For food space: theorizing alternative food networks beyond alterity

In response to calls by scholars to deepen theoretical engagement in research on Alternative Food Networks (AFNs), in this article we critically discuss and assess major theoretical approaches deployed in the study of AFNs. After highlighting the strengths and limitations of each theoretical approach, we provide an alternative framework – which we refer to as the Geographical Political Ecology of Food Systems – that integrates the contributions that have emerged in the study of the alternative geographies of food with an understanding of capitalist processes in the food system. We do this by bringing together literature on the political ecology of food systems and multiple spatialities, including Doreen Massey's understanding of space as a heterogeneous multiplicity and Eric Sheppard's conceptualization of sociospatial positionality. We utilize research on agrarian change and AFNs in Eastern Europe to elaborate this approach. We argue that this new perspective helps navigate tensions in AFN scholarship, and provides new avenues for research and action. We focus particularly on the ability of AFNs to provide a sustainable livelihood for participating farmers, thus far a neglected topic in AFN research in Europe.


Introduction
The local food sector in the United States is burgeoning; in 2015 alone, locally produced and sold food provided US$8.7 billion in revenue (USDA 2016). Not only have the number of producers engaged in local food systems been increasing, but between 2007 and 2012, farms using direct-marketing channels were also more likely to stay in business than farms that were not (Low et al. 2015). Direct-marketing channels like farmers' markets experienced impressive growth during the same period. Since 2006, the number of farmers' markets in the United States has increased by 180 percent (Low et al. 2015). This growing interest and support for local food systems and direct-marketing channels has not been confined to the United States; consumers worldwide are seeking to create or maintain connections with farmers even as the globalization of the food system continues (Abrahams 2007;Kneafsey et al. 2013;Si et al. 2015;Wilson 2016).
Together, these phenomena have been generating "alternative geographies of food" (Whatmore and Thorne 1997), which are distinctive from conventional, globalizing, and industrial food spaces (Goodman and Goodman 2007). Making up a significant component of these geographies are Alternative Food Networks (AFNs), socially embedded networks that bring together farmers and consumers, who are committed to shopping more ethically (Sage 2003;Whatmore and Clark 2008). The uniqueness of AFNs stems from the types of trusting and embedded relationships that they appear to cultivate (Holloway and Kneafsey 2000;Sage 2007). Rather than simply searching for the lowest prices, consumers in AFNs seek locally-grown, organic, or fresh food. Similarly, producers value AFNs because they provide an alternative distribution channel, which may allow them to benefit economically by bypassing intermediaries .
While initially scholars hopefully assessed the emergence of AFNs, the assertion that they differ substantially from the broader food system has recently been challenged (Goodman and Goodman 2007). In the sphere of consumption, AFNs may reinforce neoliberal subjectivities, and exclusionary processes (Bradley and Herrera 2016;Dupuis and Goodman 2005;Guthman 2008). In the realm of production, farmers may still have to participate in conventional supply chains to make ends meet (Ilbery and Maye 2005), and recent research shows that competition may be eroding profits in AFNs (Blumberg 2018;Galt 2013a). In addition, research examining the failure of farmers' markets in the US has revealed how difficult it is to sustain AFNs (Stephenson et al. 2008). Therefore, it is not surprising that AFNs have been the subject of considerable academic debate, with critics of AFNs outlining their weaknesses and limitations, and supporters promoting their strengths (Goodman 2004;Van Der Ploeg and Renting 2004).
What is often overlooked in debates about the viability of AFNs are underlying theoretical tensions that shape diverging understandings of the content and origin of alterity. For example, political economists argue that structural processes undermine the transformative potential of AFNs (Goodman 2004) and impede farmers' abilities to make sustainable livelihoods through AFNs (Galt 2013a), while scholars focusing on cultural aspects of AFNs critique political economy approaches , and stress instead the significance of Journal of Political Ecology Vol. 27, 2020 3 values and institutions in both instigating and maintaining AFNs (Van Der Ploeg and Renting 2004). From a political economy perspective, true alternatives would counter capitalism and its tendencies. In contrast, other approaches conceptualize alterity as being manifested through an array of different values, practices and social arrangements, regardless of whether they overtly challenge capital. These theoretical disagreements have a particular scholarly lineage: the embrace by scholars of multiple theoretical perspectives to highlight and understand alterity in agri-food systems came at a time when political economy approaches were broadly criticized (Whatmore and Thorne 1997). Rather than privileging one theoretical perspective over another, our first goal in this article is to critically assess major theoretical approaches deployed in the study of AFNs. We do not provide a comprehensive overview of each theoretical perspective, which would be beyond the scope of this article, but we analyze how the different theoretical approaches have been used to study and explain AFNs. We then propose an alternative framework, a Geographical Political Ecology of Food Systems, which integrates the contributions that have emerged in the study of the alternative geographies of food with an understanding of capitalist processes in the food system. Our proposed framework brings together the following areas of research and concepts: the political ecology of food systems (Galt 2013b); multiple spatialities (Leitner et al. 2008); Doreen Massey's (2005) understanding of space as a heterogeneous multiplicity as outlined in her book, For Space (2005), from which we drew inspiration for this article's title; and Eric Sheppard's (2002) conceptualization of sociospatial positionality. The proposed framework allows for the integration of political economy approaches into AFN research, but in a manner that addresses the critiques that they have faced. We utilize empirical research on agrarian change and AFNs in Eastern Europe to elaborate this approach. We argue that this new perspective helps negotiate tensions in AFN research and practice, and provides novel avenues for research and action. We focus particularly on the ability of AFNs to provide a sustainable livelihood for participating farmers, thus far a neglected topic in AFN research in Europe. We conclude by using our framework to explain how a progressive sense of region could be used to harness AFNs for sustainable development.

