Geographical Indication Building Process for Sharr Cheese (Kosovo): “Inside Insights” on Sustainability Geographical Indication Building Process for Sharr Cheese (Kosovo): “Inside Insights” on Sustainability Geographical Indication Building Process for Sharr Cheese (Kosovo): “Inside Insights” on Sustainability

: This article aims to contribute to the reﬂection on sustainability in the ﬁeld of Geographical Indications (GI). GIs are instruments for organizing collective action that have great interpretative ﬂexibility. They are mobilized by a set of qualifying actors of differing natures, with diverse and sometimes divergent interests. For this reason, we focus on how the dimension of sustainability emerges from a collective learning process. Based on the approaches developed by Organization Studies , this article describes and analyzes the process of creating a GI for Sharr Cheese, a Balkan seasonal sheep pastoral cheese highly typical of a mountain range in Kosovo Abstract: This article aims to contribute to the reflection on sustainability in the field of Geographical Indications (GI). GIs are instruments for organizing collective action that have great interpretative flexibility. They are mobilized by a set of qualifying actors of differing natures, with diverse and sometimes divergent interests. For this reason, we focus on how the dimension of sustainability emerges from a collective learning process. Based on the approaches developed by Organization Studies , this article describes and analyzes the process of creating a GI for Sharr Cheese, a Balkan seasonal sheep pastoral cheese highly typical of a mountain range in Kosovo  (  This designation is without prejudice to positions on status, and is in line with UNSCR 1244/1999 and the ICJ Opinion on the Kosovo declaration of independence.). The authors occupied an embedded research position in this learning process, from 2015 to 2019. The article describes boundary work carried out by the facilitators of collective action (brokers) within experimental spaces during the GI-building process. It analyzes how environmental accountability within the Sharr Cheese GI emerges from a strategic knowledge-brokering process and intensive institutional work. Abstract: This article aims to contribute to the reflection on sustainability in the field of Geographical Indications (GI). GIs are instruments for organizing collective action that have great interpretative flexibility. They are mobilized by a set of qualifying actors of differing natures, with diverse and sometimes divergent interests. For this reason, we focus on how the dimension of sustainability emerges from a collective learning process. Based on the approaches developed by Organization Studies , this article describes and analyzes the process of creating a GI for Sharr Cheese, a Balkan seasonal sheep pastoral cheese highly typical of a mountain range in Kosovo  (  This designation is without prejudice to positions on status, and is in line with UNSCR 1244/1999 and the ICJ Opinion on the Kosovo declaration of independence.). The authors occupied an embedded research position in this learning process, from 2015 to 2019. The article describes boundary work carried out by the facilitators of collective action (brokers) within experimental spaces during the GI-building process. It analyzes how environmental accountability within the Sharr Cheese GI emerges from a strategic knowledge-brokering process and intensive institutional work. This designation is without prejudice to positions on status, and is in line with UNSCR 1244/1999 and the ICJ Opinion on the Kosovo declaration of independence.). The authors occupied an embedded research position in this learning process, from 2015 to 2019. The article describes boundary work carried out by the facilitators of collective action (brokers) within experimental spaces during the GI-building process. It analyzes how environmental accountability within the Sharr Cheese GI emerges from a strategic knowledge-brokering process and intensive institutional work.


Introduction
A Geographical Indication (GI) is a sign of recognition of specific characteristics, reputation, or quality of a product, attributable to its geographical origin. This recognition takes the form of a collective intellectual property right, granting legal protection of the product name in markets. From a consumer point of view, a GI guarantees a given quality and a clear origin. In other words, "GI protection systems are legal regimes which facilitate the signaling of this provenance in marketplaces" [1]. They are regulated under the multilateral commercial regime of the World Intellectual Property Organization (WIPO) under Art. 22(1) of the TRIPS Agreement (1995), or by the World Intellectual Property Organization (WIPO) in accordance with the Lisbon Agreement (1958) and the Geneva Act (2015), establishing an international registration system for GIs. They are also defined within commercial bilateral trade agreements. In the European Union (EU), GIs have since 1992 enjoyed a high level of protection guaranteed by a sui generis GI system.

Introduction
A Geographical Indication (GI) is a sign of recognition of specific characteristics, reputation, or quality of a product, attributable to its geographical origin. This recognition takes the form of a collective intellectual property right, granting legal protection of the product name in markets. From a consumer point of view, a GI guarantees a given quality and a clear origin. In other words, "GI protection systems are legal regimes which facilitate the signaling of this provenance in marketplaces" [1]. They are regulated under the multilateral commercial regime of the World Intellectual Property Organization (WIPO) under Art. 22(1) of the TRIPS Agreement (1995), or by the World Intellectual Property Organization (WIPO) in accordance with the Lisbon Agreement (1958) and the Geneva Act (2015), establishing an international registration system for GIs. They are also defined within commercial bilateral trade agreements. In the European Union (EU), GIs have since 1992 enjoyed a high level of protection guaranteed by a sui generis GI system.

