REGION IN ITS COMPLEXITY: A DISCUSSION ON CONSTRUCTIVIST APPROACHES

The contemporary regional geography paradigm is characterized by emphasizing the socially constructed nature of regions. However, the discussion on the conceptualization of region is very rich, it does not reach universal conclusion. Such a universal conceptualization is probably neither possible nor desirable. This paper aims to contribute to the ongoing discussion. It examines the various approaches towards region with special focus on the conceptualization of the institutionalization of the region based on the heuristic literature review and it attempts to propose more complex (but not ultimate) theoretical conceptualization of region that should enable to bridge the dual- ity of region; addressed as an “animate”, constantly changing, phenomenon which also makes it a resource for regional actors to meet their particular goals, one that people may perceive and feel attached to while further mediating their images thereof. The paper builds upon the Lefebvre’s theory of the societal production of space and the Paasi’s theory of the institutionalization of region and presents the idea of the societal production of region. It argues that any complex approach to region must incorporate three levels of region: “given” (practice of region), “made” (representation of region) and “perceived” (idea of region). Better understanding of the concept of region is still more necessary and relevant especially due to the increasing pressure on the applicability of regional research. Thus, the paper suggests the closer insight into the interrelation of three mentioned levels of region is crucial from the view of the contemporary state of art. This text is also published in Czech as the official on-line supplement of the article. The Czech version can be downloaded here: http://www.aucgeographica


Introduction
The contemporary regional geography paradigm (e.g. Claval 2007) is characterized by emphasizing the socially constructed nature of regions (e.g. MacLeod, Jones 2007;Paasi 2010;Jonas 2012;Jones, Woods 2013;Harrison, Grove 2014;Vainikka 2015). The "constructivist" approaches can be traced back to the 1980s. Their main stream of the 1990s was primarily focused on the economic perspective of the region. Despite criticism, the economic and political focus of most studies perpetuates itself even in the contemporary discourse. Although, the main focus is centered on regional actors and their networks that are responsible for the (re)production of regions (Amin 2004;Allen, Cochrane 2007). Region and its delimitation are thus understood as constructed by those regional actor networks that make decisions about the purpose, the territory, symbols, etc. of either an emerging or a reconstituted region. Therefore, region is a (political) process (e.g. Pred 1984;Paasi 1986).
However, the discussion on the conceptualization of region is very rich, it does not reach universal conclusion. Such a universal conceptualization is probably neither possible nor desirable. This paper aims to contribute to the ongoing discussion. It examines the various approaches toward region and its delimitation with special focus on the conceptualization of the institutionalization of region based on the heuristic literature review and it attempts to propose more complex (but not ultimate) theoretical conceptualization of region that should enable to bridge the duality of region; addressed as an "animate", constantly changing, phenomenon which also makes it a resource for regional actors to meet their particular goals, one that people may perceive and feel attached to while further mediating their images thereof. The text is structured on the heuristic discussion of world literature and the reflection thereof in the Czech academic environment. The discussion leads to the proposal of complex conceptualization based on societal production of region.

New regionalism in the world literature
The regional geography paradigm is strongly influenced by the orthodoxy of the resurgence of regions (Keating 1998;Johansson 1999;Harrison, Grove 2014) in the last few decades. As stated by Fawn (2009: 5): "region, regionalism and regionalization matter". However, none of these is new to geography. Region has always been a fundamental phenomenon examined in geographical research (e.g. Vidal de la Blache 1994; Hartshorne 1939;Gilbert 1960;Grigg 1965;Haggett 1965;Hägerstrand 1970;Urry 1981;Claval 1987). The origin of (modern) regionalism is usually connected with the rise of the state system after WWI, however, its patterns can be studied even further back in history (Fawcett 1996;Semian 2015). From a certain perspective, "the regionalized world has always featured in human history" (Fawn 2009: 6-7).
