Making space for fat bodies?

1 Conceptual tenets of obesogenic environment research

Originally defined by Swinburn et al. (1999: 564) as ‘the sum of influences that the surroundings, opportunities or conditions of life have on promoting obesity in individuals and populations’, conceptually ‘obesogenic environment’ research aims to provide an ‘ecological perspective’ (Smith and Cummins, 2008) to understandings of obesity. This perspective situates the ‘causes’ of obesity in a body’s relationships and interactions with particular (physical, social and cultural) ‘environments’. As with other work in medical geography, research on obesogenic environments seeks to identify and interrogate the different qualities of ‘health-promoting’ and ‘health-depleting’ environments, in order to explain why ‘some places have more obese people than others’ (Procter et al., 2008: 323).

There is debate within this work about the best way to conceptualize the ‘environment’ (see, for example, Harrington and Elliott, 2009, on the use of ‘neighbourhood’). However, central to geographical research is a focus on the interaction between factors at multiple scales. As Smith and Cummins (2008: 520) explain, much of this work uses ‘“multilevel” conceptual and analytical models to assess the independent relationships between people, places and obesity because they allow researchers to explore influences operating at varying geographical scales, such as the home, neighbourhood, city or country’. Exemplifying this, the most commonly used framework within research on obesogenic environments – ANGELO (Analysis Grid for Environments Linked to Obesity) ⁶ – divides environmental factors into two scales: micro (such as neighbourhood recreational facilities, ‘healthy’ food availability or school-based policies on physical education); and macro (such as regional planning policies and the perception of obesity in national media). It then plots ‘obesogenic factors’ across these two scales and four ‘environments’ (physical, economic, sociocultural and political) (Harrington and Elliott, 2009). Within this model and related work, the ‘environment’ therefore exists as a range of measures or descriptors which indicate the uses, quality and physical form of spaces rather than a research site in and of itself. Bodies predominantly exist as a range of measurements such as Body Mass Index (BMI) and Waist to Hip Ratio (WHR) ⁷ and demographic descriptors such as gender, age, ethnicity and income. Borrowing from medical models of the body, such work begins from the assumption that bodies of a particular size are inherently unhealthy and that there are ‘known’ relationships between different population groups, average body size and health (see, for example, Edwards et al., 2011; Moon et al., 2007).

The relationship between bodies and environments is conceptualized with reference to the simplistic ‘energy balance model’ which underpins many medical accounts of obesity. This model, as Pearce and Witten (2010a: 4) explain, suggests that, ‘at a fundamental level, obesity arises from an imbalance between the quantity of energy consumed and the amount expended’. Obesogenic environments are therefore those which disrupt the body’s ‘natural’ energy balance’ (Foresight, 2007: 6). ⁸ Indeed, work on obesogenic environments often focuses on either food environments (energy in) or physical activity environments (energy out) and/or classifies ‘obesogenic factors’ according to this model (see, for example, the thematic organization of chapters in the edited collection by Pearce and Witten, 2010b). Within Geographies of Obesity, this relationship is further conceptualized with reference to ‘deprivation amplification’, whereby it is assumed that ‘exposure to poor quality food environments amplifies individual risk factors for obesity such as low income, absence of transport, and poor cooking skills or knowledge’ (Cummins and McIntyre, 2006: 100).

It is also apparent within this work that what actually constitutes an ‘obesogenic environment’ is often relatively non-specific (Lake and Townsend, 2006). For example, Pearce and Witten (2010a: 5) refer to ‘all factors that are external to the individual including the social, political, economic, built or biophysical spheres’. This multiplicity of factors has been identified as problematic since, as Kirk et al. (2010: 116) warn, ‘not only is it [obesogenic environment] an elusive concept that is difficult to define and conceptualize, but attempting to consider every possible environmental contribution to energy balance can quickly become overwhelming’.

In addition, the environment exists as a spatial container, with a series of characteristics conceptualized as barriers and/or resources (or lack thereof), which facilitate or impede energy flow in/out of bodies. It acts outside, and yet is intimately involved in, the production of obese or potentially obese bodies that inhabit it. While Geographies of Obesity acknowledge that care must be taken ‘not to fall into the cul-de-sac of environmental determinism’ (Smith and Cummins, 2008: 530–531), the genealogy of models such as ANGELO firmly situate this work within an environmental/infectious disease approach ⁹ , and so a deterministic model persists. This is particularly evident in the use of terms such as ‘exposure’, ‘potency’ (Pearce and Witten, 2010b) and ‘toxicity’ to explain how ‘obesogenic environments’ produce obese bodies (Kim and Kawachi, 2010; Smith and Cummins, 2008). This is also evident in some recent political ecology inspired work which explains obesity as a result of environmental toxins (Guthman, 2012).

