Translocal practices and proximities in short quality food chains at the periphery: the case of North Swedish farmers

This paper examines the social and organizational innovation processes undertaken by small-scale producers engaged in short food supply chains in the North Swedish region of Västerbotten. The study uses the notion of proximity to empirically analyse and conceptually explore these phenomena. The paper illustrates the ‘new associationalism’ mobilized by producers in order to promote knowledge exchange and learning and highlights the role of translocal practices in sustaining this transition. The study found that open and trusted interactions with consumers are central to the development of ‘quasi-organic’ practices, and that producers belong to numerous motley associations of food professionals facilitating the creation of collective meanings about near-produced quality food. The paper contributes to the rapprochement between agri-food studies and human geography to understand the formation of local food systems from an evolutionary and relational perspective.


Introduction
This study proposes to examine the relational geography of 'new associationalism' (Marsden et al. 2002) taking place in the context of the development of short food supply chains (SFSC) in a peripheral rural region of Europe. SFSC are not fundamentally novel in principle as they correspond to how food was traded in pre-conventional agriculture era. Narratives of contemporary processes of SFSC formation essentially address how these initiatives attempt to distantiate themselves from the actors, values and practices of the conventional agri-food system. The latter entails that "[r]elearning how to be local" (Morgan and Murdoch 2000, p. 167) by recapturing pre-conventional practices (Blay-Palmer 2005) requires actors to navigate intricate relational routes. In this journey, external actors (including consumers), knowledge and institutions play a central role in introducing novel approaches and perspectives to the local know how (Ilbery et al. 2004;Clark 2005). Moreover, alternative food systems rely on hybrid practices between "local actors who belong to different worlds" (Torre et al. 2018). Hence, undertaking such a knowledge-intensive transition is intrinsically related to the constitution of new partnerships with other 'like-minded' food professionals, or as Morgan and Murdoch elegantly put it, "knowing-who becomes an important part of knowing-how" (2000, p. 168). On this basis, the overall set-up of this study was enlightened by Pieterse's (2015) commentary that, even though the observable ecologies of agriculture may appear to be local, the farming practices that shape them are fundamentally translocal.
It is now well-understood that the transition towards a postcarbon food and farming system (Hinrichs 2014) will necessitate a paradigm shift from industrial agriculture to more diversified agro-ecological systems (Rockström et al. 2017). In peripheral regions, which are often by nature not productive enough to sustain even more virtuous agricultural intensification, agroecology may be instrumental in proposing a viable alternative to keep small-scale farmers in business and more generally "to develop marginal rural resources that would otherwise remain untouched" (Persson 1983, p. 60). The recognition of the social and ecological qualities of food and their valorisation through short food supply chains reinstate the role of the small farm as a vector of community development (Ilbery et al. 2004). However, 1 3 the decline in the number of small farms in peripheral places combined with steady depopulation trends engendering "a loss of knowledge capital and know-how" (Torre and Wallet 2014, p. 666) constrains their ability to 'tap into' traditional agricultural knowledge networks, e.g., farmers' cooperatives, as a way to source and adopt new practices.
I argue here that the context of social marginalisation faced by small-scale farmers at the periphery is an extreme case of the contemporary pressures put on small-scale farming after decades of promoting productivist agriculture in the developed world. Nonetheless, it is still the most common 'style of farming' (van der Ploeg 2017) and, under those circumstances, understanding how small peripheral farms undertake the 'agroecology turn' underlines the role of agency in the emergence of new associations and the promotion of knowledge exchanges and innovative practices at the (geographical and institutional) fringe of the dominant agricultural regime.
This study uses the experiences of north Swedish farmers to explore and illustrate these relational processes using a critical case design (Flyvbjerg 2006). The notion of proximity is here used to investigate empirically and explore conceptually how translocal practices are shaped in the process of establishing SFSC. Proximity analysis underscores the determinants of interactions (Torre et al. 2018) and seeks to explain the "how" rather than the "why" of new associationalism (Lamine 2012). The application of the notion of proximity thus aims at bringing clarity to our theoretical understanding of new associationalism. The study aims to answer the following research questions: What actors appears to be central to peripheral farmers' knowledge mobilization strategy? What relational arrangements are instrumental in initiating and brokering these exchanges over time? To what extent can these emerging practices be characterized as translocal?

Short food supply chains
Short food supply chains (SFSC) often seek to integrate agroecological farming techniques, aiming to 'recapture' pre-conventional style of farming (Blay-Palmer 2005), with novel food marketing practices. SFSC can thus be characterised as both future-oriented and traditional in their set-up. In the agri-food literature, SFSC have been widely used as illustrations of the transition of the contemporary agri-food system towards enhanced resocialisation and respatialisation of food (Renting et al. 2003;Ilbery and Maye 2005;Sonnino and Marsden 2006;Jarosz 2008;Maye and Kirwan 2010;Kneafsey et al. 2013).
The main issue that the development of SFSC addresses is the widening gap between producers and consumers within the conventional food industry: geographical, contemplating the increased physical distance separating where food is produced from where it is consumed (Goodman and DuPuis 2002;Jarosz 2008;Chiffoleau 2009), cultural and cognitive, comforting a lack of awareness in and a sense of anonymity about food origins (Chiffoleau 2009;Dowler et al. 2009) and organizational, with an increased number of large, internationalised 'faceless' corporations organizing the food value chain .
As a response to this distanciation, SFSC are set to provide alternative ways of structuring the food system including the reduction of number of intermediaries (Watts et al. 2005), with ideally none (Kneafsey et al. 2013), the construction of 'thickened' social relations and close-knit communities involving both producers and consumers (Milestad et al. 2010a;Mundler and Laughrea 2016;Aggestam et al. 2017), the transmission of value-laden product information (Renting et al. 2003) enabling consumers to evaluate the true value of food (Kneafsey et al. 2013) and the adhesion to shared value systems sustained via personal trust, familiarity and reciprocity (Kneen 1993;Hinrichs 2000;Watts et al. 2005;Bos and Owen 2016;Thorsøe and Kjeldsen 2016).