Introduction
A Geographical Indication (GI) is a sign of recognition of specific characteristics, reputation, or quality of a product, attributable to its geographical origin. This recognition takes the form of a collective intellectual property right, granting legal protection of the product name in markets. From a consumer point of view, a GI guarantees a given quality and a clear origin. In other words, "GI protection systems are legal regimes which facilitate the signaling of this provenance in marketplaces" [1]. They are regulated under the multilateral commercial regime of the World Intellectual Property Organization (WIPO) under Art. 22 (1) of the TRIPS Agreement (1995), or by the World Intellectual Property Organization (WIPO) in accordance with the Lisbon Agreement (1958) and the Geneva Act (2015), establishing an international registration system for GIs. They are also defined within commercial bilateral trade agreements. In the European Union (EU), GIs have since 1992 enjoyed a high level of protection guaranteed by a sui generis GI system. But this market instrument, created in Europe at the end of the 19th century, is no longer reserved for the commercial sphere. GIs are gradually being used to pursue multiple Sustainability 2021, 13, 5696 2 of 23 objectives. They are becoming recognized as a tool for rural development, supporting localized agri-food systems and family agriculture in a context of globalized food markets [2]. In the 90s, GIs were seen as a means to protect not only biodiversity but also indigenous or local knowledge that was under threat, thereby promoting the associated traditional and cultural heritage [3,4]. Nowadays, GIs are also widely used in the international aid/development sector, as multifaceted instruments responding to the different dimensions of sustainable development (economic, social, and environmental), particularly in rural areas. These wider expectations of GIs and the multiple objectives addressed by this single instrument have attracted widespread critical analysis. Some argue that the primary objective of labels of origin is to increase the economic value of a local product, seeking to sustain its development and to maintain a community and its way of life. This view relegates environmental and social dimensions to second-order objectives [5][6][7].
Flexibility in how GIs are interpreted [8,9] is allowed by a generic infrastructure [10] that is part of an international legal-institutional framework [11], albeit evolving and contested [12]. To qualify this global order, we can therefore mobilize the notion of regime in the sense of International Political Economy, citing the canonical definition of it proposed by S.D. Krasner: a set of «implicit or explicit principles, norms, rules and decision-making procedures around which actors' expectations converge in a given area of international relations» [13]. From this analytical perspective, GIs come under the nested regime [14] of intellectual property rights (IPRs), itself defined under a wider global trade regime. However, the implementation and the actual or potential functions of an instrument are not all governed by the regime to which it belongs [15]. GIs are used in different contexts and meet multiple objectives that go beyond the strict IPR regime, and that can be related to other international regimes (such as biodiversity or development regimes, etc.). In this study, the generic infrastructure ensuring GIs interpretative flexibility is based on the European sui generis protection and registration system, institutionalized by law, in which the product specifications function as a boundary object, a locus of multiple translations [16,17]. Under this framework, GIs bring together a variety of qualifying actors who mobilize differing types of knowledge and defend various and sometimes divergent interests: producers and their unions or collective organizations, technical ministries (agriculture, industry, or commerce), chambers of commerce and industry, customs, commercial courts, providers of official development aid, agronomists, veterinarians, local authorities, managers of protected natural areas, the production and export sectors, etc. Our paper highlights this fundamental characteristic of GIs (i.e., interpretative flexibility versus generic infrastructure) and posit as a methodological consequence that sustainability performance of GIs would gain to be analyzed from "inside", referring both to the collective action they emerge from and the agro-ecosystem in which they are rooted. GIs should be assessed in terms of how well the set of practices, rules, and knowledge they help to stabilize can define and account for specific sustainability dimensions, linked to a given product and its territory of origin. Here, the notion of territory embraces the definition of agro-ecosystem; i.e., "a spatially and functionally coherent unit of agricultural activity, and includes the living and nonliving components involved in that unit as well as their interactions" [18,19].
Defined as such, GIs are no longer considered as a single instrument replicable in various contexts. They are considered as a specific device (dispositif ), as "a heterogeneous arrangement of actors, of knowledge, which serve as cognitive and coordination supports for actors engaged in a collective project" [20]. We submit that GIs are able to contribute to sustainability goals when they can provide an endogenous definition of their sustainability accountability and clear mechanisms of control. An essential point of analysis is therefore the process through which sustainability goals are developed during GI-building. How are different strategic framings of a collective action process aligned in such a way that the whole device is made accountable for its sustainably performance? The answer can come from closer consideration of the design and the results of the institutional work-consisting of intentional and strategic activities [21]-that occur in experimental spaces [22] during a GI-building process. By experimental space, we mean transitional arrangements or settings where actors from different organizational fields conduct new experiments or test alternative models [23]. In particular, during the GI-building process, successive experimental spaces are opened to test, design, and create the main elements of a GI generic infrastructure (e.g., the product specifications or the design of a control plan) before they are validated and implemented in the field of origin of the different actors (e.g., farmers, producers, processors, sellers, retailers, IP authorities, etc.). This analytical lens means analyzing the boundary work prevalent in the creation of such experimental spaces during the GI-building process. By boundary work, we refer to institutional work aimed at creating, maintaining, or disrupting practices of actors from different institutional fields or established organizational routines [21]. This reflexive form of action aims at intentionally affecting institutions, through brokering activities. Therefore, we will focus on knowledge and organizational/innovation brokers [24][25][26] and their strategic activities and practices [27]. Our purpose is to make explicit and describe the whole emergence process-made of cognitive and organizational arrangements-that actually define and account for environmental sustainability of a given GI device.
We propose to illustrate this analytical/methodological proposition with the GIbuilding process of "Sharr Cheese", a traditional cheese produced in the Sharr massif in the South of Kosovo, an EU candidate country in the Western Balkans. "Sharr Cheese" was historically used to designate an indigenous seasonal cheese traditionally produced from unpasteurized sheep's milk from Sharri pastures, a biodiversity hotspot of regional importance. However, due to both recent changes in farming systems in the Sharr massif and the consolidation of the dairy sector through semi-industrial milk-processing units, "Sharr Cheese" is now being used generically to designate any hard cheese from Kosovo, regardless of the milk's origin, its nature (sheep, bovine, or mixed), production practices, or seasonality. The aim of our account is to describe, from our experience of this specific GI device, how environmental accountability could possibly emerge from boundary work during the GI-building process.
Taking an embedded-research approach presented in Section 2, we explain how we used our experience as a knowledge broker to build our research results. We detail how such experience dedicated to defining and accounting for environmental suitability in the Sharr Cheese GI-building process could be considered as an Inquiry in the Deweyan sense of this term, and we present the methodological consequences. In Section 3, we propose a narrative and descriptive recount of this inquiry, based on our brokering activities and boundary work carried out within experimental spaces. We describe how the environmental accountability of Sharr Cheese GI emerged from a strategic knowledge-brokering process and intensive institutional work. In the last section, we discuss our results in the light of recent research on GI's capacity to deliver on sustainable goals. From an analytical viewpoint, we stress the importance of articulating a microlevel analysis of strategic activities with a macrolevel understanding of the arrangements at play within the international regime, when considering environmental sustainability issues.

Embedded-Research Approach and Pragmatic Inquiry
The authors of this article were, at various times and in various ways, involved in the construction of "Sharr Cheese" GI in Kosovo. As we undertook an explicit research role within the GI-building process, with the purpose of contributing to the collective action from an sustainable agro-environmental approach, we adopted an embeddedresearch stance. Embedded researchers are not in a position of exteriority; they are not colocated, as an ethnographer studying the context can be. Embedded researchers locate themselves alongside the other actors, as part of the context. Thus, embedded research involves an element of knowledge coproduction sensitive to the context in which it is used. Researchers are practically and materially involved in a collective action. This type of engaged research applies multiple methods and is the subject of active debate on its ethical, epistemological, and practical implications [28][29][30][31][32]. This can take several contractual and practical forms, because embedded research arrangements tend to be complex in nature.
However, what fundamentally characterizes embedded research is that the researcher assumes double accountability: for the collective action in which they participate and for the knowledge production organized and exploited in the scientific programs to which they contribute. This dual accountability differentiates embedded research from expertise or "critical friendship", which can at times seem like catch-all lucrative consultancy deals between organizations and "producers of knowledge" [33].
The researcher (or group of researchers) then becomes a stakeholder in the process, making explicit their strategic intention through mediation activities and taking a knowledge-broker position. The researcher is thus an intentional and strategic actor. "In this sense the analogy to embedded journalists, is actually quite a useful one in that we believe research to be a political act we undertake to disrupt, de-stabilize and question the validity of assumptions that inform social mechanisms within our society, research is neither neutral nor objective" (ibid.). In other words, the embedded researcher is included in the strategic device from which they are experimenting and producing knowledge at the same time. From an epistemological stance, this posture can therefore legitimately claim to be an Inquiry in the sense of the pragmatism of John Dewey [34]. Indeed, the essential process of the Inquiry lies in the transformation of a given situation qualified as unsettled or indeterminate in a problematic situation [34], resolved through the progressive determination of a problem-solution, in which finding out the constituents of a given situation (description and narrative) plays a crucial and central role, to progressively propose ideas or conceptual subject-matter. As a result, "Observed facts of the case and the ideational contents expressed in ideas are related to each other, as, respectively, a clarification of the problem involved and the proposal of some possible solution: that they are, accordingly, functional divisions in the work of inquiry" (ibid, p. 178). In our study, this Inquiry on GI sustainability is thus methodologically carried out through an embedded research position, where experimenting the problematic situation and seeking for solutions (i.e., producing ideational contents) are emerging from boundary work and brokerage activities.

