1
âRaceâ as a Social Construction in Genetics
Andrew Smart, Richard Tutton, Paul Martin and George T.H. Ellison
There is an incongruity at the heart of postgenomic biomedical science. The sequencing of the human genome promised the elimination of racial and ethnic categories from biomedical science (Schwartz 2001). However, despite this, we have witnessed a resurgent interest in âraceâ, including challenges to the longstanding orthodoxy (following Lewontin 1972) that classifications of human populations by âraceâ are not supported by genetic data (Andreasen 2000; Risch et al. 2002; Rosenberg et al. 2002; Edwards 2003). While some geneticists have consigned the concept of âraceâ to the scrapheap, others appear to be reviving it. This work is occurring within a broader social and scientific landscape, one in which differences between human populations are invested (or reinvested) with significance. This attention to groups, their differences and similarities and their boundaries feeds off of, and contributes to, our ideas about identities and belonging. It has the potential to create, reinforce and/or aggravate social and ethical issues surrounding, for example, matters of self-identification, group representation, equality, stigmatization, discrimination and resource distribution. The particular discussions about âraceâ in genetics which we consider in this chapter will be shown to reinvigorate debates about what âraceâ really is and, by extension, claims about the status of collective or individual identities that draw upon (and reproduce) ideas relating to racial division.
A major stimulus for this kind of work is the potential development of âtargetedâ diagnostic and therapeutic interventions. Such âtargetingâ is predicated on the search for significant biological differences within and between âpopulationsâ, which has led to the creation of large-scale scientific projects that map genetic variation. An often-cited example is the International Human Haplotype Map (HapMap) project, which aims to improve health by charting variations in DNA sequences between population groups labelled as âYorubaâ, âHan Chineseâ, âJapaneseâ and âAmericans of northern and western European descentâ (International HapMap Consortium 2003). It should be noted, however, that beyond such high-profile examples, genetics researchers routinely classify samples with population labels (Ellison and Jones 2002; Fullwiley 2007; Smart et al. 2008b; Hunt and Megyesi 2008). Furthermore, as the HapMap example illustrates, we are not always confronted by the use of overtly âracialâ characteristics to label population groupings, but also other classifications that contribute to discourses of âraceâ such as geography, nationality, ethnicity and/or ancestry. Indeed, genetics research often adopts classification schemes from state bureaucracies (Epstein 2007; Fullwiley 2007; Ellison et al. 2007; Smart et al. 2008b). The use of such classifications sometimes relates to the research design, although (in some jurisdictions) the state requires or encourages researchers to classify participants for reasons of monitoring policies that encourage equality and âinclusivityâ (Epstein 2007; Smart et al. 2008a).
So, what is worrisome about these routine practices in genetic research which aim to promote âinclusiveâ research and may lead to potentially helpful population-specific health interventions? There is a well-documented quandary in healthcare policy and practice: categorization by âraceâ and ethnicity can be simultaneously useful for addressing health inequalities and yet harmful because it reinforces the very notions that may have initially contributed to these inequalities (Witzig 1996; Nazroo 1998; Anand 1999; Bradby 2003). Furthermore, there are well-documented difficulties in operationalization and measurement (Aspinall 1997, 2001). These quandaries have resurfaced in recent debates about genetics, âraceâ and medicine. Some questions have been asked about the validity, reliability and utility of socially-defined population categories as well as how clinical practice and public health advice will deal with group differences. For example, a heated debate exists around so-called âracialized prescribingâ, its legitimacy, utility and social costs and benefits (Satel 2002; Burchard et al. 2003; Phiminster 2003; Cooper et al. 2003; Kahn 2004; Ellison 2006). At the centre of such debates â at least for sociologists â is the question of whether genetic research will revive racial science and scientific racism (Martin et al. 2007; Rose 2007). Given that this is genetic research, particular concern has centred on the potential for reinvigorating the notion that âracesâ are naturally occurring, biologically meaningful entities (Gannett 2004; Duster 2005).
There has been a relatively stable cross-disciplinary consensus on the ontology of âraceâ, described thus by Gannett: âThe apparent consensus view among academics from diverse disciplines â the humanities, the social sciences, and the biological sciences â is that biological races do not exist, at least in humans. Biological race is a socially-constructed categoryâ (2004: 323). Nevertheless, as noted above, there have been challenges to this consensus, such that claims about the biological reality of âraceâ have been resurrected (e.g., Sarich and Miele 2004). Work in genetics thus appears to threaten the common refrain that âraceâ is a social construction. John Hartigan Jr. (2008) has produced the most developed account of this issue. While â as we might expect â he rebuffs the claims made about the biological basis of racial categories, he comes to a somewhat critical conclusion by arguing that:
Current assertions that race is socially constructed do important work keeping [the historical and contemporary] misuses of race in view, while also challenging the evidentiary ground for making claims about linkages between race and genetics. The problem with this assertion, though, is that since it typically makes its starting point the discrediting of biological claims ⊠âsocial constructionâ generally offers meagre guidance in comprehending the cultural dynamics at work in racial matters. (2008: 185)
To repel claims about the biological reality of âraceâ by claiming instead that it is a social construction is, he argues, becoming less effective (see also Hartigan 2006).