The revitalized interest in regions can be dated back to the 1970s. Being strongly intertwined with the reaction to strictly nomothetic approaches in geography, it strives to replace them with more humanistic ones, stressing the importance of human actions in the reproduction of geographical space and its organization. Place is produced and space is structured through human perception (Tuan 1990). Thus, every place has an individual character. Actually, this idea is based on Heidegger's conception (cf. place as lived and experienced space ;Heidegger 1971: 152). A special stream of regional conceptualization can be identified within humanistic approaches. It deals with vernacular regions and define region as a product of the spatial perception of average people (Jordan 1978: 293). This stream was popular especially in North American literature of the 1970s and 1980s (e.g. Meining 1972Jordan 1978;Zelinsky 1980), but it perpetuates until today (e.g. Reginster, Edwards 2001;Liesch et al. 2015;Vukosav, Fuerst-Bjeliš 2016). Nevertheless, humanistic geographers are more concerned with personal experience and an individual's position within a community, and less with the examination of communities as a complex. Thus, place has a highly personal character and any particular part of space may have different meanings for different people. In distinction therefrom, the concept of region has a collective character (Paasi 1986). Therefore, rethinking the region must equally be seen through the prism of the "cultural turn" and "spatial turn" in social science (Claval 1998;Thrift 2006) and, later, also through the "institutional turn" in economic geography (Jones, Woods 2013;Tomaney 2014) as an attempt to adapt the fundamental concepts to the changes of the forms of the society's geographical organization (Hampl 2002) caused by general development processes in society, namely globalization and transition to postindustrial society. Over time, region came to be understood more or less as a static category that was used by a majority of scholars to sort data and information (Klemenčič 2005). Without any deeper discussion region was often taken as given. The static and given category has nonetheless been brought into question and region has become a subject of geographical research.
The attempts to adapt the conceptualization of region to the dynamic societal changes of the last three decades have resulted into pluralistic approaches toward the concept itself and are regularly criticized for exaggerated constructivism and over-theorizing (Martin 2001;Barnett 2004) on the one hand, and for a vague theoretical embeddedness on the other (Schmitt-Egner 2002), namely in terms of conceptual definition of the concept of the region itself (Harrison, Grove 2014). For many authors region has become a concept hard to grasp and is often labelled as elusive (Keating 1998), chaotic (Lovering 1999) or enigmatic (MacLeod, Jones 2007). Nevertheless, rethinking the region as a social construct (Thrift 1983;Pred 1984;Paasi 1986) can be seen as the central point of the emergence of these discussions. This conceptualization became dominant in the new regionalist discourse of the 1990s (Keating 1998) wherein the region is often canonized as a development and governance tool in the hands of economists and policy-makers (Fawn 2009;Jones, Woods 2013;. This was most noticeably the case of the vision of "Europe of Regions" promoted by the European Union (Johansson 1999). Regions were taken for basic economic units with a potential to erode the integrity of national states as an essential factor for further economic and political integration (Harrison, Grove 2014). The same (economic) understanding of the new regionalism can be identified in the United States (Jonas, Pincetl 2006;Hamin, Marcucci 2008), even though with greater emphasis on city or metropolitan regions, and is well documented in other literature (Thompson 2000;Frost 2008;Paül, Haslam McKenzie 2015).
The economy-based understanding of region tended to be unsatisfactory and somewhat narrowing the complexity of the fundamental geographical concept. It is in particular contradictory to the very idea of rethinking the concept of region. Such a narrow conceptualization is often criticized for being strictly pragmatic (Hampl 2002) and ideologically reproduced (Lovering 1999), lacking any critical reflection of the social construct model which is regularly taken as given without any further theorization (Paasi 2010). The economical conceptualization generally lacks any political and power-related perspective: who constructs and reproduces the region and why (Frisvoll, Rye 2009;Paasi 2010;Jonas 2012;Harrison, Grove 2014), but also any cultural perspective: how the inhabitants perceive various regions, how they orient themselves among them and whether they can feel some attachment to them (Semian 2012;Antonsich, Holland 2014).