This literature has tended to predominantly examine the frontend of SFSC, i.e., the social, economic, organizational and spatial characteristics constituting 'close' producer-consumer relations. Understanding how the adjective 'short' may, or may not, apply to the reconfiguration of the backend of SFSC, understood as the web of supporting actornetworks engaging with producers, has been much less problematized. A case could be made that for individual producers, the efficiency of the frontend in maximizing the monetary value of food produce and minimizing the transaction costs originates from setting up an adequate backend system supporting the mobilisation and implementation of innovative practices. Enacting such agricultural transitions requires the acquisition and adoption of new sets of knowledge and skills (Winter 1997;Knickel et al. 2009). The introduction of novelties often depends on the transmission of tacit knowledge (Knickel et al. 2009;Dupré et al. 2017) facilitated by shared experiences, new cognitive frameworks, collaborative experimentations and physical interactions (Knickel et al. 2009;Milestad et al. 2010b;Marsden and Farioli 2015;Läpple et al. 2016).
The transition to SFSC thus rests on the producers' ability to mobilise, adapt and apply knowledge that is often foreign to the local farming community. Bui et al. (2016) noted that the development of quality food niches, which SFSC fall into, are fertile grounds for the introduction and application of new rules and practices deviating from conventional ones entailing new visions of farming. Such initiatives thus profoundly challenge the conventional farming order and the wisdom of traditional know how, not the least with respect to the tacit and voiced understandings of what constitutes a good farmer (Sutherland and Burton 2011). Accessing external knowledge is a key challenge to be overcome in order to engage in SFSC. Scholars have highlighted the role of non-local and even non-farming actors in enabling such knowledge intensive transitions (Feagan 2007;Lamine 2012;McKitterick et al. 2016). Another hurdle is the level of resistance from conventional farmers in letting such external knowledge settle in the vicinity of their own practices. In their study of SFSC producers in France, Chevallier et al. (2014) stressed the cognitive distance separating farmers newly engaged in SFSC and more established conventional farmers from the same locale, invoking a sense of marginalisation of the former within the local farming community. The literature on the transition to organic farming can here provide valuable pointers to that discussion, especially as there is a relatively high proportion of certified organic farms involved in SFSC (Mundler and Laughrea 2016). Padel (2001) documented the social isolation of producers converting to organic farming in their own community, whereas Blay-Palmer (2005) noted that the implementation of organic farming induced simultaneous processes of adhesion and opposition to the introduction of such 'foreign' values.

Proximity matters
Scholars have discussed how the development of SFSC may enact and valorise the potential borne by strong geographical proximity between producers and consumers (Aubry and Kebir 2013;Guiraud and Rouchier 2015;Chiffoleau et al. 2018;Dubois 2018). Referring to the notion of proximity may seem unproblematic in the case of SFSC as they are usually understood as spatially compact food provision systems (Eriksen and Sundbo 2015;Favilli et al. 2015). Scholars have nonetheless warned against the local bias (Hinrichs 2003) or local food trap (Brunori et al. 2016) in which "the local is uncritically accepted as being 'good'" (Maye and Kirwan 2010, p. 8) and perceived as endowed with predominantly positive meanings (Le ). Hinrichs argued that, while cognitive and social associations may indeed "flourish under conditions of spatial proximity, this is not automatically or necessarily the case" (2003, p. 36). This argument was later corroborated by empirical studies that have singled out how the coordination of organizationally short producer-consumer relations may be performed 'at a distance' (Renting et al. 2003;Watts et al. 2005;Aubry and Chiffoleau 2009;Kneafsey et al. 2013).
What this concretely means is that being close (Milestad et al. 2010a) can be achieved through multiple spatial and relational configurations. Trust, cognitive likeness and social kinship play a central role in activating and valorising producer-consumer proximity (Watts et al. 2005;Forney and Häberli 2016;Thorsøe and Kjeldsen 2016). Forney and Häberli (2016) further argued that the mechanism driving food respatialisation is less geographical proximity than social proximity. For less spatially compact SFSC, bridging the distance between producers and consumers requires the establishment of more structured relational arrangements (Renting et al. 2003;Aubry and Kebir 2013;Thorsøe and Noe 2016). An illustration of this is the role of regional identities in institutionalising speciality food marketing (Filippi and Torre 2003;Hinrichs 2003;Torre 2006;Feagan 2007;Halkier et al. 2017). All in all, the emergence of multiple SFSC configurations reflects the various expressions of proximity unfolding in space and time (Kneen 1993;Hinrichs 2000).
The remainder of this section will review three emblematic and well-documented examples of SFSC in order to illustrate the diversity of these expressions of proximity: farmers markets, AMAP and AOC labels.
Farmers' markets are the most recognisable "arena of exchange […] concrete and meaningful spaces in which food is exchanged" (Holloway et al. 2007, p. 9), allowing for the creation and reproduction of shared meanings about the quality and origin of food (Holloway and Kneafsey 2000;Feagan 2007;Holloway et al. 2007), despite the fact that such interactions are often brief (Le Velly and Dubuisson-Quellier 2008). These traditional venues (Guiomar 2011) are instrumental in perpetuating and anchoring the culinary and farming heritage in contemporary practices and exchanges. Farmers' markets are thus critical to the constitution of 'local food' as a cognitive construct. Dubois (2018) acknowledged that, in operational terms, farmers' market participation is still the foundational element to create the social and relational capital to engage in wider types of SFSC for individual producers.
AMAPs (Association pour le Maintien d'une Agriculture Paysanne), France's equivalent to the North American CSA-schemes, are market arrangements connecting producers and consumers within a certain territory (Le Velly and Dufeu 2016). AMAPs are consumer-based (Aubry and Kebir 2013) meaning that a collective of consumers takes the initiative to contract one or more producers to regularly supply them with fresh farm produce boxes, often on a weekly basis (Dubuisson-Quellier et al. 2011). The institutional framework for AMAP is ensured by subscribing to the national charter based on the principles of peasant agriculture (Chiffoleau 2009) and the implementation of fair pricing (Le Velly and Dufeu 2016). However, the specific contractual arrangements (product type and quantity, delivery or pricing) are negotiated directly between the consumer collective and the contracted producer(s) (Dubuisson-Quellier et al. 2011). Chiffoleau (2009, p. 220) noted that the motivations to engage in such initiatives are more political, combining "a sense of civic responsibility with ecology", than economic. Although the participation in AMAP creates a sense of belonging ) among participants, it does so without requiring face-to-face interactions, as some consumers actually never meet with the producers (Le Velly and Dubuisson-Quellier 2008). Internet-based applications have now made the coordination of AMAP more efficient . The food communities emerging from the constitution of AMAP have provided a platform for farmers who are too innovative to thrive within the conventional food industry an opportunity to develop practices aligned with their own professional aspirations and beliefs regarding food chain sustainability (Chiffoleau 2009).