There is growing recognition of the limitations of this conceptualization of the environment within Geographies of Obesity since, as Townsend and Lake (2009: 913) state, the focus is on what is in the environment rather than how the environment is used, meaning there is ‘generally little examination of what might be called “fitness for purpose”’. Such questions signify an openness to thinking beyond the built or physical environment, prompting critical engagement with problems inherent in identifying obesogenic factors and thus environments. Far from a static concept, then, the ‘obesogenic environment’ within geographical research is always and already open to questioning because of the difficulties in understanding precisely what or whom is influencing bodies and environments and how they interact. At present, however, the questions being asked are limited in that they fail to challenge the usefulness of the underlying ‘energy balance’ model or the assumed relationship between fatness and health. It is here then that we suggest engagement with Critical Geographies of Obesity/Fatness as potentially useful in furthering this research, as will be made evident through the research trajectories examined in this paper.

2 Methodological tools and debates in obesogenic environment research

In methodological terms, obesogenic environment research uses a combination of existing secondary data sets, specially commissioned surveys, observational fieldwork and mapping, and, although less common, some studies incorporate qualitative work with local communities (Pearce et al., 2009). Where data is not available at small area level, secondary data are often used in micro-simulation models and in the production of ‘synthetic’ estimates of body size based on a combination of existing body size data and demographic variables (see, for example, Edwards and Clarke, 2009; Edwards et al., 2011; Moon et al., 2007; Stafford et al., 2007). Situated within a positivist research paradigm, emphasis is placed on identifying ‘good and reliable tools and indicators’ (Elinder and Jansson, 2009: 312). The term ‘objective measures’ (Lake et al., 2010; Townsend and Lake, 2009) is used to refer to secondary data or any data that are generated directly by a researcher through, for example, using government/industry databases or telephone directories to identify the location of food outlets (Fraser et al., 2010). These are considered to be more reliable than ‘perceived measures’ (Townsend and Lake, 2009) or self-reported data which are based upon the perceptions of people who live in particular environments. However, it is also widely appreciated that such ‘objective’ measures may be limited in their ability to understand fully how people use, and understand, their environments. Thus, Townsend and Lake (2009) suggest that more mixed-methods research is needed.

Further methodological concerns relate to the difficulty in establishing causality in a field which, according to Townsend and Lake (2009: 909), ‘contains a great deal of correlation studies rather than exploring cause and effect’. In fact, as Mujajahid et al. (2008: 1356, cited in Turrell, 2010: 151) argue, ‘although neighbourhood environments are often identified as potentially important factors in understanding the obesity epidemic, little research provides evidence of this importance’ (see also Townsend and Lake, 2009: 912). As such, concerns have been raised within this scholarship about the conceptual basis for addressing the relationship between bodies and environments, and it has been suggested that alternative theoretical approaches are needed (Moon, 2010; Thornton and Kavanagh, 2010). Acknowledging these limitations draws attention to particular commonalities between Geographies of Obesity and Critical Geographies of Obesity/Fatness. In particular, an openness to explore alternative theoretical models for understanding bodies and environments is important to our proposed critical account of obesogenic environments, along with a commitment to work with, rather than control for, complexity, and to acknowledge alternative notions of causality (Moon, 2010). In light of this, in the following section of the article, we demonstrate how critique can function to draw together, as well as highlight differences between, the two areas of work on obesity/fatness within academic Geography in order to develop what we term Critical Geographies of Obesogenic Environments.