Our Embedded-Research Position
Our embedded-research position in the process of building a GI on Sharr Cheese in Kosovo took the form of brokerage activities of different kinds over a four-year period, from 2015 to 2019. These interventions, coordinated over time, were part of the same inquiry: how to build environmental accountability of Sharr Cheese during the GI-building process developing both at the local and national levels? These brokerage activities undertook various forms of interactions, involving, through their funding and operational formats, different scales of involvement: research-action activities led under the BiodivBalkans project aimed at creating GIs in the Balkans as a conservation and production of biodiversity tool (2012-2017), financed by the French Facility for Global Environment (FFEM); field studies carried out through a collective field study internship (AgroParisTech); or dedicated agro-environmental studies funded by Erasmus field study grants (2017); multiactor networking activities, supported by the French technical agricultural cooperation (GIP ADECIA) for the regional promotion of GIs; technical assistances activities, led under ad hoc European Technical Assistance and Information Exchange (TAIEX) missions on the legal, technical, and organizational aspects of the GI, carried out between 2015 and 2018; under missions in 2019 on the constitution of the Book of Specifications, funded by the European Commission's Directorate General for Development and Cooperation (DG DEVCO-EuropeAid) under a project to strengthen commercial skills in Kosovo; advocacy work and territorial animation activities led by a French NGO active in the Western Balkans (AIDA-International Association for the Development of Agri-environment), supported by several grants from the French Embassy in Kosovo. All these activities were recorded and are briefly described in Appendix A. It is therefore important to note that this inquiry did not take the shape of a single project, but resulted in a coordination of brokerage actions-an original feature of our embedded research position. This characteristic is Sustainability 2021, 13, 5696 5 of 23 particularly important given the long-term nature of a collective action aimed at building a GI.

Method
We describe the GI-building process for Sharr Cheese in Kosovo using a situated narrative and description covering the period 2015-2019. Through the description and narration of this GI-building situation of Sharr Cheese in Kosovo, we make explicit the transformation of an unsettled situation into a problematic one.
From our experience as knowledge brokers, we recount the strategic dimension of this inquiry on building environmental accountability in a GI on Sharr Cheese. Thus, our situated narrative is based on all the written records of our boundary work. During our embedded research, we recorded and organized in chronological order all the materials from our brokerage activities: minutes of meetings, workshop agendas, attendance lists, technical reports, leaflets, and e-mails. This technique is derived from the observation methods used in socio-technical networks developed by sociology of science and technology (SST) scholars [35]. It provides a processual account of translation processes and the effects of enrollments, or, in other words, the deployment, the extension, and the ramifications of a given device related to the boundary work carried out at its margins, in experimental spaces. This narration elucidates both the diachronic dimension of the strategic activity of anchoring an environmental objective within a GI device and the contingent nature of any arrangements of actors, knowledge, norms, and instruments in a given situation.
Finally, our situated narrative reports on the knowledge content of this boundary work, and how it shaped a collective learning process, stabilized arrangements, and opened up new experimental spaces, throughout the GI-building process. Therefore, we construct a situated and processual description of what is usually considered a context or an empirical setting. This situated description builds on objectified knowledge intentionally produced during the GI-building process through our boundary work, to create a common understanding within the experimental spaces, particularly on environmental issues related to Sharr Cheese production systems [36]. The descriptions were developed from qualitative and quantitative data, both primary and secondary, collected during several periods of fieldwork carried out between 2016 and 2019. A first territorial diagnosis, centered on the strategic analysis of environmental management (SEMA) [37,38], identified the main crosscutting issues between cheese production practices and the value chain with territorial and biodiversity management in the Sharr mountains [39]. Using an agrarian diagnostic approach enabled us to characterize the coexistence of different livestock production systems in the region and their contribution to the existing landscapes [40]. It also afforded us a better understanding of the farming systems involved with Sharr Cheese production and how their technical and economic performance might be impacted by the creation of a GI [41,42].

Phase 1: Opening Configuration-2015/2016
The choice of Sharr cheese as Kosovo's first GI pilot product was based on intensive boundary work at different organizational scales, made possible by the strategic convergence of different brokering activities, as well as the explicit design of a series of experimental spaces. We detail here the brokering activities directed at overcoming the difficulties of building a GI for a cheese produced by different communities in a partitioned and mountainous territory like the Sharr Mountains. Added to these difficulties was the challenge of redefining the identity of a product whose name was on the verge of generic application to a whole range of hard cheeses in Kosovo. Behind this generic use of the name "Sharr Cheese", there was also the challenge of preserving and promoting pastoral production practices and their related biodiversity as lying at the heart of seasonal cheese from the Sharr pastures. The process of creating a GI on Sharr cheese was initiated as part of Kosovo's attempt to approximate the EU acquis communautaire. The adoption of the legal and regulatory standards and rules shared by all EU Member States is a mandatory condition for entry into the EU. Kosovo, a candidate for EU membership, has been developing EU norms in the field of intellectual property rights since 2012 at its Intellectual Property Agency (IPA) under the umbrella of the Ministry of Trade and Industry (MoT). Rapid institutional progress led to a legal framework for GIs established in 2016 (Law No. 05/L-051 on Geographical Indications and Designations of Origin). This law is correlated with the entry into force of the Stabilization and Association Agreement (SAA) on 1 April 2016, mandating gradual abolition of customs duties on agricultural and food products between the EU and Kosovo, as a first step toward EU accession. Sharr Cheese was therefore listed among the potential GI pilot products. Sharr Cheese producers were engaged in a GI-building process aligning three different framings: (1) an institutional/national framing aimed at embodying the newly created legal IP framework in a GI product; (2) an environmental/regional framing identifying GIs as a potential instrument for enhancing and conserving agricultural lands of high nature value (HNV farming areas), linked to agro-silvo-pastoral farming systems [43]; and (3) a sectorial/local framing involving a dynamic Kosovan dairy value chain in the Prizren area (Southern Kosovo) with good contacts at national level, which identified GI as a potential source of government support. However, this initial situation fell far short of providing the conditions required for effective territorial collective GI-building.

