1 INTRODUCTION

A separation from the neoclassical economic ideas occurred (^{Bathelt & Glückler, 2003}) from the seminal work of ^{Amin and Thrift (2000}) that suggested another direction for economic geography borrowing concepts from other social sciences. The use of the evolutionary and relational economic perspective in the researches has been growing, opening promising discussions for the development of new theories (^{Boschma & Frenken, 2010}). The topic has attracted the attention of researchers and in April 2009, the Times Higher Education presented a data analysis by Thomson Reuters, Essential Science Indicators (ESI), highlighting "Relational and Evolutionary Economic Geography" as the third most researched topic in social sciences - with 2,232 citations in 41 scientific papers. (^{Randelli, Romei & Tortora, 2014}; ^{Domareski-Ruiz; Chim-Miki & Gândara, 2014}).

Relational approaches have been more frequent in contemporary studies and advocated by many scholar leaders (^{Sunley, 2008}). However, severe criticism and doubts still surround this approach, which is presumably new thinking, a paradigm based on theory. Some authors point out that a "wave of enthusiasm" has taken over the theoretical debates compromising the discussion and critical evaluation (^{Storper, 1997}; ^{Boggs & Rantisi, 2003}; ^{Overman, 2004}; ^{Yeung, 2005}; Sunley, 2008).

The relational approach is a dispersed set of theories and ideas that share common characteristics but differ in important respects (^{Bathelt, 2006}). These approaches have been used to analyze modes of economic coordination or governance where there are collaborative and trustful relationships that favor the exchange of knowledge (^{Jones & Hesterly; Borgatti, 1997}; ^{Dyer & Singh, 1998}; ^{Rutten, 2004 Capello & Faggian, 2005}).

According to scholars, the level of research in Relational Economic Geography (REG) still does not have enough conceptual and methodological basis to have a practical meaning. Also, REG still is incapable of generating conditions of applicability in the real-world context, particularly to become the framework of empirical research. Even so, the approach seems adequate to analyze organizational networks, having the potential to contribute to scientific research and empirical analysis. (^{Bathelt & Glückler, 2003}; ^{Boschma & Frenken, 2006}; Boschma & Martin, 2007; ^{Sanz-Ibáñez & Antón Clavé, 2014}).

This paper presents and discusses relational thinking, bringing to the academic debate the REG proposal, its origins, and background. It intends to encourage the scientific community to join efforts to consolidate a theoretical framework to allow the use of this approach in several areas. The research was set in tourism, and its primary objective is to discuss whether the use of REG complements or replaces the traditional analysis of tourism destinations that considers them as network, district, cluster or LPA (Local Productive Arrangements).

Tourism was chosen for this discussion because it is a singular economic activity based on the interdependence of actors to compose the destination as an integral product (^{Della Corte & Sciarelli, 2012}). The research is classified as qualitative with a theoretical and conceptual structure which considers the following questions: How can REG contribute to the analysis of tourism destinations? Does REG replace the networks, districts, clusters, and productive agglomerations approaches or does it complement them? Can REG be applied to any productive structure or are there limitations? Thus, this research is characterized as analytical and descriptive, with a qualitative approach based on a theoretical and conceptual review of papers on REG published in Scopus, Web of Science, Science Direct, and Scielo databases. Also, a follow-up of works cited by the authors consulted was done to provide a broad and historical view of REG as a conceptual perspective.

2 RELATIONAL THINKING AS A BASE OF RELATIONAL ECONOMIC GEOGRAPHY (REG): ORIGINS AND CONCEPTS

Since the 1990s, economic geogra-phers have conducted studies on the complexity of relations between actors and structures. Also, they have studied how these relations affect the spatial dynamics of economic activities by consolidating the so-called Relational Economic Geography (^{Amin, 1998}; ^{Dicken & Malmberg, 2001}; ^{Ettlinger, 2001}; ^{Bathelt & Gluckler, 2003}; ^{Boggs & Rantisi, 2003}). Studies with different contributions and backgrounds have emerged within Economic Geography, highlighting: the relational construction of spatial identity (Amin & Thrift, 2000) and the works related to the so-called "windows of locational opportunity" and "untraded interdependencies" (^{Scott, 1988}; ^{Storper & Walker, 1989}). Other approaches came from the contributions of economic sociology with the notions of social embeddedness and trust-based relationships (^{Granovetter, 1985}), and more recently, the social production of knowledge (^{Faulconbridge, 2006}) has been discussed.