1 Measuring (obese) bodies

As mentioned earlier, obesogenic environment research relies on the use of particular measures in order to categorize the qualities of environments and those who inhabit them. Underlying this ‘search for the right variable’ is the assumption that there must be a relationship between the environment, body size and health. The purpose of such research is to establish which measures best illustrate that relationship, despite little definitive evidence that this relationship exists. Indeed, as the following conclusion to Turrell’s (2010) systematic review of obesogenic environment research on neighbourhood physical activity environments and bodyweight illustrates, rather than using data to challenge assumptions about the relationships between environment, body size and health, research which poses a challenge to these assumptions is often ‘put aside’ in order to maintain simple models which identify obesogenic characteristics:

The evidence was mixed and inconsistent, and for each statistically significant association that was reported there was often an accompanying null finding…However, there is another way of interpreting the pattern of evidence which lends itself to the conclusion that the neighbourhood environment influences bodyweight…If we put aside the null findings (and admittedly run the risk of over-stating and simplifying things) and use the positive evidence to devise a neighbourhood that was conducive to a healthy bodyweight then it would probably have the following characteristics. (Turrell, 2010: 163–164)

Obesogenic environment research is, therefore, often reliant on microsimulation models in order to produce data at the required scale. These models use demographic variables, such as gender, age, ethnicity and class, in the production of synthetic data (see, for example, Edwards et al., 2011; Moon et al., 2007). Thus, the inadequacies of measures such as BMI/WHR to account for bodily differences across different population groups (see Ross, 2005) are actually integral to the production of synthetic data sets. This is acknowledged to some extent; for example, Moon et al. (2007) explain that it is crucial to recognize that any conclusions drawn on the basis of such ‘synthetic estimates’ ‘may give a misleading picture: we are in effect describing geographic variations in the sociodemographic profile that is [assumed to be] associated with obesity/overweight. This may not be the same as the actual geography of obesity/overweight’ (p. 29). However, such caveats fail to acknowledge the problems with using BMI in the first place (Evans and Colls, 2009); problems which may be exacerbated when race, age and gender are used to produce synthetic data.

Critically engaging with the type and use of measurement within research on obesogenic environments is therefore important to our approach. This is not only because of the problems with these measures outlined above, but also because, even where data is available within required areas, synthetic estimates maintain a certain power to constitute the bodies within those areas. For example, as we discuss elsewhere (Evans and Colls, 2009), in the case of the National Child Measurement Programme (NCMP) in the UK, estimates of obesity/overweight prevalence were used to assess the reliability of the data produced by measuring school children. Thus, the synthetic estimates were taken as more reliable than the empirical measurements. It is therefore important that geographers reflect not only on the adequacy of the data, but on the implications of their data for the populations they come to constitute.

2 Morality

Despite some attempts to acknowledge the moral discourses that surround body size (see Smith and Cummins, 2008), obesogenic environment research has been largely complicit (albeit perhaps unintentionally) in the reproduction of moralized assumptions about fat bodies and the environments they inhabit. Critical obesity researchers (including geographers) have argued that the production of obesity as a ‘problem’ has occurred through the combination of medico-scientific knowledges and ‘common sense’ or moral knowledges within which ‘pre-existing yet largely unexamined cultural understandings of fatness form the plinth of representations of scientific debate or agreement about weight’ (Boero, 2007: 51; see also Evans, 2006; Evans and Colls, 2009). While the focus on the environment within Geographies of Obesity may be driven by a desire to ‘remove moral blame from individuals from getting fat and to place it on social and economic factors’ (Kirkland, 2011: 466), important critical work is still required to understand how moral knowledges about different population groups inform the identification of ‘at risk’ places and bodies (Guthman, 2011; Shannon, 2013).

To this end, it is worth considering specific incidences where assumptions about the capacities and activities of particular classed, racialized and gendered bodies inform the selection of variables and researchers’ interpretations of data in obesogenic environment research. First, in class terms, as Kirkland (2011) and Guthman (2011) argue, the selection of variables that indicate obesogenicity are often indicators of wealth or reflect elite ideologies about what makes an environment, and by implication its inhabitants, ‘desirable’. These variables reveal ‘unstated preferences for places with the amenities often associated with urbane, privileged environments, including university towns, artsy enclaves, gentrified urban cores’ (Guthman, 2011: 86). For example, Harrington and Elliott (2009: 596) consider social cohesion to be an important ‘environmental’ determinant of overweight/obesity. Their selection of ‘proportion of homeowners vs. rental homes…as a proxy for neighbourhood social cohesion’ reveals an assumption that social cohesion is related to wealth/class. Likewise, Van Dyck et al. (2011: 973) rate seven items within the category ‘Esthetic [sic] related problems’, including ‘graffiti, unmaintained green spaces and illegal posters’. It is assumed that these factors will increase the propensity to be obese because they are seen to make the environment unattractive for people to be physically active. Giles-Corti et al. (2010: 138) also describe such environments as fearful and argue that fear produces a physiological reaction which causes weight gain by affecting the body’s ‘allostatic load’. ¹⁰ In such arguments, classed aesthetics are clear, with little consideration that graffiti or ‘illegal’ posters may be aesthetically pleasing to some, and/or may be evidence of particularly (physically) active subcultural communities. Moreover, there is no reflection on how ‘fear’ might be produced through the continued stigmatization of fat bodies, a process with which, we argue, work on obesogenic environments is complicit through reinforcing the labelling of all fat bodies as pathological.

