In this study we trace seven decades of thinking on obesity in a major medical textbook, providing a powerful instantiation of how medical objectivity arbitrates, officiates, and produces knowledge in a realm of considerable scientific and social uncertainty. The causes, consequences, and management of human fatness continue to be a realm of considerable debate in both scientific and lay arenas (Kassirer and Angell 1998, Klein 1996, Taubes 1998, Wickelgren 1998, Austin 1999). Of particular note, we demonstrate that despite a relatively stable overall pathogenic process, there can be important shifts in explanation, shifts that do not follow directly from definitive experimental results, shifts that do not reflect a simple, steady progression towards more ‘truthful’ beliefs.
Medical modelling and individual culpability
The subjection of a behaviour or condition to medical conceptualisation has been associated with both an increased and a decreased attribution of individual responsibility. In our case-example of the construction of obesity in a fixed source over time, we find elements of both tendencies. The earlier editions tend to target individual behaviours and individual accountability, while the later editions devote more attention to social and environmental context (as well as genetic components), with a concomitant diminution of individual blame and accountability. These results can be viewed as an instant-iation of a more general tension in medical and epidemiological research, a tension between population and individual levels of analysis (Pearce 1996), which bears notable overlap and resemblance to sociological debates concerning structure and agency.
The individual level of analysis focuses on individual lifestyle factors and risk behaviours, while the population perspective takes a more structural approach, contextualising individual factors within a social, economic, cultural, and political framework (Pearce 1996). For example, an excess of calories is associated with obesity, but the thrust of medical interrogation and social policy can converge on overconsumption or overproduction, focusing on individual exposures or the social structures that render individuals susceptible to such exposures. Many have argued that the emphasis on individual behavioural modifications, termed ‘healthism’ (Crawford 1980) or the lifestyle model, has been the reigning paradigm in medical and public health promotion since the 1970–80s (Crawford 1980, Reissman 1983, Coreil et al. 1985, Fitzgerald 1994, Travers 1995, Austin 1999). The medical model is thus reproached for obscuring the social structuration of ill health, preventing critique of the existing social order, and reinforcing the privatisation of potential interventions (Crawford 1980). Our findings in Cecil of a redirection from an individual to a population level of focus runs contrary to this broader tendency, however, demonstrating that medical framing itself does not preclude attention to context. Nevertheless, the text does not suggest any form of substantive sociopolitical change or rectification.
The recent emphasis on molecular or genetic factors represents yet another level of distinction. While drawing attention away from social structural context and back to the individual body, it nevertheless accounts for a condition in non-voluntaristic terms. This component of our findings more closely typifies classic depictions of the medicalisation process wherein sickness is substituted for badness with a concomitant change in imputed responsibility (Conrad and Schneider 1992). Hubbard and Wald (1997) have argued that current trends towards the geneticisation of chronic conditions hinges on the assumption that there is a hierarchy of causes presided over by the gene, simultaneously exculpating both individuals and society.
Given recent claims that obesity has been medicalised (Reissman 1983, Conrad and Schneider 1992, Sobal 1995), one might ask whether or not there is evidence of increased medicalisation in these texts. This assessment, however, is largely dependent on how medicalisation itself is conceptualised. Strong (1979) forcefully argued over two decades ago that the thesis of ‘medical imperialism’, the increased medicalisation of the social world, has been exaggerated by social scientists. One of his objections to this thesis is the empirical finding that patients are often critical of the medical services they receive and, for many problems, highly sceptical of the utility of medical consultation. More recently, Williams and Calnan (1996) have similarly argued, drawing on Gidden’s theorisation of social reflexivity and risk management in ‘late’ modernity, that medical jurisdiction is now limited by substantial ambivalence, scepticism, and disillusionment with scientific medicine amongst a lay populace of increasingly critical and technically-informed agents.
Strong (1979) also argued that in areas supposedly under threat of medicalisation, such as alcoholism, physicians themselves are often sceptical of the value of medical intervention. In his own work on physicians’ attitudes towards alcoholism, Strong (1980) found that physicians generally dislike treating alcoholics, preferring to manage problems that are more straightforwardly ‘biological’, or easily susceptible to abstraction from social context, problems for which they have clear-cut expertise in etiology, diagnosis, and effective treatment. Given that obesity shares many of these features with alcoholism on the physician side, and given the emergence of resistance in the lay populace against the medical management of obesity on the patient side (Klein 1996, Sobal 1999), one might follow Strong’s reasoning on the limits of medical imperialism and hypothesise that the medicalisation of obesity is subject to important doctor and patient constraints. A content analysis of medical texts, however, can neither confirm nor disconfirm such a hypothesis; substantiation would require empirical elaboration of the actual sensibilities and behaviours of doctors and patients.