Engaging Actors from Different Organizational Fields
Actors from different organizational fields were engaged in a learning process at different scales through the creation of a series of variously designed experimental spaces, based on GI-specific knowledge production. The first experimental space was shaped in such a way as to integrate the nascent Kosovan GI-building process into a wider regional momentum, making it clear that GIs could protect and promote agricultural products from high nature value farming systems throughout the Western Balkan countries. This boundary activity was inspired by European-French expertise on GIs and fueled by agrienvironmental mediation work being carried out under a biodiversity conservation project in the Balkans. Field diagnosis [39] was used to establish the link between the product, its origin, and its agri-environmental positive externalities. Things gradually took shape with the organization in Prizren of a "Meeting of GI practitioners in the countries of the Western Balkans" near the Sharr massif region in November 2016. For two days, technical ministries responsible for GIs, researchers, and practitioners (producer-processors) involved in GI-building for products from high nature value farming areas throughout the Balkans explained their approaches and exchanged their experiences: Sharr and Pljevalski cheeses in Montenegro, kid goat meat from Has in Albania, and Bistra Kashkaval cheese in North Macedonia. These exchanges were facilitated by a standardized format covering the basics of constructing a GI, as well as pointing out their positive environmental value. This created a common basis for understanding the instrument and the challenges of reconciling the European horizon of harmonization of trade and intellectual property rules with the national/local horizon of agricultural and rural development of mountain areas-often disconnected from these very areas' biodiversity conservation objectives. This engagement of actors from different organizational fields was bolstered by isolated experimental spaces opened throughout 2015 and 2016. A series of technical support missions (TAIEX) carried out by a French operator, the National Institute of Origin and Quality (INAO), with help from the French Embassy in Kosovo, appeared to be game-changing in the GI-building process. Unlike the usual technical workshops run by "authorities on GI", the participative design of their workshop resulted in greater involvement of Sharr Cheese producers and processors, as well as territorial actors from the Sharr Mountains (such as the Sharr National Park and Dragash Municipality) in the GI-building process. During this two-day workshop, in addition to presentations by national authorities demonstrating to the producers their strong commitment and support, sensory experiences (e.g., organoleptic test) and role-playing based on GI examples from France proved instrumental in building trust between actors from different organizational fields. This encouraged participation in the interactions, with the producers spontaneously asking questions and sharing thoughts, ideas, and fears. Enabled to experience the GI's generic infrastructure through role-playing, participants obtained a clearer picture of its potential benefits in terms of improvement of producers' incomes and preservation of the territory's specific ecological and sociocultural features. This approach also reduced the producers' strong resistance to collective action, which was anchored in a still vivid memory of the farming collectivization process, leading to mistrust of any collective organization structured by the State. It reassured participants that GI collective organization only requires producers to delegate certain responsibilities to the collective management, while all members maintain their financial autonomy and remain owners of their production means. An open discussion allowed producers to share, for the first time, their vision of the product and its specific territorial features. Finally, producers expressed their interest in creating an association to monitor work on the GI for Sharr Cheese.

Maintaining Experimental Spaces Over Time: Boundary Work through Learning Processes
However, GI-building is a long and iterative process. Maintaining an alive and dynamic territorial collective action created the space and the need for a twofold learning process, an instance of how GI experimentation spaces are marked out over time: (a) An organizational learning process requiring field actors to distance themselves from their organizational and sectorial routines in order to set up a "territorial" device, based on the GI generic infrastructure and its new manifestation in the Kosovan context. This situation encouraged the emergence of a Kosovar broker and mediator (organizational broker). The NGO KsIIP (Kosovo Institute for Intellectual Property), working on the articulation between the administrative and political sphere and European cooperation in the field of intellectual property, proved to be instrumental in maintaining experimental spaces over time. It linked the national-level creation of the GI framework with the local and territorial drive to build a Sharr Cheese GI pilot. (b) A cognitive learning process led by our research coalition (knowledge broker), producing ad hoc elements of knowledge to inform the choices involved in defining Sharr Cheese. This raised opportunities for collective subjectivation, and the emergence of a collective subject, among the producers of the Sharr massif.
These two learning processes were particularly important because Sharr Cheese was on the verge of becoming a generic name for all hard cheese in Kosovo: it did not refer to a clearly established territory of production, nor to a given community of practice. In fact, little attempt had been made to define Sharr Cheese, and there were no standardized procedures for the production of Sharr mountain hard cheese. Although cheese-making techniques in different parts of the Sharr Mountains and among different communities vary only slightly, the numerous producers needed to be made aware of this lack of definition, as a common ground for building the product specifications [44].