Alternative geographies of food
The concept of "alternative geographies of food" (Whatmore and Thorne 1997) emerged in the late 1990s, in tandem with increasing scholarly attention to cultures of consumption and the agency of nature. In agri-food studies, this concept marked a turn away from the dominant theoretical perspective based in critical political economy and its intellectual focus, the globalized capitalist agri-food economy (see Buttel 2001). While critical political economy approaches had produced insightful analyses of growing corporate dominance in the food system, Whatmore and Thorne (1997) argued that they also reproduced a certain spatial imaginary, a totalizing account of globalization. Their argument mirrored feminist interventions that challenged the scholarly reproduction of "capitalonormativity", a representation marginalizing or ignoring multiple noncapitalist economic practices (Gibson-Graham 1996). These critics of capitalonormativity in agri-food systems research further argued that an exclusive focus on global capitalism and conventional food chains deflected attention from already existing alternatives, which could be recognized, strengthened, and sustained (Whatmore and Thorne 1997). Furthermore, because discourses are themselves productive of the worlds they seek to represent, scholarship that represented food systems as globalized and exclusively capitalist helped produce a world in which seeing and supporting alternative food systems became more difficult.
In light of broader critiques of Marxist approaches that sought to formulate less restrictive accounts of the agency of nature (Bakker and Bridge 2006), consumers (Goodman and Dupuis 2002) and peasants (Edelman 1999), many scholars in agri-food studies turned to other theoretical frameworks to understand what came to be known as the "alternative geographies of food" (Whatmore and Thorne 1997). This research trajectory provided alternative theoretical approaches to political economy that had the goal of highlighting, analyzing, and even supporting agri-food systems that in some way provided an "alternative" to dominant globalized agroindustrial food systems (Whatmore and Thorne 1997). Two aspects of alterity are therefore at stake: first is the explicit turn to study practices and networks alternative to globalized, conventional agri-food systems; second, and more implicitly, is the turn away from agrarian political economy and other Marxist approaches. In the

Journal of Political Ecology
Vol. 27, 2020 4 following sections, we critically examine this research trajectory, pointing out its contributions but also how political economy approaches remain relevant in the study of AFNs.

Farming differently
Although the turn towards the alternative geographies of food involved a shift to consider new theoretical perspectives in agri-food studies, the existing literature in peasant studies had already provided empirical evidence that demonstrated the limits of capitalist logic in agri-food systems. Drawing on the work of Alexander Chayanov (1986Chayanov ( , 1991, this research identified the existence of other logics and values driving economic development in agriculture (Alavi and Shanin 1988;Shanin 2009). Theoretically, this trajectory was taken up by the Wageningen school of rural sociology, which developed the concept of "farming styles" (Van Der Ploeg 2008;Van Der Ploeg et al. 2000;Van Der Ploeg and Renting 2004). Starting at the scale of the farming household, the analysis focuses on the economic, cultural, and agroecological practices that form different farming styles.
AFNs have emerged as an area of focus for research after the crisis in corporate and entrepreneurial farming styles in the late 20th century in Western Europe (Van Der Ploeg 2008). Industrialization had led to increased production, but it had also produced a price squeeze for farmers: while prices for farmgate outputs declined (despite temporary spikes) in real terms, the prices for inputs increased. Faced with this dilemma, Van Der Ploeg argues that farmers have a choice to either deactivate (cease production), increase industrialization, or adopt an alternative farming style. One such alternative, which Van Der Ploeg calls "farming economically", (or the "peasant mode") had grown considerably in Western Europe following the crisis (Van Der Ploeg and Renting 2004). Farming economically involves cutting back on costs (including labor costs) and reducing capitalization, but also adopting such practices as organic farming, agro-tourism, heritage foods, biodiversity and landscape preservation. These practices work to remold social and material worlds through a renewed reliance on social, cultural, and ecological capital (Van Der Ploeg and Renting 2004). AFNs are critical components of this farming style because they are largely responsible for providing farmers with outlets to consumers.
In addition to the scholars working within the Wageningen school, researchers on AFNs have extensively documented the significance of farmer values, knowledge and practices in AFNs (Andreatta 2000;Best 2008;Charatsari et al. 2019;Darnhofer et al. 2005;Zagata 2010). While qualitatively analyzing alternative farming styles yields valuable descriptive information on the livelihoods of producers in AFNs, it does not help explain what has caused the crisis of the corporate farming and entrepreneurial farming styles, nor does it provide a framework to analyze more systemic processes, such as the conventionalization of the organic farming sector (Best 2008;Buck et al. 1997). In addition, as feminist agrarian political economists have argued, the farming household cannot be assumed to have homogeneous values and any changes in farming practices may result in the expansion of gendered non-commodified labor (Friedmann 1978;Ramamurthy 2000;Razavi 2009).
A detailed study on Italian Solidarity-based Purchase Groups, a type of AFN similar to community supported agriculture and box schemes, demonstrates the strengths and limitations of adopting an approach that draws upon farming styles concepts (Brunori et al. 2011). This study details how the skills, assets, organizational structures and communication practices utilized by farmers in these purchase groups differ radically from those present in conventional farming (Brunori et al. 2011). Communicating directly with consumers, a possibility which is unusual for farmers in conventional food networks, can be time-consuming and intensive, even if it brings rewards. These kinds of purchase groups are growing in Italy, which is causing farmers to feel pressured to fulfill the increasing demand. Indeed, Brunori et al. (2011) found that this tendency is challenging the whole purchase group system and it may force these groups to compromise their standards. From this case study and others, it is clear that capitalist dynamics are influencing AFNs (Galt 2013a). While farming styles research provides a template that highlights the significant heterogeneity of contemporary agriculture, it falls short in explaining the impacts and influences of capitalist dynamics on efforts to follow a peasant logic in farming.