Regional origin labels constitute visual markers that help consumers recognize the authenticity of a product according to its geographical provenance, the production methods used and a pointer of its overall quality . Origin labels are used to single out products that strongly relate to the cultural heritage and agricultural legacy of a region and "historically founded on old, fair and constant practices" (Torre 2006, p. 57). Origin labels are thus embedded in regional farming identities, practices and values (Bessière 1998;Torre 2000;Barham 2003). A typical case of regional origin labels is the French Appelation d'Origine Controlée (AOC), typically for specialty products such as wine and cheese. The participation to AOC cooperatives enhances the producers' ability to take advantage of actor proximity through the constitution of a coherent territory for collective action (Torre 2000(Torre , 2006Filippi and Torre 2003). Regional origin labels are considered organizationally 'short' as the produce is embedded with value-laden information that can be easily 'decoded' by consumers who may not have any personal experience of the region Renting et al. 2003). These spatially extended forms of SFSC are grounded on institutional proximity constituted through a set of formalized rules, norms and standards (Renting et al. 2003) that are necessary to translate the know how of traditional practices into monitorable codified knowledge. In Europe, the EU policymaking level has been instrumental in piloting and enforcing geographical designation and certification procedures.
These three examples show the diversity of the spatial and relational construction of 'short' food chains as an outcome of the mobilization of multiple social, institutional, organizational and cognitive resources. This interplay between the spatial and relational dimensions in SFSC was formalised by Renting et al. (2003) in a typology encompassing faceto-face (farmers' market), proximate (AMAP) and spatially extended (AOC) configurations. Whilst being widely different in operational terms, these arrangements do share the aim of brokering an experiential encounter, whether physical or cognitive, between producers and consumers (Le Velly and Dubuisson-Quellier 2008). Hence scholars characterise the engagement in SFSC as a relational process of reconnection (Winter 2003;Watts et al. 2005; Morris and Kirwan 2010).

Analytical framework
To theoretically and empirically explore SFSC as diverse socio-spatial constructs, the present study proposes to use the notion of proximity, developed in the field of human geography as a way to understand how economic agents coordinate their actions across space and how it affects their operations (Rallet 2002). It rests upon the analytical distinction between geographical proximity, corresponding to the physical distance separating actors, and organised proximity, characterising the organizational, social, institutional or cognitive aspects of the relationships (Torre and Gilly 2000;Torre and Rallet 2005).
Geographical proximity is central in inducing the creation of close relations between actors, although it is now well-documented that permanent co-localisation is not a necessary condition as the benefits from face-to-face interactions may be generated during occasional gatherings that are "dense in interactions" and allowing actors to "exchange information, express emotions and be present with a distant partner" (Torre 2008, pp. 875-876). So geographical proximity can be either permanent or temporary (Torre et al. 2018). In addition, recent developments in information and communication technologies (ICT) have made it possible to establish and maintain contacts 'at a distance' (Lorentzen 2008). In the context of alternative food networks, ICT has been valuable in facilitating exchanges and mobilizing 'virtual' communities of practices on the basis of shared cognitive frameworks (Filippi et al. 2011;Bos and Owen 2016;Chiffoleau et al. 2018). Online presence has revealed the ubiquitous nature of contemporary interactions with a heightened ability for individual actors to act and contribute in real-time to developments taking place in multiple locales (Torre 2009). Applying Torre (2009)'s evolutionary proximity model to the context of SFSC formation, Dubois (2018) showed the foundational role of farmers' markets as an incubator of reciprocal producer-consumer relations and the emergence of online applications (e.g., social media, box schemes and online shops) and new food venues and outlets (e.g., fine deli, gourmet restaurants, farm gate sales, food hubs) as a way to maintain and reproduce these relations.
Organized proximity takes shape at the confluence of two relational logics: one of belonging and one of similarity (Torre and Gilly 2000;Torre and Rallet 2005).
The logic of belonging relates to the effectiveness of the coordination among actors (Filippi and Torre 2003;Guiraud and Rouchier 2015) facilitated by their active participation in the same networks and collectives, such as associations or cooperatives (Torre et al. 2018). In the examples of AMAP and AOC, this is realized by joining a 'club' set up through formal membership and governed by common rules, routines and behaviours (Torre 2006;Filippi et al. 2011;Aubry and Kebir 2013). The logic of belonging denotes a cooperation approach among independent actors based on voluntary participation and without any form of subordination (Aubrée et al. 2018). How belonging is actually achieved in practice depends on the ability of actors to collectively find the most fitting "mode of organization" (Le Velly and Dubuisson-Quellier 2008) accommodating their various characteristics, aspirations and locales. The logic of belonging thus aims to create and consolidate a culture of cooperation among participants that can be sustained over time (Aubrée et al. 2018). In the case of remote farming communities in Brasil, Torre et al. (2018) for instance showed that the collective adoption of new farming practices was facilitated by non-commercial exchanges unfolding from the participation to the farmers' cooperative and its activities.
The logic of similarity refers to the adhesion of actors to a common set of representations, objectives and values making them more 'alike' in cognitive terms (Filippi and Torre 2003;Filippi et al. 2011;Aubry and Kebir 2013;Richard et al. 2014;Aubrée et al. 2018). As a result, actors tend to share the same knowledge base reinforcing their cognitive and institutional bonds (Filippi and Torre 2003). The logic of similarity also explains the commonalities in how actors adopt, organize and operate novel social practices, such as SFSC Aubrée et al. 2018). The logic of similarity suggests a sense of identification with actors recognizing in each other the constitutive elements of a coherent worldview. In the case of SFSC, Lamine (2012) nonetheless pinpointed the need to take into consideration the diversity of actors that take part in such innovative approaches. Torre et al. (2018) clarified that the logic of similarity does not rely on relations of personal acquaintance between actors, but examines how they may share similar references by adopting a similar farming model.
The present study will use geographical proximity (i.e., permanent or temporary) and organized proximity (i.e., the logics of belonging and similarity) to analyse and characterise the knowledge exchange processes mobilized by SFSC producers along the journey. Of particular interest will be the spatial, operational and temporal aspects of these interactions. Where and how do actors initially meet? How relations are maintained over time? What relational arrangements are conducive of knowledge exchanges?