Second, critical geographical research has highlighted how areas or environmental characteristics marked as ‘unhealthy’ or obesogenic are often those associated with particular racial or ethnic groups. For example, Herrick’s (2008: 2725) work in East Austin illustrates how ‘obesity has been taken up to mark racialized social and spatial difference’:

Figures [BMI data] show clearly that white residents, on average, have higher rates of obesity than Hispanic residents across the whole city. This figure is in stark contrast to the perception expressed by many interviewees that Hispanic residents are more prone to be obese. (Herrick, 2008: 2726)

Third, gendered assumptions also inform the selection of variables within research on obesogenic environments. Reflecting broader discourses that position obesity as a ‘modern condition’, variables often reflect a nostalgic understanding of what constitutes ‘traditional’ and ‘healthy’ lifestyles. In particular, obesity is often positioned as a consequence of the changing roles of ‘working women’. For example, Giles-Corti et al. (2010: 140) use the variable ‘hours worked by mothers’ not as an indicator of the physical activity that women do, but as a proxy measure for unhealthy lifestyles, implying that working mothers have a negative impact on a family’s health.

Across these examples, it can be argued that knowledge about obesogenic environments is rooted in discourses which reproduce middle-class, white, ‘nuclear’ family lifestyles as healthy lifestyles (Rawlins, 2009), with ‘other’ population groups identified as problematic (McPhail, 2009). Assumptions about race, class and gender (along with others such as age, sexuality, disability, and the intersections between these) are therefore written into research on ‘obesogenic environments’ through the ways in which places, as well as bodies, are posited as ‘other’ and ‘risky’ (Herrick, 2008). Thus, Kirkland (2011) argues that there is a hidden moralism through which the impression is created that:

some people are impervious to bad environments (the elites, who still manage their bodies properly) while others are more fully constructed by their environments (poor fat people). Members of one group move powerfully through the world determining their body sizes and health statuses; others are pitiably stuck within and determined by the environment. (Kirkland, 2011: 467–477)

3 Embodied environments

The final critique of obesogenic environment research that we want to discuss is concerned with the ‘type’ of bodies that are reproduced in such research. Indeed, we argue that a direct consequence of using bodily measures such as the BMI, or categorizations of bodily difference such as gender, ethnicity and socio-economic status, is a paucity of thinking, feeling, moving bodies in research on obesogenic environments. This means that the fleshy materialities, physical capabilities and sensuous dispositions of bodies that inhabit obesogenic environments are not considered. The use of secondary data also means that, despite an interest in physical activity and the movement of bodies, this research often requires bodies to be fixed in particular spatial boundaries (Primary Care Trust, postcode, etc.) and/or population groups (ethnicity, age, gender, class) (see Turrell, 2010). Where primary data is collected, engagement with bodies often remains limited. For example, the growth in research that uses Geographical Positioning Systems (GPS) to map bodies’ movements (Cooper and Page, 2008) is still more concerned with ‘measuring’ frequency and extent rather than questioning why, how and with whom bodies inhabit (obesogenic) environments, or what it feels like to inhabit those environments. Moreover, there is often an assumption that the environments a body encounters are those nearest to home.

While there have been some calls for more qualitative work relating to obesogenic environments (Townsend and Lake, 2009), there is still a lack of engagement in Geographies of Obesity with those populations being researched. Indeed, even when ‘why’ and ‘how’ questions are asked about relationships between people and different spaces and places, there remains a pre-determined consensus about what constitutes healthy and unhealthy bodies, behaviours and environments. There is little, or no, exploration of alternative, non-size related models of health (such as HAES), or the potential for participants in these studies to redefine health and/or well-being for themselves.