Since the late 1990s, rethinking the region as relational may have been considered a parallel stream of reflection but in many ways it has been a complementary paradigm for the aforementioned constructivist conceptualization of region (Johnson et al. 2011). Central to these discussion are, among others, works by Massey (1994), Allen et al. (1998) and Amin (2004). Nevertheless, the idea of relational perspective on region is not new at all. As Jonas (2012: 264) puts it: "in fact, several distinctive strands of relational thinking about regions have emerged from the work of radical human geographers and social theorists in the 1970s and early 1980s (Jonas 1988;Pudup 1988;Sayer 1989)". The idea of spatial organization according to relations and social or actor networks is crucial to these approaches. However, the relations and networks do not necessarily have to be bounded neither territorially nor in scale. They are increasingly dynamic and spatially diverse. Yet truly unbounded, territorially independent regions are still somewhat theoretical imagination (Allen et al. 1998) and their discovery is not absolutely necessary for further progress in theorization of the regional concept (Jonas 2013). Relational approaches emphasizing the idea of the space of flows and actor networks may be identified as crucial in contemporary discussions about regional development of fuzzy-bounded soft spaces (e.g. Deas 2006; Allmendinger, Haughton 2010) and city-regions (e.g. Harding et al. 2006;Lloyd, Peel 2006).
One can witness further theorization of the conceptualization of region in the new millennium which responds to the above-mentioned criticism. While political power relationships have acquired a dominant position, the regional economic framework usually continues to serve as a context for the regional development issue (Amin 2004). The rethinking of region is connected with multi-level governance and metagovernance (Jessop 2004). Owing to the different quality of research interest in regions, some authors refer to such paradigm as "new new regionalism" (Ballinger 2007) or "new regionalism vol. 2" (Harrison, Grove 2014). Such approaches see region as a somewhat contested unit that does not necessarily have to be bounded in space. However, this does not mean that regions cannot be territorially recognizable. Usually, general localization is widely accepted but the specific delimitation may differ according to the purpose of each particular region and its individual perception. Allen, Cochrane (2007) designate regional plurality as regional assemblage, addressing region as an assemblage of actor-networks with an influence on decision-making processes within a particular region. Regional actors do not necessarily have to be embedded in that region and their relational impact may reach beyond any territorial fixity. That is where the authors think beyond simple overlapping and intersecting of various regional initiatives. Thus, region possesses the intermediary character in all three scale modalities (size, level and function; cf. Howitt 1998) that integrates a region into the system and therefore it cannot be approached as a single piece of "the Russian doll" (Allen et al. 1998).

Constructed nature of region
As mentioned above, the contemporary paradigm is strongly influenced by the orthodoxy of the resurgence of region and the so-called new regionalism, or rather by one of its more recent variation. Nevertheless, this does not mean that the other conceptual approaches are not present. Contemporary geographical research basically distinguishes three modalities of understanding regions: (1) Probably the most common approach to region in the contemporary science is taking region as given. However, such regions, too, are constructed and researchers tend to use them merely as a category for sorting data and information (Klemenčič 2005).
