents under the terms of the Affordable Care Act, and we saw how the companies discussed above are working to ensure low co-pays for customers with insurance and to be included in state healthcare initiatives. Non-invasive genetic testing will therefore likely become available to most expectant parents over the coming years.

Rick Santorum, who has a child with Trisomy 18/Edwards Syndrome, took up prenatal testing provision in the Affordable Care Act during last year’s presidential election campaign:

One of the things that you don’t know about ObamaCare in one of the mandates is they require free prenatal testing,… Why? Because free prenatal testing ends up in more abortions and, therefore, less care that has to be done, because we cull the ranks of the disabled in our society. That too is part of ObamaCare — another hidden message as to what president Obama thinks of those who are less able than the elites who want to govern our country. (quoted in Rafferty and Montanaro 2012)

! 381! !

It should go without saying the attribution of eugenic intention in this particular piece of rightwing electioneering was wholly misplaced. But perhaps so too is the liberal response, best expressed by University of Chicago public health professor, Harold Pollack:

I’m writing these words with my smiling brother-in-law Vincent sitting next to me, admiring the green lunchbox that we just bought him. Vincent lives with intellectual disabilities caused by fragile X syndrome. I find the above comments indescribably insulting. Santorum’s comments are only made uglier by their utter lack of foundation. There is no evidence whatsoever that liberals–let alone President Obama–are less solicitious [sic] or caring about the disabled than other Americans. I’ve never heard any liberal health policy wonk promote genetic technologies to “cull the ranks of the disabled” or as part of any cost-cutting plan. That ugly meme is completely made up. By any reasonable measure, the proliferation of genetic diagnostic technologies coincides with great progress in public acceptance and support for people with disabilities. Certainly liberals are willing to spend more money on disability services. (Pollack 2012)

Pollack is undoubtedly correct that the primary aim of including prenatal testing in the

Affordable Care Act has nothing to do with ‘culling the ranks’ of the disabled, and he rightly dismisses the preposterous notion that ‘liberals’ who support universal health care are less concerned about the well-being of people with intellectual and other disabilities.

However, even a stopped watch is right twice a day, and in this case Santorum et al. might actually be on to something. Why? Because despite the fact that Pollack is correct about unprecedented modern support for people with disabilities, and despite the emphasis on anticipatory care seen here and with respect to the prenatal tests for sex chromosome aneuploidies seen above, there remains a strong empirical basis to the claim that many fetuses prenatally diagnosed with syndromes that have developmental implications, even when mild, will be terminated. Furthermore, we have seen how first-trimester testing is likely to substantially increase termination rates. Pollack’s observation that the proliferation of genetic testing technology ‘coincides’ with increased support of disability

! 382! ! rights and services does not obviate the fact that most fetuses diagnosed with genetic disorders, especially early in pregnancy, are aborted.

In other words, even in the absence of a centralized eugenic plan, the aggregate effect of parents exercising the right to know about and act on genetic mutations detected in utero may have, as a secondary effect, eugenic consequences. Discussion about the implications of prenatal testing for Down syndrome, cystic fibrosis and other conditions that have been ‘geneticized’ are no doubt very important. However, going beyond those few cases of geneticization to consider the many more genomically designated conditions that are already being tested for prenatally raises at least two important issues: first, it brings the population-level effects of earlier, dramatically expanded prenatal testing into sharp relief; second, it forces us to address the fact that our knowledge about the clinical implications of genetic anomalies is partial, and that what we do know is ridden by ascertainment bias.

This admittedly speculative exercise is justified, I argue, for two reasons. On the one hand, it represents a potentially crucial shift in the conditions facing the networks mobilized around genomically designated conditions. On the other hand, it is a productive middle path between the longstanding and extensive discussions of prenatal testing for

Down syndrome, Edwards Syndrome and specific single gene mutations like those for

Cystic Fibrosis or Hemophilia referenced above and the bioethical debates about ‘designer babies’ that are not well-grounded in extant trajectories of technoscientific development.

As Steinbock put it in just such a discussion in The Lancet (2008:1295), “Genetic enhancement is, for the present, science fiction.” The rapid development of non-invasive prenatal genetic testing is an important new development for genomically designated

! 383! ! conditions, and analyzing it as such brings the potential implications of an already- emergent and marketable technology into stark relief. If genetics is going to change the population in the foreseeable future it will not be through making smarter, stronger, more attractive children, but through parents choosing to terminate pregnancies when some kind of genetic abnormality is identified cheaply, easily and early. Some of these abnormalities cause conditions that are consistently severe, others are variable and often mild, and many more remain open questions. Cumulatively, these genetic disorders are not rare at all, and most of them are already available for commercial prenatal testing. The new wave of tests, however, brings to the fore an ethical question that does not conform to our usual binary of

‘pro-life’ or ‘pro-choice’. Instead, it challenges those of us who believe in the rights of parents to make informed reproductive decisions to consider a near future where legitimate individual decisions have eugenic effects at the level of the population.

It will become easier and, crucially, cheaper to order this new wave of prenatal genetic testing kits in the coming years, especially if insurance companies decide it is in their interests to cover them in order to avoid the expensive medical care often required by genetic disorders. After all, while a child with 22q11.2DS may require little or even no medical attention whatsoever, a substantial minority requires hundreds of thousands of dollars’ worth of care in infant heart operations alone. Sequenom’s success in ensuring low co-pays for their test suggests the inevitable: insurers will do the math. However, as testing becomes cheaper and easier, providing accurate information about the abnormalities they detect and skilled genetic counseling for parents facing reproductive dilemmas will, by contrast, require years of investment, training and research. There are no easy solutions, but as it stands now we risk sleepwalking into a scenario where many thousands of parents

! 384! ! every year are told, just months into a pregnancy, that their fetus has a genetic disorder that may or may not cause serious developmental and medical problems. This will not only be enormously trying for expectant parents – it will also have profound implications for the way we deal with abnormality and developmental difference as a society. Finding abnormal genomes is becoming relatively easy. Helping people understand and make decisions based on prenatal genetic testing results with uncertain implications, by contrast, represents an urgent and largely overlooked challenge. Therefore, it is worth asking whether, borrowing from Troy Duster (2003), this new way of prenatal genetic testing might combine with genomic designation to effect a new trapdoor to eugenics.

What next for genomic designation?

Genomic designation has made enormous gains in the literature over the last half century, from The Lancet’s 1960 injunction to search for phenotypes associated with trisomy of all 22 autosomal chromosomes to the delineation of medical conditions on the basis of deletions, duplications, copy-number variations and polymorphisms that are not visible under even the most powerful microscope. A recent Genetics in Medicine study by

Lisa Shaffer et al. (2007), whose work and role in Signature Genomics was discussed above, suggests that the practice of what I am calling genomic designation may become more routinized or ‘black-boxed’ (Latour 1987). Using a high-resolution genomic microarray, Shaffer and colleagues identify “a novel microdeletion syndrome of 1q41q42” in seven of 10,000 subjects referred to their lab for various forms of developmental delay.

Although no clinical features other than developmental delay (the basis for sample selection) were shared by all seven cases, the authors argue that precisely the move away

! 385! ! from a reliance on phenotypic features represents a key advantage. Shaffer et al. contrast their approach from the ‘traditional’ discovery of microdeletion syndromes that “depended on the serendipitous ascertainment of a patient with established clinical features and a chromosomal rearrangement…” (607), arguing that the “assum[ption] that the clinical features are distinctive enough to establish phenotypic relations among patients… is limited by human subjectivity.” (610) Adopting a ‘genotype-first’ approach, in contrast, introduces “an objective means of collecting a cohort of patients. Thus, a common phenotype among a group of patients may be delineated only after a common genotype has been isolated.” (611) While Shaffer et al. exaggerate the novelty of their ‘genotype-first’ sequence, and it is not clear whether the ‘syndromes’ they designate will ever give rise to networks of social and clinical practice, this move away from the ‘serendipity’ required to discover a syndrome population represents a potentially important innovation in the practice of genomic designation. Similarly, the emergence of the field of behavioural neurogenetics, which “starts with a known biologically validated gene or chromosomal abnormality and uses it to define the study cohort” (Feinstein 2009:1050), suggests that genomic designation will increasingly be used to pry open questions related to the brain, behaviour and cognition.

Indeed in between the time I circulated and defended this dissertation the NIMH announced a major turn in funding priority aimed to help psychiatry move away from the symptom based classification of the DSM towards a new biologically grounded nosology.

As the director put it, “Biology never read [the DSM],” (quoted in Belluck and Carey

2013) and so his agency has decided to fund research that will help “transform diagnosis by incorporating genetics, imaging, cognitive science, and other levels of information to

! 386! ! lay the foundation for a new classification system.” (Insel 2013) The New York Times reports past and present directors of the NIMH and others saying that they hoped psychiatry would follow cancer research in the move to classify disease according to genetic markers in such a way that cut across the symptom-based categories, and quoted one former director as saying that the DSM had become a “scientific nightmare.” He explained how “Many people who get one diagnosis get five diagnoses, but they don’t have five diseases — they have one underlying condition.” (Steven E. Hyman, quoted in

Belluck and Carey 2013) Of course this is precisely the rationale in favor of genomic designation that I saw time and again at 22q11.2DS conferences. In other words, since I circulated the dissertation just under a month ago, genomic designation appears to have won a major institutional ally in the field of psychiatric research. In sum, we have seen how genomic designation seems poised to play an ever-greater role in both esoteric and commercial biomedical research.

But can the same be said for clinical practice? A debate about the ‘genotype-first’ approach to the discovery of syndromes in a series of articles and letters spanning the New

England Journal of Medicine and Genetics in Medicine represents the most explicit discussion of genomic designation to be found in the biomedical literature. Following a paper outlining the discovery of recurrent abnormalities at site 1q21.2 with highly variable phenotypic implications (Mefford et al. 2008), the debate concerned the role of genomic testing in clinical practice (Cody 2009; Ledbetter 2009, 2009, 2008; Saul and Moeschler

2009). As David Ledbetter put it, new syndromes tend to be discovered on the basis of genetic markers following technological breakthroughs, and “This ‘genotype first’ method of syndrome identification is currently undergoing an exciting new renaissance because of

! 387! ! cytogenetic arrays…” (2008:1728). Ledbetter, and longstanding and leading figure in medical genetics, goes on (ibid: 1729-30):

What are the clinical implications of this and other similar reports of microdeletions associated with mental retardation, autism, or other physical and developmental disabilities in children? The total number of disorders is now far too large for a pediatrician, or even a pediatric geneticist, to make a specific clinical diagnosis before genetic testing. However, since whole-genome cytogenetic arrays are now available in many clinical laboratories, a pediatrician has the option of ordering this test as an adjunct to or replacement of a standard karyotype and can expect a much higher yield of clinically significant results. Clinicians, like researchers, can now shift to a "genotype first" model of diagnosis for children with unexplained developmental abnormalities. (my emphasis)

The debate represents, in part, a jurisdictional dispute regarding the discovery and diagnosis of clinically relevant entities. What his paper and the debate more generally makes clear is the growing willingness to use exploratory, non-targeted genomic tests in order to diagnose clinically non-specific genetic disorders, especially in children. Indeed a recent ‘consensus statement’ published in The American Journal of Human Genetics by the International Standard Cytogenetic Array Consortium advocated the use of non- targeted, higher-yield chromosomal microarray as a ‘first-tier clinical diagnostic test’, in place of karyotyping, “for patients with unexplained DD/ID, ASD, or MCA [multiple congenital anomalies]." (Miller et al. 2010:757) In the biomedical literature then, we see a growing acceptance of the idea that people with unexplained forms of developmental and congenital abnormality should be subjected to those tests that are most likely to provide a genetic ‘diagnosis’.

But what is to be done with these ‘genotype-first’, phenotypically inchoate diagnoses? And what about cases like 22q11.2DS or Williams Syndrome where it is less clear that the genotype came first, but very clear that it is now directing classification and practice? What kinds of networks will need to be forged in order to vivify Ledbetter’s

! 388! ! assertion that ‘Clinicians, like researchers, can now shift to a “genotype first” model of diagnosis’? Most importantly of all, what will this resurgence of genomic designation in the biomedical literature mean for the many fields that Fleck described as ‘exoteric’, and what will it therefore mean for the people who are so classified? While avid biomedical interest and advances in genomic technologies increase the volume and range of molecular observations that can give rise to genomic designation, professional jurisdictions (Abbott,

1988), patient and parent advocacy groups, state and commercial investment and clinical and broader public orientations towards genetic information are merely the most obvious social factors that will shape its trajectory. As we have seen throughout Chapters 3-6, there is great variability across both cases and historical periods when it comes to the way genomically designated ‘syndromes’ are mobilized beyond the esoteric fields of human genetics in which they are first reported.

The drive to identify genetic mutations with greater ease and increasing fervor, and to take genomic information as indicating something essential to the self, powerfully suggests the need to take stock of genomic designation while it is still a relatively marginal phenomenon. Hacking has written of “[t]he genetic imperative[:] the drive to find genetic markers in humans” (2006: 99), and he also notes that, despite humanist and scientific arguments to the contrary, ‘people can hardly avoid thinking of their genetic inheritance as part of what constitutes them, as part of who they are, as their essence’ (p. 92). As Hacking puts it (2006: 88), “knowledge of genetic ‘identities’ will forge social ones, creating new communities of shared recognition based on partial science. That is not intrinsically bad, but it is still a phenomenon that can be grossly abused.” Such a warning takes on even greater significance with genomic designation, especially as it grows in numbers and in the

! 389! ! range of human kinds implicated. Concepts like biological citizenship, biomedicalization

(Clarke et al. 2003), geneticization, the molecular gaze and biosociality have rightly received extensive scholarly attention, but they have all focused on the intersection of phenotypically designated categories and genetic understandings of them. Genomic designation, however, does not just recast or reduce human traits and maladies, essentialize, stigmatize or destigmatize abnormal persons: rather than simply shaping our understanding of existing categories of human difference along genetic lines, it actually produces new kinds of people. However, that does not mean that geneticists and allied biomedical experts attain unquestioned authority over the way we understand difference and abnormality. On the contrary, the practical implications and indeed the very clinical profiles of genomically designated conditions have been shown to be decisively shaped by both prevailing historical conditions and divergent networks of actors assembled around them.

Finally, there is no reason why genomic designation must remain limited to medicalized populations (see Shostak, Conrad, and Horwitz 2008 on the contingent, path- dependent relationship between genetics findings and medicalization). The commercialization of genetic testing for ethnicity suggests that genomic designation could potentially give rise to categories of identity formation and self-practice that are not likely to gain respectability within the scientific community (Nelson, 2008; Abu El Haj, 2007)

After all, if astrology can convince people that they are of a particular group based on the alignment of heavens at the moment of their birth, then our genomic inheritance seems primed to do so with far greater power. Meanwhile, the combination of genomic specificity at the level of SNPs and haplotypes, the dynamics of ‘biocapital’ (Rose, 2006; Sunder

Rajan, 2006) and new communications technologies (Castells, 1997) makes this prospect

! 390! ! more likely still. As we saw with 22q11.2DS, knowledge about a genetic mutation can reconfigure the boundary between the normal and the pathological and bring non-clinically significant traits to the attention of experts, parents and patients, in turn becoming part of the condition itself. Indeed, we saw how individuals have been diagnosed with 22q11.2DS, following the diagnosis of a relative, even in the absence of any relevant medical phenotype and gone on to reinterpret subclinical traits in terms of that diagnosis. In sum, the networks formed around genetic mutations like the 22q11.2 microdeletion have the capacity to both draw upon and reconfigure existing frameworks for identifying and labeling difference and abnormality. While this dissertation has focused largely on childhood abnormality and developmental difference, we have seen in conditions like

XYY, 22q11.2DS and Williams Syndromes that genomic designation has the capacity to inform understanding of and treatment for many different forms of human variation.

And yet, for all of their successes in recent years, even the most robust networks of research, treatment and advocacy for genomically designated conditions are facing a new threat to their continued growth. It is telling that Down syndrome advocates have begun to recognize the specter of non-invasive prenatal genetic testing for their population and are seeking to intervene at the point of diagnosis with information and support, i.e. immediate entrée to their network, in the hope that it will lead would-be parents to embrace the idea of having an intellectually disabled child with a genetic disorder. For conditions like XYY and 22q11.2DS increased terminations may be counteracted by increased ascertainment, because unlike Down syndrome only a minority of people with those conditions are ever actually diagnosed. Still, it is not hard to see how a new wave of non-invasive prenatal

! 391! ! genetic testing could undermine all the gains made by genomically designated syndromes in recent years by dramatically reducing their future population prevalence.