Phase 2: Learning Configuration and Strategic Knowledge Brokering-2017/2018
We then began a long process of collecting and creating shared knowledge about Sharr Cheese, determining the special features linking it to the territory where it is produced. The knowledge yielded acted as a mirror of collective practices, raising awareness among producers both of their diversity (differing production systems linked to the fragmentation of agri-ecosystems in the Sharr massif) and of what they shared. We describe here how this objectivation-subjectivation process was also a strategic space for knowledge brokers like us to anchor Sharr Cheese specifications within the definition of a resilient mountain social-ecological system. Key to this objectifying was going beyond the description of a fragmented mountainous territory to highlight the biodiversity of the Sharr massif as a common heritage maintained through pastoral practices. The name "Sharr Cheese" itself refers to the pastoral massif of Sharr (or Šar Planina/Malësia I Sharrit), which extends over the entire southern part of Kosovo, including 80 km of borders with Albania and Macedonia (Figure 1a). Stretching for nearly 800 km 2 , its altitude varies between 300 and 2600 m, and more than half of the area is located above 1500 m. The Sharr massif has a predominantly alpine and continental climate. The region is extremely rich in endemic, rare, and threatened flora and fauna, ranking it among the richest Balkan and European regions in terms of biodiversity. Of its roughly 2500 species of flora, 323 are endemic plants, which makes this territory unique in the Balkans. Alpine or subalpine pastures support some of the most species-rich habitats in Europe, including a high proportion of endemic species and rare glacial relicts [45]. In addition, several protected species that are rare in other mountain ranges in Europe still live in the Sharr mountains, such as the brown bear, lynx, wolf, and gray eagle [46]. This biodiversity was recognized and protected with the creation in 1986 of the Sharr National Park (SNP), the borders of which were extended in 2012 and which today covers an area of 53,469 ha, or 1/3 of the Kosovar Sharr massif [47]. This fragmented mountainous territory can be divided into four distinct sub-regions characterized by distinctive agri-ecological settings, as well as different territorial development dynamics (Figure 1b). The pedoclimatic characteristics of the Sharr massif and its relative conservation over time also make it an exceptional area in terms of biodiversity and alpine landscapes (estimated at 28,000 ha) (Figure 1c), linked to agropastoral activity [48].  This synergy between pastoral practices and Sharr mountain biodiversity therefore needed to be related to the long-term evolution of agrarian systems. Undergrazing and abandonment of summer pasture follow a long-term trend toward transformation of agrarian systems. A turning point was the marked agricultural abandonment that began in the 1970s under the Yugoslav regime. At that time, due to the country's industrialization, sheep farming was declining and was concentrated in Socially Owned Enterprises (SOE) like the SOE Progress (created in 1953), with a sheep herd of several thousand heads wintering in the Prizren plain and summering in the high pastures of the Sharr. Then, in 1962, the SOE Sharprodimi concentrated all sheep breeding activity in the municipality of The positive synergy between agropastoral production systems based on meadows and high pasture biodiversity is common to all Sharr massif subunits. It was therefore a natural starting point to explore and describe the synergies between the Sharr pastures' high biodiversity and agropastoral activity within a common socio-ecological system. Pastoral livestock systems use different agro-ecological layers according to the altitude and the fertility of the soils throughout the year. The use of these layers is governed by a set of agreements of different natures: the use of high-altitude plains, or glacial fluvial terraces, including the grazing of arable land and permanent meadows, is framed by rules on grazing on common land and defenses; the medium limestone or crystalline brunipodzol mountains, rendzinas, and rankers, ranging from 800 to 1800 m, are managed as village commons (utrina), allowing local grazing all year round except during the snowcovered periods; and the high mountain pastures (bjeshk), more than 1800 m above sea level on lithomorphic soils, are grazed in summer, from May to September, subject to annual grazing permits.
However, this positive environmental synergy is at risk. Over the past decades, several thousand hectares of pasture have been lost through afforestation and scrubbing due to undergrazing [48]. Although no specific indicators have been developed in Kosovo (i.e., there is no monitoring of biological diversity priority components for meadows and high pastures), modification of pastoral open landscapes through successions and changes in the composition of biological diversity is widespread in Europe. Both a decline in the populations of large numbers of species on a continental level, and its impact, are well documented [49][50][51][52].
This synergy between pastoral practices and Sharr mountain biodiversity therefore needed to be related to the long-term evolution of agrarian systems. Undergrazing and abandonment of summer pasture follow a long-term trend toward transformation of agrarian systems. A turning point was the marked agricultural abandonment that began in the 1970s under the Yugoslav regime. At that time, due to the country's industrialization, sheep farming was declining and was concentrated in Socially Owned Enterprises (SOE) like the SOE Progress (created in 1953), with a sheep herd of several thousand heads wintering in the Prizren plain and summering in the high pastures of the Sharr. Then, in 1962, the SOE Sharprodimi concentrated all sheep breeding activity in the municipality of Dragash and managed a sheep herd of between 5000 and 12,000 heads, depending on the period (and absorbed the SOE Progress flock). The economic crisis of the 1980s, the dissolution of Yugoslavia, and the end of the self-management model, as well as massive emigration linked to the conflicts of the late 1990s, completely disrupted livestock activities in the Sharr, as in most of the region. Livestock disappeared during the war in 1999. Postwar reconstruction programs favored nonpastoral cattle-rearing systems capable of providing a stable annual income, using high-productivity dairy breeds. These farms were also supported by an agricultural policy favoring modernization-mechanization of production systems (investment aid), as well as a system of direct aid strongly promoting cattle for their meat and, since 2012, for milk production.
Objectifying these agrarian system dynamics, combined with analyzing the technicaleconomic dynamics of recent developments in Sharr mountain breeding systems, was instrumental in explaining the coexistence of different types of Sharr cheeses and the issues involved in the product specifications. It provided a clear ranking of the different hard cheeses using the name "Sharr Cheese", according to their contribution to maintaining the biodiversity of the Sharr pastoral open landscape. Today, the Sharr massif accounts for nearly a quarter of the livestock sector in Kosovo and the current trend is therefore toward coexistence of sheep and cattle farming systems, or even combined facilities for certain livestock production systems (Table 1). In the four municipalities of the Sharr territory, there are nearly 43,000 sheep and 20,000 cattle [39]. Eleven different production systems (see Table 1) are distinguished according to four criteria: herd composition (bovine, ovine, or mixed); production specialization (milk or meat); flock size; and herd management (transhumant, pastoralist, or stabling) [42]. The diversity of these production systems explains the coexistence of different types of Sharr cheeses, in the following increasing order of contribution to biodiversity: • Dairy Sharr cheese (semi-industrial) made from cow's milk, annual production, mainly from bovine milk (C) and small mixed bovine (J) systems.
• Farm cow milk Sharr cheese, seasonal, from transhumant cattle breeding systems (A) and mixed cattle systems with or without sheep facilities (G) as well as selfconsumption cattle systems (I; K). Thus, from an environmental perspective, the optimum specifications for Sharr Cheese would define a pastoral, seasonal product of breeding systems using the full range of pastoral open landscape (meadows and alpine grasslands) present at different levels of vegetation.