Values, practices, institutions
The focus on farmers, their values, and livelihoods has been marginal with respect to the amount of research conducted on consumers in AFNs (Alkon and McCullen 2011;Barta 2017;Dixon and Richards 2016;Freedman et al. 2016;Slocum 2007;Thilmany et al. 2008). Consumer perspectives, values, and practices, as well as the possibilities and limitations of consumer agency, have been a significant focus in agri-food studies following the cultural turn (Barta 2017;Buttel 2001). In addition to identifying the demographics of consumers who participate in AFNs (Freedman et al. 2016), researchers have also applied different theoretical perspectives to analyze why certain consumers are excluded from spaces fostered by AFNs (Alkon and McCullen 2011;Slocum 2007). Less often have researchers connected consumer and producer values and practices across the supply chain forged by AFNs (see Kirwan 2004 for an exception).
One of the major approaches to analyze the connections between consumers and producers in AFNs is convention theory (Goodman et al. 2012;Ponte 2016;Wills and Arundel 2017). Convention theory is a framework that emerged in part to address the fact that political economy left "little theoretical space to discern much deviation from the precepts of 'capitalist ordering' (either on the part of producers or consumers)" (Morgan et al. 2006: 17). Originally introduced in France and further developed by Storper and Salais (1997), convention theory attempts to bring to light the range of values, practices, and institutions that form the basis of economic action. Globally, economies are diverse because no production system is devoid of conventions, the "humanly constructed orders of routines, cognitive frameworks, institutions, practices, and objects" (Storper and Salais 1997: 12). Conventions are not simply the values that structure production-consumption networks, they also reflect institutional contexts, social and cultural constructions of quality, and even different understandings of profitability (Storper and Salais 1997). Murdoch and colleagues (2000) are largely responsible for applying convention theory to the study of AFNs. They define AFNs as new associational networks that engender different conventions of quality and value. In this view, the recent emergence and growth of AFNs reflects growing interest by producers and consumers in conventions unavailable in the conventional food system. In contrast to the latter's standardized products, AFNs function by communicating knowledge of the production process, including information about place of origin, and by communicating a distinguishable quality characteristic of the product. The translation of these values is what engenders success in starting, building, and maintaining the supply chain or network. Convention theory assumes that the AFN will function as long as conventions of quality and value meet both consumer needs and farmer producing capabilities.
Building upon these insights, scholars have undertaken work in identification, description, and analysis of AFNs, particularly in the Western Europe (Corsi et al. 2018;Renting et al. 2003;Roep and Wiskerke 2006;Venn et al. 2006). For example, Roep and Wiskerke (2006) highlight the importance of establishing forms of local embeddedness by cultivating a supportive institutional environment and by initiating strategic alliances between chain partners. These are components of alternative food production systems that may help explain why AFNs function well in some places and not in others, where institutional environments and possibilities for strategic alliances are weaker.
Convention theory provides important insights into the multitude of informal practices, values, and institutional contexts that form the bases for economic action. By unveiling these, scholars can help distinguish between the various conventions that help AFNs thrive and those under which they flounder. For example, defining "quality" can be a contested process, involving negotiations between different actors in the supply chain . For example, quality governance through labeling and standards (with the specific purpose of capturing added-value) has had the adverse effect of marginalizing small-scale producers (Neilson 2007).
In studies using convention theory, the underlying (if under-examined) assumption is that AFNs are founded upon a farmer's ability to capture extra value added through alternative distribution channels . In applications of convention theory to food studies, aspects of the economic system that impact all worlds of production (albeit differently) have not garnered much analytical attention, even though there is empirical evidence that processes like competition traverse these different "worlds" or sets of conventions (see Guthman 2004). Just like conventional producers, organic farmers compete with each other and feel pressured

Journal of Political Ecology
Vol. 27, 2020 6 to maximize profits (Guthman 2004). Moreover, convention theory does not provide a framework to explain the exclusion of consumers who practice the conventions of AFNs, but do not have the money to fully participate in this food system. More generally, a focus exclusively on the AFN itself risks ignoring those who have already been excluded and the reasons for their exclusion (see Zitcer 2015).

Doing economies differently
Rather than examining the institutional and value-laden contexts in which economic transactions occur as a source of alterity, other scholars have focused on the alterity of economic transactions that sustain AFNs (Connelly et al. 2011;Wilson 2013;Wilson 2014). Local food systems, such as AFNs, may function as part of a social economy because they foster local circulations of economic value and they help address social needs, such as the desire for consumers to obtain fresh produce and to have a connection with farmers . Similarly, they may exhibit qualities of a moral economy, in which economic transactions are guided by values of social justice, sustainability, or fairness (Connelly et al. 2011).
While the idea of a moral economy is often used to convey the sense that economic relations are guided by moral positions, the concept has a richer and more complex history. Originally articulated by E.P. Thompson, it was developed in agrarian studies by James Scott (1976) to explain the roots of resistance and rebellion in transforming rural societies. Scott argued that precapitalist peasant communities operated following a subsistence ethic, whereby elites and the state guaranteed support for the minimum level of subsistence to which everyone had a right. When these guarantees were ignored, political mobilizations among peasants occurred. In other words, it was not exploitation that caused rebellion, but an abrogation of the moral economy that tacitly existed between peasants, the elite, and the state. In Scott's case study, the existing moral economy was challenged by a market-based capitalist and colonial economy.
In more recent scholarship on AFNs, scholars have often paired the concept of a moral economy with characteristics that underpin economic transactions in AFNs. In one study of AFNs in Manchester, Psarikidou and Szerszynski argued that a moral economy was at work through "relations of solidarity and justice with proximate and distant others, regard for land and for the global environment, concern for social inclusion, interest in the well-being of the disadvantaged, and the reskilling of everyday life" (2012: 36). Multiple studies exist that link AFNs to moral economies, but most of this research problematically assumes that AFNs are insulated from capitalist dynamics (see Leiper and Clarke-Sather 2017). Research on farmer livelihoods in AFNs has shown the opposite, that even when elements of a moral economy infuse economic transactions between farmers and producers, farmer livelihoods are not protected from the effects of competition (Galt 2013a). Studies on the moral economy that lack Scott's detailed examination of the tensions between the capitalist and moral economies fail to adequately analyze farmer livelihoods in AFNs.
The diverse economies framework offers a different way of conceptualizing how alterity arises from the economic activity of AFNs (Gibson-Graham 1996). This approach is based on a poststructural perspective on class, which understands classes not as fixed categories or defined groups with certain characteristics, but as processes that appropriate and distribute surplus labor. Although capitalism is often considered a dominant and omnipresent mode of production, Gibson-Graham argue that capitalocentrism has foreclosed the possibility of seeing different economic practices. Indeed, according to Gibson-Graham, capitalist class processes are one of many diverse economic practices that structure the economy, but they are far from being the dominant practices. Instead, diverse economies are composed of: market, alternative market, and nonmarket transactions; wage, alternative paid, and unpaid labor; and capitalist, alternative capitalist, and non-capitalist enterprises. The concept of "alternative" is linked to transactions, forms of labor, and types of enterprises that distinguish themselves from the wage labor produced in capitalist firms for private accumulation. Gibson-Graham's representation of a diverse economy is not anti-capitalist, nor is it intended to displace capitalist firms (as is evident from its inclusion of green capitalist and socially responsible firms). Rather, Gibson-Graham have used action research methods to first help communities grasp the importance of the diverse economic practices that make their community, and to then collaboratively develop viable non-capitalist community economies that generate and distribute surplus locally.
Enacting diverse economies in food systems involves destabilizing what constitutes the capitalist food economy and rereading the food economy for difference, actions that can pave the way for alternative, noncapitalist economic becomings. Although Gibson-Graham's action research methods appear to be a substantial part of their theoretical framework (as part of enacting diverse economies), scholars of AFNs who draw on Gibson-Graham mostly draw upon their theory of diverse economies to examine possibilities for enacting diverse food economies (Holloway et al. 2007). For example, in their study of collective purchasing, Little et al. (2010) analyze the "'becomingness' of ethical purchasing" (2010: 1797), for its ability to shed light on why consumers get involved in AFNs (i.e., to enact another food economy).
Gibson-Graham's vision has been subject to various critical assessments by other scholars (Aguilar 2005;Glassman 2003;Jancius 2006;Laurie 2005), but what is most problematic about Gibson-Graham's approach is the theorization of economic difference. They categorize economic practices in discrete spheres according to surplus allocation, with definite boundaries separating capitalist and non-capitalist, market and non-market spheres. This categorization creates divisions between practices that are actually interrelated. For example, food provision is a task that combines capitalist and non-capitalist economic practices; food can be bought as a commodity, and that same food can be processed at home and given away for free to family members and friends. Spikes in food prices may also have impacts on non-capitalist food practices, by forcing greater selfsufficiency or reliance on food aid. As demonstrated by these examples, non-capitalist and capitalist spheres are interrelated in complex ways.
While Gibson-Graham's framework provides an explanation for why AFNs have appeared, and how they are sustained, it does not account for the dynamics affecting producer livelihoods. Their critique of capitalocentrism runs the danger of sidelining an understanding of capitalist dynamics, jeopardizing experiments in alternative economies that themselves are shaped in complex ways by capitalism. By bracketing off what constitutes non-capitalism, a representational strategy that seeks to make alternatives more visible, diverse economies frameworks have not been able to account for capitalist encroachment on AFNs. Meanwhile, such encroachment affects farmer livelihoods and impacts farmers' ability to sustain AFNs (see Andrée et al. 2010).