In this chapter, we look in more detail at the appearance of the notion that âraceâ is a social construction in the field of genetics. Contributors to an influential genetics journal, Nature Genetics, used the expressions âsocially defined constructâ, âsocial constructâ and âsocially constructedâ in a Special Issue entitled Genetics for the Human Race (further details of which are given below). The appearance of this vocabulary initiated a debate within our research group about whether or not the use of such phrases was to be expected, what the authors meant when they used these terms and why they were using them. In order to start answering these questions, we undertook a detailed analysis of the usage of constructionist terms in the Special Issue. However, this analysis raised new questions in our minds about what it means more broadly for anyone to claim that anything is a social construction. We therefore begin by outlining three aspects of social construction that appeared to be the most relevant to our interests; these will comprise our analytical framework during the subsequent discussion.
An Analytical Framework: What Does it Mean to Say Something is a Social Construction?
There is an extensive literature on social construction spanning different disciplines, and what follows here is certainly not a review of this broad and complex idea. To gain an overview, we have relied on the work of philosopher Ian Hacking (1999). We recognize that Hacking is not the final arbiter on ideas about social construction and that some of the arguments reported below are refuted by his critics. Nevertheless, his philosophical groundwork clarifies many of the key characteristics of social constructionism.
Hacking begins by explaining how claims that something is a social construction involve challenging a notion that something is determined by âthe nature of thingsâ, thereby emphasizing that it is not inevitable but that it âwas brought into existence or shaped by social events, sources, historyâ (1999: 6â7). Saying that something is a social construction means uncovering contingency where none was previously thought to have existed by attending to specific social, cultural and historical contexts. Hacking then proceeds to draw attention to notable characteristics and points of difference that exist in claims that something is a social construction, three of which proved particularly useful for our analysis.
Objects, Ideas and their Interactions
Hacking (1999) explains that the sorts of things that can (and have been) described as socially constructed are extraordinarily wide-ranging. He notes that claims to social construction can be ambiguous and that âthey have in mind several interacting itemsâ (1999: 33):
It is plain in the case of gender. What is being constructed? The idea of gendered human beings (an idea), and gendered human beings themselves (people); language; institutions; bodies. Above all, âthe experiences of being female.â (1999: 28)
His path through these ambiguities draws an analytical distinction between claims that have been attached to different things. Most importantly for our current purpose is to note his attempted division between âideasâ and âobjectsâ, and his argument that it is different to claim that an idea of something is a social construction than to claim that an object that is âin the worldâ is a social construction (1999: 21). Nevertheless, this is a dichotomy which he acknowledges is âfuzzy at the edgesâ (1999: 21), because ideas such as classifications or groupings have âextensions â classes, sets, and groups ⊠[which] are collections in the world, and so count as âobjectsââ (1999: 22).
Hacking goes on to argue that there can be interactions between these ideas and the objects to which they refer (e.g., between a classification [idea] and the people subjected to that classification [object]). For example: âWays of classifying human beings interact with the human beings who are classified ⊠People think of themselves as a kind, perhaps, reject the classification ⊠Moreover, classifications do not exist only in the empty space of language but in institutions, practices, material interactions with things and other peopleâ (1999: 31). Some classifications are thus âinteractiveâ, that is, they involve âconscious interactions between kind and personâ (1999: 32). In such cases, he says, there is potential for âlooping effectsâ whereby awareness of the classification âcan have consequences for the very group, for the kind of people that is invokedâ (1999: 34).
Scope and Commitment
Hacking (1999) argues that accepting one thing as a social construction does not necessitate thinking that everything else is. Thus, claims that a specific thing is a social construction can be regarded as âlocalâ while, in contrast, âuniversalâ constructionism would be the claim âthat every object whatsoeverâ is socially constructed (1999: 24). However, Hacking questions whether such an extreme universal position has ever really been advocated. He argues that the primary use of labelling something as a social construction is to raise consciousness about social, cultural and historical contingency in ways that undercut assumptions about inevitability. Consequently, he says, such claims are often (but not always or necessarily) seen to offer a potential for liberation and are accompanied by arguments about value judgments and the need for change. He proceeds to outline six âgrades of commitmentâ to social constructionism â these are labelled as historical, ironic, reformist, unmasking, rebellious and revolutionary â which range from merely having a particular attitude to scholarly activity through to using it as the basis for a form of political activism (1999: 19).