I offer this possible future, in part, to dispel a teleological interpretation of my analysis of genomic designation. While it is certainly true that important gains have been made and that many conditions seem very favorable to the continuing advancement of genomically designated kinds of people, those successes and that trajectory may ultimately prove to be fragile. Even if observable genetic mutations continue to matter in a way that they did not in the 1960s and 70s, they may also be mobilized towards very different ends than the ones sought by patient advocates. The XYY-criminality nexus examined in

Chapter 3 should not be treated as just an historical oddity – after all, the history of genetics and human difference is a varied and often disquieting story. Furthermore, this dissertation has been almost entirely limited to highly developed Western countries with their liberal political regimes even as genetic testing is starting to gain significant traction elsewhere, notably in China where Sequenom immediately launched the first version of

MaterniT21TM. Clearly then, the future of genomic designation will depend not only upon the networks of research and advocacy organized around genetic mutations, but also on the shifting historical conditions and prevailing ways of understanding and acting on abnormality in which they are embedded.

Thinking through reiterated facticity, we have seen how the very same object of knowledge – a genetic mutation – can take the form, in turn, of an esoteric finding in a human genetics journal, a ‘syndrome’ in the literature, an observation of uncertain and negotiated significance in a ‘bioclinical collective’, or a robust kind of person underwritten

! 392! ! by an extensive and complex network of material and semiotic actors. The same object of knowledge, in short, can be a very different kind of fact across historical and institutional contexts, as well as over the course of its career. Even though genomic designation emerged as a discursive formation in human genetics almost as soon as it was possible to observe abnormal chromosomes, we have seen how much more is required for it to give rise to what I called micro-apparatuses of knowledge and power that can rearrange existing elements and correspond to historically given needs. While we saw XYY Syndrome take on spectacular but ultimately doomed significance within the broader apparatus of criminality and deviance studies, and Brunner and Williams Syndrome pointed towards alternative ways of mobilizing genetic mutations, it was the model pioneered by Fragile X that appears to hold the most promise today. Indeed both XYY and Williams Syndromes have moved towards collaborative research and health-advocacy of the kinds outlined in

Chapter 6. In this way, the kind of fact forged by Fragile X researchers and advocates has provided a model and a repertoire of collective action that others can creatively adopt and adapt for assembling the complex networks that can turn knowledge about a genetic mutation into a real kind of person.

The meaning of a genetic mutation is in large part a sociological phenomenon, and we should therefore expect the kinds of people given to us by genomic designation to continue to change along with the heterogeneous networks that make them matter. As we have seen, fixity at the level of the genome forecloses certain processes of looping but also enables others. In between eradication and becoming a widespread and even dominant form of classification there are myriad possibilities. While we might be rightly wary of the way genomic designation can bias expectations and create self-fulfilling prophecies, lead

! 393! ! to overtreatment and anxiety or represent a new trapdoor to eugenics, we should not expect genomically designated conditions to simply peter out as categories of knowledge and practice in the medium term. Rather, we should pay close attention to the hybrid networks that are forming around genetic mutations and the way they can impact how we understand and act on human difference. In so doing, we will better understand how knowledge about our genomes can reshape what it means to be ill, different and, ultimately, human.

! 394! ! ! !

Bibliography

Abbott, Andrew. 1988. System of Professions: Essay on the Division of Expert Labour. University of Chicago Press. Abrahams, Brett S., and Daniel H. Geschwind. 2008. “Advances in Autism Genetics: On the Threshold of a New Neurobiology.” Nature Reviews. Genetics 9(5):341–55. Abrams, N., and E. Pergament. 1971. “Childhood Psychosis Combined with XYY Abnormalities.” The Journal of Genetic Psychology 118(1st Half):13–16. Aglan, M. S., A. K. Kamel, and N. A. Helmy. 2008. “Partial Trisomy of the Distal Part of 10q: a Report of Two Egyptian Cases.” Genetic Counseling (Geneva, Switzerland) 19(2):199–209. Alfi, O., G. N. Donnell, B. F. Crandall, A. Derencsenyi, and R. Menon. 1973. “Deletion of the Short Arm of Chromosome No.9 (46,9p-): a New Deletion Syndrome.” Annales De Génétique 16(1):17–22. Allan, W., C. N. Herndon, F. C. Dudley, and others. 1944. “Some Examples of the Inheritance of Mental Deficiency: Apparently Sex-linked Idiocy and Microcephaly.” Am J Ment Defic 48:325–34. Allan, William, and C. Nash Herndon. 1944. “RETINITIS PIGMENTOSA AND APPARENTLY SEX-LINKED IDIOCY: In a Single Sibship.” Journal of Heredity 35(2):41–43. Allderdice, P. W. et al. 1969. “The 13q-deletion Syndrome.” American Journal of Human Genetics 21(5):499–512. Alpman, Asude et al. 2005. “Ring Chromosome 20 Syndrome with Intractable Epilepsy.” Developmental Medicine and Child Neurology 47(5):343–46. Anon. 1968a. “Criminal Law: Question of Y.” Time, October 25. Retrieved September 1, 2011 (http://www.time.com/time/magazine/article/0,9171,900408,00.html). Anon. 1968b. “Genetics: Of Chromosomes & Crime.” Time, May 3. Retrieved September 28, 2010 (http://www.time.com/time/magazine/article/0,9171,841254,00.html). Anon. 1968c. “Nature or Nurture?” The New York Times, April 23. Retrieved November 20, 2012 (http://select.nytimes.com/gst/abstract.html?res=F20C14FB355D147493C1AB178 FD85F4C8685F9). Anon. 1976. The XYY Man.

! 395! ! ! ! Anon. 2009. “Scientist Debunks ‘Warrior Gene’.” New Zealand Herald, September 12. Retrieved April 11, 2012 (http://www.nzherald.co.nz/nz/news/article.cfm?c_id=1&objectid=10596821). Anon. 2011. “Fragile X Syndrome - PubMed Health.” Retrieved April 27, 2012 (http://www.ncbi.nlm.nih.gov/pubmedhealth/PMH0002633/). Ariani, Francesca et al. 2004. “Real-time Quantitative PCR as a Routine Method for Screening Large Rearrangements in Rett Syndrome: Report of One Case of MECP2 Deletion and One Case of MECP2 Duplication.” Human Mutation 24(2):172–77. Armstrong, David. 1995. “The Rise of Surveillance Medicine.” Sociology of Health & Illness 17(3):393–404. Armstrong, David. 2011. “Diagnosis and Nosology in Primary Care.” Social Science & Medicine 73(6):801–7. Autism Research Review International. 1987. “Autism, Fragile X Connection Investigated.” Autism Research Review International 1(1):6–7. Autism Research Review International. 1989. “Link Between Autism, Genes Is Explored.” Autism Research Review International 3(2):1–2, 6. Autism Research Review International. 1991. “Fragile X Gene Discovery Announced.” Autism Research Review International 5(2):1–2. Bailey, A. et al. 1995. “Autism as a Strongly Genetic Disorder: Evidence from a British Twin Study.” Psychological Medicine 25(01):63–77. Baio, Jon. 2012. Prevalence of Autism Spectrum Disorders: Autism and Developmental Disabilities Monitoring Network, 14 Sites, United States, 2008. Morbidity and Mortality Weekly Report. Surveillance Summaries. Volume 61, Number 3. Centers for Disease Control and Prevention. 1600 Clifton Road, Atlanta, GA 30333. Tel: 800-311-3435; Tel: 404-639-3311; Web site: http://www.cdc.gov. Retrieved February 6, 2013 (http://www.eric.ed.gov/ERICWebPortal/detail?accno=ED530639). Bales, Abigail M., Christina A. Zaleski, and Elizabeth W. McPherson. 2010. “Newborn Screening Programs: Should 22q11 Deletion Syndrome Be Added?” Genetics in Medicine 12(3):135–44. Ballif, B., J. Coppinger, and L. G. Shaffer. 2008. “Identification of Inherited and de Novo Deletions and Duplications of the 16p11.2 Autism Susceptibility Region.” Annals of Neurology 64:S82–S83.

! 396! ! ! ! Ballif, Blake C. et al. 2007. “Discovery of a Previously Unrecognized Microdeletion Syndrome of 16p11.2-p12.2.” Nat Genet 39(9):1071–73. Barbaro, Michela et al. 2009. “Characterization of Deletions at 9p Affecting the Candidate Regions for Sex Reversal and Deletion 9p Syndrome by MLPA.” European Journal of Human Genetics: EJHG 17(11):1439–47. Bassett, Anne S. et al. 2011. “Practical Guidelines for Managing Patients with 22q11.2 Deletion Syndrome.” The Journal of Pediatrics 159(2):332–339.e1. Baty, Bonnie J., Brent L. Blackburn, and John C. Carey. 2005. “Natural History of Trisomy 18 and Trisomy 13: I. Growth, Physical Assessment, Medical Histories, Survival, and Recurrence Risk.” American journal of medical genetics 49(2):175– 88. Bauer, Diane. 1972. “Maryland Tests for Criminal Potential.” in Experimentation With Human Beings; The Authority of the Investigator, Subject, Professions, and State in the Human Experimentation Process., edited by Jay Katz. Russell Sage Foundation. Bauer, Diane et al. 1980. “Special Supplement: The XYY Controversy: Researching Violence and Genetics.” The Hastings Center Report 10(4):1–31. Bearman, Peter. 2008. “Introduction: Exploring Genetics and Social Structure.” American Journal of Sociology 114(s1):v–x. Beaver, Kevin M., Matt DeLisi, Michael G. Vaughn, and J. C. Barnes. 2010. “Monoamine Oxidase A Genotype Is Associated with Gang Membership and Weapon Use.” Comprehensive Psychiatry 51:130–34. Beckwith, Jon. 1975. “Harvard XYY Study.” Science (New York, N.Y.) 187(4174):298–99. Beckwith, Jon, and Jonathan King. 1974. “The XYY Syndrome: a Dangerous Myth.” New Scientist (1971) 64(923):474–76. Bejjani, Bassem A., and Lisa G. Shaffer. 2006. “Application of Array-Based Comparative Genomic Hybridization to Clinical Diagnostics.” J Mol Diagn 8(5):528–33. Belluck, Pam, and Benedict Carey. 2013. “Psychiatry’s New Guide Falls Short, Experts Say.” The New York Times, May 6. Retrieved May 10, 2013 (http://www.nytimes.com/2013/05/07/health/psychiatrys-new-guide-falls-short- experts-say.html). Berg, Jonathan S., Lorraine Potocki, and Carlos A. Bacino. 2010. “Common Recurrent Microduplication Syndromes: Diagnosis and Management in Clinical Practice.” American Journal of Medical Genetics Part A 152A(5):1066–78.

! 397! ! ! ! Best, Rachel Kahn. 2012. “Disease Politics and Medical Research Funding Three Ways Advocacy Shapes Policy.” American Sociological Review 77(5):780–803. Betancur, Catalina. 2011. “Etiological Heterogeneity in Autism Spectrum Disorders: More Than 100 Genetic and Genomic Disorders and Still Counting.” Brain Research 1380:42–77. Betancur, Catalina, Takeshi Sakurai, and Joseph D. Buxbaum. 2009. “The Emerging Role of Synaptic Cell-adhesion Pathways in the Pathogenesis of Autism Spectrum Disorders.” Trends in Neurosciences 32(7):402–12. Beuren, A. J., J. Apitz, and D. Harmjanz. 1962. “Supravalvular Aortic Stenosis in Association with Mental Retardation and a Certain Facial Appearance.” Circulation 26:1235–40. Bishop, Dorothy V. M. 2008. “Forty Years on: Uta Frith’s Contribution to Research on Autism and Dyslexia, 1966–2006.” Quarterly Journal of Experimental Psychology (2006) 61(1):16–26. Bishop, Dorothy V. M. 2010. “Which Neurodevelopmental Disorders Get Researched and Why?” Plos One 5(11). Bishop, Dorothy V. M. et al. 2011. “Autism, Language and Communication in Children with Sex Chromosome Trisomies.” Archives of Disease in Childhood 96(10):954– 59. Black, Herbert. 1968. “1 Man in 300 May Inherit ‘Violent’ Gene.” Boston Globe (1960- 1981), August 7. Retrieved January 10, 2013 (http://search.proquest.com/docview/365892145/abstract/13B8C284F75EED4D23/ 9?accountid=10226). Blomquist, Hans K:Son et al. 1985. “Frequency of the Fragile X Syndrome in Infantile Autism.” Clinical Genetics 27(2):113–17.

Bockman, Johanna, and Gil Eyal. 2002. “Eastern Europe as a Laboratory for Economic Knowledge: The Transnational Roots of Neoliberalism.” The American Journal of Sociology 108(2):310–52. BOGHOSIANSELL, L. et al. 1994. “MOLECULAR MAPPING OF THE EDWARDS- SYNDROME PHENOTYPE TO 2 NONCONTIGUOUS REGIONS ON CHROMOSOME-18.” AMERICAN JOURNAL OF HUMAN GENETICS 55(3):476–83. Böök, Jan A. 1964. “Some Mechanisms of Chromosome Variations and Their Relation to Human Malformations.” The Eugenics Review 56(3):151–57.

! 398! ! ! ! Borg, I., J. D. Delhanty, and M. Baraitser. 1995. “Detection of Hemizygosity at the Elastin Locus by FISH Analysis as a Diagnostic Test in Both Classical and Atypical Cases of Williams Syndrome.” Journal of Medical Genetics 32(9):692–96. Borgaonkar, D. S. 1972. “Chromosome Survey of Maryland Boys.” P. 13a in American Journal of Human Genetics, vol. 24. Philadelphia: Philadelphia. Borgaonkar, Digamber S. 1975. Chromosomal Variation in Man: a Catalog of Chromosomal Variants and Anomalies. Johns Hopkins University Press. Bosk, Charles L. 1992. All God’s Mistakes: Genetic Counseling in a Pediatric Hospital. 1st ed. University Of Chicago Press. Boucher, Geoff. 1994. “A Williams Child Parent’s Educated Guess on Origin of Fairy and Elf Folk Tales.” Los Angeles Times Articles, August 21. Retrieved April 26, 2012 (http://articles.latimes.com/1994-08-21/news/mn-29668_1_williams-syndrome). Bourdieu, Pierre. 1984. Distinction: A Social Critique of the Judgement of Taste. Harvard University Press. Bourret, Pascale. 2005. “BRCA Patients and Clinical Collectives New Configurations of Action in Cancer Genetics Practices.” Social Studies of Science 35(1):41–68. Bourret, Pascale, Peter Keating, and Alberto Cambrosio. 2011. “Regulating Diagnosis in Post-genomic Medicine: Re-aligning Clinical Judgment?” Social Science & Medicine 73(6):816–24. Bourret, Pascale, Andrei Mogoutov, Claire Julian-Reynier, and Alberto Cambrosio. 2006. “A New Clinical Collective for French Cancer Genetics A Heterogeneous Mapping Analysis.” Science, Technology & Human Values 31(4):431–64. Bowker, Geoffrey C., and Susan Leigh Star. 2000. Sorting Things Out: Classification and Its Consequences. MIT Press. British Medical Journal. 1979. “What Is to Be Done with the XYY Fetus?” British Medical Journal 1(6177):1519–20. Brkanac, Zoran, Wendy H. Raskind, and Bryan H. King. 2008. “Pharmacology and Genetics of Autism: Implications for Diagnosis and Treatment.” Personalized Medicine 5(6):599–607. Brody, Jane E. 1968. “Rate of Crime-Linked Genetic Flaw Is Revised; Abnormality Found in Men Shows Evidence of Having Much Higher Incidence.” The New York Times, August 7, 34. Brody, Jane E. 1974a. “HARVARD BACKS GENETIC STUDY; Research Involving Young With XYY Chromosome Had Been Criticized Genetic Aberration An Extra

! 399! ! ! ! Chromosome ‘A Sensible Request’.” The New York Times, December 14. Retrieved January 11, 2013 (http://select.nytimes.com/gst/abstract.html?res=9C07E3DB1E31EF34BC4C52DF B467838F669EDE&scp=2&sq=xyy+brody&st=p). Brody, Jane E. 1974b. “Scientists’ Group Terms Boston Study of Children With Extra Sex Chromosome Unethical and Harmful; Wider Fears Cited.” The New York Times, November 15. Retrieved November 20, 2012 (http://select.nytimes.com/gst/abstract.html?res=F60D14FB3F59137B93C7A8178 AD95F408785F9). Brody, Jane E. 1975. “BABIES’ SCREENING IS ENDED IN BOSTON; Extra Chromosome Study Stirred Controversy on Its Ethics and Value.” New York Times, June 20, 6. Brown, A. et al. 1996. “Two Patients with Duplication of 17p11.2: The Reciprocal of the Smith-Magenis Syndrome Deletion?” American Journal of Medical Genetics 63(2):373–77. Brown, S., J. Russo, D. Chitayat, and D. Warburton. 1995. “The 13q- Syndrome: The Molecular Definition of a Critical Deletion Region in Band 13q32.” American Journal of Human Genetics 57(4):859–66. Brown, W. M. Court; Et al. 1964. Abnormalities of the Sex Chromosome Complement in Man!: Privy Council Medi9cal Research Council Special Report Series No. 305. Her Majesty’s Stationery Office. Brown, W. Ted et al. 1982. “Autism Is Associated with the fragile-X Syndrome.” Journal of Autism and Developmental Disorders 12(3):303–8. Brown, WT et al. 1982. “ASSOCIATION OF FRAGILE-X SYNDROME WITH AUTISM.” Lancet 1(8263):100–100. Brunner, H. G. et al. 1993. “X-linked Borderline Mental Retardation with Prominent Behavioral Disturbance: Phenotype, Genetic Localization, and Evidence for Disturbed Monoamine Metabolism.” American Journal of Human Genetics 52(6):1032–39. Brunner, H., M. Nelen, X. Breakefield, H. Ropers, and B. van Oost. 1993. “Abnormal Behavior Associated with a Point Mutation in the Structural Gene for Monoamine Oxidase A.” Science 262(5133):578–80. Brunner, Han G. 1996. “MAOA Deficiency and Abnormal Behavior: Perpectives on an Association.” Pp. 155–67 in Genetics of Criminal and Antisocial Behaviour. New York, NY: Wiley.