Our findings, however, do inform this issue in another way. Conrad and Schneider (1980) responded to Strong by arguing that he takes an unnecessarily narrow view of medicalisation by focusing on ‘what doctors actually control and do’ at the level of doctor-patient interaction. They advocate a broader conceptual frame wherein medicalisation can occur on three levels: the conceptual, the institutional, and the doctor-patient interaction levels (Conrad and Schneider 1980, Conrad 1992). Our findings do operate at the level of the conceptual, the level of the formulation of ideas, definitions, and explanations. A content analysis, however, cannot speak to how physicians (or non-physicians) actually conceptualise obesity in everyday practice. Nevertheless, one might still ask if there is evidence of increased conceptual medicalisation of obesity within the confines of this textbook. Despite such delineation of the question, there is still no clear-cut answer, for even at the conceptual level of medicalisation, the answer is further dependent on which factors are taken to be the most salient indicators of such a process. While some would view the invocation of a lifestyle model as evidence of medicalisation (Crawford 1980), others would argue that the lifestyle model is opposite to medicalisation, because it ‘turns health into the moral’, while medicalisation is properly conceptualised as that which ‘turns the moral into the medical’ with the proposition of biomedical causes and interventions (Conrad 1992). The invocation of genetic factors and a disease model would thus constitute evidence of medicalisation for the latter definition of this process. As indicated by these various debates, medicalisation is itself the object of claims-making.
Again, these texts are not generalisable to the medical profession as a whole, and medicalisation is a sociocultural process that is not limited to, or necessarily dependent on, the activities of the medical personnel (Conrad 1992). Sobal (1995) has undertaken such broader analysis and concluded that there has indeed been an overall increase in the medicalisation of fatness over the past century with some recent claims towards demedicalisation. These texts do demonstrate, however, that subjection of a condition to medical conceptualisation can entail attention to individual as well as context, or agent as well as structure. Furthermore, they bring to light the fact that emphasis on the individual, the usual depiction of a medical model, can be associated with individual liability as well as reprieve. This suggests that commonly invoked dichotomies such as individual vs. structure, and medical (organic) vs. social (non-organic), oppositions which are often conceived as overlapping, are actually cross-cutting. Strong (1979) and Conrad/Schneider (1980) also debated the consequences of a social, or non-organic, model of health on individual freedoms. Strong argued that a social model of health could entail closer monitoring and policing of behaviour such as people’s nutrition or leisure activities. Conrad and Schneider argued, in response, that the dangers of a social model are contingent on who controls that model and the supervision of programmes charged with ameliorating the problem. Drawing on our findings, we can, perhaps, clarify, or at least reframe, this debate by proposing that the implication of a social model of health depends, first and foremost, on whether the ‘social’ denotes emphasis on (and targeting of) individual behaviours or the structuration of such behaviours by sociopolitical organisation.
Social and historical contours
The transition in etiological focus from individual behaviours to generic environmental determinants represents a process of positive assertion rather than passive observation, and such change must be situated in the context of broader processes of social and cultural organisation. We suggest some selected and initial points of inquiry here, as the avenues of relevant exploration are many in this regard, and a comprehensive analysis is beyond the scope of our current study.
At the turn of the century, Western medicine inherited not only a legacy of secularisation, but also the scientific and bureaucratic rationalisation of society with industrial capitalism. Turner (1996) has argued that the project of modernity was critically linked to an ideology of rational, self-controlled mastery over the desires of the labouring body, with religious moral authority transferred to secular institutions such as the medical profession. Drawing on Foucault’s (1977) analysis of the systemisation of rational surveillance over the body and the subordination of desire to reason, Turner regards the growth of nutritional management as a rationalisation of conduct and an important technique aimed at improving labour efficiency. In an extensive cultural history of dieting in the United States, Schwartz (1986) has argued that in the first decades of the 20th century, impelled by movements ranging from Taylor’s scientific management in the workplace to domestic and nutritional science in the home, the body is cast in terms of an economy and efficiency to be centrally and willfully regulated. During WWI, Herbert Hoover, as head of the US Food Administration, mounted a propaganda campaign for domestic food conservation with slogans such as ‘Food Will Win the War’, equating individual excess with treason (Schwartz 1986). During the Depression, Americans were again asked to show restraint in food consumption, this time by Hoover’s presidential administration (Schwartz 1986). These notions resonate with early models in Cecil that concurrently frame obesity as a function of individual misconduct and as a deterrent to labour productivity and societal well-being.