The analysis will also emphasize the role of translocal practices in the development of SFSC, revealing the constitution of shared experiences of near-produced quality food across multiple locales. A translocal practice here refers to experiential knowledge external in scope and/or origin that are introduced, adopted and embedded into the existing farming know-how through a collaborative process involving actors from multiple locations.

The case study approach: method, context and relevance
This paper uses a case study approach analysing the experiences of independent producers engaged to various extents in SFSC in the county of Västerbotten in north Sweden. The empirical material was collected during semi-structured interviews conducted by the author with nine small-scale farmers ( Table 1). The farms were first identified on the Gårdnära.se website, which maps Swedish farms engaged in near-produced (and often organic) quality food practices. Farmers were then contacted by email to enquire for an interview. Three other interviews were performed with key regional informants to contextualise recent agricultural and rural developments in the County.
The farming landscape of Västerbotten is dominated by dairy and beef cattle, with a farm-size structure consisting of large-to-mid-sized farms on the coastal areas along the Gulf of Bothnia and mid-to-small-sized farms in the inland areas towards the Norwegian border (Andersen 2017; Dubois and Carson 2019). Small-scale farming is essentially oriented towards organic and local sales, which has showed sluggish development due to limitations in slaughtering, distribution and marketing capabilities (Västerbotten 2011). Quality food in Sweden is primarily associated with organic food standards and the respect of sanitary rules, and organic food is widely available through conventional food retailers (Hochedez 2008). In 2010, about 5.5% of agricultural holdings and 14.3% of arable land were categorized as organic (Yngwe 2014). The transition towards quality food has increased the demand from farmers to meet and share their experiences with the aim of improving their own practices (Yngwe 2014). Although SFSC were poorly developed in Sweden ten years ago (Hochedez 2008), Swedish regions have more recently supported such initiatives, often induced by an increased awareness and interest of consumers in buying near-produced quality food (Aggestam et al. 2017).
This research uses a 'critical case' design based on insights from a limited number of in-depth cases. Flyvbjerg (2006) defined the critical case approach as illustrating aspects of a general problem (in our case the development of SFSC) based on experiences from a specific entity. Although inherently limited in scope and extent, focusing on narrow cases raises important issues that contribute to the overall understanding of the constitution of the agricultural innovation space (Fielke and Wilson 2017). The present study also feeds into a growing section of agri-food studies interested in examining the systemic role of quality food niches and associated experiences in spearheading 1 3 and informing a more wide-ranging transition towards sustainable food regime transformation (Brunori et al. 2012;Marsden 2013;Hinrichs 2014;Ingram et al. 2015;Bui et al. 2016;von Oelreich and Milestad 2017).
The peripheral position of our case study from the 'hot spots' of European productivist agriculture limits a priori the resistance 'from within' to the establishment of SFSC and creates a fertile ground for introducing new practices Ilbery and Maye 2005). Peripherality creates the necessary "spaces to act" allowing for experimentations (Marsden 2010, p. 242). Scholars also emphasized the potential role of SFSC in accompanying the emergence of quality food niches in such peripheral settings (Renting et al. 2003;Ilbery et al. 2004). At the same time, the long physical distances separating actors in the region also hampers their ability to exchange experiences and develop a shared farming know how locally. The case study set-up thus proposes a unique opportunity to delve deeper into the innovation processes that permits the re-embeddedness of quality food niches at the geographical and institutional margins of contemporary agriculture.

Results
The analysis of experiences from north Swedish farmers in setting up SFSC operations revealed two fields of knowledge that appear at the core of the producers' innovation and marketing initiatives. The following sections give an account of how proximities are mobilized and new associations between actors brokered across locales and walks of life.

Beyond certification: the emergence of quasi-organic food niches
The creation and sustaining of personal and systemic trust to promote food sustainability is a well-documented feature in SFSC studies (Thorsøe and Kjeldsen 2016).   Blay-Palmer (2005) noted the instrumental role of the introduction of EU certification standards in 1991 in accompanying the rise of organic practices in different locales throughout Europe. Certification created a formalized system of codified knowledge, including rules, norms and standards, guaranteeing the same level of 'organicity' and product quality independently of the immediate agricultural context. Both producers and consumers mutually adhere to these values, which shifted the coordination of organic food from a 'nucleated' regime, built around personal trust, to an industrial regime, guaranteed by institutional trust (Gigon and Crevoisier 1999). The universality of organic certification standards means that its application does not need to cater for the producers' previous farming experiences, although some of these practices may already be in place. So, from the moment producers become certified and apply organic standards, this external knowledge becomes part the pool of farming practices of the locale.
The main organic certification actor in Sweden is KRAV, founded in 1985. KRAV standards admittedly go beyond the minimum standards set out by EU regulations, and includes other societal and environmental concerns such as animal welfare, energy efficiency or climate-impact and social responsibility considerations (KRAV 2016;von Oelreich and Milestad 2017). Swedish farmers that are KRAV-certified thus automatically adopt practices that go beyond what organic certification knowledge institutionally stands for.
The relation of our interviewed farmers to KRAV is complex: if all tend to consider their production as organic, only few of them are certified to date. Those who are stated that being KRAV-certified created a market advantage that translated into a price premium and thus generated increased monetary added-value. Ingrid, a former dairy farmer turned beef producer located in the inland, saw certification as a motivation and goal to be attained, partly linked to the prospects of getting a higher financial compensation from Sweden's agri-environmental schemes. Concretely, becoming certified did not necessitate adopting and implementing new practices on her farm. In that sense, certification did not induce an extra cost in operational terms. Moreover, Ingrid sees KRAV-labelling as a competitive advantage allowing her to better showcase her produce as only few fellow inland meat producers are certified.