Intensive Enrollment Process and Conflictual Choices: Experimental Spaces at Risk
Building the specification for Sharr Cheese involved more than simply transcribing and codifying traditional know-how within a GI framework guaranteed by the State. This process of defining a product and its features linked to a specific place of origin was more like a postmodern recomposition of a narrative representing a collective subject. There were choices to be made.
Our objective account was therefore presented to the producers-who had contributed to our diagnostic process through extended interviews-for criticism, adjustment, and validation. This reporting-back phase involved significant extension of the experimental spaces over 2017, in each of the subregions of the Sharr massif area, as well as in Prizren and Prishtina, with the State agencies and technical ministries responsible for the GI regulatory and institutional framework. Our reports covered both the history and the dynamics of the agrarian-pastoral system; the different production systems; the challenges, difficulties, and opportunities of the territory; and the possibility of creating a GI and its potential benefits and consequences. They constructed a stabilized and objective description of the coexistence of livestock production systems in the Sharr massif, their interactions with agri-ecosystems, and the dynamics at work. This descriptive basis was collectively validated, while enabling different actors to express different positions of interest in defining the product, the breeding practices (especially pastoral), the appellation area, and the processing of the product.
In these experimental spaces on the Sharr Cheese specifications, several factors were found to be in conflict. (1) The territory. The first proposals for the definition of the GI territory led to two opposing views: an exclusive definition, with only regions mainly composed of pastoral areas and mountain pastures (Gora, or Gora + Opoja), considered as the traditional heart of the production of Sharr Cheese; and an inclusive definition, which also included all four subregions of the Sharr massif, and part of the Prizren plain, home to the large wintering transhumants (H), as well as a large group of bovine production systems and their semi-industrial processing units. (2) The product. A broad definition of Sharr Cheese (mixed bovine/ovine, annual, dairy or farm cheese) was widely supported by all technical ministries and funding bodies as beneficial to overall livestock activities. The sectors (mainly bovine-milk) organized around semi-industrial or artisanal dairies were also presented as a basis for collective organization capable of handling the GI's management and promotion. The exclusive definition of the product made it pastoral, seasonal, sheep only, and without pasteurization. This definition was mainly supported by the herders of Gora, but also by the large transhumant herders of the plain of Prizren.
(3) The communities. Even trickier was the creation of an intellectual property right to be shared by four communities that, in the recent past, had been engaged in violent conflicts. Bosnians, Gorans, Serbs (Slavic language speakers), and Albanians are spread over the entire Sharr massif: Serbian communities mainly in the Štrpce/Shtërpcë valley, Albanians around the Opoja plain, Gorans south of Dragash in the Gora, and Bosnians in the Zhupa. There is also a small Albanian community in the municipality of Štrpce/Shtërpcë and in Zhupa ( Figure 2). dairies were also presented as a basis for collective organization capable of handling the GI's management and promotion. The exclusive definition of the product made it pastoral, seasonal, sheep only, and without pasteurization. This definition was mainly supported by the herders of Gora, but also by the large transhumant herders of the plain of Prizren. (3) The communities. Even trickier was the creation of an intellectual property right to be shared by four communities that, in the recent past, had been engaged in violent conflicts. Bosnians, Gorans, Serbs (Slavic language speakers), and Albanians are spread over the entire Sharr massif: Serbian communities mainly in the Štrpce/Shtërpcë valley, Albanians around the Opoja plain, Gorans south of Dragash in the Gora, and Bosnians in the Zhupa. There is also a small Albanian community in the municipality of Štrpce/Shtërpcë and in Zhupa (Figure 2). Moreover, certain types of farming systems particularly suited to the geomorphological conditions of the Sharr subregions are also associated with a given community. Thus, a definition of product specifications based on farming practices was likely to arouse community resentment. Moreover, the nascent leadership of the Sharr Cheese Producers Association was clearly driven by Albanian producers from Prizren (dairies and large transhumant herders). Relations with Slavic-speaking communities were often difficult. The Gorans, isolated in the Dragash area, found it difficult to take part in the meetings and organizational activities involved in setting up the Sharr Cheese Producers Association. Bosnian breeders and producers themselves (not necessarily Albanian speakers) also had difficulty integrating into collective action. The most complex case was that of the Serbian producers located within the Štrpce/Shtërpcë enclave, which, like the other Serbian enclaves in Kosovo, maintains a parallel administration that reports to Belgrade (Serbia).

Phase 3: Distancing Work-2018/2019
Opening up choices and multiple definition-setting options, this process was potentially letting a "Sharr Cheese" collective perform its own boundary work. However, stakeholders in the GI-building process had difficulty distancing themselves from their original organizational field. Rigid and conflicting positions of interest linked to market position in the value chain, sector membership, or community identity prevented a more positive construction of product specifications. To defuse this tense situation, boundary work continued and intensified through close coordination between the NGO AIDA (follow-up on the environmental activity of the BiodivBalkans project), INAO, and the French Embassy, in conjunction with KsIIP and the Sharr Cheese Producers Association. Moreover, certain types of farming systems particularly suited to the geomorphological conditions of the Sharr subregions are also associated with a given community. Thus, a definition of product specifications based on farming practices was likely to arouse community resentment. Moreover, the nascent leadership of the Sharr Cheese Producers Association was clearly driven by Albanian producers from Prizren (dairies and large transhumant herders). Relations with Slavic-speaking communities were often difficult. The Gorans, isolated in the Dragash area, found it difficult to take part in the meetings and organizational activities involved in setting up the Sharr Cheese Producers Association. Bosnian breeders and producers themselves (not necessarily Albanian speakers) also had difficulty integrating into collective action. The most complex case was that of the Serbian producers located within the Štrpce/Shtërpcë enclave, which, like the other Serbian enclaves in Kosovo, maintains a parallel administration that reports to Belgrade (Serbia).

Phase 3: Distancing Work-2018/2019
Opening up choices and multiple definition-setting options, this process was potentially letting a "Sharr Cheese" collective perform its own boundary work. However, stakeholders in the GI-building process had difficulty distancing themselves from their original organizational field. Rigid and conflicting positions of interest linked to market position in the value chain, sector membership, or community identity prevented a more positive construction of product specifications. To defuse this tense situation, boundary work continued and intensified through close coordination between the NGO AIDA (follow-up on the environmental activity of the BiodivBalkans project), INAO, and the French Embassy, in conjunction with KsIIP and the Sharr Cheese Producers Association.

Distancing Work
The first attempt at distancing consisted of maintaining an open deliberative process, especially over the product specifications. A bilingual technician-facilitator for the Association was hired to maintain an inclusive and open relationship with producers from different production systems and different communities (Albanians, Gorans, Bosnians, and Serbs). His knowledge of the languages spoken (Albanian and Slavic) and of the field, acquired during the six months of the investigation and the reporting-back sessions, seemed to us a major asset. The second means of distancing consisted in sticking to the national GI-building process. Thus, the Kosovar organizational broker (KsIIP) consistently worked toward formalizing the local dynamic of collective action, in particular by sup-porting the creation of the Sharr Cheese Producers Association and its official registration. At the same time, KsIIP worked in conjunction with the European cooperation sector to exploit European know-how in the protection of GIs, leading to the implementing decrees necessary to make a sui generis system functional.

Engaging Participants through Sensory Experience: Cheese Tasting (2018 and 2019)
The second aspect of our mediation was focused on the product specifications. Working from the characterization of the territory, its agri-ecological subsets, and the typology of production systems, we collected cheese samples representing the full range of production systems (more or less pastoralists and farmhouse), and also the diversity of the territories and communities present in the Sharr massif.
Two organoleptic tasting workshops were organized over two consecutive seasons of Sharr cheese production. These blind tastings were carried out in each of the four communities, following the establishment of a qualification grid for the taste qualities of the cheese in three languages (English, Albanian, Slavic) ( Figure 3) [53]. Sources: authors (a) [36] and (b) [53].

Phase 4: Anchoring Work: Building Environmental Accountability of a GI through Its Specifications
These experimental spaces thus enabled actors from different institutional and cognitive horizons to establish choices in conditions favorable to deliberation. The challenge now was to reconnect these areas of innovation to the organizational fields. GI generic infrastructure, of which the specifications and the control plan are two key elements, enables the agreements obtained in experimental spaces to be stabilized.