Similar to Foucault’s (1977) description of the emergence of detailed examinations and taxonomies for discipline of the body, Schwartz (1986) describes an early 20th century progression within multiple sectors of the US, ranging from the insurance industry to criminology, wherein bodies are increasingly weighed and charted, with weight functioning as an index of moral character. Such speculation about the connection between body shape and character has been related to anxiety over the arrival of new immigrants from Southern and Eastern Europe (Saukko 1999). For example, obesity and an attendant portrayal of slack personality were often associated with immigrant groups such as Italians and Jews during this era (Saukko 1999). Again, it is in the context of such broader processes that obesity is attributed to moral failure and personal deficiency in the early editions of Cecil.
Armstrong (1983) has argued that in the course of the 20th century the clinical gaze comes to exceed the space bounded by body and clinic, moving into the surrounding social sphere. In the later editions of Cecil, individual accountability recedes in favour of vigilance over the surfeit of conveniences found in Western consumer culture. Many have postulated that the modern cult of slimming and health consciousness relates to a moral and aesthetic rebellion against the gross excesses of consumer culture, and point to the contradictions imposed by late capitalism, a political economy that at once requires ever higher levels of consumption and the regulation of desires to cultivate production (Crawford 1984, Schwartz 1986, Bordo 1993, Stearns 1997). In the postmodern economy of advertising, information technology, and service industries, the labouring body has become the desiring body, a body in search of personal satisfaction through an emphasis on leisure and seemingly insatiable consumption practices (Turner 1996). Cultural imaginings and ethical precepts with respect to individual, physical bodies may reflect anxieties concerning the social body (Douglas 1966). Fatness, therefore, may function as a site for social reflection, or social diagnosis, an index of our relation to consumption in advanced consumer capitalism.
Criticism of the food industry emerges in Cecil in the 1960s, a time of civic unrest in the US and the rise of postindustrial political sentiments advocating distrust in government, industry and consumerism. Levenstein (1993) has traced how elements of the New Left, consumers’ rights activists and environmentalists seemed joined together in attacking the food industry, particularly with respect to the marketing and proliferation of processed foods. Attacking the food industries was becoming ‘a mini-industry in its own right’ with a willing media and a receptive middle class audience that was losing faith in the food industries and the government (Levenstein 1993). The postwar period in America was also marked by a psychiatric turn to psychoanalysis and the role of childhood traumas, along with popular and academic anxiety about suburban conformity, with particular attention to the conduct of the mother in middle class families (Saukko 1999). In Cecil, it is during this time period that mothers and American homes are charged with enforcing early patterns of overeating.
The representation of obesity also informs, and is informed by, other social tensions. In the US, obesity disproportionately affects women of lower socioeconomic status and certain racial and ethnic groups (Sobal and Stunkard 1989, Kuczmarski et al. 1994). The framing of obesity, therefore, does not operate independently of the framings of race, class, and gender. For example, the female body has been symbolically configured as a site of excess and irrationality (Bordo 1993), resonant with a view of obesity as a failure of rational self-control. Obesity is also correlated with poverty, and we have noted the linkage of bodily constitutional types with various immigrant groups earlier this century. More generally, poverty itself is enmeshed within arguments concerning the role of individual responsibility versus victimisation caused by unfavourable social circumstances. The association of diseases with poverty must be placed in the context of multiple long-standing, historical debates on the causal nature of these relationships, the direction of causation, the emphasis on collective versus individual malady, and the efficacy of economic versus medical intervention (Schwartz 1984, Eyler 1992).
One might also consider the influence of shifting theoretical paradigms within biomedicine. World War II has been described as a watershed for the beginning of the chronic disease era for epidemiology (Susser 1985, Susser and Susser 1996). The force of the germ theory paradigm was fading, with infectious disease mortality greatly reduced by higher living standards, vaccines, and chemotherapy. Additionally, the population distribution was shifting toward older age, and chronic diseases, such as coronary heart disease and lung cancer, so-called ‘diseases of civilisation’, achieved prominence. A new theoretical paradigm and research framework emerged from the standpoint that chronic diseases have environmental causes, many of which are preventable. This post-war reorientation of focus with respect to causative attribution is certainly consistent and coincident with the turn to environment we describe for the explanation of obesity in Cecil. Reflecting more recent trends in biomedical research, the turn to genetics in the latest edition signals the increasingly hegemonic status of molecular epidemiology and genetic-based explanations.
The post-war period was also marked by the institutionalisation and rapid growth of federal funding from the National Institutes of Health (NIH), which was internally subdivided into organ- and disease-based categorical institutes. This organisational shift, in part, prompted an increase in the subspecialisation of research and training (Howell 1989). In the early 1970s, the American Board of Internal Medicine acknowledged Endocrinology and Metabolism as a new subspecialty area with obesity falling under its purview. In Cecil, authors for the first two entries are affiliated with the general specialty of internal medicine, whereas authors for the later entries are affiliated with the more narrow subspecialty of endocrinology. Advocacy for the delineation and legitimisation of a new subspecialty may, in some instances, motivate a reconfiguration of clinical phenomena into more concrete disease categories and direct research funds towards organic-based entities (Lawrence 1992). Hence, with respect to our findings, the recasting of obesity as disease and the appeal to genetically-mediated, organic causes may be associated with wider professional or organisational transformations.