Table 1 is extracted from ^{Yeung (2005}) that proposed 'rethinking' REG. It synthesizes the relational frameworks analyzed by geographic, economic, and management concepts, leading to a classification of spatial manifestations and concepts currently used in the regional development and competitiveness studies.

^{Sunley (2008}) argues that the origins of relational thinking in economic geography can be traced to sociology from the rooting of networks in economic life. The author stresses this thinking has grown in popularity because the contemporary capitalism has become increasingly relational. The changes caused by late capitalism are characterized by the restructuring and globalization of productive processes, generating new forms of coordination between companies (^{Sanz-Ibáñez & Antón Clavé, 2014}). ^{Dicken, Kelly, Olds, and Yeung (2001}) consider that the relational perspective is a starting point for the empirical work. Thus, REG stands out as a tool for analyzing regions and how the relational assets can generate advantages for their development (^{Coe et al., 2004}).

The Relational thinking generated from theoretical-empirical approaches and studies can be grouped into four major schools of thought. The emergence of these schools gathered scholars from Economic and Regional Geography that started relational studies. Some examples are shown in Table 2.

In this way, the relational thinking was gradually consolidating the REG. ^{Bathelt and Glückler (2003}) presented a paper summarizing the background called the second evolutionary stage of REG. This stage followed a transition in Germany in which Economic Geography contributed to consolidate the so-called new REG paradigm. Economic Geography in the German school was influenced by two important paradigms: Länderkunde (the science of description and regionnal synthesis) and Raumwissenschaft (Spatial Science), which generated the first transition. That phase was strongly influen-ced by the works of ^{Isard (1956}, 1966) from the American Economic Geography (Bathelt & Glückler, 2003).

Table 3 shows a line of changes in which space stops being an object and becomes a perspective. Thus, it became an object of knowledge. As well, the conception of action and theories used in research shifted the axis. The object of research becomes a decontextualization of the principles of socioeconomic changes in the space, supported by an epistemology of critical and evolutionary realism, abandoning the search for ge- neral explanatory or deterministic laws.

The changes presented in Table 3 continued to form the second evolutionary stage in REG research. ^{Bathelt and Glückler (2003}), drawing on ^{Storper’s (1997}) work, introduced four core concepts or ions (according to the authors’ nomenclature) for analysis in Economic Geography, namely, Organization, Evolution, Innovation, and Interaction. In this reformulation, Storper (1997) conceptualized the so-called "Holy Trinity" (Technology, Organization, and Territory) which served as the starting point for the REG paradigm proposed by Bathelt and Glückler (2003).

Five dimensions are highlighted in the second phase of REG (see Table 3): Design of space (space as perspective); Object of knowledge (contextual economic relations - social practice, process); Conception of action (network theory and embeddedness); Epistemological perspective (evolutionary perspective), and Research objective (de-contextualization of principles of socioeconomic exchange in spatial perspective - that is, alteration of explanatory framework of socioeconomic changes based on spatial perspective) (^{Bathelt & Glückler, 2003}).

Although researchers acknowledge that the context influences the economic action and present studies focused on the dynamics of spaces within "Relational Thinking" (^{Bathelt, 2006}), even so, few researchers directly conceptualize REG in the papers currently published. Most of the papers specify fundamentals, epistemological bases, and origins using different conceptions and theories to derive its relational analyses. The lack of conceptualization reinforces the main criticism on REG, that the relational concept has a high level of abstraction (^{Sunley, 2008}; ^{Sanz-Ibáñez & Antón Clavé, 2014}). Most authors refer to it as relational thinking, relational perspective or relational approaches. Nevertheless, they coincide that the dynamics of relationships is the object of REG, and that object must be included in space. Table 4 presents a brief synthesis of how authors have expressed the relational thinking.

According to ^{Bathelt and Glückler (2003}) the objective research in REG is based on three major propositions:

Thus, fundamentally, REG examines the space from the actions that human relations generate on the production systems which, in turn, influence the organization and the territorial development. ^{Sanz-Ibáñez and Antón Clavé (2014}) based on ^{Bathelt and Glückler (2011}), and ^{Storper (1997}) point out that REG focuses on: organization, concern with the social and spatial, with the division and integration of work; evolution, positive and negative impacts of historical structures, processes and events in current decisions; innovation, processes creation and diffusion of knowledge, as well as, the effects of technological changes; interaction, the interactions among economic agents and the formal and informal institutions.