Likewise, KRAV-certification is an important part of the business model for John's farm. He recalled a concrete example, that he could take up to 120 kr/kg for his strawberries at the Farmers Market in Umeå. KRAV certification made it possible for John to apply a premium on his strawberries' price and thus generate higher revenue. John also considers KRAV-labelling as a statement showing his commitment to the core values of the organic movement. John nonetheless considers KRAV as a minimum standard though, as he claims his produce quality is "better than KRAV". For Ingrid and John, certification mostly validates the scope of their farming acquis, i.e., the range of practices that they already implement throughout their production process, and thus does not induce an extra-cost in terms of adaptation. In that respect, certification aims at validating producers' practical knowledge in relation to a complex body of codified knowledge and translating these practices into a set of indicators that can be objectively monitored by external controllers. For the consumers, certification permits an instant recognition of food quality thanks to the recognisable label apposed on the packaging and/or at the farmers' market stand. Organicity, in this case, is brokered using an artefact that is outside the scope, but nonetheless strongly embedded, in producer-consumer relations. As in the case of AOC, KRAV-certification is effective for generating values about food quality that can be mediated 'at a distance'. For individual producers like Ingrid and John, certification is an important market-based resource for generating larger revenues from hand-to-hand transactions, such as farmers market or farm gate sales. It also unlocks opportunities to generate new revenue streams by expanding towards types of SFSC that can be activated 'at a distance', e.g., internet-based box schemes. In that sense, certification can be deemed as a relational device allowing for the enlargement of the local geographies of food sustainability and quality (Higgins et al. 2008).
This view of certification as a path towards increased economic profitability was corroborated by Erika and her husband who aim at certifying their lamb production in the medium term. To fulfil KRAV criteria, they first needed to acquire more land in order to produce enough own produced organic fodder to feed their animals. Erika mentioned that becoming certified is "much more worth", suggesting price premium as an expected benefit, and that adaptation should not be a major effort in operational terms as their practices are already "as close to organic as they can be", but is rather a matter of paperwork. The motivation to adopt certification is essentially the monetization of farmers' current practices. For such farmers, certification constitutes an incremental move of an institutional and organizational nature rather than a radical shift in their farming practices. As such, certification does not contribute to the introduction of novelty in the local farming know how: certification is an act of adhesion by a producer to a global system of values and practices that 'backs up' and transcends their own views and actions.
Benny and his wife Kerstin are newcomers in farming. Their swine breeding farm is not yet certified, but they intend to soon start the process as they already fulfil all the requirements. A current hindrance is that in order for their farm to become certified, the animals have to be slaughtered at a certified abattoir. However, the closest certified abattoirs, geographically speaking, are located in either neighbouring counties, i.e., Norrbotten in the north or Jämtland 1 3 in the south. This geographical distance clashes with values that Benny and his wife strongly associate with their version of organic food engagement: being certified would imply transporting live animals over long distances and thus causing unnecessary animal stress and 'food miles' impacts.
In this case, we see the limits of certification as an institutional arrangement applicable to all locales. In remote rural settings, such as where Benny's farm is located, actors of the food chain are scarce and scattered geographically inducing long transportation. The ability for individual small producers to adhere to organic conventions thus requires from other territorial actors, both upstream (supply of animal feed) and downstream (abattoir, food artisans) in the value chain, to adhere as well to the same value system. This prerequisite makes the shift to organics more complex to operationalize due to the absence of a critical mass of actors adhering to the same value system in close geographical proximity. This is even more so the case for producers operating SFSC as they already navigate outside the mainstream organic food system. For geographically isolated producers, the logic of belonging comes into play as a tighter cooperation with other organic actors within reasonable geographical distance is necessary to sustain their distribution system.
Hence, the existence of a relatively dense territorial network of 'like-minded' actors constitutes a favorable ground for the logic of similarity to thrive. As such preconditions are not met in our case study region, most of our interviewed farmers have chosen to work with organicity using an alternative modus operandi than certification. This modus operandi consists in conveying critical value-laden information about the product quality and singularity directly to the consumers (Le Velly and Dubuisson-Quellier 2008).
While waiting to proceed with certification, Benny and his wife Kerstin have chosen to be open to their customers in order to make their farming practices transparent and intelligible. They believe that their customers know that their practices are "pretty organic". To convey information about their practices, they use online social media (Facebook) to showcase their daily activities, especially using visual representation (pictures). Through social media, they are able to connect through multiple circles of social networks (friends of friends) and thus expand the realm of social proximity and personal trust and mediate the ecological and social values of their produce, such as animal welfare, landscape stewardship or preservation of the local cultural heritage. The use of appraisal and reputation through social capillarity for validating their "pretty organic" practices enables Benny and his wife during this transition phase to already command a premium price on their products and to recruit and consolidate a loyal base of 'like-minded' consumers (i.e., cognitive and social proximity).
The proximity between producers and consumers is in this case shaped through reciprocal exchanges of information and the creation of a mutual understanding of the product singularity. This proximity is achieved through shared cognitive frameworks (Kaltoft 2001) and a tacit moral contract guiding ethical relationships (Venn et al. 2006). Whereas certification addresses essentially the ecological quality of food, the mobilization of social and cognitive proximity through direct consumer-producer interactions allows to trace "a route to actors behaving with kindness, care and generosity" (Tregear 2007) and replaces quasi-organic practices within wider rural social practices. Overall, the experimental nature of organic thinking and its reliance on experiential knowledge (Aeberhard and Rist 2009) is a fertile ground for the introduction of novelty into the local farming know how.
In this context, some interviewed producers questioned the pertinence of certification in pursuing organicity within the framework of SFSC. Peter, who is a small-scale animal breeder located in the inland, sees certification as an administrative burden and, in the end, that it is not worth the effort, especially for "small market" producers as himself. Peter's farm is not certified but applies traditional farming techniques that are showcased on the farm's blog. Kevin is a beef meat producer whose farm is located close to Umeå. His farm is not certified, and will never be according to him. He considers that he is so serious and open to his customers about his breeding practices that he does not need anybody else to monitor and validate them. He sees KRAV as a brand "like Coca-Cola". Kevin claims that "everything I do is organic" although he acknowledges that he is not allowed to explicitly label his produce as such.
The progressive constitution of a body of shared experiences of food quality with consumers makes producers less constrained by the standards and norms from quality and certification labels . In several of our cases, quasi-organic producers were able to nonetheless command a price premium thanks to the 'open channel' established between them and their consumers, making it possible for the latter to recognise and adhere to these specific ecological, cognitive and societal values. Hence, the development of SFSC made possible the emergence of quasi-organic practices at the geographical and institutional margins of the organic food system. Each of these sociospatial constellations conveys its own version of organicity resulting from the singularity of the coordination processes taking place between producers and consumers.