Within this approach to region, various scopes of regions can be identified, ranging from natural (delimited and based on natural features) to ad hoc ones (established for a particular purpose). Unfortunately, instead of further scrutinizing the concept of region, most authors use it as a statistical or institutional unit, i.e. as an analytical tool. As stated by Paasi (2010Paasi ( : 2297, this approach goes on gaining strength, as it is well linked to applied research wherein region is not the primary subject of the study. (2) (a) The traditional geographical approach sees region as a construct of geographical research, searching for "a relatively autonomous complex of mutually intertwined and qualitatively hybrid components of environment" (Hampl 2002: 334). Regions are thus bounded, contiguous subdivisions of broader units (Paasi 2010). Traditional vernacular regions can be counted among them (e.g. Jordan 1978;Zelinsky 1980). (b) Additionally, the synthesizing scholar quest for geographical organization can also be ranked into this category. The core of these approaches is, generally speaking, geographical organization modeling according to spatial patterns of various social processes. In this sense, they are still attached to the former ones, since they look for functional regions through the integrity of socio-economic relations (Hampl, Marada 2015). (3) The most recent understanding of region puts the concept of region in relation to social discourse (Paasi 2010). Region is not only influenced by, but also influences the reproduction of socio-economic, cultural, political and power relations in space. Such relational understanding conceptualizes the region as a product of networks, interactions and articulations of live "everyday" connections (Allen et al. 1998) and thus can emerge across the imposed constructed boundaries, and even across scales. This division allows to conclude that region is a social construction. It can be constructed by scholars in an attempt to identify and comprehend the spatial organization on Earth's surface. It is a result of external frameworks and internal conditionality. In view thereof, a region may be constructed by various actors in an attempt to enforce their interests or demonstrate their power. This particular part varies a lot. By these terms one can understand: (a) a simple statistical regionalization whose main purpose is to compare diverse parts of the world and measure the differences; (b) administrative regionalization aimed at decentralizing the existing political power in search for better governance; but also (c) activities of regional actors advancing (by way of founding ad hoc regions) their vision of development trajectories (Semian et al. 2016). A region may also be constructed by citizens through their "everyday" activities, relations and perception of the enveloping information of "regional" character (Sørensen 2008).
Region as a social construction is reproduced as a historically contingent process (Pred 1984;Wilson 2007), no matter whether it is endowed with a long historical narrative path or the result of a recent ad hoc initiative. Thus, region is a process. It is changeable in space and time. Region is constructed (formed), constantly reproduced through the process of institutionalization and may disappear some day through deinstutionalization (Paasi 1986;Raagmaa 2002;Zimmerbauer, Paasi 2013). It means that every region is perpetually in a state of being formed and reproduced through both material and symbolic processes (Granier 2007). Once a region acquires a place in the regional system, it becomes part of the reproductive and transformative process of society. Regions thus influence and are simultaneously influenced by political, economic, social and cultural processes, i.e. the basic mechanisms of all societal changes. Paasi (1986) provides a guideline suggesting to comprehend the process of "becoming" as that of regional institutionalization. Despite having been contested by many scholars, including Paasi himself (cf. e.g. Frisvoll, Rye 2009;Paasi 2010;Riukulehto 2015;Vainikka 2015), this framework remains to be a very useful tool to deconstruct a region, both theoretically and methodologically, when conducting research. Paasi (1986) distinguishes four phases of the process of institutionalization of regions: (1) the assumption of territorial shape of a region; (2) the development of symbolic shape of a region; (3) the development of regional institutions (institutional shape); (4) the establishment of region as part of a regional system.
The order of these phases is purely theoretical. In practice, these phases can take place simultaneously or in a different order varying according to different purposes (types) of region (cf. e.g. Kašková, Chromý 2014). However, each phase and its reproduction mutually influence the reproduction of all other phases (e.g. Messely et al. 2014). Apart from that, some scholars try to establish some chronological order among the four phases; e.g. Zimmerbauer (2011) sees as pivotal the necessity for the emergence of territorial and symbolic shapes. These should provide a basis for further development of regional institutions, and, potentially, clear the way for regions to get embedded in a regional system.
Nevertheless, there are other modalities showing how to approach the process of institutionalization of region both theoretically and methodologically. The understanding of region as an institution can be dated back to Paul Vidal de la Blache (1994: 26), who argues that region is substantially shaped by human action. In his paper on "American West", Donald Meining (1972: 161) suggests that regional research should focus on four regional features: (1) population; (2) circulation; (3) political areas; and (4) culture. According to Meining, region must be studied in all its complexity, including not only its socio-cultural features, but also all the relations within and between regions. However, under the term "political areas" he understood merely the basic administrative territories that can be replaced, in a more recent perspective, with a region defined as a contested arena of political decisions.