A key point for the development of tourism as an economic activity is entrepreneurship. ^{Vale, Amâncio and Lima (2006}) consider three categories of entrepreneurs related to organizational networks or productive arrangements: class-ical entrepreneur (an independent enterprise), utilitarian collective entrepreneur (action in a sectorial network with a single goal), and sectoral collective entrepreneur (action in a sectoral network with multiple goals). A better performance in tourism activity can be obtained based on sectoral collective entrepreneurs.

Entrepreneurship is a contextualized and contingent human action and, from the four ions defined by ^{Bathelt and Glückler (2003}) - Innovation, Organization, Interact-ion, and Evolution - generates economic activity for territorial development. In this sense, economic science identifies entrepreneurship as a vital element for development, also, Schumpeter (1961) considered that entrepreneurs play a leading role in the economic evolution. Figure 1 shows the transition from relational thinking towards the paradigm of REG according to Bathelt and Glückler (2003) proposition.

Source: The authors (2018)

Figure 1 Relational Economic Geography (REG) proposition 

Following this line, space as a socially constructed entity by human action is economic action. Thus, contextual economic relations are the process of social practices (intentions, strategies, and activities of actors) which in turn is the object of knowledge of REG (Butler, 2003).

Human action gains prominence because it creates or limits opportunities. The actions occur in a space that accumulated knowledge, habits, and other elements, therefore influencing future decisions, i.e., path dependence. However, this set, (human action + context + path dependence) still suffers another influence, i.e., the contingency of human action itself and the context, since the economic activity is not always predictable (^{Bathelt & Glückler, 2011}; ^{Sanz-Ibáñez & Antón Clavé, 2014}).

3 TOURISM DESTINATIONS ANALYZED AS PRODUCTIVE AGGLOMERATIONS (NETWORKS, DISTRICTS, CLUSTER, LPA) FROM REG PERSPECTIVE

It is consensual among researchers that tourism is a space activity, socially constructed. Therefore, it evolves (^{Seaton & Bennett, 1996}; ^{Ringer, 1998}; Saarinen, 2001; ^{Shaw & Williams, 2004}). Also, the formation of organizational networks and governance mechanisms are considered ways of supporting the local development in tourism destinations (^{Della Corte & Sciarelli, 2012}).

Tourism destinations have both economic and social functions, with high complexity and interdependence of actors. Also, tourism activities are characterized by great diversity. The analysis of the operational way of the tourism destination and its typology of productive agglomeration contribute to developing planning and management tools more suited to the local and regional reality. Therefore, it improves the level of distribution of tourism activity benefits and the sustainability of available resources (^{Buhalis, 2000}).

Studies on the evolution and performance of destinations using conventional models of tourism geography have evidenced weaknesses as explanatory tools, emphasizing the need for new categories and perspective of analysis (^{Brouder & Eriksson, 2013}; ^{Williams, 2013}; ^{Sanz-Ibánez & Antón Clavé, 2014}).

Tourism destinations can be organized in different ways. They represent productive agglomerations that convert local synergies into competitive advantages (Costa & Souto-Maior, 2006). At this point it becomes important to look at the boundaries between some typologies such as networks, districts, cluster, and local productive arrangements (LPAs), to verify whether REG can replace the analysis of tourism destinations based on these theories or if it complements them.

A tourism destination is commonly seen as a set of institutions and actors, located in a physical or virtual space, but with relations that challenge the traditional dichotomy of production-consumption. Indeed, it is a unit of action and production, where several stakeholders interact producing the tourism experience (Saraniemi & Kylänen, 2010; ^{Pearce, 2014}). In tourism research, the term agglomeration is usually used, since the development of this activity at the micro level generates some firms territorially located and often articulated seeking the development of the activity. This categorization comes from ^{Porter (1999}), which defines cluster as a geographically concentrated grouping of interrelated firms and correlated institutions in an area, linked by common and complementary elements. That is, tourism agglomerations are what Porter calls clusters. However, due to its comprehensive nature, industrial districts, and local productive arrangements (LPAs) have also been included in tourism cluster studies.

The border between networks, districts, clusters, and LPAs is sometimes difficult to establish. The fine line between an LPA and a cluster can be considered the greater geographical and sectoral concentration that the clusters have in comparison with productive agglomerations, as well as, the better collective ability to deal with the market (^{Schmitz, 1997}; Silva, 2014). LPAs are agglomerations of organizations, with large numbers of small firms, which use a joint action, cooperating to achieve better competitiveness and development (^{Pyke & Sengenberger, 1993}; ^{Caporali & Volker, 2004}; ^{Costa et al., 2012}).