Translocal communities of quality food practices
Individual producers have a propensity to engage in multiple types of SFSC simultaneously (Thorsøe and Noe 2016;Dubois 2018). The consolidation of SFSC operations thus requires the acquisition of a broad range of entrepreneurial skills (Mundler and Laughrea 2016;Aggestam et al. 2017) related to distribution, logistics, packaging, marketing or food processing. Scholars have also extensively documented the involvement of producers in a wide range of networks or communities of practices in relation to rural and agricultural innovation processes (Oreszczyn et al. 2010;Favilli et al. 2015;Marsden and Farioli 2015;Cross and Ampt 2017). Agricultural practices are context-dependent and the local farming know how is too complex and large to be effectively codified (Torre 2000). Tacit knowledge is acknowledgedly "very difficult to acquire and transmit" (Dupré et al. 2017, p. 404). Tacit knowledge exchanges may be brokered through shared experiences, cognitive references and physical interactions (Knickel et al. 2009;Marsden and Farioli 2015;Läpple et al. 2016).
The latter seems to indicate a priori that geographical proximity is instrumental to the diffusion of innovation in local agriculture. The specificity of our case study context, characterised by low geographical proximity, allows us to examine the role of temporary geographical proximity in the mobilization of knowledge and relational resources in the development of SFSC.
In the household, Benny is in charge of the expansion of the farm's delicatessen activities. At the time of the interview, a small processing shed adjoined to the farm shop was in construction. Not having prior experiences in food processing, Benny acknowledged that he needs to learn about the different moments of the processing chain, from slaughtering, to manufacturing products (e.g., sausages) and to developing appropriate cool chain logistics. Most of this learning takes place at artisan food workshop arranged by the County administration or other public organisations and advisory services, especially in the neighbouring County of Jämtland which has a much longer tradition of artisan food than Västerbotten. At one of these occasions, Benny recalled that a German charcuterie-master instructed the participants on traditional German sausage-making techniques. Benny's sausage-making business requires coordination with food actors outside the region. His meat is delivered at three different charcuterie-makers in Sweden: one relatively nearby, one near Östersund (Jämtland) and one further down south in Mid-Sweden. The manufactured sausages are sent back to Benny to be sold locally. Benny got acquainted with these different actors during organised workshops. In Benny's case, the participation to these workshops fulfils two purposes: expanding his entrepreneurial skills in non-farming activities and getting him acquainted with a diversity of food actors.
These workshops are illustrative examples of how such temporary meeting places function as incubators of interactions that may evolve over time into deeper relations affecting the scale and mode of operations of SFSC. Actors taking part in these workshops already share from the outset, albeit unbeknown to them at first, the same values about quality food. Face-to-face interactions allow them to recognize and express this shared value-system: temporary geographical proximity is here instrumental in unlocking the untapped cognitive proximity borne by these geographically and socially distant actors. Similarity here does not mean sociocultural homogeneity of these actor groupings: the diversity of origins, both geographical and professional, is favourable to the appropriation of external knowledge and the co-creation of new meanings and practices. Aubrée et al. (2018) noted the importance of "crossed sessions" between producers and artisan food actors in promoting quality food transitions. These actors usually have little, if any, acquaintance with each other in social and professional terms and have no experience of working together (Aubrée et al. 2018). The joint participation to "crossed sessions" gradually changed attitudes and mentalities about cross-professional cooperation and create new opportunities for actors to collectively organize short food chains.
Other producers in our case study region have experiences of working with artisan food actors. Kevin and Sigrid cooperate with a Danish charcuterie-master who operates from a location south of Umeå. The charcuterie-master subsequently supplies products to local supermarkets as well. Jointly to starting her farm shop, Ingrid started cooperating with a food producer located in Strömsund, a small inland locality in Jämtland, who makes sausage using Ingrid's meat. The sausages are then sold by Ingrid directly at her farm shop. Ingrid got acquainted with this person during a study trip about food processing in Jämtland that the Västerbotten County Administration Board once arranged.
The mobilization of external knowledge through interactions with non-farming and extra-local actors is instrumental to the expansion of SFSC activities. This expansion concerns the diversification of marketing channels used, the enlarged geographical outreach of these operations, the increased range of products available and the enhanced ability to generate revenue streams and add monetary value to the foodstuff without increasing production volumes. Overall, the engagement in SFSC changes how producers view the role of intermediaries in the food system. First, cooperation is voluntary, reciprocal and mutually beneficial and not operated by arms' length price bargaining but based on 'synced' value systems. Second, although the engagement of SFSC for small producers aims to reduce their subordination to dominant intermediaries of the conventional food industry and thus gain greater autonomy (Forney and Häberli 2016), the examination of the 'backend' relations unveils the constitution of a complex system of inter-dependencies among smaller food actors. The inter-dependencies are instrumental in unlocking the 'dormant' permanent geographical proximity between producers and consumers and creating new monetary value instead of capturing a share of it (Praly et al. 2009).
Peter and his wife have established their self-sufficient, mixed-farming farm, located in Vilhelmina, around what he calls 'landscape values'. Peter strives to reintroduce farming techniques that were once part of the farming know-how of the area, but that have eventually been abandoned and now mostly forgotten. Peter is active in the ÅTER-network, a network of producers seeking to preserve and disseminate traditional farming practices. Every year he and his wife make a road-trip to southern Sweden to meet other network participants. They share their experiences of traditional farming lifestyles on a dedicated blog. Peter himself browses other producers' blogs and social media to find inspiration for new activities. Peter recalled how his wife got the idea to use egg-shell based colouring to colour lamb wool, in order to create a more personalized product, from the blog of another farmer located in mid-Sweden. After some online correspondence, Peter's wife travelled south to spend a week at this farmer's place to learn and developing proficiency in this unique craft. Angela, an organic vegetable producer located on the outskirts of Umeå, travels every year to Karlstad in southern Sweden where the annual convention of a farmers' association takes place. The rest of the year, she maintains contact with other producers "on the tablet".
Within the contemporary food system, producers attempting to "recapture" pre-conventional practices (Blay-Palmer 2005, p. 563) often stand isolated in geographical, cognitive, social or organizational terms (or all of the above). The experience of Peter and his wife epitomizes "how local, tacit knowledge can be relearned" (Morgan and Murdoch 2000, p. 171) by mobilizing farming practices that are already in use elsewhere. Tapping into such experiential knowledge by importing and introducing them to a new locale creates bridges between the farming know how of communities that are otherwise physically disconnected. The construction of translocal practices is facilitated by temporary geographical proximity, i.e., through the active participation and networking at conventions, fairs or at the occasion of study trips.