Conversely, Critical Geographies of Obesity/Fatness can be firmly situated within wider debates in academic Geography concerned with ‘the body’ (see Longhurst, 2001) which centre both the fleshy materialities of fatness (Colls, 2007) and the everyday emotional experiences of being a sized body (Hopkins, 2008, 2012; Longhurst, 2005, 2010, 2012). This work has demonstrated the value of understanding how and why sized bodies inhabit different spaces, as well as exploring the personal consequences of categorizing bodies as obese. For example, Hopkins (2012) explores the self-consciousness and fear of judgement from other people that young fat people experience when eating in public; Colls (2006) documents the emotional experiences surrounding body size for fat women when shopping for clothes; and Colls (2012) reveals the importance of ‘size accepting’ spaces for fat people to feel safe away from potential discrimination in everyday public life. Moreover, Fat Studies research drawn from outside geography draws attention to the availability (or not) of spaces that are physically accessible for fat bodies (see Bias, 2012, on public transport); while other work acknowledges that fat people are physically active (see Ellison, 2009, and Schuster and Tealer, 2009, on fat aerobics) – presenting a direct challenge to assumptions to the contrary that inform research on ‘obesogenic environments’.

From these examples, it becomes obvious that by not fully engaging with ‘how’ fat and potentially fat bodies actually experience and narrate their bodies and environments, obesogenic environment research is not fully engaging with a range of ‘factors’ such as discriminatory practices, stigma (see Cooper, 1998; Puhl and Heuer, 2009) and accessibility, which affect relationships between bodies, and between bodies and environments (Longhurst, 2010). There are potential connections here between research in health geography ¹¹ concerned with place-based stigma (Keene and Padilla, 2010; Pearce, 2012) as a barrier to health (Pearce, 2012) and Fat Studies research which considers how stigma prevents fat people from accessing health care (Solovay, 2000). However, we suggest that caution is needed in exploring these connections. Our critical engagement with obesogenic environment research, thus far, has demonstrated the inherent problems with labelling particular bodies and environments as obese and/or obesogenic. We are therefore wary of any research which may involve the identification of particular places as problematic because this could exacerbate their stigmatization. For example, Guthman (2011) suggests that identifying particular areas as ‘obesogenic’ may lead to disinvestment in those areas or, conversely, to gentrification which may cause the displacement of those populations who previously lived there. Thus, it is vital that geographers researching obesity/fatness remain vigilant to the often unintended, but possibly harmful, consequences of categorizing and labelling particular places as stigmatized and/or obesogenic.

1 (Re)conceptualizing obesogenic environments

First, a ‘critical geography of obesogenic environments’ requires a reconceptualization of the term itself. Throughout this article we have argued that it is vital that geographical research avoids, or at least considers the repercussions of, reproducing a hegemonic pathologization of fat bodies as inherently unhealthy. In contrast, as Fat Studies and HAES scholars argue, geographical work must acknowledge that it is possible to be fat, fit and healthy (Bacon and Aphramor, 2011). Such an acknowledgement would mean not relying upon flawed measures of body size as proxies for health, and avoiding the reproduction of racial, classed and gendered assumptions when assessing the quality of environments. It also requires untethering the relationships between bodies and environments from an energy balance model which narrowly conceives of a body as produced through its energy consumption and expenditure. In so doing, this approach offers the potential to answer calls from within Geographies of Obesity for a much-needed debate about the conceptual and theoretical models on which this work is premised (Moon, 2010). This would facilitate important and original geographical contributions to interdisciplinary knowledge about the relationship between health and place by critically engaging with the contradictions, complexities and inconsistencies within existing research concerning causation, definitions and the use of variables (as discussed earlier). It would therefore allow the development of new understandings of the relationships between health and environments, which are not reducible to simplistic measures of body size.