Posthumanist approaches: materiality, embodiment, networks
A disproportionate focus on human perspectives has led to the neglect of the agency of non-human actants in AFNs, as proponents of post-humanist approaches, such as Actor Network Theory (ANT) point out (Whatmore and Thorne 1997). According to these approaches, AFNs are alternative because of the way they reconfigure networks of human and non-human actants. For example, Whatmore and Thorne (1997) used Bruno Latour's notions of hybrid networks and John Law's conceptualization of "modes of ordering" to counter the orthodox notion that globalization is an expansive and homogenizing colonization of space: conventional food networks extend across space but do not colonize all spaces, and the networks are themselves ridden with instabilities. ANT provides the tools to conceptualize space as constituted by a multitude of networks and assemblages given shape by various actors. The effects of these networks are not pre-given but are relational achievements of all the actors involved in any given network.
ANT thus provides a useful ontology to challenge homogeneous accounts of the globalization of agrofood systems, to make visible the range of food networks operating according to alternative modes of ordering, and to highlight the fact that any network is an effect of the weaving together of humans and non-humans. In addition, ANT provides conceptual tools and a relational understanding of power that does not privilege either productive or consumptive spheres (Hopkinson 2017;Lockie and Kitto 2000). By allowing for a broadened understanding of agency, the realm of politics is expanded, without presupposing which actors (human and nonhuman) assert political agency.
By focusing on how networks operate, ANT also illuminates the effort required to maintain network relationships. Attention to the mechanisms of connection or disconnection along networks is therefore a central focus of research on AFNs. For example, Le Velly and Dufeu (2016) use ANT to show how market agencements, hybrid assemblages of human and non-human actors, are necessary to make a fish CSA function in a coastal city of France. This assemblage does not involve a simple and unidirectional relationship between Journal of Political Ecology Vol. 27, 2020 8 humans and non-humans. French food safety regulations influence the organization of the assemblage, but these regulations are also a product of the materiality of the fish itself. ANT provides the theoretical tools to understand how AFNs are sustained by enrolling multiple actors in sometimes complex assemblages. For example, ANT can provide a description of how certain actors (consumers) enroll actants (money) to purchase goods through AFNs. However, it does not help explain why some people have access to money to participate in AFNs and why others do not. It does not explain why certain farmers may have the capacity to enroll actants into networks to form AFNs, while others may struggle.
ANT is not the only posthumanist theory that has been applied to the study of AFNs. Building upon feminist theorists of the body (Grosz 1994;Probyn 2000), who urged scholars to analyze both the discursive and the material influences in the making of the body, geographers have increasingly highlighted the importance of the "visceral" in food research (Goodman 2016). According to Hayes-Conroy: "to speak of visceral (re)actions is to simultaneously refer to the relationality of affect, the textuality of emotion, and the experiential qualities of feeling" (Hayes-Conroy 2009: 34). In other words, AFNs engender alterity because they cultivate visceral reactions that are not produced when consuming food procured through conventional food networks.
The visceral reactions that consumers experience by eating fresh and local foods, foods they highly desire, have to do with the biology of eating and taste, which many supporters of AFNs assume to be homogeneous. However, taste is highly variable, and even for individuals it cannot be assumed to be a static property of a living body. Because taste is a complex interaction between the social and biological world, some tastes can be learned and unlearned. Although studies of AFNs have highlighted the importance of the visceral in attracting and retaining consumers, Hayes-Conroy (2009) argues that for consumers who do not participate in alternative food initiatives, a visceral perspective can also be used to understand how taste is a barrier to participating in AFNs. For example, the taste of fast food may evoke fond memories of childhood, whereas the taste of kale may be unfamiliar. Therefore, attention to the materiality of the body in its encounters with food may explain who is attracted to or excluded from AFNs (Alkon and McCullen 2011; Slocum 2007).
While visceral approaches highlight the importance of emotion, affect, and the feeling body, these theoretical perspectives do not by themselves help explain how and why some bodies develop taste for foods likely to be sold through AFNs, while others do not. In the US, structural and political economic processes have led to the creation of a food system that produces an abundance of cheap sugar and sodium-laden foods that are disproportionately consumed by the poor (Drewnowski and Specter 2004). From a farmer's perspective, visceral approaches may explain why people may be attracted to farming, an occupation that involves working outdoors. However, the visceral aspects of agricultural employment or the visceral pleasures of selling healthy produce directly to consumers can only explain one dimension of farmer livelihoods in AFNs. In light of the systemic issues that produce inequality in the food system and the limitations of visceral approaches, Goodman (2016) has argued that visceral approaches should engage more directly with concerns around food justice and austerity.