! 400! ! ! ! Buiting, K. et al. 1992. “A Putative Gene Family in 15q11-13 and 16p11.2: Possible Implications for Prader-Willi and Angelman Syndromes.” Proceedings of the National Academy of Sciences of the United States of America 89(12):5457–61. Bumiller, Kristin. 2009. “The Geneticization of Autism: From New Reproductive Technologies to the Conception of Genetic Normalcy.” Signs 34(4):875–99. Burn, J. 1999. “Closing Time for CATCH22.” Journal of Medical Genetics 36(10):737–38. Caglayan, Ahmet O. 2010. “Genetic Causes of Syndromic and Non-syndromic Autism.” Developmental Medicine & Child Neurology 52(2):130–38. Callon, M., and V. Rabeharisoa. 2003. “Research ‘in the Wild’ and the Shaping of New Social Identities.” Technology in Society 25(2):193–204. Callon, Michel. 1986. “Some Elements of a Sociology of Translation: Domestication of the Scallops and the Fishermen of St Brieuc Bay.” Power action and belief a new sociology of knowledge 32(4):196–223. Callon, Michel. 1995. “Four Models for the Dynamics of Science.” Pp. 29–63 in Handbook of Science and Technology Studies, edited by Sheila Jasanoff, Gerald E. Markle, James C. Peterson, and Trevor J. Pinch. London: SAGE Publications, Inc. Callon, Michel, and Vololona Rabeharisoa. 2004. “Gino’s Lesson on Humanity: Genetics, Mutual Entanglements and the Sociologist’s Role.” Economy and Society 33(1):1. Callon, Michel, and Vololona Rabeharisoa. 2008. “The Growing Engagement of Emergent Concerned Groups in Political and Economic Life: Lessons from the French Association of Neuromuscular Disease Patients.” Science Technology Human Values 33(2):230–61. Cambrosio, Alberto, Peter Keating, Pascale Bourret, Philippe Mustar, and Susan Rogers. 2009. “Genomic Platforms and Hybrid Formations.” Pp. 502–20 in Handbook of Genetics and Society: Mapping the new genomic era, edited by Paul Atkinson, Peter E. Glasner, and Margaret Lock. Routledge. Cambrosio, Alberto, Peter Keating, and Andrei Mogoutov. 2004. “Mapping Collaborative Work and Innovation in Biomedicine A Computer-Assisted Analysis of Antibody Reagent Workshops.” Social Studies of Science 34(3):325–64. Canguilhem, Georges. 2008. Knowledge of Life. 3rd ed. Fordham University Press. Carlson, Tucker. 2012. “Eugenics, American Style.” Slate, February 21. Retrieved January 12, 2013 (http://www.slate.com/articles/news_and_politics/politics/2012/02/rick_santorum_ prenatal_testing_and_abortion_tucker_carlson_s_classic_essay_on_prenatal_testin g_and_the_abortion_of_down_syndrome_babies_.html).

! 401! ! ! ! Carter, C. O., J. L. Hamerton, P. E. POLANI, A. GUNALP, and S. D. WELLER. 1960. “Chromosome Translocation as a Cause of Familial Mongolism.” Lancet 2(7152):678–80. Caspi, Avshalom et al. 2002. “Role of Genotype in the Cycle of Violence in Maltreated Children.” Science 297(5582):851–54. Chapman, By Kate. 2009. “Maori ‘Warrior Gene’ Research Slammed.” Stuff.co.nz, September 11. Retrieved April 25, 2012 (http://www.stuff.co.nz/national/2855426/Maori-warrior-gene-research-slammed). Check Hayden, Erika. 2012. “Fetal Tests Spur Legal Battle.” Nature 486(7404):454–454. Chedd, Graham. 2001. “Growing Up Different.” Scientific American Frontiers. Retrieved October 5, 2011 (http://www.pbs.org/saf/1205/resources/transcript.htm). Cicchetti, Dante, and Sheree L. Toth. 2006. “Building Bridges and Crossing Them: Translational Research in Developmental Psychopathology.” Development and Psychopathology 18(03):619–22. Clarke, Adele E., Janet K. Shim, Laura Mamo, Jennifer Ruth Fosket, and Jennifer R. Fishman. 2003. “Biomedicalization: Technoscientific Transformations of Health, Illness, and U.S. Biomedicine.” American Sociological Review 68(2):161–94. Cody, Jannine. 2009. “Reply to Letter from Drs. Ledbetter, Saul, and Moeschler.” Genetics in Medicine 11(9):682–83. Cohen, David et al. 2005. “Specific Genetic Disorders and Autism: Clinical Contribution Towards Their Identification.” Journal of Autism and Developmental Disorders 35(1):103–16. Cohen, I. L. et al. 1991. “Why Are Autism and the fragile-X Syndrome Associated? Conceptual and Methodological Issues.” American Journal of Human Genetics 48(2):195–202. Collier, J. G. 1966. “THE YY SYNDROME.” The Lancet 287(7445):1036. Collins, Francis S., Eric D. Green, Alan E. Guttmacher, and Mark S. Guyer. 2003. “A Vision for the Future of Genomics Research.” Nature 422(6934):835–47. Collins, Harry, Robert Evans, and Michael E. Gorman. 2010. “Trading Zones and Interactional Expertise.” Pp. 7–24 in Trading Zones and Interactional Expertise: Creating New Kinds of Collaboration. Cambridge, MA: MIT Press. Collins, Harry, Robert Evans, and Mike Gorman. 2007. “Trading Zones and Interactional Expertise.” Studies In History and Philosophy of Science Part A 38(4):657–66.

! 402! ! ! ! Cordeiro, Lisa, Nicole Tartaglia, David Roeltgen, and Judith Ross. 2012. “Social Deficits in Male Children and Adolescents with Sex Chromosome Aneuploidy: A Comparison of XXY, XYY, and XXYY Syndromes.” Research in Developmental Disabilities 33(4):1254–63. Court Brown, W. M. 1968. “Males with an XYY Sex Chromosome Complement.” Journal of Medical Genetics 5(4):341–59. Court Brown, W. M., W. H. Price, and P. A. Jacobs. 1968. “The XYY Male.” BMJ 4(5629):513–513. Courtens, Winnie, Inge Schramme, and Annick Laridon. 2008. “Microduplication 22q11.2: A Benign Polymorphism or a Syndrome with a Very Large Clinical Variability and Reduced penetrance?—Report of Two Families.” American Journal of Medical Genetics Part A 146A(6):758–63. Crampton, Peter, and Chris Parkin. 2007. “Warrior Genes and Risk-taking Science.” The New Zealand Medical Journal 120(1250):U2439. Crane, Diana. 1969. “Social Structure in a Group of Scientists: A Test of the ‘Invisible College’ Hypothesis.” American Sociological Review 34(3):335–52. Csardi, Gabor, and Tamas Nepusz. 2006. “The Igraph Software Package for Complex Network Research.” InterJournal, Complex Systems 1695:38. Culliton, Barbara J. 1975. “XYY: Harvard Researcher Under Fire Stops Newborn Screening.” Science 188(4195):1284–85. D’Angelo, Carla S. et al. 2010. “Extending the Phenotype of Monosomy 1p36 Syndrome and Mapping of a Critical Region for Obesity and Hyperphagia.” American Journal of Medical Genetics Part A 152A(1):102–10. Daston, Lorraine J., and Peter Galison. 2010. Objectivity. Zone Books. Davies, Gail. 2010. “Captivating Behaviour: Mouse Models, Experimental Genetics and Reductionist Returns in the Neurosciences.” The Sociological Review 58(s1):53–72. Davies, J., A. Jaffé, and A. Bush. 1998. “Distal 10q Trisomy Syndrome with Unusual Cardiac and Pulmonary Abnormalities.” Journal of Medical Genetics 35(1):72–74. Day, Philip. 2010. “Born to Rage?” Explorer. Retrieved (http://www.imdb.com/title/tt1742902/). Della Cava, Marco R. 2010. “Henry Rollins, ‘Born to Rage’ Hunt Anger’s Genetic Roots.” USA Today, December 13. Retrieved October 6, 2011 (http://www.usatoday.com/life/television/news/2010-12-14-ragetv14_ST_N.htm).

! 403! ! ! ! Dershowitz, Alan M. 2007. Préemption: A Knife That Cuts Both Ways. W. W. Norton & Company. Devlin, L., and P. J. Morrison. 2004. “Accuracy of the Clinical Diagnosis of Down Syndrome.” The Ulster Medical Journal 73(1):4–12. Dhar, S. u. et al. 2010. “22q13.3 Deletion Syndrome: Clinical and Molecular Analysis Using Array CGH.” American Journal of Medical Genetics Part A 152A(3):573– 81. DiGeorge, A. M. 1968. “Congenital Absence of the Thymus and Its Immunologic Consequences: Concurrence with Congenital Hypoparathyroidism.” Birth Defects 4(1):116–21. Dixon, Darrin P. 2008. “Informed Consent or Institutionalized Eugenics? How the Medical Profession Encourages Abortion of Fetuses with Down Syndrome.” Retrieved March 23, 2013 (http://works.bepress.com/darrin_dixon/1). Dobbs, David. 2007. “The Gregarious Brain.” The New York Times, July 8. Retrieved January 16, 2011 (http://www.nytimes.com/2007/07/08/magazine/08sociability- t.html?_r=2&pagewanted=all). Donnai, D., and A. Karmiloff-Smith. 2000. “Williams Syndrome: From Genotype through to the Cognitive Phenotype.” AMERICAN JOURNAL OF MEDICAL GENETICS 97(2):164–71. Down, John Langdon. 1866. “Observations on an Ethnic Classification of Idiots.” London Hospital Reports 3:259–62. Drew, Liam J. et al. 2011. “The 22q11.2 Microdeletion: Fifteen Years of Insights into the Genetic and Neural Complexity of Psychiatric Disorders.” International Journal of Developmental Neuroscience In Press, Corrected Proof. Retrieved October 28, 2010 (http://www.sciencedirect.com/science/article/B6T01-514Y289- 1/2/c50fa0a773a58d37136b9d82191252dc). Driscoll, Deborah A. et al. 1992. “Deletions and Microdeletions of 22q11.2 in Velo-cardio- facial Syndrome.” American Journal of Medical Genetics 44(2):261–68. Dumars, K. W., C. Gaskill, and N. Kitzmiller. 1964. “LE CRI DU CHAT (CRYING CAT) SYNDROME.” American Journal of Diseases of Children (1960) 108:533–37. Durand, Christelle M. et al. 2007. “Mutations in the Gene Encoding the Synaptic Scaffolding Protein SHANK3 Are Associated with Autism Spectrum Disorders.” Nature Genetics 39(1):25–27. Durkheim, Émile. 2001. The Elementary Forms of Religious Life. Oxford University Press.

! 404! ! ! ! Duster, Troy. 2003a. Backdoor to Eugenics. Psychology Press. Duster, Troy. 2003b. Backdoor to Eugenics. 2nd ed. Routledge. Edwards, J. H., D. G. HARNDEN, A. H. CAMERON, V. M. CROSSE, and O. H. WOLFF. 1960. “A New Trisomic Syndrome.” Lancet 1(7128):787–90. Edwards, Jim. 2011. “Death of an Insider: R&D Chief Passes Before She Can Testify in Stock Fraud Case - CBS News.” cbsnews.com. Retrieved February 13, 2012 (http://www.cbsnews.com/8301-505123_162-42847754/death-of-an-insider-rd- chief-passes-before-she-can-testify-in-stock-fraud-case/). Einfeld, Stewart L., Bruce J. Tonge, and Vaughan W. Rees. 2001. “Longitudinal Course of Behavioral and Emotional Problems in Williams Syndrome.” American Journal on Mental Retardation 106(1):73–81. Einfeld, Stewart, Helen Molony, and Wayne Hall. 1989. “Autism Is Not Associated with the Fragile X Syndrome.” American Journal of Medical Genetics 34(2):187–93. El-Haj, Nadia Abu. 2007. “The Genetic Reinscription of Race.” Annual Review of Anthropology 36(1):283–300. Elghezal, Hatem et al. 2007. “Ring Chromosome 20 Syndrome Without Deletions of the Subtelomeric and CHRNA4--KCNQ2 Genes Loci.” European Journal of Medical Genetics 50(6):441–45. Ellis, J. R., O. J. Miller, L. S. Penrose, and G. E. B. Scott. 1961. “A Male with XXYY Chromosomes.” Annals of human genetics 25(2):145–51. Ensenauer, Regina E. et al. 2003. “Microduplication 22q11.2, an Emerging Syndrome: Clinical, Cytogenetic, and Molecular Analysis of Thirteen Patients.” American Journal of Human Genetics 73(5):1027–40. Epstein, Charles J. 2002. “From Down Syndrome to the ‘Human’ in ‘Human Genetics’.” American Journal of Human Genetics 70(2):300–313. Epstein, Steven. 1996. Impure Science: AIDS, Activism, And the Politics of Knowledge. University of California Press. Epstein, Steven. 2007a. Inclusion: The Politics of Difference in Medical Research. 1st ed. University Of Chicago Press. Epstein, Steven. 2007b. “Patient Groups and Health Movements.” Pp. 499–539 in The Handbook of Science and Technology Studies, edited by Edward J. Hackett, Olga Amsterdamska, Michael E. Lynch, and Judy Wajcman. The MIT Press.

! 405! ! ! ! Van Esch, Hilde et al. 2005. “Duplication of the MECP2 Region Is a Frequent Cause of Severe Mental Retardation and Progressive Neurological Symptoms in Males.” American Journal of Human Genetics 77(3):442–53. Eyal, Gil. 2013. “For a Sociology of Expertise: The Social Origins of the Autism Epidemic.” American Journal of Sociology 118(4):863–907. Eyal, Gil, B. Hart,, E. Onculer, N. Oren, and N. Rossi. 2010. The Autism Matrix. Polity. Fan, Hei-mun Christina, and Stephen R. Quake. 2010. “Noninvasive Diagnosis of Fetal Aneuploidy by Sequencing.” Retrieved February 1, 2012 (http://www.freepatentsonline.com/y2010/0138165.html). Fang, J. S. et al. 2008. “Cytogenetic and Molecular Characterization of a Three-generation Family with Chromosome 5p Terminal Deletion.” Clinical Genetics 73(6):585–90. Feenstra, Ilse et al. 2007. “Genotype-phenotype Mapping of Chromosome 18q Deletions by High-resolution Array CGH: An Update of the Phenotypic Map.” American Journal of Medical Genetics. Part A 143A(16):1858–67. Feinstein, Carl. 2009. “Stranded, Part II.” Journal of the American Academy of Child & Adolescent Psychiatry 48(11):1049–50. Fincher, David. 1992. Alien3. Finucane, Brenda. 2011. “Interview Brenda Finucane, Indianapolis IN.” Finucane, Brenda, Barbara Haas-Givler, and Elliott W. Simon. 2003. “Genetics, Mental Retardation, and the Forging of New Alliances.” American Journal of Medical Genetics 117C(1):66–72.