âSocialâ versus âŠ
Some claims about social construction are framed in oppositional terms. Certain things are argued to be either social constructions or ârealâ, âbiologicalâ or ânaturalâ. Much of the furore surrounding social construction can be traced back to the suggestion, or implication, that the things that natural scientists studied, how they studied them and the outcomes of their studies were not inevitable. Hacking explains that the âscience warsâ debates were âheavily loaded with the words [objective, ideological, factual and real]â (1999: 23). For him, an opposition between things that are real and things that are social constructs is not always necessary; certain things can be both (1999: 29, 101). He also argues that the under certain circumstances things (such as certain mental illnesses) can be both social constructions and biological or natural âkindsâ (1999: 119â24).
Our Analytical Frame and the Social Construction of âRaceâ
Hacking (1999: 16) only briefly examines âraceâ as a social construction to illustrate how the function of such claims is to undercut ideas that âoneâs race is a part of oneâs âessenceââ. He returns to the subject of âraceâ elsewhere, when he argues for the continuing relevance of âthe category of raceâ because âraces in some contexts are not only statistically significant but also statistically usefulâ (2005: 108) and when he describes the quandaries that arise from contemporary genetic research (2006). However, our purpose here is not to review Hackingâs views on âraceâ, but rather to highlight aspects of his general discussion of social construction which we found useful for analysing the Special Issue. To recap, these were: differences in respect to the focus of claims, including the interactions between âideasâ and âobjectsâ; differences in respect to the scope of, and commitment to, claims; and the (oppositional) form of some claims. For now, three initial thoughts can be noted about how these general ideas concerning social construction relate to âraceâ in the context of genetic research.
First, to claim that âraceâ is a social construction can hide great complexity and diversity (in focus, scope, commitment, etc.). This implies not just potential differences in usage between people, or within and across disciplines and subdisciplines, but also that different meanings can be implied or read into what â on the surface â seems like a simple statement. Secondly, ideas about âinteractive classificationsâ, âlooping effectsâ and fuzzy boundaries between objects and ideas are all pertinent for thinking about âraceâ as a classification of humankind. These underline that when we talk about âracesâ in humanity, we engage in debates about identity politics, which frequently (but not exclusively) invoke a long and sometimes brutal history of discrimination and stigmatization. Thirdly, claims that something is a social construction rather than ârealâ, âbiologicalâ or ânaturalâ are particularly relevant to debates about âraceâ, which often follow this pattern. We will return to these issues in our discussion.
Social Construction in the Special Issue of Nature Genetics
Our analysis of the appearance of the notion that âraceâ is a social construction in âgeneticsâ centres on the November 2004 Special Issue of Nature Genetics, titled Genetics for the Human Race.1 We acknowledge from the outset that this is a somewhat narrow âsampleâ of perspectives from âgeneticsâ and, had our approach been different, we may have found additional and alternative views. Nonetheless, our aim is to be exploratory and discursive rather than strictly empirical. In this section we provide further details about the Special Issue and describe the range of positions that were evident in respect to the notion that âraceâ is a social construction. We make some analytical comments as we reveal the position of each set of authors, but a more comparative analysis will follow in the section of this chapter titled Discussion.
Background: The Special Issue
Our analysis of the Special Issue Genetics for the Human Race reflects its importance as a landmark publication in debates about âraceâ and genetics, a significance also recognized by other discussions of this field (Hacking 2005, 2006; Rose 2007; Hartigan 2008). It garners such attention because Nature Genetics is a high-status academic journal and the contributors to the Special Issue are well-recognized figures in the debate. The discussions that are published reveal not only the ongoing quandaries about âraceâ and genetics but also the importance of the topic for leading scientific figures (including Francis Collins, the Director of the United States National Institute of Health) and the field of genetics more generally. Nevertheless, the Special Issue should be recognized as existing within a history of interest about population classifications in Nature Genetics. It pursued a number of themes raised in previous Nature Genetics editorials (2000, 2001 and 2004a), such as: how âraceâ and ethnicity should be defined, classified and operationalized; the value of alternative concepts (like ancestry); and the importance of public engagement on this topic.
The Special Issue originated from a workshop, âHuman Genome Variation and âRaceââ, held on 15 May 2003 at the National Human Genome Center (NHGC), Howard University, Washington DC. Howard University has historical links to Black politics in the U.S.A., something which is reflected in the aims of the Centre. The NHGC is âdedicated to the engagement of African Americans and other people of African ancestry into the mainstream of human genome researchâ, which has led to criticism that âthe center perpetuates race-based science and medicineâ (Royal and Dunston 2004: S6). The workshop b...