Beyond alterity: a geographical political ecology of food systems
The theoretical perspectives discussed above provide unique insights into what constitutes alterity in AFNs. For Whatmore and Thorne (1997), alterity is equated with a "mode of ordering", a certain way of establishing relationships that is not primarily based on making the most money or farming the cheapest product. Convention theory highlights how certain synergies are necessary for different worlds of production to function: what shapes the alterity of AFNs are not only the connections between consumer and producers, but the different institutions, norms, and informal practices that enable these connections. On the farm, Van Der Ploeg (2008) categorizes alterity in the form of those farming practices that adopt a peasant logic. Scholars using the diverse economies approach to study AFNs point to the alterity of non-capitalist processes. All of the approaches presented above provide insight on how AFNs are sustained, but each by itself has significant weaknesses, especially when accounting for the dynamics affecting farmer livelihoods in AFNs. In what follows we present an alternative approach, which we refer to as a Geographical Political Ecology of Food Systems. This approach integrates contributions that have emerged from the scholarship on the alternative geographies of food with an understanding of capitalist processes in the food system.
Agrarian political economy provides conceptual tools to analyze the effects of participating in AFNs on farmer livelihoods (Galt 2013a), but it has also faced criticism with respect to its lack of consideration for local ecological factors and farmer agency Galt (2013b). While much of agrarian political economy has focused on class analysis (Bernstein 2010), researchers working in dialogue with this tradition, including feminist agrarian political economists, have provided useful insights on the significance of power relations, gender and identity in farmer livelihoods and households (Berry 1993). Galt (2013b) argues that such contributions from scholarship in agrarian political economy could be harnessed and integrated into a "political ecology of agrifood systems." Political ecologists view ecological systems as power-laden, and embedded in broader political, economic, and social systems (Robbins 2012). Emerging from studies of environmental degradation, conservation, and land uses, often in agri-food contexts, political ecology has directed particular attention to how the production or consumption of environmental resources -as well as access to or participation in various regimes of land use, agricultural relationships, or food production processes -may produce or reinforce existing inequalities (Robbins 2012). Political ecologists have been critical of apolitical or asocial research on environmental issues, especially including those related to the food system. Centrally, political ecology research on livelihoods has highlighted how everyday practices and relations produce material changes in landscapes (Batterbury 2001;McCusker and Carr 2006). Political ecology approaches span from interpretive research on values, to analyses of socio-ecological relations and political economies (Galt 2013b).
Although Galt (2013b) presents a coherent framework for a political ecology of food systems, we argue that this research agenda could be strengthened with the integration of multiple spatial dimensions (see Leitner et al. 2008). Thus far, the dominant focus of scholarship on AFNs has been on network formations (Murdoch 2000;Renting et al. 2003;Venn et al. 2006), scale (Born and Purcell 2006;Guptill and Wilkins 2002;Selfa and Qazi 2005) and place (Feagan 2007). Less often have sociospatial positionality (Sheppard 2002) or space, as conceptualized by Massey (2005), been integrated into AFN research. Geographers have developed various conceptualizations of space that stress that space should not be understood as a bounded entity or a fixed backdrop in front of which events occur (Crang and Thrift 2000). Rather, space is a product of material and social relations, which themselves span and remake space (Massey 1994;Nagar et al. 2002). Building upon this understanding, Massey (2005) offered three propositions to conceptualize space. First, space is constituted through relations ranging from the global to the local. For example, globalization does not flow to enclosed peripheries, as is often portrayed in hegemonic geographical imaginations. Instead, "global" flows mingle with more local trajectories, stories, and practices. Thus, places constitute and are constituted by power-geometries (Massey 2005). Second, space is composed of a heterogeneous multiplicity of coexisting trajectories, or flows made by humans and non-humans. Space is a "field of multiple actors, trajectories, stories with their own energies-which may mingle in harmony, collide, even annihilate each other" (Massey 2007: 22). At the farm, multiple trajectories across space come together to produce a harvest of food, from the soil that is the product of geologic time to the human worker whose presence is brief, but influential. Third, space is always in process, and therefore, any account of space is by definition partial.
Massey's conception of space as heterogeneous multiplicity allows for an understanding of how capital structures space and weaves together diverse places by redirecting and reworking existing trajectories through its development. Geographers have demonstrated how capital plays a role in the making of uneven development (Bair et al. 2013;Castree 1999;Harvey 1989). The contradictory characteristics of capital's power-geometries riddle space with processes of growth and contraction, revaluation and devaluation, and integration and disarticulation with commodity chains (Bair and Werner 2011;Hough 2011;Ramamurthy 2011;Werner 2016). The remaking of space is critical to capitalist restructuring, but space, as a multiplicity that is always in process, cannot be reduced to capital's logic and the social domination it produces. Understanding capitalism as a powerful but not omnipresent force in space helps account for why and how AFNs may be forced to grow or conventionalize.
Massey's conception of space as a heterogeneous multiplicity provides a useful starting point to frame the emergence of AFNs. While AFNs have prospered in some metropolitan areas, research in the US has demonstrated regional inequality in farmer access to AFNs (Jarosz 2008). For example, the trajectories that constitute the landscapes of industrial agriculture in the US are power-laden and globalizing, making it difficult for farmers to start AFNs . In rural areas affected by depopulation, maintaining a consumer