Finucane, Brenda M., Deborah Konar, Barbara Haas! Givler, Michael B. Kurtt, and Charles I. Scott Jr. 1994. “The Spasmodic Upper!body Squeeze: a Chalacteristic Behavior in Smith!Magenis Syndrome.” Developmental Medicine & Child Neurology 36(1):78–83. Fischbach, Gerald. 2007. “Summary of Simons Foundation Autism Research Initiative.” Retrieved (http://iacc.hhs.gov/events/2007/transcript_113007.pdf). Fleck, Ludwik. 1981. Genesis and Development of a Scientific Fact. University Of Chicago Press. Folstein, Susan, and Michael Rutter. 1977a. “Genetic Influences and Infantile Autism.” Nature 265(5596):726–28. Folstein, Susan, and Michael Rutter. 1977b. “Infantile Autism: A Genetic Study of 21 Twin Pairs.” Journal of Child Psychology and Psychiatry 18(4):297–321.

! 406! ! ! ! Ford, C. E., K. W. Jones, P. E. Polani, J. C. De Almeida, and J. H. Briggs. 1959. “A SEX- CHROMOSOME ANOMALY IN A CASE OF GONADAL DYSGENESIS (TURNER’S SYNDROME).” The Lancet 273(7075):711–13. Fortunato, Santo. 2010. “Community Detection in Graphs.” Physics Reports 486(3–5):75– 174. Foucault, M. 2001. Order of Things: An Archaeology of the Human Sciences. 2nd ed. Routledge. Foucault, Michel. 1973. The Birth of the Clinic: An Archaeology of Medical Perception. Pantheon Books. Foucault, Michel. 1977. “The Confession of the Flesh.” Pp. 194–228 in Power/Knowledge: Selected Interviews and Other Writings, 1972-1977, edited by Colin Gordon. New York, NY: Pantheon. Foucault, Michel. 1982. The Archaeology of Knowledge & The Discourse on Language. Reprint. Vintage. Foucault, Michel. 1990a. Politics, Philosophy, Culture: Interviews and Other Writings, 1977-1984. Psychology Press. Foucault, Michel. 1990b. The History of Sexuality, Vol. 1: An Introduction. Vintage. Foucault, Michel. 2002. Archaeology of Knowledge. Routledge. Foucault, Michel. 2003. “Society Must Be Defended”: Lectures at the Collège de France, 1975-1976. Reprint. Picador. Foucault, Michel. 2009. Security, Territory, Population: Lectures at the Collège de France 1977--1978. 1st ed. edited by Michel Senellart. Picador. Foucault, Michel. 2010. The Birth of Biopolitics: Lectures at the Collège de France, 1978- -1979. Reprint. Picador. Fox, Nick J. 2011. “Boundary Objects, Social Meanings and the Success of New Technologies.” Sociology 45(1):70–85. Franklin, Sarah. 2003. “Re-thinking Nature-culture: Anthropology and the New Genetics.” Anthropological Theory 3(1):65–85. Franzke, A. W. 1975. “Letter: Telling Parents About XYY Sons.” The New England journal of medicine 293(2):100–101. Frazzetto, G. et al. 2007. “Early Trauma and Increased Risk for Physical Aggression During Adulthood: The Moderating Role of MAOA Genotype.” PLoS One 2(5):e486.

! 407! ! ! ! Freese, Jeremy. 2008. “Genetics and the Social Science Explanation of Individual Outcomes.” American Journal of Sociology 114(S1):S1–S35. Freese, Jeremy, and Sara Shostak. 2009. “Genetics and Social Inquiry.” Annual Review of Sociology 35(1):107–28. Frith, Uta. 2003. Autism: Explaining the Enigma. Wiley-Blackwell. Fujimura, Joan H., Troy Duster, and Ramya Rajagopalan. 2008. “Introduction: Race, Genetics, and Disease Questions of Evidence, Matters of Consequence.” Social Studies of Science 38(5):643–56. Fullwiley, Duana. 2007. “Race and Genetics: Attempts to Define the Relationship.” BioSocieties 2(02):221–37. Gajecka, Marzena, Katherine L. Mackay, and Lisa G. Shaffer. 2007. “Monosomy 1p36 Deletion Syndrome.” American Journal of Medical Genetics. Part C, Seminars in Medical Genetics 145C(4):346–56. Galasso, Cinzia et al. 2008. “Deletion 2q37: An Identifiable Clinical Syndrome With Mental Retardation and Autism.” Journal of Child Neurology 23(7):802–6. Galison, Peter Louis. 1997. Image and Logic: A Material Culture of Microphysics. 1st ed. University Of Chicago Press. Garber, K. 2007. “Autism’s Cause May Reside in Abnormalities at the Synapse.” Science 317(5835):190–91. Gawde, H. et al. 2006. “Chromosome 22 Microdeletion by F.I.S.H. in Isolated Congenital Heart Disease.” The Indian Journal of Pediatrics 73(10):885–88. Geerts, M., J. Steyaert, and JP Fryns. 2003. “The XYY Syndrome: A Follow-up Study on 38 Boys.” Genetic Counseling 14(3):267–79. Genealogy by Genetics, Ltd. 2010. “The Warrior Gene.” thewarriorgene. Retrieved March 29, 2012 (http://thewarriorgene.com/TheWarriorGene.html). Gibbons, A. 2004. “Tracking the Evolutionary History of a‘ Warrior’ Gene.” Science 304(5672):818. Gieryn, Thomas F. 1983. “Boundary-Work and the Demarcation of Science from Non- Science: Strains and Interests in Professional Ideologies of Scientists.” American Sociological Review 48(6):781–95. Giles, Annabel. 2008. “The XYY Factor: How a Rare Chromosome Disorder Brought My Son a World of Pain.” The Daily Mail, November 1. Retrieved

! 408! ! ! ! (http://www.dailymail.co.uk/health/article-1082293/The-XYY-Factor-How-rare- chromosome-disorder-brought-son-world-pain.html). Gillberg, Christopher. 1983. “IDENTICAL TRIPLETS WITH INFANTILE-AUTISM AND THE FRAGILE-X SYNDROME.” British Journal of Psychiatry 143(SEP):256–60. Gillberg, Christopher. 1998. “Chromosomal Disorders and Autism.” Journal of Autism and Developmental Disorders 28(5):415–25. Gillberg, Christopher, and Peder Rasmussen. 1994. “Brief Report: Four Case Histories and a Literature Review of Williams Syndrome and Autistic Behavior.” Journal of Autism and Developmental Disorders 24(3):381–93. Gillberg, Christopher, Inger Winnergard, and Jan Wahlström. 1984. “The Sex Chromosomes-one Key to Autism? An XYY Case of Infantile Autism.” Applied Research in Mental Retardation 5(3):353–60. Glaser, Sarah E., and Steven R. Shaw. 2011. “Emotion Regulation and Development in Children with Autism and 22q13 Deletion Syndrome: Evidence for Group Differences.” Research in Autism Spectrum Disorders 5(2):926–34. Glass, B. 1971. “Science: Endless Horizons or Golden Age?” Science (New York, N.Y.) 171(3966):23–29. Goh, Kwang-Il et al. 2007. “The Human Disease Network.” Proceedings of the National Academy of Sciences 104(21):8685–90. Goizet, C. et al. 2000. “Case with Autistic Syndrome and Chromosome 22q13.3 Deletion Detected by FISH.” American Journal of Medical Genetics 96(6):839–44. Goldberg, R., R. Marion, M. Borderon, A. Wiznia, and R. J. Shprintzen. 1985. “Phenotypic Overlap Between Velo-cardio-facial Syndrome and Di George Sequence.” American Journal of Human Genetics 37(Suppl):A43–A84. Gould, Stephen Jay. 1996. The Mismeasure of Man. REV EXP. W. W. Norton & Company. Griffiths, Dorothy, and Robert King. 2004. Demystifying Syndromes: Clinical and Educational Implications of Common Syndromes Associated with Persons with Intellectual Disabilities. NADD. Grisart, B. et al. 2009. “17q21.31 Microduplication Patients Are Characterised by Behavioural Problems and Poor Social Interaction.” Journal of Medical Genetics 46(8):524–30.

! 409! ! ! ! Gropman, Andrea L., Wallace C. Duncan, and Ann C. M. Smith. 2006. “Neurologic and Developmental Features of the Smith-Magenis Syndrome (del 17p11.2).” Pediatric Neurology 34(5):337–50. De Grouchy, J., M. ARTHUIS, C. SALMON, M. LAMY, and S. THIEFFRY. 1964. “[THE CRYING CAT SYNDROME: A FURTHER CASE.].” Annales De Génétique 7:13–16. De Grouchy, J., J. Bonnette, and C. Salmon. 1966. “[Deletion of the Short Arm of Chromosome 18].” Annales De Génétique 9(1):19–26. De Grouchy, J., M. Lamy, St Thieffry, M. Arthuis, and C. H. Salmon. 1963. “Dysmorphie Complexe Avec Oligophrenie: Deletion Des Bras Courts D’un Chromosome 17– 18.” CR Acad Sci 258:1028. De Grouchy, J., A. Rossier, and N. Joab. 1967. “Une Nouvelle Observation D’abberation Chromosome 18p- [A New Observation of Chromosome Aberrations 18p].” Annales De Génétique 10(4):221–23. Grouchy, Jean de, and Catherine Turleau. 1977. Clinical Atlas of Human Chromosomes. Wiley. Gustavson, K. H., W. Arancibia, U. Eriksson, and L. Svennerholm. 1986. “Deleted Ring Chromosome 22 in a Mentally Retarded Boy.” Clinical Genetics 29(4):337–41. Guzman, Maria Luisa et al. 2012. “Growth in Chilean Infants with Chromosome 22q11 Microdeletion Syndrome.” American Journal of Medical Genetics Part A 158A(11):2682–86. Habel, Alex, Michael-John McGinn, Elaine H. Zackai, Nancy Unanue, and Donna M. McDonald-McGinn. 2012. “Syndrome-specific Growth Charts for 22q11.2 Deletion Syndrome in Caucasian Children.” American Journal of Medical Genetics Part A 158A(11):2665–71.

Hacking, Ian. 1995. “The Looping Effects of Human Kinds.” Pp. 351–94 in Causal cognition: A multidisciplinary debate, Symposia of the Fyssen Foundation., edited by D. Sperber, D. Premack, and A. J. Premack. New York, NY, US: Oxford University Press. Hacking, Ian. 1998. Rewriting the Soul. Princeton University Press. Hacking, Ian. 2006a. “Genetics, Biosocial Groups & the Future of Identity.” Daedalus 135(4):81–95. Hacking, Ian. 2006b. “Making Up People.” London Review of Books, August 17, 23–26.

! 410! ! ! ! Hacking, Ian. 2006c. “What Is Tom Saying to Maureen?” London Review of Books, May 11, 3–7. Hacking, Ian. 2007. “Kinds of People: Moving Targets.” Proceedings of the British Academy 51:285–318. Hagerman, Randi J. et al. 1986. “An Analysis of Autism in Fifty Males with the Fragile X Syndrome.” American Journal of Medical Genetics 23(1!2):359–74. Hagerman, Randi J. et al. 2009. “Advances in the Treatment of Fragile X Syndrome.” Pediatrics 123(1):378–90. Hagerman, Randi Jenssen. 1999. Neurodevelopmental Disorders: Diagnosis and Treatment. Oxford University Press. Hagerman, Rj, and Aw Jackson. 1985. “Autism or Fragile-X Syndrome.” Journal of the American Academy of Child and Adolescent Psychiatry 24(2):239–40. Halder, Ashutosh, Manish Jain, Isha Chaudhary, and Madhulika Kabra. 2010. “Prevalence of 22q11.2 Microdeletion in 146 Patients with Cardiac Malformation in a Referral Hospital of North India.” BMC Medical Genetics 11:101. Hall, J. G. 1993. “CATCH 22.” Journal of Medical Genetics 30(10):801–2. Hamamy, Hanan A., and Sophie Dahoun. 2004. “Parental Decisions Following the Prenatal Diagnosis of Sex Chromosome Abnormalities.” European Journal of Obstetrics & Gynecology and Reproductive Biology 116(1):58–62. Hamerton, J. L. 1976. “Human Population Cytogenetics: Dilemmas and Problems.” American Journal of Human Genetics 28(2):107–22. Hamerton, John L. 1963. Chromosomes in Medicine. New York: Medical Advisory Committee of the National Spastics Society, in Association with Wm. Heinemann (Medical Books) Ltd. Hamerton, John L. 1971. Human Cytogenetics: Clinical Cytogenetics. Academic Press. Hammond, Peter. 2007. “The Use of 3D Face Shape Modelling in Dysmorphology.” Archives of Disease in Childhood 92(12):1120–26. Hansen, Bent Sigurd. 2005. “Something Rotten in the State of Denmark: Eugnics and the Ascent of the Welfare State.” Pp. 9–76 in Eugenics And the Welfare State: Sterilization Policy in Demark, Sweden, Norway, and Findland, edited by Gunnar Broberg and Nils Roll-Hansen. MSU Press. Harmon, Amy. 2007a. “After DNA Diagnosis: ‘Hello, 16p11.2. Are You Just Like Me?’” The New York Times, December 28. Retrieved December 22, 2008

! 411! ! ! ! (http://www.nytimes.com/2007/12/28/health/research/28dna.html?_r=2&em&ex=1 198990800&en=584ecb49c0c7776f&ei=5087%0A&oref=slogin). Harmon, Amy. 2007b. “Prenatal Test Puts Down Syndrome in Hard Focus.” The New York Times, May 9. Retrieved March 23, 2013 (http://www.nytimes.com/2007/05/09/us/09down.html). Harper, Peter. 2006. First Years of Human Chromosomes: The Beginnings of Human Cytogenetics. 1st ed. Scion Publishing Ltd. Hauge, Xueya et al. 2008. “Detailed Characterization of, and Clinical Correlations in, 10 Patients with Distal Deletions of Chromosome 9p.” Genetics in Medicine: Official Journal of the American College of Medical Genetics 10(8):599–611. Hauschka, Ts, Aa Sandberg, Mn Goldstein, Je Hasson, and Gf Koepf. 1962. “An XYY Man with Progeny Indicating Familial Tendency to Non-Disjunction.” American Journal of Human Genetics 14(1):22–&. Hayden, Erika Check. 2008. “Biological Tools Revamp Disease Classification.” Nature 453(7196):709. Haydu, Jeffrey. 1998. “Making Use of the Past: Time Periods as Cases to Compare and as Sequences of Problem Solving.” American Journal of Sociology 104(2):339–71.

Heath, Deborah, Rayna Rapp, and Karen!Sue Taussig. 2004a. “Genetic Citizenship.” Pp. 152–67 in A Companion to the Anthropology of Politics, edited by David Nugent and Joan Vincent. Malden, MA: Blackwell.

Heath, Deborah, Rayna Rapp, and Karen!Sue Taussig. 2004b. “Genetic Citizenship.” 152– 67. Hedgecoe, A. 1998. “Geneticization, Medicalisation and Polemics.” Medicine, Health Care and Philosophy 1(3):235–43.

Hedgecoe, A. M. 2000. “The Popularization of Genetics as Geneticization.” Public Understanding of Science 9(2):183. Hedgecoe, Adam. 2001. “Schizophrenia and the Narrative of Enlightened Geneticization.” Social Studies of Science 31(6):875–911. Hedgecoe, Adam. 2003. “Expansion and Uncertainty: Cystic Fibrosis, Classification and Genetics.” Sociology of Health & Illness 25(1):50–70. Hedgecoe, Adam. 2008. “From Resistance to Usefulness: Sociology and the Clinical Use of Genetic Tests.” Biosocieties 3(02):183–94.