This alternative stance can be contextualized within Fat Studies and HAES research in which health and well-being are divorced from weight/size (see Rothblum and Solovay, 2009). This does not necessitate a movement away from a concern with health inequalities per se since, as Bacon and Aphramor (2011: 1) explain, ‘the primary intent of HAES is to support improved health behaviours for people of all sizes without using weight as a mediator’. What obesogenic environment research does tell us is that there are a number of issues that affect the everyday lives of differently sized bodies living in particular environments. Indeed, both Geographies of Obesity and Critical Geographies of Obesity/Fatness are committed to understanding and intervening in particular economic, political and social factors that might affect a body’s capacity to be well, and how these factors map onto broader structures of social disadvantage. Therefore, a HAES-informed approach would enable a Critical Geography of Obesogenic Environments to document the range of factors that can affect (the health of) bodies in ways that may or may not relate to their bodyweight or size. For example, these might include food pricing, food availability, and access to recreation spaces, footpaths and cycle lanes, while taking into account how different bodies experience, move within, and interact in different spaces.

We argue that this approach could also be facilitated by aligning obesogenic environment research with the social model of disability which, as Parr and Butler (1999: 4) explain, recognizes that ‘it is society’s organization which “disables” people with physical and/or mental limitations so that they are marginalized socially, economically and politically’. While there are debates within Fat Studies about the potential of disability theory and politics to facilitate a challenge to the stigmatization of fat bodies (see Cooper, 1997, 2010; Kirkland, 2008), and about problems with the social model of disability in its inability to acknowledge the ‘true pain’ of disability (Hall, 2000), there is potential here for this theoretical framework to envisage a reconceptualization of obesogenic environments – not as environments that make bodies (fat), but as environments that make (fat) bodies problematic. Thus, rather than seeing all bodies of a particular size as unhealthy, this shift would involve recognizing that bodies can be healthy, regardless of size, and that what matters in terms of inequalities is how physical, social and legal barriers may prevent (fat) bodies from being well/healthy. For example, Longhurst’s (2010: 212) work with fat women who live in Hamilton, New Zealand, offers an insight into the kind of geographies that this re-theorization might enable through her documentation of how ‘environments, both emotional and material, can be disabling for fat people’.

This is a fundamentally different approach to one that is only interested in understanding how environments cause fat. The intention is not to facilitate fat bodies’ access to public space or particular foods in the hope that this activity might make them thin, but rather, as the quote from Kirkland (2008) opening this section indicates, it involves facilitating fat people’s access to, and use of, public space through recognizing our rights to public space as fat citizens.

2 Researching fat bodies and environments

The second research trajectory that we propose is concerned with methodologies. In particular, we would like to comment on the ‘weight bias’ that often emerges through the research process. Puhl and Heuer (2009: 941) use this term to describe ‘inequalities in employment settings, health care facilities, and educational institutions often due to widespread negative stereotypes that overweight and obese persons are lazy, unmotivated, lacking in self-discipline, less competent, non-compliant and sloppy’. While they use this term to describe how fat people are discriminated against in their everyday lives, as we have demonstrated in this article, such a ‘bias’ is also present in obesogenic environment research, evident in the assumptions and models that inform decisions about the variables used to produce synthetic data and identify obesogenic environments. It is also a product of the marked distance between those producing knowledge about obesogenic environments, and the bodies of those about whom this knowledge is being produced. In short, geographies of obesogenic environments, although concerned with fat, are thoroughly disembodied.

We argue that Critical Geographies of Obesogenic Environments require a methodological shift facilitated by engaging with work from feminist, queer, postcolonial and disability theory. Such work highlights the importance of the multiplicities of embodied experience, and challenges any ‘distinction and distancing between the reality out there (which we map), and the in here (our bodies or selves)’ (Binnie, 1997: 224). For example, within Critical Geographies of Obesity/Fatness (and associated work in Fat Studies) there is an emerging body of work which draws on autoethnography (Longhurst, 2012; Murray, 2010) to highlight the everyday experiences and contradictions of being a fat (academic) body. ¹² In short, regardless of method, it is necessary to reflect not only on our own bodies as researchers, but on how research may be informed by a personal investment in particular ‘grids of right and wrong’ about what researchers ‘do to make their bodies and identities’ (Guthman, 2009: 1125).