Journal of Political Ecology
Vol. 27, 2020 10 base for local food may be difficult (Hinrichs 2003). As a result, farms that participate in AFNs are clustered around metropolitan areas in the US (Jarosz 2008;Low and Vogel 2011). However, within these areas AFNs do not serve all consumers equitably: farmers' markets in low-income urban areas of the US may face problems related to profitability (Fang et al. 2013). In addition, the meanings and values associated with AFNs may differ across space and therefore affect consumer or farmer capacity to form AFNs (Forssell and Lankoski 2015;Selfa and Qazi 2005). Therefore, both from the perspective of consumption and production, places have diverging capacities to foster viable and sustainable AFNs, and there is complexity in how these capacities are inter-related. California, for instance, hosts a thriving local food movement in part because of its "unsurpassed diversity in growing conditions such that an enormous number of crops can be produced year-round within the state's borders" (Guthman 2008(Guthman : 1180. The heterogeneous multiplicity of trajectories that span space and constitute California produce these conditions, and therefore make "eating local a task of little sacrifice" (Guthman 2008(Guthman : 1180. From a farmers' perspective, a spinach grower on the central coast of California can claim the "local" and gain the support of spatially proximate and wealthy consumers. While the spinach itself may be sold locally, it may also be shipped to locations all over the US where it then competes with other "local" spinach. For farmers, the capacity to form AFNs depends on access to places with sufficiently wealthy and interested consumers, but also to limited competition from other farmers . Building on feminist scholarship, Eric Sheppard has developed the concept of sociospatial positionality to understand the "shifting, asymmetric, and path-dependent ways in which the futures of places depend on their interdependencies with other places" (2002: 308). The relations and trajectories constituting space are not free-flowing and random, but they may even exhibit path-dependent characteristics that affect development possibilities. As a result, the prospect that places have to build their own development trajectories does not depend exclusively on local actions: places are not equally endowed with the possibility to challenge existing power relations and realize alternative futures in the food system. The possibility of developing an AFN in one place depends on interdependencies with other places; for example, a farmer in a rural location might grow delicious produce, but this matters little if she cannot find customers willing to pay for it. Sociospatial positionality also shapes visceral reactions, as places with diverse local and global connections may afford its residents access to a variety of tastes.
In the following section we mobilize a geographical political ecology of food systems framework that utilizes an understanding of space as a heterogeneous multiplicity of trajectories, and of development as influenced by sociospatial positionality to explain the emergence and persistence of AFNs in Eastern Europe. Although this region includes a significant portion of the European Union's farms (Eurostat 2018), its food systems have yet to garner significant attention in political ecology (see Brawner 2015 for an exception). However, its residents' longstanding practices of subsistence food production, as well as old and new AFNs, have recently become the focus of scholarly attention (Smith and Jehlička 2013;Yotova 2018). Because this region continues to be affected by growing disparities and rural poverty (Žakevičiūtė 2019), political ecological perspectives could provide a framework for new rural development pathways.

The political ecology of food systems in Eastern Europe
In Eastern Europe, food systems have been transformed radically by the collapse of state-socialism, the implementation of shock therapy and neoliberalism in the 1990s, and the more recent expansion of the European Union (Caldwell 2009;Hann 2002;Jung et al. 2014). While the Soviet-backed regimes and their successors differed with respect to the implementation of agricultural collectivization and the nationalization of land and resources, and in the 1990s, decollectivization, privatization and liberalization, these processes have left lingering imprints on rural landscapes, land use and food systems (Creed 1998;Duijzings 2013;Hann 2003;Lieskovský et al. 2014;Schwartz 2005;Verdery 1996). Although the negative environmental impacts of Sovietstyle development were well-documented, most notably in the Chernobyl disaster (Phillips 2002), new problems have arisen and some persistent environmental issues have lingered since the collapse of the Soviet Union and its supported regimes in the former Warsaw Pact countries (Galbreath 2013;Pavlínek and Pickles 2002;Smith et al. 2000). The privatization of natural resources, from agricultural land to forests, was supposed