! 412! ! ! ! Herguner, Sabri, and Nahit Motavalli Mukaddes. 2006. “Autism and Williams Syndrome: A Case Report.” World Journal of Biological Psychiatry 7(3):186–88. Herva, R., I. Saarinen, and L. Leikkonen. 1977. “The R(20) Syndrome.” Journal of Medical Genetics 14(4):281–83. Hilts, Philip J. 1992. “U.S. Puts a Halt To Talks Tying Genes to Crime.” New York Times, September 5. Retrieved January 13, 2013 (http://www.nytimes.com/1992/09/05/us/us-puts-a-halt-to-talks-tying-genes-to- crime.html). Hinkel, G. K. et al. 1997. “Overgrowth and retarded development associated with chromosome 22q13 deletion.” Monatsschrift Kinderheilkunde 145(9):892–96. Hogart, A. et al. 2009. “Chromosome 15q11–13 Duplication Syndrome Brain Reveals Epigenetic Alterations in Gene Expression Not Predicted from Copy Number.” Journal of medical genetics 46(2):86–93. Hogarth, Stuart, Gail Javitt, and David Melzer. 2008. “The Current Landscape for Direct- to-Consumer Genetic Testing: Legal, Ethical, and Policy Issues.” Annual Review of Genomics and Human Genetics 9(1):161–82. Holroyd, J., and D. Mcarthur. 1976. “Mental-Retardation and Stress on Parents - Contrast Between Downs-Syndrome and Childhood Autism.” American Journal of Mental Deficiency 80(4):431–36. Hook, E. B., and D. M. Schreinemachers. 1983. “Trends in Utilization of Prenatal Cytogenetic Diagnosis by New York State Residents in 1979 and 1980.” American Journal of Public Health 73(2):198–202. Horgan, John. 2011. “Born Outlaws? A Criminally Feeble Proposition.” The Chronicle of Higher Education, July 17. Retrieved October 3, 2011 (http://chronicle.com/article/Born-Outlaws-A-Criminally/128197/).

Insel, Thomsas. 2013. “Transforming Diagnosis.” NIMH Director’s Blog. Retrieved (http://www.nimh.nih.gov/about/director/2013/transforming-diagnosis.shtml). Jacob, François. 1993. The Logic of Life. New edition. Princeton University Press. Jacobs, P. A. 1982. “The William Allan Memorial Award Address: Human Population Cytogenetics: The First Twenty-five Years.” American Journal of Human Genetics 34(5):689–98. Jacobs, P. A., and J. A. Strong. 1959. “A Case of Human Intersexuality Having a Possible XXY Sex-determining Mechanism.” Nature 183(4657):302–3.

! 413! ! ! ! Jacobs, Patricia A., MURIEL Brunton, MARIE M. Melville, R. P. Brittain, and W. F. McClemont. 1965. “Aggressive Behaviour, Mental Sub-normality and the XYY Male.” Nature 208(5017):1351–52. Jacobs, Patricia A., W. H. Price, W. M. Court Brown, R. P. Brittain, and P. B. Whatmore. 1968. “Chromosome Studies on Men in a Maximum Security Hospital.” Annals of Human Genetics 31(4):339–58. Jacobs, PatriciaA., A. G. Baikie, et al. 1959. “EVIDENCE FOR THE EXISTENCE OF THE HUMAN ‘SUPER FEMALE’.” The Lancet 274(7100):423–25. Jacobs, PatriciaA., W. M. Court Brown, A. G. Baikie, and J. A. Strong. 1959. “THE SOMATIC CHROMOSOMES IN MONGOLISM.” The Lancet 273(7075):710. Jarvik, Lissy F., Victor Klodin, and Steven S. Matsuyama. 1973. “Human Aggression and the Extra Y Chromosome: Fact or Fantasy?” American Psychologist 28(8):674–82. Järvinen-Pasley, Anna et al. 2008. “Defining the Social Phenotype in Williams Syndrome: A Model for Linking Gene, the Brain, and Behavior.” Development and Psychopathology 20(01):1–35. Jasanoff, Sheila. 2013. States of Knowledge: The Co-Production of Science and the Social Order. Taylor & Francis. Jeon, Kwon Chan, Lei-Shih Chen, and Patricia Goodson. 2012. “Decision to Abort after a Prenatal Diagnosis of Sex Chromosome Abnormality: a Systematic Review of the Literature.” Genet Med 14(1):27–38. Johnston, A. W., M. A. Ferguson-Smith, S. D. Handmaker, Howard W. Jones, and Georgeanna S. Jones. 1961. “The Triple-X Syndrome.” British Medical Journal 2(5259):1046–52. Jones, Wendy et al. 2000. “II. Hypersociability in Williams Syndrome.” Journal of Cognitive Neuroscience 12(supplement 1):30–46. Jorde, Lynn B. 2012. “2011 Presidential Address: From Classroom to Courtroom to Clinic—Closing the Gaps in Human Genetics Education.” The American Journal of Human Genetics 90(3):387–89. Kanner, L. 1943. “Autistic Disturbances of Affective Contact.” Acta Paedopsychiatrica 35(4):100–136. Karayiorgou, Maria, Tony J. Simon, and Joseph A. Gogos. 2010. “22q11.2 Microdeletions: Linking DNA Structural Variation to Brain Dysfunction and Schizophrenia.” Nature Reviews Neuroscience 11(6):402–16.

! 414! ! ! ! Kay, Lily E. 1996. The Molecular Vision of Life: Caltech, the Rockefeller Foundation, and the Rise of the New Biology. Oxford University Press, USA. Keating, Peter, and Alberto Cambrosio. 2006. Biomedical Platforms: Realigning the Normal and the Pathological in Late-Twentieth-Century Medicine. 1st ed. The MIT Press. Kent, Jon. 2011. Embraceable. Kerr, A. 2004. “Giving up on Geneticization: a Comment on Hedgecoe’s ‘Expansion and Uncertainty: Cystic Fibrosis, Classification and Genetics’.” Sociology of health & illness 26(1):102–6. Kerr, Anne. 2000. “(Re)Constructing Genetic Disease.” Social Studies of Science 30(6):847 –894. Kerr, Anne. 2004. “Giving up on Geneticization: a Comment on Hedgecoe’s ‘Expansion and Uncertainty: Cystic Fibrosis, Classification and Genetics’.” Sociology of Health & Illness 26(1):102–6. Kerr, Anne, Sarah Cunningham-Burley, and Amanda Amos. 1998a. “Eugenics and the New Genetics in Britain: Examining Contemporary Professionals’ Accounts.” Science, Technology & Human Values 23(2):175–98. Kerr, Anne, Sarah Cunningham-Burley, and Amanda Amos. 1998b. “The New Genetics and Health: Mobilizing Lay Expertise.” Public Understanding of Science 7(1):41 – 60. Kevles, Daniel J. 1985. In the Name of Eugenics: Genetics and the Uses of Human Heredity. University of California Press. King, Marissa, and Peter Bearman. 2009. “Diagnostic Change and the Increased Prevalence of Autism.” Int. J. Epidemiol. 38(5):1224–34. Kingsbury, Kathleen. 2009. “Which Kids Join Gangs? A Genetic Explanation.” Time, June 10. Retrieved April 25, 2012 (http://www.time.com/time/health/article/0,8599,1903703,00.html). Kirk, E. P. et al. 2009. “The Clinical Variability of the MECP2 Duplication Syndrome: Description of Two Families with Duplications Excluding L1CAM and FLNA.” Clinical Genetics 75(3):301–3. Kitzman, Jacob O. et al. 2012. “Noninvasive Whole-Genome Sequencing of a Human Fetus.” Science Translational Medicine 4(137):137ra76–137ra76. Kline, A. D. et al. 1993. “Molecular Analysis of the 18q- Syndrome--and Correlation with Phenotype.” American Journal of Human Genetics 52(5):895–906.

! 415! ! ! ! Knox, Richard. 1974. “Scientists Oppose Survey of Chromosome Disorder.” Boston Globe, November 16, 3. Kogan, Michael D. et al. 2009. “Prevalence of Parent-Reported Diagnosis of Autism Spectrum Disorder Among Children in the US, 2007.” Pediatrics 124(5):1395– 1403. Kolata, Gina. 2012. “Scientists to Seek Clues to Violence in Genome of Gunman in Newtown, Conn.” The New York Times, December 24. Retrieved December 26, 2012 (http://www.nytimes.com/2012/12/25/science/scientists-to-seek-clues-to- violence-in-genome-of-gunman-in-newtown-conn.html). Kolevzon, Alexander. 2012. “Clinical Trial in 22q13 Deletion Syndrome(Phelan- McDermid Syndrome).” Retrieved (http://clinicaltrials.gov/show/NCT01525901). Koolen, D. A. et al. 2008. “Clinical and Molecular Delineation of the 17q21.31 Microdeletion Syndrome.” Journal of Medical Genetics 45(11):710–20. Koolen, David A. et al. 2006. “A New Chromosome 17q21.31 Microdeletion Syndrome Associated with a Common Inversion Polymorphism.” Nature Genetics 38(9):999– 1001. Koroshetz, Walter. 2007. “Summary of Autism Spectrum Disorder (ASD) Related ActivitiesNational Institute of Neurological Disorders and Stroke.” Retrieved (http://iacc.hhs.gov/events/2007/transcript_113007.pdf). Kouser, Mehreen. 2011. “Shank Synaptic Genes in Autism: Human Genetics to Mouse Models and Therapeutics.” Autism Speaks Official Blog. Retrieved January 16, 2013 (http://blog.autismspeaks.org/2011/05/24/shan/). KREMER, EJ et al. 1991. “MAPPING OF DNA INSTABILITY AT THE FRAGILE-X TO A TRINUCLEOTIDE REPEAT SEQUENCE P(CCG)N.” SCIENCE 252(5013):1711–14.

Kripke, Saul A. 1980. Naming and Necessity. Wiley. Krueger, Dilja D., and Mark F. Bear. 2011. “Toward Fulfilling the Promise of Molecular Medicine in Fragile X Syndrome.” Annual Review of Medicine 62(1):411–29. Kumar, Ravinesh A. et al. 2008. “Recurrent 16p11.2 Microdeletions in Autism.” Human Molecular Genetics 17(4):628–38. Lamy, Maurice, and Jean de Grouchy. 1967. L’ homme et l’hérédité. Hachette. Lancet. 1966. “The YY Syndrome.” The Lancet 287(7437):583–84.

! 416! ! ! ! Larson, R. S., and M. G. Butler. 1995. “Use of Fluorescence in Situ Hybridization (FISH) in the Diagnosis of DiGeorge Sequence and Related Diseases.” Diagnostic Molecular Pathology: The American Journal of Surgical Pathology, Part B 4(4):274–78. Latour, Bruno. 1988. Science in Action: How to Follow Scientists and Engineers through Society. Harvard University Press. Latour, Bruno. 1993. The Pasteurization of France. Harvard University Press. Latour, Bruno. 1999. Pandora’s Hope: Essays on the Reality of Science Studies. Cambridge, Mass.: Harvard University Press. Latour, Bruno. 2007. Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford University Press, USA. Law, John. 1992. “Notes on the Theory of the Actor-network: Ordering, Strategy, and Heterogeneity.” Systemic Practice and Action Research 5(4):379–93. Lea, R., and G. Chambers. 2007. “Monoamine Oxidase, Addiction, and the ‘warrior’ Gene Hypothesis.” Journal of the New Zealand Medical Association 120(1250):U2441. Lea, Rod, D. Hall, M. Green, and Geoffrey Chambers. 2005. “Tracking the Evolutionary History of the Warrior Gene in the South Pacific.” Auckland, Australia. Ledbetter, D. H. 2009. “‘New Microdeletion Syndromes: Complex, but No New Paradigms’.” Journal of Medical Genetics 46(8):576. Ledbetter, David H. 2008. “Cytogenetic Technology -- Genotype and Phenotype.” N Engl J Med 359(16):1728–30. Ledbetter, David H. 2009. “Response to Saul and Moeschler ‘How Best to Use CGH Arrays in the Clinical Setting’.” Genetics in Medicine 11(5):371–72. Lederberg, Joshua. 1968. “‘Criminal Genetic Types’ Pose a Research Dilemma.” The Washington Post, October 19, A15. Leicht, E. A., and M. E. J. Newman. 2008. “Community Structure in Directed Networks.” Physical review letters 100(11):118703. Lejeune, J. et al. 1960. “A PROPOSED STANDARD SYSTEM OF NOMENCLATURE OF HUMAN MITOTIC CHROMOSOMES.” The Lancet 275(7133):1063–65. Lejeune, J. et al. 1963. “[3 CASES OF PARTIAL DELETION OF THE SHORT ARM OF A 5 CHROMOSOME.].” Comptes Rendus Hebdomadaires Des Séances De l’Académie Des Sciences 257:3098–3102.

! 417! ! ! ! Lejeune, J., R. Turpin, and M. Gautier. 1959. “Le Mongolisme, Premier Exemple D’aberration Autosomique Humaine.” Ann Genet 1:41–49. Lenhoff, Howard M. 1999. “A Real-World Source for the ‘Little People’: A Comparison of Fairies to Individuals with Williams Syndrome.”Pp. 150–60 in Nursery Realms: Children in the Worlds of Science Fiction, Fantasy, and Horror, edited by Gary Westfahl and George Edgar Slusser. London: University of Georgia Press. Lenhoff, Howard M., Paul P. Wang, Frank Greenberg, and Ursula Bellugi. 1997. “Williams Syndrome and the Brain.” Scientific American 277(6):68–73. Leroi, Armand Marie. 2006. “The Future of Neo-eugenics. Now That Many People Approve the Elimination of Certain Genetically Defective Fetuses, Is Society Closer to Screening All Fetuses for All Known Mutations?” EMBO Reports 7(12):1184–87. Levitas, A. et al. 1983a. “Autism and the Fragile X Syndrome.” Journal of Developmental and Behavioral Pediatrics: JDBP 4(3):151–58. Levitas, A. et al. 1983b. “Autism and the Fragile X Syndrome.” Journal of Developmental and Behavioral Pediatrics: JDBP 4(3):151–58. Lindee, M. Susan. 2008. Moments of Truth in Genetic Medicine. 1st ed. The Johns Hopkins University Press. Lippman, Abby. 1991a. “Prenatal Genetic Testing and Screening: Constructing Needs and Reinforcing Inequities.” American Journal of Law & Medicine 17:15. Lippman, Abby. 1991b. “The Geneticization of Health and Illness: Implications for Social Practice.” Endocrinologie 29(1-2):85–90. Lippman, Abby. 1994. “The Genetic Construction of Prenatal Testing.” Women and prenatal testing 13–14. Lippman, Abby. 1998. “The Politics of Health: Geneticization Versus Health Promotion.” The politics of women’s health: exploring agency and autonomy 64–82. Lipson, A., B. Emanuel, P. Colley, K. Fagan, and DA Driscoll. 1994. “‘CATCH 22’ Sans Cardiac Anomaly, Thymic Hypoplasia, Cleft-palate, and Hypocalcemia: cAtch 22. A Common Result of 22q11 Deficiency?” Journal of Medical Genetics 31(9):741– 741. Lipson, A. H. et al. 1991. “Velocardiofacial (Shprintzen) Syndrome: An Important Syndrome for the Dysmorphologist to Recognise.” Journal of Medical Genetics 28(9):596–604.