In addition, Peter and Angela's examples show how internet connectivity changes how farming knowledge is created, exchanged and applied. Blay-Palmer (2005) previously showed how the internet was an importance source of information for innovators in organic farming. In its current evolution, the internet is thus particularly adapted to facilitate mutual learning processes based on the exchange of tacit knowledge within geographically distributed networks of practices, hence the notion of virtual reconnection proposed by Bos and Owen (2016). These new arrangements promote new governance models based on the appraisal of interactive, participatory, inclusive and transparent approaches promoting translocal learning and constituting a knowledge support system (Lamine 2012;Chiffoleau et al. 2018).
The recognition of shared landscape values plays an incubating role for the establishment of new modes of governance among farmers aiming to improve their collective capacity of action and contributing to organizational innovations in local agriculture. An interesting case in Sweden is the creation of genetic banks. Genetic banks are bottom-up coalitions of producers seeking to preserve and maintain healthy populations of native animal breeds. Peter is engaged in the development of a genetic bank for Lapland goats. The involvement of an animal population geneticist from the Swedish University of Agricultural Sciences (SLU) was required to validate the knowledge base to set up the Lapland goat genetic bank. Through the genetic bank, Peter keeps contact with other Swedish farmers that breed the same goat species. Although genetic banks cover the whole Swedish territory, groupings of farmers are responsible for coordinating operations locally. The socio-spatial characteristic of these actor-networks is of a distributed network of teams that is brought together by a shared worldview (cognitive proximity) and common rules (institutional proximity), but for which operational elements (organizational proximity) may vary with the locale.
Kerstin, Benny's wife, is engaged in the Landtsvinet association, a nation-wide genetic bank for the Linderöd pigs, a typical Swedish, but not specifically north Swedish, pig breed. Kerstin had no prior experiences of breeding Linderöd pigs. Kerstin recalled that other producers in the network were sceptical that this breed could thrive outdoors in the harsh climate of Västerbotten's mountainous inland. Kerstin and Benny chose to breed this rare Swedish breed in order to distinguish themselves from other regional producers. The genetic bank issues a certificate and establishes a contract with the producer, with a strict set of rules for them to follow. After being part of the association for some years, Kerstin was co-opted by the association chairman to be part of the Board and he entrusted her to take a more prominent role in the association. The association has yearly meetings in Malmö. The participation in the genetic bank made it possible for Kerstin to export a small number of Linderöd pigs to Finland. She described a chaotic process which involved many back-and-forth exchanges with the representatives of the Swedish Board of Agriculture and a veterinarian, as none of these actors had prior experiences in dealing with such a process.
Another farmer who has been actively engaged in farmer associations is John. John and his wife aim to develop a garden concept farm. He wishes to establish a genetic bank for apple trees. The bank is not formally established yet at it requires formal approval from scientists at SLU in Alnarp near Malmö. He nonetheless considers his farm as being a biobank as it already contains multiple species of apple trees. He is adamant in reintroducing different apple tree species locally and plants trees on his customers' property through his own 'planting school'. John admits that, as a diversified farm, he needs to be like a 'spider in the web'.
He has been a Board member at The Swedish Horticultural Society since 2014. He is responsible for activities related to organic farming. Through his engagement he has been able to expand his contact network, help organize activities for regional producers, but also voice his ideas across the country. For instance, he is striving to develop an e-plant system at national level for registering different types of plants on a web-portal.
The constitution of genetic banks illustrates the processes of 'social and governance innovation' taking place on the outskirts of the strongholds of the conventional food regime (Marsden 2013;von Oelreich and Milestad 2017). Marsden (2013, p. 125) defined such sociotechnical niches as "small networks of actors supporting innovation on the basis of expectations and visions". These sociotechnical experiments (Hargreaves et al. 2013) are constructed on the appraisal of social, ethical and cultural rules (Seyfang and Smith 2007). Genetic banks show how farmers develop alternative ways of coordinating their actions based on shared understanding of the values inherent to quality food and institutionalizing them within the spectrum of contemporary farming practices through socially embedded organizational innovations.

Discussion: the works of proximity in shaping food niches at the periphery
In this exploratory study, our aim was to flesh out what 'close' backend relations actually entail in the development of local food niches. The use of the notion of proximity to explore and analyse the experiences of north Swedish farmers in establishing SFSC operations revealed the multiplicity of learning and innovation processes mobilized in order for farmers to realize their own version of a near-produced quality food system. To recapitulate, we found that: (1) an open and trusted interaction with consumers allows SFSC producers to implement quasi-organic practices, i.e., production practices that are organic in principle, but do not formally comply with certification standards, creating locally negotiated 'patches' of organic values and unlocking new marketing opportunities, and (2) producers are engaged in motley associations of food professionals aiming to the creation and valorisation of "cooperative know-how" (Tregear and Cooper 2016, p. 106) and the construction of collective value-based meanings of quality food.
The analytical distinction between, on the one hand, permanent and temporary geographical proximity and, on the other hand, the logics of belonging and similarity specifically unveiled the role of non-farming, extra-local actors in introducing and embedding new practical knowledge into the local farming know how. The development of quality food niches is driven by the constitution of place-based, yet geographically distributed, associations of actors and takes place at the interplay of the logics of belonging and similarity, i.e., by enmeshing relational processes of cognitive, social, institutional and organizational proximities.
The selection of our case study made possible the characterization of such translocal practices. Indeed, our smallscale 'organic' SFSC producers are peripheral on (at least) four accounts: first, they are located far away from the 'hot spots' of the European conventional large-scale agriculture (cf. Murdoch et al. 2000); second, they are located far away from large metropolitan areas and their abundant pool of potential consumers (cf. Jarosz 2008; Aubry and Kebir 2013); third, the farming landscape is relatively scattered with long distances between individual farms, the more so the further inland one gets; and fourth, producers engaging in SFSC are still in minority in their respective rural communities, making the local support system 'thin'. To sum up, our cases emphasize how new farming styles and practices are introduced and consolidated in the absence of geographical proximity: reaching out to and mobilizing food actors beyond the immediate surroundings thus becomes a necessary approach to sustain innovative behaviour. This view challenges the common understanding in agri-food studies of proximity as co-localization. The analysis of north Swedish SFSC experiences allowed us to explore and characterise more closely the relational devices mobilized to promote innovations in local agriculture through the constitution of geographically distributed teams of 'like-minded' actors.