Journal of Political Ecology
Vol. 27, 2020 11 to lead to more efficient use of resources and sustainable development; political ecologists writing about the region have documented the opposite tendency (Dorondel 2016;Stahl 2012). In the food system, large-scale farms have threatened biodiversity (Sutcliffe et al. 2015), and the concentration of pollutants from these farms continues to wreak harm on the environment (Skorupski 2012). In the Eastern European states that are now members of the European Union, growing corporate concentration and consolidation throughout the food system has been fostered by a Common Agricultural Policy that was adopted with little consideration of the needs of the region's agricultural sector (Chaplin et al. 2004;Swain 2016), or consumer needs and preferences for healthy or nutritious food more broadly (Walls et al. 2016). Paradoxically, the application of the sophisticated EU regulatory framework in food safety has not enhanced consumer trust in the food system (Aistara 2015). Motivated by multiple social and environmental concerns, consumers and farmers are helping AFNs grow in Eastern Europe (Balázs et al. 2016, Benedek andBilewicz and Śpiewak 2018;Grivins and Tisenkopfs 2015;Mincyte 2012;Smeds 2015;Spilková and Perlín 2013;Syrovátková 2016;Syrovátková et al. 2015). This phenomenon has garnered much scholarly interest because of the novelty of some AFNs and also because the region has a strong history of self-provisioning and farmer-consumer linkages (Mincyte 2011;Varga 2018). The political ecology of land-use in the region helps explain why these older AFNs have persisted, and also why certain new AFNs, such as farmers' markets in Lithuania (Blumberg 2015) and the Czech Republic (Spilková and Perlín 2013), gained rapid popularity for farmers and consumers. In the Soviet era, most states pursued policies of agricultural collectivization, but out of necessity, states also allocated land to households for subsidiary agriculture in both urban and rural areas (Bellows 2004;Blumberg and Mincyte 2019a;Mincyte 2009). While Eastern European states followed diverging policy programs to decollectivize farms and privatize agricultural land starting in the 1990s, a general land use pattern emerged based on a bimodal farm structure with the plurality or majority of farms being small in scale and with fewer large-scale corporate or cooperative farms (Mathijs and Noev 2004). These small-scale farms continued the practice of supplying urban residents with food through markets and informal networks in the 1990s, and were ready participants when initiatives to create farmers' markets traversed the region after 2008 (Blumberg 2015).
New AFNs that resemble models utilized in Western Europe exist in a complex and intertwined manner with the more abundant number of AFNs that have stronger local histories based in self-provisioning, informal food networks, and public markets. For example, in Lithuania the recent popularity of new farmers' markets can be explained by the fact that the practice of shopping at public markets was already widespread and valued (Blumberg 2015). In other words, the consumer conventions of shopping at the public market, with all the positive and negative associations that entails, were transferred easily to new farmers' markets. Thinking of space as a heterogeneous multiplicity of trajectories provides a conceptual framework to analyze the emergence and persistence of diverse AFNs in Eastern Europe, many of which, as noted earlier, are rooted in the household provisioning practices that were an iconic part of the informal economy of the Soviet era. However, they were not independent or isolated from the formal state sector; indeed, they were intertwined with the economic relations that spanned the Soviet space economy. For example, farm workers used collective farm machinery on their own private plots, and public markets maintained by state authorities provided them with a place to sell their produce to urban residents. An understanding of the heterogeneity of space and the existence of multiple, intertwined and complex trajectories helps explain how even under the Soviet regime, multiple formal and informal food networks existed and flourished.
The wide variety of informal household economic practices that proved to be critical in sustaining livelihoods throughout the Soviet era gained new importance following the implementation of neoliberal marketization reforms in the 1990s (Smith et al. 2008;Staddon 2009;Smith and Stenning 2006). While AFNs helped guarantee survival and ensure procurement of valued food products, scholars of this time period have emphasized that these networks were mobilized as 'survival strategies', which were often labor-intensive and required the stressful cultivation of informal ties and networks. But even under the stress of neoliberal shock therapy, which led to an expansion of the informal economy, formal economic processes and practices coexisted and intermingled with informal economic practices. In her study of households in Moscow, Pavlovskaya (2004) found that access to informal networks considerably improved the economic position of households, in comparison to households without similar access. She also revealed how household economies relied on a gender-based division of labor, which disadvantaged women because of their disproportionate household responsibilities, such as in care work and food provisioning (Pavlovskaya 2004). In other words, differences like gender, race, class and ethnicity shape and are shaped by the making of space; for example, while spatial flows may disadvantage women, women are also less likely to be able to harness the power to reshape spatial trajectories to their advantage.
A debate has emerged on whether to consider household self-provisioning as a survival strategy utilized to cope with poverty (Seeth et al. 1998;Humphrey and Mandel 2002), or as a practice that is undertaken due to other social or cultural reasons (Czegledy 2002). Thinking of space as a multiplicity of trajectories, always in process and produced through power-geometries that privilege and advantage different people and places, provides a framework to understand how multiple values can coexist across space. Self-provisioning can be an economic necessity, and a means by which to cultivate social capital or express cultural values. For example, with the growth of the conventional retail food sector, subsistence food practices and other AFNs in the region have been revalued; consumer disillusionment with economic transition and the aftermath of the financial crisis have led to renewed appreciation of subsistence food production and direct ties with farmers (Blumberg 2015;Spilková and Perlín 2013).
Researchers on AFNs in the region are now saying that these older, subsistence-oriented and informal food practices and networks offer possible pathways for sustainable development (Smith and Jehlička 2013;Yotova 2018). What is interesting about this suggestion is that it contrasts with previous political efforts that viewed informal food networks and subsistence production as an impediment to progress and sustainable development in food systems in Eastern Europe (Mincyte 2011). The concept of sociospatial positionality helps explain this development in agri-food policy for Eastern Europe. Now heightened by the integration of some CEE states within the EU, uneven development within the region is being reproduced, while a persistent, but differentiated East-West disparity remains (Bohle and Greskovits 2006). For those states that gained EU membership, the process was not unproblematic. Central and Eastern European applicant states were treated less favorably than previously accepted applicants; as a result, the accession process was also marked by tension and an unequal power dynamic. While applicant states were subject to censure on a wide variety of topics, from human rights to agriculture, critics pointed out how the old EU member states themselves fell short of EU ideals (Clark and Jones 2011). Hierarchies of power and wealth were reinforced with the production of hierarchies of knowledge (Böröcz and Kovács 2001). In other words, just as much as EU accession involved adopting new policies, it also involved applying measures to assess achievements towards certain assumed goals or norms, which because they existed in the EU, were assumed to be universal.
The power-geometries and sociospatial positionality that defined the Eastern European countries as "supplicants" (Swain 2004: 200) during EU accession were apparent in the agricultural and food sector in at least three ways. First, the accession states had to agree to receive less funding from the Common Agricultural Policy (CAP) in comparison to the old EU member states, even though all farmers would be competing in the same market after accession in 2004. Second, the CAP was not modified to address the needs of Eastern European farmers, especially its numerous small-scale farmers who utilize AFNs. Accession states were also encouraged to use development funding to improve competitiveness over furthering multifunctionality (Râmniceanu and Ackrill 2007). Third, stringent food safety regulations also needed to be implemented, and little consideration was given for their potential adverse effects on local producers (Dunn 2003).
Although the rural economies of the accession states differed, in comparison with the old EU member states, they had economies that were more dependent on agriculture as a proportion of GDP and they had higher percentages of their populations engaged in agriculture as a livelihood. Researchers have documented how informal AFNs became popular for small-scale producers who were excluded from conventional supply chains due to the implementation of food safety regulations to which they could not afford to adhere (Blumberg and Mincyte 2019b;Dunn 2003;Mincyte 2011). For example, Harboe Knudsen's (2010) ethnographic work in a Lithuanian marketplace documents how increased food safety requirements actually meant that dairy sellers who met the standards were allowed to stay inside the market, whereas those who did not meet the standards just moved right outside the boundaries of the market and continued to sell their dairy products. Although not sanctioned by the authorities, the informal markets outside the market hall were able to attract customers and often offered lower prices. Recognizing the significance of their own past and present dairy consumption