Sustainability's Technical Anchoring: Building the Book of Specifications
The recent Sharr Cheese Producers Association, created in March 2019, has kept this exclusive definition of the product, including only cheese from sheep farming systems with a strong pastoral component. Thus, Sharr Cheese is defined as a seasonal hard sheep's cheese produced between April and October from whole sheep's milk. Flocks must be composed exclusively of sheep belonging to a local breed (Sharplaninian Pramenka/Šarplaninska breed) and/or Merinos breed, and/or cross-bred from the above. A series of strict requirements relate to the management of the herds and their diet, which are key to the specific qualities of Sharr Cheese. This limits the source to farms with a strong pastoral component, which use all the resources of the different agro-ecological stages, depending on the season. Systems based on extensive supplies of fodder and concentrates are excluded, while summer grazing sheep farming systems are included (B; D; F; H). These criteria can be met in a large geographical area that includes the four subsets, but they require compliance with pastoral practices and a diet without supplements dur- For the participants, this was an unprecedented sensory experience taking place in a friendly atmosphere. Eating together and discussing how to define the taste of "anonymous cheeses" was a strong and founding experience. First, it opened the way to positive appraisal of cheeses produced by others. Second, it was an opportunity to recognize the existence of "fake" Sharr cheeses. Moreover, producers experiencing these tasting groups were in a much better position to participate in a collective action recognizing both the potential diversity of this designation of origin and its core identity-an identity worth defending through a GI. This work also raised the gender issue, albeit modestly, thanks to interviews with Albanian and Goran woman and the participation of Serbian women in the local product identification roundtable. Unanimously, participants in this work on organoleptic characteristics definitively ruled out cow's milk and all nonseasonal products from nonpastoral breeding systems (and therefore industrial production from cow's milk or mixed dairy products from the plain) (Table 1). Sharr Cheese, therefore, has now recovered its traditional definition of farm, seasonal, and pastoral cheese, and only sheep's milk can claim the designation.

Phase 4: Anchoring Work: Building Environmental Accountability of a GI through Its Specifications
These experimental spaces thus enabled actors from different institutional and cognitive horizons to establish choices in conditions favorable to deliberation. The challenge now was to reconnect these areas of innovation to the organizational fields. GI generic infrastructure, of which the specifications and the control plan are two key elements, enables the agreements obtained in experimental spaces to be stabilized.

Sustainability's Technical Anchoring: Building the Book of Specifications
The recent Sharr Cheese Producers Association, created in March 2019, has kept this exclusive definition of the product, including only cheese from sheep farming systems with a strong pastoral component. Thus, Sharr Cheese is defined as a seasonal hard sheep's cheese produced between April and October from whole sheep's milk. Flocks must be composed exclusively of sheep belonging to a local breed (Sharplaninian Pramenka/Šarplaninska breed) and/or Merinos breed, and/or cross-bred from the above. A series of strict requirements relate to the management of the herds and their diet, which are key to the specific qualities of Sharr Cheese. This limits the source to farms with a strong pastoral component, which use all the resources of the different agro-ecological stages, depending on the season. Systems based on extensive supplies of fodder and concentrates are excluded, while summer grazing sheep farming systems are included (B; D; F; H). These criteria can be met in a large geographical area that includes the four subsets, but they require compliance with pastoral practices and a diet without supplements during the lactation period. However, an effective control plan checking that all of these production criteria are respected and guaranteeing the product's special features and quality remains to be defined.

Sustainability's Institutional Anchoring: Building Synergies between IPR Framework and Other Rural, Environmental, and Agricultural Policies
However, at the national level, it is proving difficult to finalize a functional administrative infrastructure in line with the sui generis system of GI protection [54]. Collaboration between the Ministry of Agriculture, Forest and Rural Development (MoA) on the GI registration (file review commission) and accreditation procedure and the Food and Veterinary Agency (FVA) on the external control system for GIs, is taking much longer than expected [55], with competition among the ministries' technical prerogatives. This weak convergence of public action around the GI instrument is worsened by sectoralized public policies [56] and a corresponding compartmentalization of funding for technical assistance and external aid. All this limits consideration of environmental and territorial development issues beyond temporary learning and innovation areas. The synergy between existing public policy strategies and tools and the promotion of GI products (e.g., direct agricultural aid for pastoral farms, support for rural development (2nd pillar)) remains to be developed, as well as the shared governance that would include the Sharr National Park, which manages all the grazing permits on the public mountain pastures of the Sharr massif (i.e., about 70%) and monitors the biodiversity of its different ecosystems.

Sustainability's Market Anchoring: Reinforcing Value Chains via Origin-Based Promotion
A third anchoring dimension of the Sharr Cheese GI relates to market positioning. The exclusive definition of Sharr Cheese ruled out hard cheeses produced by semi-industrial dairies. These cheeses are mainly produced from pasteurized cow's milk at low cost for supermarkets and grocery stores (from EUR 2.5/kg consumer price), as well as for part of the catering sector. Their supply radius is much wider than the Sharr massif, and relies on a network of collection points also covering products from the agricultural plains of Dukajin and Metohija. Production (traceability, hygiene) is standardized in response to the stated preferences of a section of young, urban consumers who are sensitive to issues of food product traceability and hygiene (food safety and quality) [57]. For health reasons, they also favor a less fatty and less salty bovine cheese than the traditional Sharr Cheese. The packaging of cheese in a plastic jar, ready to eat, is also popular with smaller "urban lifestyle" households spending less time on meal preparation [39].
The choices made in defining the Sharr Cheese GI were thus aimed at enhancing the value of a traditional product whose reputation is linked to its pastoral origin and is clearly identified by consumers in Kosovo and beyond (diaspora, and regional con-sumers). However, because the definition includes cheeses produced by dairies subject to the sanitary standards in force in the dairy sector, the labelling can distinguish between (1) "farmhouse"/"domestic product", and (2) "made in bacilo" or "made in stan" for which pasteurization is prohibited. By formalizing features linked to the pastoral and Kosovar origin of the cheese, this marketing strategy also relies on the "patriotism regarding domestic milk and cheese" expressed by Kosovar consumers. These consumers are also very sensitive to the origin of dairy products, even though they remain relatively unfamiliar with designations of origin [58,59]. The strategy is based on the strong popularity of local markets and specialized dairy shops, and on direct selling, the main outlets for the sheep, pastoral, and seasonal Sharr Cheese. The restaurants in the Zhupa valley also provide an important market for this cheese.