Lastly, one might consider how knowledge and representation of an object is refashioned as the object itself changes. The proportion and absolute number of overweight Americans has increased dramatically over the time period covered by these textbooks. Recent studies show that in the last decade alone obesity increased from 12.0 per cent to 17.9 per cent (Mokdad et al. 1999), and that over half the adult population is now overweight or obese (Flegal et al. 1998). Perhaps an increase in numbers afflicted can itself render a phenomenon less deviant, less a matter of individual particularities, and precipitate recourse to more systemic effects. Indeed, obesity is now framed in the discourse of public health and cast in the terminology of ‘epidemics’ and even ‘pandemics’ (Egger and Swinburn 1997, Mokdad et al. 1999). Recently, authors in the British Medical Journal proposed an ‘ecological’ approach to obesity wherein obesity is regarded as a normal response to an abnormal environment (Egger and Swinburn 1997), and others in Science proposed that we ‘cure’ the environment to reverse the obesity epidemic (Hill and Peters 1998), echoing the findings of our study.
We do not wish to suggest that individuals are no longer held accountable for their own weight status. The diet, fitness and beauty industries are certainly subtended by, and dependent on, a discourse of individual agency, and obesity continues to be a highly stigmatised state with prejudice and reprobation against obese persons widespread within medicine and society at large. Obesity is often described as the last realm of socially permissible discrimination, and there continues to be enormous social and medical pressure to lose weight (Kassirer and Angell 1998). Moreover, dietary modification is but one component of a modern medical regimen promoting healthy lifestyles as a solution to problems ranging from cancer to sexually transmitted diseases (Fitzgerald 1994). As Foucault (1988) has suggested, the site of disciplinary power has moved, in recent times, from institutionalised surveillance to ‘technologies of the self’. On the other hand, the rubric of victimisation has been mobilised in identity politics by activist groups seeking non-discriminatory rights, and groups in the US such as NAAFA (National Association to Advance Fat Acceptance) have recently employed such a platform for political discourse. Sobal (1995) has argued that such developments can be viewed as a demedicalisation of obesity and the deployment of a political model in place of a medical model; the pattern of medicalisation and demedicalisation has much precedent (Fox 1988).
Textual representation and science
Lastly, we conclude by reflecting on the act of writing, an act which is a decidedly social practice. Over the last two decades, critical theoretical insight from multifarious disciplinary commitments has revealed the limits of representation, problematising various traditional assumptions subtending the production of knowledge. Points of contention are many, but examples include: the relation of the subject to his or her object of research (Bourdieu 1977, Cifford 1986, De Certeau 1986); the ideal of an Archimedean point of value-free epistemological privilege (Smith 1987, Haraway 1991); and the objective and subjective dichotomy between scientific and literary textual practice (Clifford 1986, De Certeau 1986, Latour 1987, Agger 2000). The entries from Cecil are written in accordance with rhetorical conventions typical, and indeed expected, of scientific texts; the discursive posture signals de-authored transparent representation, or mere reflection, of an external world of accumulated facts. This mode of representation, however, obscures the actions and conditions of its production. Scientific texts are the product of deliberate authorial choices; findings are narrated, and particular problematics are selected over others, while assumptions and ellipses are deferred (Agger 2000).
Our analysis of Cecil effectively illustrates this process, with each permutation on obesity highlighting the authoriality (Agger 2000) and argumentation that determines representation, as well as the underdetermined, contingent nature of all versions2. The editions we present are indeed written by five different authors, and their perspectives are shown to be situated and partial (Haraway 1991). Authorial choices are embedded in broader social and historical contexts, and we have offered some initial speculation in this regard. Detailed and substantive inquiry into the local, individual circumstances of each author is beyond the scope of this project, but we do note their institutional and specialty affiliations. The question arises as to whether the changes we find represent change with time or simply differences between authors. Our answer is that they are undoubtedly secondary to both, and one cannot be unequivocally privileged over the other. Moreover, the relatively monotonic nature of the changes suggests that it is not merely author idiosyncrasy that drives the changes, but rather that there is also a real shift in the perception of obesity that the authors are capturing3.
Using Cecil and obesity as a specific case-example, our work demonstrates that despite a relatively stable overall physiological process, substantive changes can occur between levels of explanatory analysis with an ongoing dialectic between these explanations and constructions of culpability and remediation. Given that each level of analysis and focal point for explanation has distinctive ideological and public health implications, we find that the intelligible body, a body explained, is always the useful body, a body subject to sociopolitical regulation (Foucault 1977).