Journal of Political Ecology
Vol. 27, 2020 13 practices, which did not meet the new requirements, consumers continued to purchase milk from the unsanctioned sellers outside the market (Harboe Knudsen 2010). At the same time, these AFNs were not considered to be a foundation for sustainable development, but rather an archaic practice only utilized by marginalized consumers and farmers (Mincyte 2011). This idea was part of a broader narrative that imagined that Eastern Europe had to "catch up" to Western Europe through transition policies, which would herald development in a linear manner. The catching up narrative is deeply problematic because it denies the existence of the heterogenous multiplicity of space and sociospatial positionality: treating nation-states as isolated units ignores their interconnectedness and how existing transnational flows help constitute economies in those states. For example, both before and after European Union integration, Eastern European states were more closely linked to eastern markets for exports and imports, such as those in Russia. Following the implementation of successive Russian embargoes on a variety of EU foodstuffs, some prices for farmers within the EU fell as produce flooded the market. In Lithuania, some largescale farmers responded by abandoning conventional markets and starting new AFNs. This had the impact of heightening competition with farmers who already marketed through AFNs (Blumberg 2018). Sociospatial positionality affects AFNs and farmer livelihoods in complex ways, in part because of crises caused by capitalist dynamics.
An understanding of space as a heterogeneous multiplicity helps shed light on how multiple processes affect AFNs and farmer livelihoods more specifically, including dynamics in the conventional food sector. AFNs are not a relic of the past, but they are part of the multiple modernities that constitute space. In other words, although they may resemble past practices, they are products of the present. The multiple trajectories that make modernities are imbued with power relations, as is evident by the fact that some trajectories of food have become marginalized while others have gained dominance. The sociospatial positionality of Eastern European states during the EU accession process helps explain why existing agricultural and rural policy was not tailored to the needs of Eastern European farmers. However, now most of these states are EU members and they have the power to shape future CAP policies. Nevertheless, even as full-fledged EU members, small and poorer EU member states face constraints in exercising full-fledged participation in negotiations over EU policies (Panke 2010). Because subsidies form a large share of income for farmers across the EU, CAP policies have the potential to improve or worsen farmer livelihoods in AFNs.
An analysis based on understanding space as a heterogeneous and power-laden multiplicity explains why rural development policies emanating from the West have been privileged in Eastern Europe in their effort to "catch up" to the West. Scholars have critiqued the catching up narrative and the concept of a linear, stageist transition (Burawoy and Verdery 1999), and they have provided evidence that alternative food production, procurement and marketing practices provide possible sustainable development pathways (Ančić et al. 2019;Blumberg and Mincyte 2019a;Blumberg 2018;Pungas 2019;Smith and Jehlička 2013;Spilková and Vágner 2018;Yotova 2018). For example, writing about permaculture gardens in Bulgaria, Brawner argues that because of its flexibility and its resemblance to traditional Bulgarian gardening practices, permaculture offers a contrast to "Western-centric notion of progress, creating oppositional frameworks for development" (2015: 441). However, this scholarship has yet to harness an alternative sociospatial conceptualization, which could form the basis of new agri-food policies that could better support AFNs in different regions.

Conclusion
The beginning of the 21st century has witnessed a growing desire by consumers, as well as scholars, to envision and create an alternative food system. Research on AFNs has been growing in tandem with their increasing popularity. In recognition of the need to deepen theoretical engagement in AFN research (Brunori et al. 2011), in this article we have reviewed major theoretical approaches deployed in the study of AFNs, and we have situated the emergence of AFN research in a turn to consider the alternative geographies of food. This intellectual movement involved critiquing capitalocentrism in agri-food research, and embracing other theoretical approaches besides agrarian political economy. What is clear is that theorizing alterity has done much to enable an understanding of what is new and different about AFNs and to highlight the values espoused by those involved in AFNs, but theories of alterity have fallen short on providing the tools to understand how actually existing AFNs succeed or struggle within a capitalist economy.
Recognizing the enduring contributions of agrarian political economy, as well as its limitations, we build upon Galt's (2013b) proposal for a "political ecology of food systems" by using a framework that integrates multiple spatialities (Leitner et al. 2008), including Massey's (2005) understanding of space as a heterogeneous multiplicity and Sheppard's (2002) conceptualization of sociospatial positionality. We utilize this framework to analyze the emergence and diversity of AFNs in Eastern Europe, primarily focusing on the impacts of EU accession and membership.
Specifically, we have argued that understanding AFNs involves shifting focus beyond their assumed alterity towards their sociospatial context. Massey's (2005) conception of space as a power-laden and heterogeneous multiplicity that is always in process helps conceive of AFNs as influenced by various flows and trajectories that traverse space, from conventional food chains to capital flows. In Eastern Europe, AFNs have been a survival strategy, enabled by the shock therapy of neoliberal transition in the 1990s, but they have also become elements of a "quiet sustainability", which has been overlooked in the policy that has been applied in Eastern Europe (Smith and Jehlička 2013). Conversely, food systems have been a central vehicle for European integration, from the Common Agricultural Policy to the transnational expansion of supermarket chains. But they have also been vehicles to foster exclusionary values, such as reactionary nationalism linking "pure" blood with land (Brüggemeier et al. 2005).
With the rise of the euro-skeptic far right in many European nation-states (see Matijevic and Boni 2019), an understanding of space as a heterogeneous multiplicity, of place as open and unbounded, and of sociospatial positionality as shaping development outcomes could help further a more progressive integration politics. Writing in the 1990s, Massey noted that there was a "recrudescence of some very problematical senses of place, from reactionary nationalisms, to competitive localisms, to introverted obsessions with 'heritage,'" and that therefore there was a need to "hold on to that notion of geographical difference, of uniqueness, even of rootedness if people want that, without being reactionary" (Massey 1994: 4). She suggests that what is needed is a progressive sense of place, a sense that places are not enclosed entities with stable essences; instead, they are always in process as meeting points for the multiple relations that span space. A progressive sense of place would be outward-looking and recognize the interconnections across space that make places. Building upon this notion, Dupuis and Goodman argue that "an inclusive and reflexive politics in place would understand local food systems not as local 'resistance' against a global capitalist 'logic' but as a mutually constitutive, imperfect, political process in which the local and the global make each other on an everyday basis" (Dupuis and Goodman 2005: 369).
Similarly, in Eastern Europe, European integration of the food system has often been represented as a supranational force that imposes itself on a resistant and sometimes reactionary "local." Conversely, in the West, this has often been taken up in a narrative that assumes Eastern European backwardness and the simultaneous need to embrace policies to "catch-up" to Western Europe (Dzenovska 2013). While power-geometries do shape the flows that traverse space, an understanding of space as an on-going and heterogeneous multiplicity disallows stageist narratives that position some places behind others in time-space. From the new farmers' markets to older forms of self-provisioning networks and public markets, the global, European, and local mutually constitute and make the practices that animate diverse AFNs in Eastern Europe. Rather than being opposed to the local, Europeanization is shaping the very dialogue that is producing scholarly connections across Eastern Europe that draw attention to commonalities and differences in food provisioning through AFNs. Cadieux (2016) has proposed that a progressive sense of regions would further political ecological research, and we suggest that this could be utilized to further research and action on AFNs in Eastern Europe. A progressive sense of region would recognize the interconnections that have shaped the region, and it could be used to create alternative development pathways for AFNs that are not built open Western models, but harness the region's existing practices to build sustainability in the food system.