A Situated Approach to Environmental Sustainability in GI-Building
Our study contributes first and foremost to the recent literature on GIs' sustainability. GIs' capacity to deliver on sustainable development goals depends on their generic infrastructure identifying a product with specific resources in its locality of origin, via a collective specification process. Such collective intellectual property rights make it possible to recognize, protect, and therefore enhance the sustainability characteristics (i.e., economic, social, or environmental) of a given production system or a designated area of origin. However, some previous research suggests that it is difficult to assess ex ante the capacity of GIs [60] to generate this virtuous circle [61]. Others note encouraging results through ex post evaluation, but highlight the methodological challenge of drawing robust general conclusions about GIs' impacts and sustainability [62][63][64].
GIs as IPR instruments, while conferring legal protection on place-based quality products, may not suffice to provide the desired effects [65], whether on environmental issues, cultural heritage [6,66], rural and inclusive development issues [67], or economic performance [68,69]. A more context-specific approach may be needed [70]. Many scholars have turned to studying collective action and GIs' local governance, revealed as key factors in determining GIs' (positive and negative) impacts on sustainability issues, as well as on public good provision [1]. Studies have addressed the role of formal and informal institutions in the GI-building process [71,72], the controversies arising over GI management either by a collective of producers or by the public authorities, [73] or actors' configurations along value chains that favor positive collaboration or exclusion [74][75][76]. Our research focus follows this trend toward detailed analysis of the social and collective dimension of GIs.
GI-building requires intense collective organizational effort [77], involving stakeholders from different organizational fields. In Kovoso, a potential candidate for EU membership, a particularly intense collective effort is required to transpose the EU institutional and legal normative IPR framework into national regulations. In addition, the endogenous dynamics around product specifications in rural areas also demands institutional attention: informal agricultural production value chains need to organized, and the gaps in institutional backing for rural areas filled. This makes those considered intermediaries, translators, or knowledge brokers [26] central to the GI-building process and its collective action dynamics. Acting as organizational entrepreneurs, they provide the intensive boundary work [22] necessary for the GI generic infrastructure, especially in countries with no previous institutional experience. Moreover, they act as knowledge brokers [78][79][80], building the knowledge of the territory of origin required for the product specifications.
Further knowledge brokering is required to introduce an environmental dimension into the GI-building process, moving toward environmental description and accountability. Our results show that this GI environmental dimension is not given ex ante, but emerges from an intense and contested colearning process. This endogenous dimension of environmental sustainability in the GI marks it out from other voluntary standards and ecolabels (organic, sustainable) governed by a set of predefined criteria or indicators. The external definition of sustainability runs counter to the specific, context-dependent nature of GI environmental performance. The literature on voluntary standards and their impacts has debated whether it is better to opt for an external definition or what can be referred to as a "club approach", or emerging definition of sustainability [81][82][83]. A voluntary emerging definition of environmental sustainability coming from a collective learning process could lead to effective application of well-designed and appropriate sustainability criteria at key phases in production and processing. However, this highly demanding process is very sensitive to the nature and the design of the knowledge brokering activity. Finally, our results argue for developing a processual and situated approach to analyzing environmental sustainability in GI-building. More specifically we highlight the importance of producing, through an in itinere reflexive analysis, actionable knowledge on GI sustainability. We demonstrated how a Sharr Cheese GI device could be accountable for its environmental contribution to pastoral open landscape biodiversity-even if the establishment of a control plan is crucial in the next phase. This actionable knowledge produced throughout the GI-building process answers both strategical needs of the collective action process and an environmental intentionality from a territorial-agro-ecosystemic perspective. This actionable knowledge is highly sensitive to its context of production and evolutive. For these reasons, GIs specifications need to be seen as a living document that can be modified and amended over time. In this perspective, existing work on environmentally related amendments of GI specifications [84,85] should be complemented by a processual and situated analysis of the specifications' evolution at the GI-device level. In the vein of a reflexive governance approach to GI, such analysis could be useful in developing place-based pathways to sustainability [86].

Strategic Environmental Analysis of GI-Building
Our results also suggest that strategic analysis of the GI-building process from an environmental perspective can be enhanced by investigating not only macrolevel strategic positioning, but also microlevel social activities and practices [87]. Building from the organization literature on innovation, we mobilized the notion of experimental space as a source of environmental innovation and institutionalization. Our results highlight the constant interplay between the GI generic infrastructure and the different phases of boundary work within GI-building at the device level: enrolling/boundary work, learning process, distancing work, and anchoring work. They show how GI generic infrastructure is instrumental in anchoring new knowledge and organizational change emerging from experimental spaces, particularly through the product specifications, a phase pointed out in previous research as crucial to the success of innovation in institutionalization dynamics [88]. In other words, acting as boundary objects, the specifications and the control plan can be seen as robust assemblies radiating out beyond the space in which they were produced. As they are enforced by public authorities, as well as a broad range of actors, they turn collective action into a stable GI device that can be seen in relation to the different international regimes it embodies.
Our research also highlights the need to consider both the processual and the substantial dimension of strategic environmental action. This is the reason why our research focuses on the design of environmental purposive action within the GI-building process. Our situated narrative of GI-building for Sharr Cheese in Kosovo exemplifies the processual, exploratory, and strategic dimensions of institutional work, where other narratives adopted an external analytical stance [72,89]. This approach opens the black box of the collective learning process, where knowledge and action are the two faces of the same coin. Our results thus offer both a substantial and processual account of the boundary work. This cross-disciplinary approach is original in its content and its methods, building from management sciences, geography, economics, and agronomy. It confers to boundary work both a functional (strategic) dimension in building a collective action toward environmental objectives, and an epistemological dimension in producing actionable knowledge adapted to a given context. Our methodological choice follows pragmatic lines in the sense of John Dewey, particularly of his theory of Inquiry as the principle of the consubstantiality of action and knowledge [34]. Inquiry as an iterative, unstable and nonlinear process enables a situation to be defined so that it appears unified; i.e., the resulting understanding of the situation is coherent and shared between the participants, to "extend the subject's capacities for action" [90] in ambiguous and uncertain contexts.

Conclusions
Although GIs are flexible instruments that are open to multiple interpretations, they remain mainly anchored in the IPR nested regime embedded in a global trade macroregime. However, this can include environmental and social dimensions not directly related to the GI infrastructure, but arising from intentional and strategic activity at the liminalities, through experimental spaces. This is why our research focused on boundary work during the GI-building process in which we participated as embedded researchers.
While the principal rationale behind GIs is the context-specific identity of a product and a situation, there are few processual narratives of GI-building in the academic literature. Using this "inside insight" on environmental sustainability, we sought to illustrate the (agro-ecological) features of a given territory of origin, as well as the unstable nature of the strategic assemblies of knowledge, actors, norms, and instruments at work in the GI-building process. Each GI-building process could be considered as a situated device. An attempt to integrate environmental objectives into the design of a GI through product specifications could be taken as an inquiry, where experience and ideation results from an important boundary work, carried out in experimental spaces, during a contested collective learning process.
We also point out the strategic, even political, dimension of knowledge brokerage needed to protect, from an IPR regime point of view, the tangible and intangible attributes of a product related to its origin. Thus, we can conclude with Gangjee that "it would be a mistake to assume that formal legal recognition as a GI inevitably ensures provenance and authenticity to the extent necessary in order to achieve developmental goals or satisfy consumer expectations. In many cases, something more is required" [1]. In our work, this "something more" is further examined though the notions of knowledge-brokerage activities and boundary work in experimental spaces. What this amounts to can be considered an Inquiry on "how to build environmental accountability within the GI-building process". Again, to use Deweyan vocabulary, the question is not so much the evaluation (assessment) of a given instrument (i.e., a GI) but its valuation-in the sense of giving it meaning and value-though collective experience.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to contractual restrictions. in the restitutions, round tables as well as tasting workshops, along with representatives of local authorities, Sharr National Parc, and dedicated ministries who were actively discussing our results, propositions, and perspectives on the interest building a geographical indication on an emblematic product from Kosovo.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results".