We also suggest that Critical Geographies of Obesogenic Environments be developed further through participatory research. The aim here would not be to find an additional variable to understand better obesity causation, or to provide qualitative data within a pre-given framework of ‘healthy’ and ‘unhealthy’ lifestyles (Townsend and Lake, 2009). Instead, such research would document how people narrate their everyday embodied lives in relation to their own understandings of ‘health’ in ways which may or may not reveal the enabling/disabling effects of particular (physical, social, political) environments on bodies of different sizes. It could also uncover their grasp on environmental inequalities that are not related to weight and do not demonize or stigmatize fat bodies (in line with the conceptual shift outlined above). We are not arguing here for exclusively qualitative work. There is real potential for critical quantitative geographies and Participatory GIS (PGIS) to contribute to this work (Shannon, 2013), since ‘a Participatory GIS celebrates the multiplicity of geographical realities rather than the disembodied, objective and technical “solutions” which have tended to characterize many conventional GIS applications’ (Dunn, 2007: 616). This approach should allow (fat) people to define and document for themselves the elements of their ‘environment’ which may affect how they access and move around in it, and how this makes them feel.

3 Re-theorizing body-environment relations

Finally, we want to suggest that integral to the development of Critical Geographies of Obesogenic Environments is a re-theorization of how body-environment relations are conceptualized. This article has reiterated throughout that, although wary of reproducing an environmentally deterministic account, much work on obesogenic environments contains elements of this. In short, the environment, however broadly conceived, is seen to determine body size. This is reflected in the evolutionary arguments that are used to justify the necessity of an ‘ecological’ or ‘environmental’ account of obesity. For example, with reference to Egger and Swinburn (1997), Smith and Cummins (2008: 519) suggest that ‘explanations for the rapid rise in obesity must originate from changes in the environment, as the distribution of genetic traits predisposing individuals to become overweight or obese remains unchanged over short biological time scales’. Central to this argument is the assertion that obesity is a specifically ‘modern’ condition, caused by living in environments which are ‘toxic’ because they disrupt ‘natural’ human-environment relationships. Particular understandings of nature and modernity are, therefore, central to these arguments, and are premised on a model in which, as Grosz (1999: 382) explains, ‘cities have become (or may always have been) alienating environments, environments which do not allow the body a “natural”, “healthy”, or “conducive” context’.

While there have been calls within Geographies of Obesity to re-theorize causation within body-environment relations (Moon, 2010), we suggest that what needs to be questioned is whether a causative model of this relationship can ever productively avoid the pitfalls of environmental determinism. Indeed, we argue that it is necessary to open up ideas about ‘nature’ and the spatiotemporal relations between bodies and environments to alternative, non-causal, theoretical frameworks which also politicize these relationships.

Here, there is clear potential for engagements with broader intra- and interdisciplinary work to make an important contribution to obesity/fatness research. Historically, a strong resonance can be found with early critiques of geographies of mental health which, as Dean (1984: 292) argued, were ‘rooted in the school of human ecology which controlled not only the methods of analysis…but the explanations that could be entertained’ (see also, Smith, 1978). Here, a shift away from spatial ecology approaches led to second- and third-wave geographies of mental health which enabled more theoretically sophisticated, ethically sensitive and methodologically rich research (Wolch and Philo, 2000). ¹³

There are also important connections to be made with geographical work informed by urban political ecology (see Heynen et al., 2006), which Guthman (2011: 9) argues can be useful for deepening an understanding of obesity as an ‘ecological condition’ (see also Heynen, 2006; Marvin and Medd, 2006). She states that such an approach:

demands that we pay attention to the broader political-economic and cultural context in which individual decision affecting ecologies – even internal, bodily ecologies – are made. It also requires we pay attention to the role of corporate behaviour, state regulation, and the political economy more generally…[and] to how knowledge of obesity as a biological condition is constructed and interpreted. (Guthman, 2011: 9)

Elsewhere, in interdisciplinary Fat Studies scholarship, significant work is being developed which draws on feminist and queer theory to question the evolutionary narratives through which obesity is constructed as a ‘modern problem’ (see White, 2012a, 2012b). There is also an important body of geographical work on urban health politics (Brown, 2009; Craddock, 1999). Combined, these different but related theoretical approaches allow for a critical, geographical and politicized interrogation of ideas concerning what constitutes the ‘natural’ in relation to bodies, environments and body-environment relations whereby:

the natural is not the inert, passive, unchanging element against which culture elaborates itself but the matter of the cultural, that which enables and actively facilitates cultural variation and change, indeed that which ensures that the cultural, including its subject-agents, are never self-identical, that they differ from themselves and necessarily change over time…[T]he other, culture, providing the latest torsions, vectors, and forces in the operations of an ever-changing, temporally sensitive nature. (Grosz, 2005: 47, original emphasis)