! 418! ! ! ! LIU, KAYUET, NOAM ZERUBAVEL, and PETER BEARMAN. 2010. “Social Demographic Change and Autism.” Demography 47(2):327–43. Lo, Y. M. et al. 1997. “Presence of Fetal DNA in Maternal Plasma and Serum.” the lancet 350(9076):485–87. Lock, Margaret. 2005. “Eclipse of the Gene and the Return of Divination.” Current Anthropology 46(s5):S47–S70. Loscalzo, Joseph, Isaac Kohane, and Albert-Laszlo Barabasi. 2007. “Human Disease Classification in the Postgenomic Era: A Complex Systems Approach to Human Pathobiology.” Molecular Systems Biology 3(124). Retrieved June 8, 2012 (http://www.nature.com/msb/journal/v3/n1/full/msb4100163.html). Losowsky, M. S. 1961. “Hereditary Mental Defect Showing the Pattern of Sex Influence.” Journal of Mental Deficiency Research 5:60–62. Louise Slaughter. 2008. Genetic Information Nondiscrimination Act. Retrieved April 26, 2012 (http://thomas.loc.gov/cgi-bin/bdquery/z?d110:H.R.493:). Lovett, Edward. 2012. “Williams Syndrome Grows in Awareness, Research.” ABC News. Retrieved April 26, 2012 (http://abcnews.go.com/blogs/health/2012/03/16/williams-syndrome-grows-in- awareness-research/). LOWERY, MC et al. 1995. “STRONG CORRELATION OF ELASTIN DELETIONS, DETECTED BY FISH, WITH WILLIAMS-SYNDROME - EVALUATION OF 235 PATIENTS.” AMERICAN JOURNAL OF HUMAN GENETICS 57(1):49–53. Lubs, H. A. 1969. “A Marker X Chromosome.” American Journal of Human Genetics 21(3):231–44. Lynch, D. R. et al. 1995. “Cerebellar Atrophy in a Patient with Velocardiofacial Syndrome.” Journal of Medical Genetics 32(7):561 –563. Lyons, Richard D. 1968a. “Genetic Abnormality Is Linked to Crime; Genetics Linked to Violent Crimes.” The New York Times, April 21, 1. Lyons, Richard D. 1968b. “Ultimate Speck Appeal May Cite a Genetic Defect.” The New York Times, April 22. Retrieved November 20, 2012 (http://select.nytimes.com/gst/abstract.html?res=F20C10FA355D147493C0AB178 FD85F4C8685F9). Manning, M. A. et al. 2004. “Terminal 22q Deletion Syndrome: A Newly Recognized Cause of Speech and Language Disability in the Autism Spectrum.” Pediatrics 114(2):451–57.

! 419! ! ! ! Manning, Melanie A. et al. 2004. “Terminal 22q Deletion Syndrome: a Newly Recognized Cause of Speech and Language Disability in the Autism Spectrum.” Pediatrics 114(2):451–57. Mansfield, Caroline, Suellen Hopfer, and Theresa M. Marteau. 1999. “Termination Rates after Prenatal Diagnosis of Down Syndrome, Spina Bifida, Anencephaly, and Turner and Klinefelter Syndromes: a Systematic Literature Review.” Prenatal Diagnosis 19(9):808–12. Martin, J. Purdon, and Julia Bell. 1943. “A PEDIGREE OF MENTAL DEFECT SHOWING SEX-LINKAGE.” Journal of Neurology and Psychiatry 6(3-4):154–57. McCabe, Linda L., and Edward R. B. McCabe. 2011. “Down Syndrome: Coercion and Eugenics.” Genetics in Medicine 13(8):708–10. McDermott, Rose, Dustin Tingley, Jonathan Cowden, Giovanni Frazzetto, and Dominic D. P. Johnson. 2009. “Monoamine Oxidase A Gene (MAOA) Predicts Behavioral Aggression Following Provocation.” 106(7):2118–23. McDonald-McGinn, D. M., E. H. Zackai, and D. Low. 1997. “What’s in a Name? The 22q11.2 Deletion.” American Journal of Medical Genetics 72(2):247–49. McDonald-McGinn, Donna. 2011. “Interview with Donna McDonald-McGinn, Philadelphia PA.” McKusick, Victor A. 1966. Mendelian Inheritance in Man: A Catalog of Autosomal Dominant, Autosomal Recessive, and X-Linked Phenotypes. Baltimore, MD: JHU Press. McKusick, Victor A. 1993. “Medical Genetics.” JAMA: The Journal of the American Medical Association 270(19):2351 –2356. McKusick, Victor A. 1998. Mendelian Inheritance in Man: A Catalog of Human Genes and Genetic Disorders. Baltimore, MD: JHU Press. McLennan, Yingratana, Jonathan Polussa, Flora Tassone, and Randi Hagerman. 2011. “Fragile X Syndrome.” Current Genomics 12(3):216–24. Mefford, Heather C. et al. 2008. “Recurrent Rearrangements of Chromosome 1q21.1 and Variable Pediatric Phenotypes.” N Engl J Med 359(16):1685–99. Meyer-Lindenberg, Andreas et al. 2006. “Neural Mechanisms of Genetic Risk for Impulsivity and Violence in Humans.” Proceedings of the National Academy of Sciences 103(16):6269–74.

! 420! ! ! ! Meyer-Lindenberg, Andreas, and Daniel R. Weinberger. 2006. “Intermediate Phenotypes and Genetic Mechanisms of Psychiatric Disorders.” Nature Reviews Neuroscience 7(10):818–27. Mikkelsen, M., G. Fischer, J. Hansen, B. Pilgaard, and J. Nielsen. 1983. “The Impact of Legal Termination of Pregnancy and of Prenatal Diagnosis on the Birth Prevalence of Down Syndrome in Denmark.” Annals of human genetics 47(Pt 2):123–31. Miller, David T. et al. 2010. “Consensus Statement: Chromosomal Microarray Is a First- Tier Clinical Diagnostic Test for Individuals with Developmental Disabilities or Congenital Anomalies.” The American Journal of Human Genetics 86(5):749–64. Miller, Fiona Alice, Catherine Ahern, Jacqueline Ogilvie, Mita Giacomini, and Lisa Schwartz. 2005. “Ruling in and Ruling Out: Implications of Molecular Genetic Diagnoses for Disease Classification.” Social Science & Medicine 61(12):2536–45. Miller, Fiona Alice, Megan E. Begbie, Mita Giacomini, Catherine Ahern, and Erin A. Harvey. 2006. “Redefining Disease? The Nosologic Implications of Molecular Genetic Knowledge.” Perspectives in Biology and Medicine 49(1):99–114. Miller, Kimberley A. 2008. “FISH Diagnosis of 22q11.2 Deletion Syndrome.” Newborn and Infant Nursing Reviews 8(1):e11–e19. Moeschler, John B., T. K. Mohandas, Arnold B. Hawk, and Walter W. Noll. 2002. “Estimate of Prevalence of Proximal 15q Duplication Syndrome.” American Journal of Medical Genetics 111(4):440–42. Moessner, R. et al. 2007. “Contribution of< I> SHANK3 Mutations to Autism Spectrum Disorder.” The American Journal of Human Genetics 81(6):1289–97. Mol, Annemarie. 2002. The Body Multiple: Ontology in Medical Practice. Duke University Press. Money, J., A. Franzke, and D. S. Borgaonkar. 1975. “XYY Syndrome, Stigmatization, Social Class, and Aggression: Study of 15 Cases.” Southern medical journal 68(12):1536–42. Money, John, Charles Annecillo, Barry Van Orman, and Digamber S. Borgaonkar. 1974. “Cytogenetics, Hormones and Behavior Disability: Comparison of XYY and XXY Syndromes.” Clinical Genetics 6(5):370–82. Morell, V. 1993. “Evidence Found for a Possible ‘Aggression Gene’.” Science 260(5115):1722–23. Morelle, Rebecca. 2007. “3D Face Scans Spot Gene Syndromes.” BBC News, September 9. Retrieved November 18, 2009 (http://news.bbc.co.uk/1/hi/sci/tech/6982030.stm).

! 421! ! ! ! Morgan, Thomas Hunt. 1915. The Mechanism of Mendelian Heredity. Holt. Morton, J. E. et al. 1997. “Fragile X Syndrome Is Less Common Than Previously Estimated.” Journal of Medical Genetics 34(1):1–5. Moss, Joanna, Chris Oliver, Lisa Nelson, Caroline Richards, and Scott Hall. 2013. “Delineating the Profile of Autism Spectrum Disorder Characteristics in Cornelia de Lange and Fragile X Syndromes.” American Journal on Intellectual and Developmental Disabilities 118(1):55–73. Mukaddes, Nahit Motavalli, and Sabri Herguner. 2007. “Autistic Disorder and 22q11.2 Duplication.” The World Journal of Biological Psychiatry: The Official Journal of the World Federation of Societies of Biological Psychiatry 8(2):127–30. N.A. 1968. “XYY Chromosome Defense, The.” Georgetown Law Journal 57:892–922. Naber, John M., Carl A. Huether, and Beth A. Goodwin. 1987. “Temporal Changes in Ohio Amniocentesis Utilization During the First Twelve Years (1972–1983), and Frequency of Chromosome Abnormalities Observed.” Prenatal Diagnosis 7(1):51– 65. Natoli, Jaime L., Deborah L. Ackerman, Suzanne McDermott, and Janice G. Edwards. 2012. “Prenatal Diagnosis of Down Syndrome: a Systematic Review of Termination Rates (1995–2011).” Prenatal Diagnosis 32(2):142–53. Navon, Daniel. 2011. “Genomic Designation: How Genetics Can Delineate New, Phenotypically Diffuse Medical Categories.” Social Studies of Science 41(2):203 – 226. Navon, Daniel. 2012. “Genetic Counseling, Activism and ‘Genotype-First’ Diagnosis of Developmental Disorders.” Journal of Genetic Counseling 21(6):770–76. Navon, Daniel, and Uri Shwed. 2012. “The Chromosome 22q11.2 Deletion: From the Unification of Biomedical Fields to a New Kind of Genetic Condition.” Social Science & Medicine 75(9):1633–41. Nelkin, Dorothy. 1996. “The Social Dynamics of Genetic Testing: The Case of Fragile-X.” Medical Anthropology Quarterly 10(4):537–50. Nelson, A. 2008. “Bio Science: Genetic Genealogy Testing and the Pursuit of African Ancestry.” Social Studies of Science 38(5):759–83. Newsweek. 1970. “Congenital Criminals.” Newsweek, 98–99. Nickerson, E., F. Greenberg, M. T. Keating, C. McCaskill, and L. G. Shaffer. 1995. “Deletions of the Elastin Gene at 7q11.23 Occur in Approximately 90% of Patients with Williams Syndrome.” American Journal of Human Genetics 56(5):1156–61.

! 422! ! ! ! Nielsen, Johannes, Kirsten Riber Christensen, Ursula Friedrich, Eva Zeuthen, and Olaf Østergaard. 1973. “Childhood of Males with the XYY Syndrome.” Journal of Autism and Developmental Disorders 3(1):5–26. Novas, Carlos. 2006. “The Political Economy of Hope: Patients’ Organizations, Science and Biovalue.” BioSocieties 1(03):289–305. OBERLE, I. et al. 1991. “INSTABILITY OF A 550 BASE PAIR DNA SEGMENT AND ABNORMAL METHYLATION IN FRAGILE X-SYNDROME.” SCIENCE 252(5009):1097–1102. Opitz, John M., Allan T. Segal, Hallgrim Klove, Charles Mathews, and Robert L. Lehrke. 1965. “X-linked Mental Retardation. Study of a Large Kindred with 20 Affected Members.” The Journal of Pediatrics 67(4):713–14. Osborne, L. R. et al. 1996. “Identification of Genes from a 500-kb Region at 7q11. 23 That Is Commonly Deleted in Williams Syndrome Patients.” Genomics 36(2):328–36. Otto, James H., and A. Towle. 1973. Modern Biology: 1973. Book World Promotions. Overhauser, J. et al. 1994. “Molecular and Phenotypic Mapping of the Short Arm of Chromosome 5: Sublocalization of the Critical Region for the Cri-du-chat Syndrome.” Human Molecular Genetics 3(2):247–52. Pagon, Roberta A. et al. 2009. “22q11. 2 Duplication.” Retrieved March 19, 2013 (http://www.ncbi.nlm.nih.gov/books/NBK3823/). Panofsky, Aaron. 2011. “Generating Sociability to Drive Science: Patient Advocacy Organizations and Genetics Research.” Social Studies of Science 41(1):31–57. Parens, Erik, Audrey R. Chapman, and Nancy Press, eds. 2006. Wrestling With Behavioral Genetics: Science, Ethics, And Public Conversation. JHU Press. Park, W. Wallac. 1966. “THE YY SYNDROME.” The Lancet 288(7479):1468. Parker, Charles E., Jamshed Mavalwala, John Melnyk, and Charles H. Fish. 1970. “The 48, XXYY Syndrome.” The American Journal of Medicine 48(6):777–81. Parsons, Talcott. 1991. The Social System. Routledge. Patau, Klaus, DavidW. Smith, Eeva Therman, StanleyL. Inhorn, and HansP. Wagner. 1960. “MULTIPLE CONGENITAL ANOMALY CAUSED BY AN EXTRA AUTOSOME.” The Lancet 275(7128):790–93. Peca, Joao et al. 2011. “Shank3 Mutant Mice Display Autistic-like Behaviours and Striatal Dysfunction.” Nature 472(7344):437–42.

! 423! ! ! ! Pennington, Carla. 2011. “Dr. Phil - Born to Rage?” Born to Rage. Retrieved October 6, 2011 (http://www.drphil.com/shows/show/1626). Penrose, L. S., J. R. Ellis, and J. D. Delhanty. 1960. “Chromosomal Translocations in Mongolism and in Normal Relatives.” Lancet 2(7147):409–10. Phelan, D. M. C. 2003. “Deletion 22q13 Syndrome.” Orphanet. Phelan, Jo C. 2005. “Geneticization of Deviant Behavior and Consequences for Stigma: The Case of Mental Illness.” Journal of Health and Social Behavior 46(4):307–22. Phelan, Katy. 2011. “Interview with Katy Phelan, New York NY.” Phelan, M. C. et al. 1992. “Cytogenetic, Biochemical, and Molecular Analyses of a 22q13 Deletion.” American Journal of Medical Genetics 43(5):872–76. Phelan, M. C. et al. 2001. “22q13 Deletion Syndrome.” American Journal of Medical Genetics 101(2):91–99. Phelan, M. C., R. C. Rogers, and L. K. Byrd. 1993. “ALBRIGHT HEREDITARY OSTEODYSTROPHY AND DEL(2)(Q37) IN 2 UNRELATED INDIVIDUALS.” American Journal of Human Genetics 53(3):S484. Phelan, M. C., R. C. Rogers, and R. E. Stevenson. 1988. “A de Novo Terminal Deletion of 22q.” AMERICAN JOURNAL OF HUMAN GENETICS 43:A118. Phelan, Mary. 2008. “Deletion 22q13.3 Syndrome.” Orphanet Journal of Rare Diseases 3(1):14. Phelan, Mary C. et al. 1995. “Albright Hereditary Osteodystrophy and Del(2)(q37.3) in Four Unrelated Individuals.” American Journal of Medical Genetics 58(1):1–7. Phelan, Mary C. et al. 2001. “22q13 Deletion Syndrome.” American Journal of Medical Genetics 101(2):91–99.

Phelan-McDermid Syndrome Foundation. 2013. Phelan-McDermid Syndrome Foundation Research Update, February 2013. Phelan-McDermid Syndrome Foundation. Retrieved March 15, 2013 (http://share.snacktools.com/FC7FB7BA9F7/fu5arsfm). PIERETTI, M. et al. 1991. “ABSENCE OF EXPRESSION OF THE FMR-1 GENE IN FRAGILE-X SYNDROME.” CELL 66(4):817–22. Pinker, Steven. 2007. The Language Instinct: How the Mind Creates Language. 1 Reprint. Harper Perennial Modern Classics. Polani, P. E., J. H. Briggs, C. E. Ford, C. M. Clarke, and J. M. Berg. 1960. “A Mongol Girl with 46 Chromosomes.” Lancet 1(7127):721–24.