On the other hand, ^{Dini (1997}) distinguishes industrial districts from networks because he considers networks have a limited number of firms, participants are identified, the composition has low variation and participants are not necessarily in the same territory. Particularly, it is necessary to emphasize the excluding nature of networks when considering tourism destinations in their entirety. The United Nations Development Organization defines a business network as a permanent strategic alliance between a limited and clearly defined group of independent firms, which collaborate to achieve common medium and long-term goals, aimed at developing the competitiveness of the participants (ONUDI, 1999). In this sense, a network is a set of preferred, selected, members. Therefore, it has an exclusionary character that does not apply to the analysis of an integral tourism destination (^{Camagni, 1991}; ^{Ceglie et al, 1999}).

On the other hand, theoretical-empirical studies on districts have been the focus of researchers since the 1970s. In that period, a major economic recession led several countries to seek solutions many of which relied on the formation of an industrial district (^{Pyke et al., 1990}). In this sense, there are many organizations located in a geographical area (co-location), with different structures and sizes participating in the different production stages for the generation of a homogeneous product (Pyke et al., 1990). Another factor making tourism destinations closer to industrial districts is the massive participation of small firms, characteristic emphasized by several authors of both industrial districts and tourism (^{Hjalager, 2000}). Also, two other prominent characteristics in the district theoretical framework bring it closer to tourism destinations: the existence of public and private local institutions that support the economic agents, and the existence of a common cultural and social context that links these economic agents, generating codes of common behavior (Pyke et al., 1990).

^{Hjalager (2000}) was one of the authors that contributed to the establishment of the typology "Tourism Districts." Her study identifies five determinants to qualify a tourism district: interdependence of firms; flexible firm boundaries; cooperative competition; trust in sustained collaboration; a community culture with supportive public policies. Likewise, from the perspective of networks, tourism districts as productive space arrangement is also grounded in REG.

The associative character of cooperative companies seeking competitiveness is explicitly a feature included in relational thinking. Also, the collective learning generated by the group leads to path dependence and spatial constraints put the participants in a specific context (^{Bathelt, 2006}).

The re-reading of this productive reality through REG lens enhance some characteristics. The existence of co-location is pointed out by the REG scholars as a relational facilitator (^{Bathelt, 2006}). The agents subject to a special context leads them to operate under specific institutional and social conditions from which they cannot be easily separated. This feature is also considered within the REG approach (^{Polanyi, 1957}; ^{Granovetter, 1985}). Within the five ^{Hjalager’s (2000}) determinants defined for tourism districts, ^{Amin and Cohendet (2004}) consider, in terms of relational thinking, that relational proximity enables close social interaction and becomes a source of competitiveness.

Cluster classification deserves a special attention since this construct is one of the most studied productive forms. The concept defined by Porter (1988) finds a practical problem in the tourism activity. Activity within tourism destinations is diffuse, which makes it difficult to accurately measure the grouping and categorization, thus raising doubts about tourism cluster classification (Amato, 1999; Ivars Baidal et al., 2014). However, the territorial and endogenous model of regional development is still considered the most appropriate. It is possible to choose productive specialization (tourism) as a representative element of the explanations of local and regional economic development (^{Silva, 2006}).

This imprecision has led to the use of cluster in a broad sense for any productive arrangement, and many times confusing it with the concept of district. The two forms have similarities, but a cluster entails a degree of geographic concentration, which in tourism destinations does not always occur. The most used measure to verify this concentration has been the Location Quotient (LQ) and the number of sectorial jobs to define the importance of the activity for a region.

However, the literature does not indicate any reference value for a tourism destination to be considered a cluster. Another fact that weighs on this perspective is highlighted by Amato (1999). The author points out the difficulty of "clusterization" since it is a productive chain composed of a series of complementary services to the main tourism products, making it even more difficult to separate clearly the categories or agglomerates (clustering).

Ivars Baidal et al. (2014) emphasize that this configuration requires preconditions, such as regional or innovation policies, company initiatives or associations, the existence of a unifying project, or a person with leadership and mobilization capacity. In any case, the cluster perspective is based on a collective efficiency (Porter, 1988) generated from the interrelationship between the agents, circumscribed in a geographic space. There is a coincidence with the perspective of REG since relationships are a key condition for obtaining this collective efficiency.