The examination of quasi-organic practices illustrated the fuzzy boundary that exists between the formalized knowledge of 'institutional' organic farming standards and norms and the reflexive value system embedded in each producer's own interpretation of organicity. Whilst most of our case study producers are not certified, they tend to benchmark their practices against these standards, which manifests itself empirically through the use of expressions such 'as close as organic', 'better than KRAV', "'pretty' organic" or "everything I do is organic". So even though certification standards are formally external to our farmers' practices, they act as a normative reference point that influences how producers implement their vision of near-produced quality food. Marketing quasi-organic food necessitates, from the producer's standpoint, substantial efforts in establishing and maintaining a high level of social and cognitive proximities with the consumers. Direct interactions, whether face-toface or online, allow producers to make explicit aspects of their know how related to food quality and societal values and explain how their practices are similar or different than established organic certification standards. The combination of social and cognitive proximities thus allows for the transmission of tacit 'value-laden information' about food quality that producers use to showcase and valorise their produce without being dependent on the enforcement of external standards. However effective, the time-consuming and slow-paced character of the quasi-organic pathway, requiring numerous and regular face-to-face interactions and co-optation to expand the customer-base (Dubois 2018) create some limitations on the scale and scope of revenuegenerating activities undertaken by individual farmers. In comparison, SFSC producers choosing to pursue certification may benefit from the institutional and organizational proximities that these schemes induce. These producers may then expand their marketing activities towards what Renting et al. (2003) labelled spatially extended forms of SFSC, as complementary revenue streams that may be more easily operated and monitored 'at a distance'.
Additionally, this study unearthed a range of socio-technical spaces in which SFSC producers SFSC take part. This 'new associationalism', a notion introduced by Marsden et al. (2002) to frame the examination of emerging actor networks originating from agri-food diversification, is a clear manifestation of increased needs for producers to acquire new skills and competences and develop new ways of coordinating their expanding operations. Face-to-face interactions broker the constitution of these informal learning networks, but, unlike classic farming cooperatives, these are not confined to a certain subregional geometry (Watts et al. 2005;Forney and Häberli 2016). The study also corroborates Clark's (2005, p. 490) understanding of these arrangements as heterarchical, i.e., based on interdependencies rather than subordination, and as "evolving […] complex assemblages of 'local' and 'at a distance' relations". In this context, Torre (2009)'s evolutionary proximity model provides a potent heuristic device for characterising the key moments of the mobilization of organized and geographical proximities in the constitution of these learning networks (see Fig. 1).
According to Torre (2009), the first moment is initiated through serendipitous encounters taking place at the occasion of specific events or arrangements (e.g., conventions, fairs) organized by third parties. For our producers, this phase typically takes place at farmers' market or during collective workshops. This is principally a phase of introduction and recognition: producers get acquainted with either consumers or professionals unbeknown to them up to that point; through repeated, albeit brief, encounters and informal exchanges, producers are able to identify 'like-minded' actors (i.e., cognitive proximity) and establish privileged bonds with them (i.e., social proximity). The recognition phase triggers the introduction of novelties as it exposes producers to new knowledge from non-farming and/or extralocal actors. Acquiring such skills often requires directly liaising with a given individual and creating a social and cognitive contact interface making the transmission of tacit knowledge possible.
Torre (2009) labelled the second phase ubiquities. This phase consists in direct interactions operated 'at a distance' and undertaken 'on-the-fly' over a long period of time. Exchanges during this phase do not require face-toface interactions. The formation of more or less formalized groupings of actors is enabled by the adhesion to a common set of rules governing future interactions. The absence of face-to-face interactions thus induces a creative space for individual actors to reflect on the benefits, practicalities and extent of their engagement in such networks. The role of ICT, and more specifically its more recent evolutions (e.g., mobile broadband, social media), is critical to enable this stage. Here, our case study context, combining long physical distances and generally high level of ICT endowment in a globally technology-friendly society/economy, showed the added-value of these distant-but-direct interactions to consolidate these communities of practices. The development of more formalised organisational arrangements makes it easier to recruit new members because it brings more visibility and transparency.
The ubiquities phase aims to consolidate the foundation of social acquaintance and cognitive likeness initiated during the recognition phase. However, with repeated interactions, Fig. 1 Torre's evolutionary proximity model applied to the 'new associationalism' of quality food niches in the Swedish north exchanges become more routinised thus switching the coordination to other modes of organized proximity: from social and cognitive proximities, the interactions progressively rely on more or less formalized cooperation (organizational proximity) and tacitly agreed code of conduct (institutional proximity). The ubiquities phase is the stage during which the logic of belonging progressively enmeshes with the logic of similarity resulting from the initial recognition phase.
The third phase in Torre's model is of mobilities. It encompasses punctual meetings that are directly organized by the actors through mutual agreement and characterised by face-to-face, and often one-on-one, interactions (Torre 2009). Unlike encounters under the recognition phase, these meetings do not take place on neutral ground, and are often undertaken as study visits on one of the actor's home ground. Through direct observation, these occurrences allow the two parties to substantiate the values on which the relationship is embedded. These meetings are especially adapted for more in-depth exchanges of tacit knowledge, the transmission of know-how using a learning-by-doing modus operandi and the co-creation of shared experiences: they are the privileged sites for the constitution of translocal practices.

Conclusions
By investigating the new associations among food actors belonging to alternative food chains, the present study responds to Tregear and Cooper's (2016, p. 109) call for research to investigate "cooperative know-how […] to other types of rural collaborations beyond producer cooperatives". The evolutionary proximity model developed by Torre (2009) applied to our specific cases of learning networks in the context of geographically isolated and institutionally marginal producers informs us about how the adhesion to certain values facilitates the emergence of a new culture of collaboration (Aubrée et al. 2018). Although this culture of collaboration is firmly embedded in the territory, it is enacted by geographical distributed teams of 'like-minded' individuals and sustained by mutually agreeing to common sets of rules and behaviours. These new associations are vectors of the creation of translocal practices embedded in the farming practices of multiple locales.
The transition to food sustainability cannot afford to leave aside untapped agricultural assets (Dubois and Carson 2019). Producers located at the 'periphery' still have the possibility to make a decisive contribution to achieving this overarching goal locally. This study illustrated the diversity of new practices that emerge from the establishment of short quality food chains in those places and showed how new associations indeed have the potential to realize Hinrichs (2014, p. 148)'s vision of niche formation as "making mere innovative ideas feasible, perhaps even achievable opportunities".