! 424! ! ! ! Pollack, Andrew. 2011. “Sequenom’s Test for Down Syndrome Raises Hopes and Questions.” The New York Times, October 17. Retrieved January 28, 2012 (http://www.nytimes.com/2011/10/18/business/sequenom-test-for-down-syndrome- raises-hopes-and-questions.html). Pollack, Harold. 2012. “Santorum: He’s Not Funny Any More.” The Reality-Based Community. Retrieved (http://www.samefacts.com/2012/02/religion-and- politics/santorum-hes-not-funny-any-more/). Portnoī, M. F. 2009. “Microduplication 22q11. 2: a New Chromosomal Syndrome.” European journal of medical genetics 52(2-3):88–93. Potocki, Lorraine et al. 2000. “Molecular Mechanism for Duplication 17p11.2[mdash] the Homologous Recombination Reciprocal of the Smith-Magenis Microdeletion.” Nat Genet 24(1):84–87. Potocki, Lorraine et al. 2007. “Characterization of Potocki-Lupski Syndrome (dup(17)(p11.2p11.2)) and Delineation of a Dosage-sensitive Critical Interval That Can Convey an Autism Phenotype.” American Journal of Human Genetics 80(4):633–49. Potocki, Lorraine, Christine J. Shaw, Pawel Stankiewicz, and James R. Lupski. 2003. “Variability in Clinical Phenotype Despite Common Chromosomal Deletion in Smith-Magenis Syndrome [del(17)(p11.2p11.2)].” Genetics in Medicine: Official Journal of the American College of Medical Genetics 5(6):430–34. Prasad, Chitra et al. 2000. “Genetic Evaluation of Pervasive Developmental Disorders: The Terminal 22q13 Deletion Syndrome May Represent a Recognizable Phenotype.” Clinical Genetics 57(2):103–9. Price, Derek J. De Solla. 1965. “Statistical Studies of Networks of Scientific Papers.” P. 187 in Statistical Association Methods for Mechanized Documentation: Symposium Proceedings, vol. 269. Retrieved March 13, 2013 (http://books.google.com/books?hl=en&lr=&id=r56ZrbfTdkYC&oi=fnd&pg=PA1 87&dq=price+1965+citation&ots=2uXqr9Yf0s&sig=HBjOIZksh_SB2PKSfvALC FlktW8). Price, W. H., J. A. Strong, P. B. Whatmore, and W. F. McClemont. 1966. “Criminal Patients with XYY Sex-chromosome Complement.” Lancet 1(7437):565–66. Price, W. H., and P. B. Whatmore. 1967. “Criminal Behaviour and the XYY Male.” Nature 213(5078):815–815. Pyeritz, R., J. Beckwith, and L. Miller. 1975. “Letter: XYY Disclosure Condemned.” The New England journal of medicine 293(10):508.

! 425! ! ! ! Pyeritz, Reed E., Herb Schreier, Chuck Madansky, Larry Miller, and Jon Beckwith. 1977. “The XYY Male: The Making of a Myth.” Pp. 86–100 in Biology as a Social Weapon, edited by The Ann Arbor Science for the People Editorial Collective. Minneapolis: Burgess Pub. Co. Quelin, C. et al. 2009. “Twelve New Patients with 13q Deletion Syndrome: Genotype– phenotype Analyses in Progress.” European Journal of Medical Genetics 52(1):41– 46. Rabeharisoa, V. et al. 2012. “The Dynamics of Causes and Conditions: The Rareness of Diseases in French and Portuguese Patients’ Organizations’ Engagement in Research.” CSI WORKING PAPERS SERIES. Retrieved December 9, 2012 (http://hal.archives-ouvertes.fr/docs/00/70/20/88/PDF/WP_CSI_026.pdf). Rabeharisoa, Vololona, and Pascale Bourret. 2009. “Staging and Weighting Evidence in Biomedicine: Comparing Clinical Practices in Cancer Genetics and Psychiatric Genetics.” Social Studies of Science 39(5):691–715. Rabeharisoa, Vololona, and Michel Callon. 2002. “The Involvement of Patients’ Associations in Research.” International Social Science Journal 54(171):57–63. Rabinow, Paul. 1992. “Studies in the Anthropology of Reason.” Anthropology Today 8(5):7–10. Rader, Karen. 2004. Making Mice: Standardizing Animals for American Biomedical Research, 1900-1955. 1st ed. Princeton University Press. Rafferty, Andrew, and Domenico Montanaro. 2012. “Santorum Says Obama Looks down on Disabled, Encouraging More Abortions.” NBC News. Retrieved March 23, 2013 (http://firstread.nbcnews.com/_news/2012/02/18/10444238-santorum-says-obama- looks-down-on-disabled-encouraging-more-abortions). Ramocki, Melissa B. et al. 2009. “Autism and Other Neuropsychiatric Symptoms Are Prevalent in Individuals with MeCP2 Duplication Syndrome.” Annals of Neurology 66(6):771–82. Ramocki, Melissa B., Y. Jane Tavyev, and Sarika U. Peters. 2010. “The MECP2 Duplication Syndrome.” American Journal of Medical Genetics. Part A 152A(5):1079–88. Rapp, Rayna. 2000. Testing Women, Testing the Fetus: The Social Impact of Amniocentesis in America. 1st ed. Routledge. Rappoport, Stanley, and William D. Kaplan. 1961. “Chromosomal Aberrations in Man.” The Journal of Pediatrics 59(3):415–38.

! 426! ! ! ! Ratcliffe, S. 1999. “Long Term Outcome in Children of Sex Chromosome Abnormalities.” Archives of Disease in Childhood 80(2):192–95. Reilly, C. 2012. “Behavioural Phenotypes and Special Educational Needs: Is Aetiology Important in the Classroom?” Journal of Intellectual Disability Research 56(10):929–46. Reinhold, Robert. 1968. “ABORTIONS LINKED TO GENETIC DEFECT; Fetal Tests Said to Be Used as Basis for Operations.” The New York Times, September 2, 19. Reiss, A. L., C. Feinstein, K. N. Rosenbaum, M. A. Borengasser-Caruso, and B. M. Goldsmith. 1985. “Autism Associated with Williams Syndrome.” The Journal of Pediatrics 106(2):247–49. Reiss, Allan L. et al. 2004. “An Experiment of Nature: Brain Anatomy Parallels Cognition and Behavior in Williams Syndrome.” The Journal of Neuroscience 24(21):5009– 15. Reiss, Allan L., and Lisa Freund. 1992. “Behavioral Phenotype of Fragile X Syndrome: DSM!III!R Autistic Behavior in Male Children.” American Journal of Medical Genetics 43(1!2):35–46. Renpenning, H., J. W. GERRARD, W. A. ZALESKI, and T. TABATA. 1962. “Familial Sex-linked Mental Retardation.” Canadian Medical Association Journal 87:954–56. Repetto, G. M. et al. 2009. “Clinical Features of Chromosome 22q11.2 Microdeletion Syndrome in 208 Chilean Patients.” Clinical Genetics. Retrieved October 22, 2009 (http://www.ncbi.nlm.nih.gov/pubmed/19737282). Reuters. 2013. “Illumina to Buy Verinata Health for $350 Mln Plus Milestone Payments.” Reuters, January 7. Retrieved March 22, 2013 (http://www.reuters.com/article/2013/01/07/illumina-verinatahealth- idUSL4N0AC29X20130107). Ricci, N., and P. Malacarne. 1964. “AN XYY HUMAN MALE.” The Lancet 283(7335):721. Rimland, Bernard. 1964. Infantile Autism: The Syndrome and Its Implications for a Neural Theory of Behavior. Taylor & Francis. Rineer, Suzanne, Brenda Finucane, and Elliott W. Simon. 1998. “Autistic Symptoms Among Children and Young Adults with Isodicentric Chromosome 15.” American Journal of Medical Genetics 81(5):428–33.

! 427! ! ! ! Ritvo, Edward R., and B. J. Freeman. 1977. “National Society for Autistic Children Definition of the Syndrome of Autism.” Journal of Pediatric Psychology 2(4):146– 48. Robin, Nathaniel H. 2006. “It Does Matter: The Importance of Making the Diagnosis of a Genetic Syndrome.” Current Opinion in Pediatrics 18(6):595–97. Robinson, W. P. et al. 1996. “Delineation of 7q11.2 Deletions Associated with Williams- Beuren Syndrome and Mapping of a Repetitive Sequence to Within and to Either Side of the Common Deletion.” Genomics 34(1):17–23. Ropers, H. Hilger, and Ben C. J. Hamel. 2005. “X-linked Mental Retardation.” Nature Reviews Genetics 6(1):46–57. Rose, Nikolas. 1998a. Inventing Our Selves: Psychology, Power, and Personhood. Cambridge University Press. Rose, Nikolas. 1998b. “Medicine, History and the Present.” Pp. 48–72 in Reassessing Foucault: Power, Medicine and the Body, edited by Colin Jones and Roy Porter. New York, NY: Routledge. Rose, Nikolas. 2008. “Race, Risk and Medicine in the Age of ‘Your Own Personal Genome’.” BioSocieties 3(4):423–39. Rose, Nikolas [1]. 2007. “Molecular Biopolitics, Somatic Ethics and the Spirit of Biocapital.” Social Theory & Health 5:3–29. Rose, Nikolas S. 2007. The Politics of Life Itself. Princeton University Press. Ross, Judith L. et al. 2012. “Behavioral and Social Phenotypes in Boys With 47,XYY Syndrome or 47,XXY Klinefelter Syndrome.” Pediatrics 129(4):769–78. Rossen, Michael L., and Harvey B. Sarnat. 1998. “Why Should Neurologists Be Interested in Williams Syndrome?” Neurology 51(1):8 –9. Roubertoux, Pierre L., and Bernard Kerdelhué. 2006. “Trisomy 21: From Chromosomes to Mental Retardation.” Behavior Genetics 36(3):346–54. Rousseau, F. et al. 1991. “Direct Diagnosis by DNA Analysis of the Fragile X Syndrome of Mental Retardation.” The New England Journal of Medicine 325(24):1673–81. Rowan, Rob, and Dennis A. Powers. 1991. “A Molecular Genetic Classification of Zooxanthellae and the Evolution of Animal-Algal Symbioses.” Science 251(4999):1348–51. Royce, Kenneth. 1973. The XYY Man. HarperCollins Publishers. Royce, Kenneth. 1977. The XYY Man: The Concrete Boot. Mayflower.

! 428! ! ! ! Ruangdaraganon, N., C. Tocharoentanaphol, P. Khowsathit, T. Sombuntham, and B. Pongpanich. 1999. “Chromosome 22q11 Deletion Syndrome: The First Three Cases Reported in Thailand.” Journal of the Medical Association of Thailand = Chotmaihet Thangphaet 82 Suppl 1:S179–185. Rutter, Michael. 1978. “Diagnosis and Definition of Childhood Autism.” Journal of Autism and Developmental Disorders 8(2):139–61. Rutter, Michael. 2000. “Genetic Studies of Autism: From the 1970s into the Millennium.” Journal of Abnormal Child Psychology 28(1):3–14. Sabol, S. Z., Stella Hu, and D. Hamer. 1998. “A Functional Polymorphism in the Monoamine Oxidase A Gene Promoter.” Human Genetics 103(3):273–79. Sacks, Oliver. 2007. Musicophilia: Tales of Music and the Brain. 1st ed. Knopf. Sandberg, A. A., G. F. Koepf, T. Ishihara, and T. S. Hauschka. 1961. “AN XYY HUMAN MALE.” The Lancet 278(7200):488–89. Sandberg, Avery A., Takaaki Ishihara, Lois H. Crosswhite, and George F. Koepf. 1963. “XYY Genotype — Report of a Case in a Male.” New England Journal of Medicine 268(11):585–89. Sarasua, Sara M. et al. 2011. “Association Between Deletion Size and Important Phenotypes Expands the Genomic Region of Interest in Phelan–McDermid Syndrome (22q13 Deletion Syndrome).” Journal of Medical Genetics 48(11):761 – 766. Saul, Robert A., and John B. Moeschler. 2009. “How Best to Use CGH Arrays in the Clinical Setting.” Genetics in Medicine 11(5):371. De Saussure, Ferdinand. 2011. Course in General Linguistics. Columbia University Press. Scambler, P. J. et al. 1992. “Velo-cardio-facial Syndrome Associated with Chromosome 22 Deletions Encompassing the DiGeorge Locus.” Lancet 339(8802):1138–39. Scambler, Peter. 2011. “Interview with Peter Scambler, London UK.” Scambler, Peter J. et al. 1991. “Microdeletions Within 22q11 Associated with Sporadic and Familial DiGeorge Syndrome.” Genomics 10(1):201–6. Schaefer, G. Bradley, and Nancy J. Mendelsohn. 2008. “Genetics Evaluation for the Etiologic Diagnosis of Autism Spectrum Disorders.” Genetics in Medicine: Official Journal of the American College of Medical Genetics 10(1):4–12.

! 429! ! ! ! Schaffer, Rebecca, Kristine Kuczynski, and Debra Skinner. 2008. “Producing Genetic Knowledge and Citizenship through the Internet: Mothers, Pediatric Genetics, and Cybermedicine.” Sociology of Health & Illness 30(1):145–59. SCHMITT, H. et al. 1994. “INVOLVEMENT OF 22Q13.3 IN CHROMOSOMAL- ANOMALIES.” ONCOLOGY REPORTS 1(5):881–84. Schopler, Eric, Robert Reichler, Robert DeVellis, and Kenneth Daly. 1980. “Toward Objective Classification of Childhood Autism: Childhood Autism Rating Scale (CARS).” Journal of Autism and Developmental Disorders 10(1):91–103. Sequenom. 2010. Sequenom: 2009 Summary Annual Report. Retrieved February 13, 2012 (http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0C CQQFjAA&url=http%3A%2F%2Fphx.corporate- ir.net%2FExternal.File%3Fitem%3DUGFyZW50SUQ9NDExNjU3fENoaWxkSU Q9NDIyMjE3fFR5cGU9MQ%3D%3D%26t%3D1&ei=m3g5T8D_MoLs0gHouJG xAg&usg=AFQjCNF6u_YDpbhckyVtP6uJBP00fErokw). Sergovich, F., G. H. Valentine, A. T. L. Chen, R. A. H. Kinch, and M. S. Smout. 1969. “Chromosome Aberrations in 2159 Consecutive Newborn Babies.” New England Journal of Medicine 280(16):851–55. Sewell, William H. 1996. “Three Temporalities: Toward an Eventful Sociology.” Pp. 245– 80 in The Historic Turn in the Human Sciences, edited by Terrence J. McDonald. Ann Arbor: University of Michigan Press. Retrieved March 28, 2013 (http://books.google.com/books?hl=en&lr=&id=a7noy6lcU7IC&oi=fnd&pg=PA24 5&dq=sewell+jr.+independence+equivalence+comparative&ots=Q_JzDQ63Ac&si g=4r2ZrB1SEPmSk412Ua3u31yZd9Y). Sewell, William H. Jr. 1992. “A Theory of Structure: Duality, Agency, and Transformation.” The American Journal of Sociology 98(1):1. Shaffer, Brian L., Aaron B. Caughey, and Mary E. Norton. 2006. “Variation in the Decision to Terminate Pregnancy in the Setting of Fetal Aneuploidy.” Prenatal Diagnosis 26(8):667–71. Shaffer, Lisa G. et al. 2007. “The Discovery of Microdeletion Syndromes in the Post- genomic Era: Review of the Methodology and Characterization of a New 1q41q42 Microdeletion Syndrome.” Genetics in Medicine: Official Journal of the American College of Medical Genetics 9(9):607–16. Shaffer, Lisa G. et al. 2007. “The Identification of Microdeletion Syndromes and Other Chromosome Abnormalities: Cytogenetic Methods of the Past, New Technologies for the Future.” American Journal of Medical Genetics Part C: Seminars in Medical Genetics 145C(4):335–45.

! 430! ! ! ! Shaffer, Lisa G. 2009. ISCN 2009: An International System for Human Cytogenetic Nomenclature (2009): Recommendations of the International Standing Committee on Human Cytogenetic Nomenc. 1st ed. edited by Lisa G. Shaffer, Marilyn L. Slovak, and Lynda J. Campbell. S. Karger Publishing. Shaffer, Lisa G. et al. 2011. “The Development of a Rapid Assay for Prenatal Testing of Common Aneuploidies and Microdeletion Syndromes.” Prenatal Diagnosis 31(8):778–87. Shah, Saleem Alam. 1970. Report on the XYY Chromosomal Abnormality. National Institute of Mental Health, Center for Studies of Crime and Delinquency. Shakespeare, TOM. 1998. “Choices and Rights: Eugenics, Genetics and Disability Equality.” Disability & Society 13(5):665–81. Shapira, S. K. et al. 1997. “Chromosome 1p36 Deletions: The Clinical Phenotype and Molecular Characterization of a Common Newly Delineated Syndrome.” American Journal of Human Genetics 61(3):642–50. Shaw, ALISON, JOANNA Latimer, PAUL Atkinson, and KATIE Featherston. 2003. “Surveying ‘Slides’: Clinical Perception and Clinical Judgment in the Construction of a Genetic Diagnosis.” New Genetics & Society 22(1):3. Shostak, Sara, Peter Conrad, and Allan V. Horwitz. 2008. “Sequencing and Its Consequences: Path Dependence and the Relationships Between Genetics and Medicalization.” American Journal of Sociology 114(S1):S287–S316. Shprintzen, R. J. et al. 1978. “A New Syndrome Involving Cleft Palate, Cardiac Anomalies, Typical Facies, and Learning Disabilities: Velo-cardio-facial Syndrome.” The Cleft Palate Journal 15(1):56–62. Shprintzen, Robert J. 1998. “The Name Game.” Perspectives on Speech Science and Orofacial Disorders 8(1):7–11.