In the LPA approach, the territory has a specific scope of analysis and action, where the productive, innovative, and cooperative process occurs (Cassiolato, Lastres & Maciel, 2013). This type of agglomeration has an important degree of embeddedness, can have different forms of governance, generates and shares knowledge, providing dynamism and innovation to the region in which it is located (Silva, 2014). ^{Hoffmann and Campos (2013}) highlight four features of LPAs and that have been much discussed by the scholars: trust, skilled labor, cooperation, and the role of institutions. From a relational thinking perspective, this is human action in its context generating economic action (^{Bathelt, 2006}). Thus, REG offers an angle to study the sum of the action of these resources in the territorial dimension and the reflecting of human activity on local competitiveness.

Table 5 presents tourism destination characteristics considering whether they are met in the analysis through the different forms of productive agglomerations. Also, it shows if the REG background would be able to achieve these characteristics, considering its theoretical foundations applied to tourism destination analysis.

The need to adapt some traditional concepts to tourism is evident, as is the case of agglomeration, networks, district, or cluster, and to translate some theories to the practice of tourism destinations, mainly to give clear limits regarding clustering and conformation of districts (^{Judd, 1995}; Weidenfeld, Butler & Williams, 2010; ^{Sanz-Ibáñez & Antón Clavé, 2014}). Given the above, it can be stated that REG's contribution to the analysis of tourism destinations is visible. However, there are questions that remain such as: Does REG replace the perspective of networks, districts, clusters, and arrangements or does it complement it? Can REG be applied in any of these productive arrangements or are there limitations?

It can be seen in Table 5 that the more the analysis of the destination tends to the micro scale, the better the characteristics of tourism destinations are met. Besides, REG demonstrates that it can include all the characteristics present in tourism destinations thus allowing for a complete analysis.

Some authors have sought a specific typology of agglomeration for tourism, coining terms, and definitions such as: Tourism District (^{Hjalager, 2000}; ^{Baggio, 2008}); Tourism Cluster (^{Costa et al, 2012}); Local Tourism System (^{Capone, 2004}, 2006; ^{Maulet, 2006}; ^{Lazzeretti & Capone, 2008}); or Dynamic Tourism Destination (^{Sanz-Ibáñez & Antón Clavé, 2014}). Thus, it is observed that tourism destinations move towards their own typology, and identity as a productive arrangement and unit of spatial analysis, which apparently seems to be a mixture of networks, districts, clusters, and LPAs, but certainly based on relational thinking, since it establishes a network of relations between actors that compete for the division of the market, while cooperating to develop it (^{Della Corte & Sciarelli, 2012}).

The classic destination analysis is based on models of tourism competitiveness such as those of ^{Porter (1990}), ^{Dwyer and Kim (2003}), ^{Ritchie and Crouch (2003}), as well as the analysis by resources and capacities based on Barney (1989), and by productive agglomerations. However, these perspectives of analysis face the challenge of including a more active relationship between the different actors that form this sub-sector.

The internal and external cooperation-competitive relations (coopetition) to the tourism destination lead to the improvement of collective efficiency, the formation of learning and innovation regions (Ivars Baidal et al., 2014), but, in this sense, a typology of productive systems for the tourism destinations is still not clearly defined.

Complementing the previous discussions should be noted that ^{Pearce (2014}) recently has proposed an integrative conceptual framework of tourism destinations, by synthetizing elements from industrial districts, networks, clusters, systems, and social structures. His proposal integrates the geographic dimension (space and place), the mode of production dimension (structure, behavior, and actors), and the dynamic dimension (structure and driving factors). Notably, the author emphasizes that the relationship between these dimensions and their factors generate the complexity, adaptability, and evolution of destinations. However, the author warns that there is still a long way to go in refining each of the elements that make up his proposal, he reinforces that it is a path that points to a proper classification for tourism destinations (Pearce, 2014) and, as can be seen, transits through the fundamentals of Evolutionary and Relational Economic Geography.

4 ANALYSIS OF RESULTS

The arguments and rationale presented show that relational analysis is necessary for regional or local economic studies, and especially for studies of tourism destinations. The REG has in its center the human action and is based on the contingency, context, and path dependence, which is based on interaction, innovation, organization, and evolution ions. These elements conform the entrepreneurship, which generates the economic action immersed in an environment of coopetition for the formation and development of tourism destinations as an integral product. The sum of these considerations indicates the integration of this knowledge using REG for regional analyses.