Shwed, Uri, and Peter S. Bearman. 2010. “The Temporal Structure of Scientific Consensus Formation.” American Sociological Review 75(6):817 –840. Silverman, Chloe. 2011. Understanding Autism: Parents, Doctors, and the History of a Disorder. Princeton University Press. Simon, Elliott W. 2012. “Interview with Eliot Simon, Elwyn PA.” Simon, Tony J. 1997. “Reconceptualizing the Origins of Number Knowledge: A ‘non- numerical’ Account.” Cognitive Development 12(3):349–72. Simon, Tony J. 2012. “Interview with Tony Simon, Orlando FL.”

! 431! ! ! ! Singer, Emily. 2002. “Study Links Antisocial Behavior to Abuse, Variant Gene.” Los Angeles Times Articles, August 2. Retrieved April 8, 2012 (http://articles.latimes.com/2002/aug/02/science/sci-abuse2). Singh, Jennifer. 2010. “Autism Spectrum Disorders: Parents, Scientists and Interpretations of Genetic Knowledge.” University of California San Fransisco. Singh, Jennifer, Judy Illes, Laura Lazzeroni, and Joachim Hallmayer. 2009. “Trends in US Autism Research Funding.” Journal of Autism and Developmental Disorders 39(5):788–95. Singleton, Vicky, and Mike Michael. 1993. “Actor-Networks and Ambivalence: General Practitioners in the UK Cervical Screening Programme.” Social Studies of Science 23(2):227–64. Skinner, Debra, and Rebecca Schaffer. 2006. “Families and Genetic Diagnoses in the Genomic and Internet Age.” Infants & Young Children January/March 2006 19(1):16–24. Skotko, Brian G. 2009. “With New Prenatal Testing, Will Babies with Down Syndrome Slowly Disappear?” Archives of Disease in Childhood 94(11):823–26. Smith, A. C. et al. 1986. “Interstitial Deletion of (17)(p11.2p11.2) in Nine Patients.” American Journal of Medical Genetics 24(3):393–414. Smith, David W., Klaus Patau, Eeva Therman, and Stanley L. Inhorn. 1960. “A New Autosomal Trisomy Syndrome: Multiple Congenital Anomalies Caused by an Extra Chromosome.” The Journal of Pediatrics 57(3):338–45. St George. 1998. “What Studying Genetically-based Disorders Can Tell Us About Ourselves.” Trends in Cognitive Sciences 2(6):203–4. Star, Susan Leigh. 2010. “This Is Not a Boundary Object: Reflections on the Origin of a Concept.” Science, Technology & Human Values 35(5):601–17. Star, Susan Leigh, and James R. Griesemer. 1989. “Institutional Ecology, `Translations’ and Boundary Objects: Amateurs and Professionals in Berkeley’s Museum of Vertebrate Zoology, 1907-39.” Social Studies of Science 19(3):387 –420. Steele, M. W., and W. R. Breg Jr. 1966. “Chromosome Analysis of Human Amniotic-fluid Cells.” Lancet 1(7434):383–85. Steinbock, Bonnie. 2008. “Designer Babies: Choosing Our Children’s Genes.” The Lancet 372(9646):1294–95. Stevens, Cathy A., John C. Carey, and Ann O. Shigeoka. 1990. “Di George Anomaly and Velocardiofacial Syndrome.” Pediatrics 85(4):526–30.

! 432! ! ! ! Stock, Robert W. 1968. “The XYY And The Criminal; The XYY and the Criminal.” The New York Times, October 20. Retrieved November 20, 2012 (http://select.nytimes.com/gst/abstract.html?res=F50817F93E5F127A93C2AB178 BD95F4C8685F9). Stokes, Jon. 2006. “Maori ‘Warrior Gene’ Claims Appalling, Says Geneticist.” New Zealand Herald, August 10. Retrieved April 11, 2012 (http://www.nzherald.co.nz/health/news/article.cfm?c_id=204&objectid=10395491 ). Stromme, P., PG Bjornstad, and K. Ramstad. 2002. “Prevalence Estimation of Williams Syndrome.” JOURNAL OF CHILD NEUROLOGY 17(4):269–71. Sunder Rajan, Kaushik. 2006. Biocapital: The Constitution of Postgenomic Life. Duke University Press Books. Swidler, Ann. 1986. “Culture in Action: Symbols and Strategies.” American Sociological Review 51(2):273–86. Swidler, Ann. 2001. Talk of Love: How Culture Matters. University of Chicago Press. Tabor, Holly K., and Martine D. Lappé. 2011. “The Autism Genetic Resource Exchange: Changing Pace, Priorities, and Roles in Discover Science.” Pp. 56–71 in Achieving Justice in Genomic Translation, edited by Wylie Burke and Kelly Edwards. New York: Oxford University Press. Tartaglia, N. R. et al. 2006. “Attention Deficit Hyperactivity Disorder and Autism Spectrum Disorders in Males with XXY, XYY and XXYY Syndromes.” Journal of Intellectual Disability Research 50:787–787. Tartaglia, Nicole et al. 2008. “A New Look at XXYY Syndrome: Medical and Psychological Features.” American Journal of Medical Genetics Part A 146A(12):1509–22.

Tartaglia, Nicole R., Susan Howell, Ashley Sutherland, Rebecca Wilson, and Lennie Wilson. 2010. “A Review of Trisomy X (47,XXX).” Orphanet Journal of Rare Diseases 5(1):8.

Taussig, Karen!Sue, Rayna Rapp, and Deborah Heath. 2003. “Flexible Eugenics: Technologies of the Self in the Age of Genetics.” Pp. 58–76 in Genetic Nature/Culture: Anthropology and Science Beyond the Two-Culture Divide, edited by Alan H. Goodman, Deborah Heath, and M. Susan Lindee. University of California Press. Tecott, L. H. 2001. “Mouse Molecular Genetic Technologies: Promise for Psychiatric Research.” Archives of General Psychiatry 58(11):995–1004.

! 433! ! ! ! Telfer, Mary A., David Baker, Gerald R. Clark, and Claude E. Richardson. 1968. “Incidence of Gross Chromosomal Errors Among Tall Criminal American Males.” Science 159(3820):1249 –1250. Terry, Sharon F., Patrick F. Terry, Katherine A. Rauen, Jouni Uitto, and Lionel G. Bercovitch. 2007. “Advocacy Groups as Research Organizations: The PXE International Example.” Nat Rev Genet 8(2):157–64. The 1000 Genomes Project Consortium. 2010. “A Map of Human Genome Variation from Population-scale Sequencing.” Nature 467(7319):1061–73. The Trans-NIH Fragile X Research Coordinating Group and Scientific Working Groups. 2008. National Institutes Of Health Research Plan On Fragile X Syndrome and Associated Disorders. U.S. Department Of Health And Human Services (DHHS), National Institutes Of Health (NIH). Retrieved (http://www.nichd.nih.gov/publications/pubs/upload/NIH_Research_Plan_on_Fragi le_X_and_Assoc_Disorders-06-2009.pdf). Thieffry, S., M. ARTHUIS, de GROUCHY, M. LAMY, and C. SALMON. 1963. “[Deletion of the Short Arms of Chromosome 17-18: Complex Deformities with Oligophrenia.].” Archives Françaises De Pédiatrie 20:740–45. Tilly, Charles. 1993. “Contentious Repertoires in Great Britain, 1758-1834.” Social Science History 17(2):253–80. Tilly, Charles. 2003. The Politics of Collective Violence. Cambridge University Press. Tilly, Charles. 2004. Social Movements, 1768-2004. Paradigm Publishers. Timmermans, Stefan, and Marc Berg. 1997. “Standardization in Action: Achieving Local Universality through Medical Protocols.” Social Studies of Science 27(2):273–305. Timmermans, Stefan, and Mara Buchbinder. 2010. “Patients-in-waiting: Living Between Sickness and Health in the Genomics Era.” Journal of Health and Social Behavior 51(4):408–23. Timmermans, Stefan, and Mara Buchbinder. 2011. “Improving Expanded Newborn Screening.” Journal of Health and Social Behavior 52(2):279 –281. Timmermans, Stefan, and Mara Buchbinder. 2012. “Expanded Newborn Screening: Articulating the Ontology of Diseases with Bridging Work in the Clinic.” Sociology of Health & Illness 34(2):208–20. Timmermans, Stefan, and Steven Epstein. 2010. “A World of Standards but Not a Standard World: Toward a Sociology of Standards and Standardization*.” Annual Review of Sociology 36(1):69–89.

! 434! ! ! ! Timmermans, Stefan, and Steven Haas. 2008. “Towards a Sociology of Disease.” Sociology of Health & Illness 30(5):659–76. Tjio, Joe Hin, and Albert Levan. 1956. “The Chromosome Number of Man.” Hereditas 42(1-2):1–6. Turleau, Catherine. 2008. “Monosomy 18p.” Orphanet Journal of Rare Diseases 3:4. Turner, G. et al. 1996. “Prevalence of Fragile X Syndrome.” American Journal of Medical Genetics 64(1,):196–97.

Turner, Gillian, and John M. Opitz. 1980. “X!linked Mental Retardation, X!linked Mental Retardation.” American Journal of Medical Genetics, American Journal of Medical Genetics 7, 7(4, 4):407, 407–15, 415. Turpin, Raymond, and Jérôme Lejeune. 1965. Les Chromosomes humains. Gauthier- Villars. Turpin, Raymond, and Jérôme Lejeune. 1969. Human Afflictions and Chromosomal Aberrations. Pergamon Press. Venkitaraman, A. R. 2002. “Cancer Susceptibility and the Functions of BRCA1 and BRCA2.” Cell 108(2):171–82. VERKERK, AJMH et al. 1991. “IDENTIFICATION OF A GENE (FMR-1) CONTAINING A CGG REPEAT COINCIDENT WITH A BREAKPOINT CLUSTER REGION EXHIBITING LENGTH VARIATION IN FRAGILE-X SYNDROME.” CELL 65(5):905–14. Verp, Marion S. et al. 1988. “Parental Decision Following Prenatal Diagnosis of Fetal Chromosome Abnormality.” American Journal of Medical Genetics 29(3):613–22. Vignetti, P., L. Capotorti, and E. Ferrante. 1964. “XYY CHROMOSOMAL CONSTITUTION WITH GENITAL ABNORMALITY.” The Lancet 284(7359):588–89. Vorstman, J. A. S. et al. 2005. “Identification of Novel Autism Candidate Regions through Analysis of Reported Cytogenetic Abnormalities Associated with Autism.” Mol Psychiatry 11(1):18–28. Wade, Nicholas. 2009. “Genes Show Limited Value in Predicting Diseases.” The New York Times, April 16. Retrieved February 13, 2012 (http://www.nytimes.com/2009/04/16/health/research/16gene.html). Walzer, S., and P. S. Gerald. 1975. “Social Class and Frequency of XYY and XXY.” Science (New York, N.Y.) 190(4220):1228–29.

! 435! ! ! ! Walzer, S., P. S. Gerald, and S. A. Shah. 1978. “The XYY Genotype.” Annual Review of Medicine 29(1):563–70. Warburg, M., and M. Mikkelsen. 1963. “13-15 TRISOMY WITH SEVERE OCULAR MALFORMATIONS.” Archives of Ophthalmology 69(3):420–420. Warburg, Mette. 1960. “Anophthalmos Complicated by Mental Retardation and Cleft Palate.” Acta Ophthalmologica 38(4):394–404. Warkany, Josef. 1960. “Etiology of Mongolism.” The Journal of Pediatrics 56(3):412–19. Watson, MS et al. 1984. “FRAGILE X IN A SURVEY OF 75 AUTISTIC MALES.” New England Journal of Medicine 310(22):1462–1462. Webb, T. P., S. Bundey, A. Thake, and J. Todd. 1986. “The Frequency of the Fragile X Chromosome Among Schoolchildren in Coventry.” Journal of Medical Genetics 23(5):396–99. Weber, Max. 2011. Methodology of Social Sciences: Max Weber. Reprint. edited by Edward Shils and Henry Finch. Transaction Publishers. Weiss, Lauren A. et al. 2008. “Association Between Microdeletion and Microduplication at 16p11.2 and Autism.” N Engl J Med 358(7):667–75. Weiss, Philip. 1975. “Ending the Test for Extra Chromosomes.” The Harvard Crimson, September 15. Retrieved December 21, 2012 (http://www.thecrimson.com/article/1975/9/15/ending-the-test-for-extra- chromosomes/). Wentzel, C., M. Fernström, Y. Öhrner, G. Annerén, and A. C. Thuresson. 2008. “Clinical Variability of the 22q11. 2 Duplication Syndrome.” European journal of medical genetics 51(6):501–10. Whitmarsh, Ian, Arlene M. Davis, Debra Skinner, and Donald B. Bailey. 2007. “A Place for Genetic Uncertainty: Parents Valuing an Unknown in the Meaning of Disease.” Social Science & Medicine 65(6):1082–93. Williams, J. C. P., B. G. Barratt-Boyes, and J. B. Lowe. 1961. “Supravalvular Aortic Stenosis.” Circulation 24(6):1311–18. Wilson, Edward O. 1980. Sociobiology. Harvard University Press. Wing, Lorna, and Judith Gould. 1979. “Severe Impairments of Social Interaction and Associated Abnormalities in Children: Epidemiology and Classification.” Journal of Autism and Developmental Disorders 9(1):11–29.

! 436! ! ! ! Wong, A. C. et al. 1997. “Molecular Characterization of a 130-kb Terminal Microdeletion at 22q in a Child with Mild Mental Retardation.” American Journal of Human Genetics 60(1):113–20. Wong, A. C. C., C. J. Bell, J. P. Dumanski, M. L. Budarf, and H. E. McDermid. 1995. “Molecular Characterization of a Microdeletion at 22q13.3.” AMERICAN JOURNAL OF HUMAN GENETICS 57:A130. Wong, Edgar Chan, Kyle A. Ferguson, Victor Chow, and Sai Ma. 2008. “Sperm Aneuploidy and Meiotic Sex Chromosome Configurations in an Infertile XYY Male.” Human Reproduction 23(2):374–78. Woolf, Steven H. 2008. “The Meaning of Translational Research and Why It Matters.” JAMA: The Journal of the American Medical Association 299(2):211. Writer, Victor Cohn Washington Post Staff. 1968. “A Criminal by Heredity?: Testing for ‘Criminal Heredity’ Urged by Three Genticists.” The Washington Post, Times Herald (1959-1973), August 7, A1. Wu, Yuan-Qing et al. 1998. “Delineation of the Common Critical Region in Williams Syndrome and Clinical Correlation of Growth, Heart Defects, Ethnicity, and Parental Origin.” American Journal of Medical Genetics 78(1):82–89. Wulfsberg, E. A., J. Leana-Cox, and G. Neri. 1996. “What’s in a Name? Chromosome 22q Abnormalities and the DiGeorge, Velocardiofacial, and Conotruncal Anomalies Face Syndromes.” American Journal of Medical Genetics 65(4):317–19. Wulfsberg, EA, J. LeanaCox, and G. Neri. 1997. “What’s in a Name? The 22q11.2 Deletion - Reply.” AMERICAN JOURNAL OF MEDICAL GENETICS 72(2):248– 49. Yagi, Hisato et al. 2003. “Role of TBX1 in Human Del22q11.2 Syndrome.” The Lancet 362(9393):1366–73.

Yong, Ed. 2010. “Dangerous DNA.” New Scientist 205(2755):34–37. Yunis, J. J., and O. Sanchez. 1974. “A New Syndrome Resulting from Partial Trisomy for the Distal Third of the Long Arm of Chromosome 10.” The Journal of Pediatrics 84(4):567–70. Yunis, Jorge J. 1977. New Chromosomal Syndromes. Academic Press. Zeaiter, Zaher, Zhongxing Liang, and Didier Raoult. 2002. “Genetic Classification and Differentiation of Bartonella Species Based on Comparison of Partial ftsZ Gene Sequences.” Journal of Clinical Microbiology 40(10):3641–47.

! 437!