Human action permeates and forms the foundations of REG that support the formation and evolution of integrated tourism destinations. However, the way REG will materialize in destination planning, as an analysis instrument and tool to assist in the decision-making process, is still undefined. The understanding of coopetition relation-ships can be a way of analyzing tourism destinations considering the perspectives of REG and enabling the development of tools for empirical relational analysis. This form of behavior fits into the formation of value networks that include competitors, complementors, customers, and suppliers (Brandenburg & Nalebuff, 1996).

Another issue arising from the discussion is that the foundations of REG interact with each other, acting in their formation, being a cycle without clear beginning. Tourism destinations are built upon natural, cultural, social, political, and economic resources which affect the context and trajectory that, in turn, condition the human action. But, can this order of formation be reversed?

Another question is to know whether REG should be used as a new competing paradigm in the analysis of tourism destinations or considered as a complementary paradigm to existing ones; and how this complement could be used in the theoretical-empirical analyses adding to knowledge. Finally, it is questioned whether mature theories should be revised to include social relations in their theoretical foundations.

These questions are in line with numerous criticisms of REG, which suggests that despite its acclaimed use, its evolution towards a consolidated theory has not occurred. The same situation arises with the concepts of agglomerations such as cluster/district/networks / LPAs that continue to receive criticism for the thin line that separates them and the lack of clear criteria to differentiate these typologies and apply them in the analysis of tourism destinations.

The best way of framing tourism destinations regarding the typology of agglomeration is still not clear. However, REG can be used to analyze destinations regardless of their form of productive organization since it encompasses all destination characteristics. Moreover, the object of analysis becomes the human actions and relationships, and these would occur regardless of the productive agglomeration considered.

5 FINAL REMARKS

This paper has critically examined relational thinking, bringing to the academic debate the REG proposal, its origin and foundation, with the purpose of encouraging the scientific community to join the effort to develop a methodology of analysis and empirical application that would enable the use of this approach. This study carried out a compilation of papers and presents a consistent literature review on relational theory, articulating these concepts with tourism. We have discussed the use of REG in tourism destination studies since some destination characteristics rely on the interdependence of the actors to become an integral product (^{Della Corte & Sciarelli, 2012}).

In this debate we presented some critical points of REG that are pointed out in the study developed by ^{Sanz-Ibáñez and Antón Clavé (2014}), such as: lack of specificity and the extreme abstraction pointed out by ^{Sunley (2008}); little theorizing about power relations (^{Yeung, 2005}); focus on firms rather than individuals in micro-level analysis (Ettlinger, 2003); and low emphasis on non-local flows and relationships (Yeung, 2005).

However, it should be noted that se-veral authors have worked within relational perspectives. The role of the local and global interaction between economic and non-economic actors is undoubtedly a factor that can either help or hinder the development of places (^{Bathelt & Glückler, 2011}).

It seems that REG can contribute to the analysis of tourism destinations, however, more in-depth theoretical grounding, analysis tools, and specific methodologies are needed. In this way, this diffuse approach could achieve the status of a paradigm and make a useful contribution to regional and local analysis, as well as management support for public policies. REG can also contribute to a much-debated topic among scholars, i.e., the establishment of a typology of productive agglomeration with evident characteristics as to its form of classification, clearing up doubts if we consider the destination from the perspective of a network, district, cluster, or LPA.

The entrepreneurial action generated by the four ions defined by ^{Bathelt and Glückler (2003}) - Innovation, Organization, Interaction, and Evolution, whose apex is made by human action based on the foundations of REG (context, contingency and path-dependence) system that can lead to regional competitiveness.

However, whether is the context that forms and develops the tourism destination or the other way around, being the human action and path dependence reflex of this context, are questions to be further addressed by REG. It seems that it is a cycle that can be started from any point, depending on the previous political, natural, economic, and social condition that the environment offers.

To conclude, we suggest that from relational thinking further studies prioritize the creation of a typology with its own criteria for classification and analysis of tourism destinations, in order to establish mechanisms for their practical application. This can be achieved either by adapting the various existing theories in the industrial sector and the economy, or by creating new forms, but keeping in mind the peculiarities of tourism.

Thus, it is considered that REG can be used to analyze a tourism destination in any form of agglomeration because it has space as a perspective and not as an object or causal factor. This means that the object of knowledge is the economic context of relationships, a fact that occurs in any productive arrangement. Finally, it can be concluded that the REG can be a theoretical basis for creating tools for analysis, classification, and monitoring the influence of relational proximity and its influence on the development of localities and regions. These tools can contribute both to take REG out of the abstraction and to give substance to a new theorization applied to tourism.