Other authors, too, have listed criteria crucial for the emergence of region. Some of them approach the region from a strictly socio-cultural position. Bill Lancaster (2007: 24) regards common space, language, culture, economy, political movements, traditions, and relationship to the nation-state as key unifiers of a region. His list represents a typical example of scrutinizing region as a community which has revived the humanistic tradition in geography (Tuan 1990). In the same publication, Charles Phythian-Adams elaborates the crucial regional features in more depth (Phythian-Adams 2007: 8-9), listing seven key features for defining a regional community: (1) concentration of population; (2) hierarchical structure of settlement; (3) intra-dependence of region; (4) self-identifying, but interlocking neighborhoods; (5) regional economic or political interest counterweight against national power structures; (6) a demographic continuity (of indigenous families); and (7) a regional sense of belonging together.
Likewise, scientists based in different (non-Anglophone) scholarly traditions deconstruct region in very similar ways (Riukulehto 2015). Borders, landscape, language, group solidarity, and administrative autonomy are the main criteria defining region mentioned by Desiderio Fernández Manjón (2010: 68-69) based on José Ortega y Gasset's assertion that any human being is defined by her or his circumstances (as cited in Riukulehto 2015: 10). Even though the criteria listed above differ from one study to another, a closer look reveals many significant intersections among the aforementioned concepts.

Reflection of constructivist approaches in the Czech literature
Similarly to the world literature, region is often reduced to a mere category or context for the study of various phenomena in Czech geographical literature (e.g. Kůsová 2013;Novotná et al. 2013;Ženka, Pavlínek 2013). In such researches, region is scrutinized as a given, unchangeable and static entity and its nature is not further theorized.
Moreover, the strong and persistent influence of nomothetic approaches to a region can be witnessed in Czech geographical literature. The nomothetic delimitation and evaluation of regions represents one of the most important branches of contemporary regional research in Czechia (e.g. Halás, Klapka 2010;Kraft et al. 2014;Hampl, Marada 2015). It is based on the construction of regions according to various relations in space (e.g. commuting, transport, etc.). This kind of regional studies has a long tradition in Czech science and its roots can be traced back to the doyen of Czech human geography Jaromír Korčák, who broadly discussed the possibility of delimitation of functional regions. Approaching them from the cultural-geographical perspective, he called them přirozené krajiny [organic landscapes] (Korčák 1934: 421). Korčák's work triggered a diversion to the nomothetic exploration of the concept of region in Czech geographic literature although Korčák (1934: 433) himself admitted that the regions he defined were very similar to those delimited by Václav Dědina. Dědina, who was influenced by Paul Vidal de la Blache, defined regions with the use of the "basin and valley concentration" method (Dědina 1921as cited in Korčák 1934. Despite the aforementioned, approaches of new regionalism gained strength in Czech regional research in the last decades. Their central point is the conceptualization of region as a historically contingent process. However, this concept has been introduced already 30 years ago, the publication of Pavel Chromý's the paper titled Formování regionální identity: nezbytná součást geografických výzkumů [Formation of regional identity: A necessary part of geographical research] (2003) can be considered a milestone in the application of this concept in the Czech academic environment. The author de facto introduces the Anssi Paasi's concept of the institutionalization of region (Paasi 1986) into the Czech literature and milieu. Many other papers, books and theses using this conceptual background or methodological framework have been published ever since. Most Czech (and Slovak) authors focus on the analysis of the inhabitants-territory relationship and its measuring (Nikischer 2013;Bucher, Ištoková 2015) or on the study thereof in a specific context such as second housing (Fialová et al. 2010) and peripheral regions (Chromý, Janů 2003;Chromý, Skála 2010). Special attention is also paid to its differentiation based on the continuity of socio-cultural development, especially in the context of areas affected by the expulsion of Czech Germans after WWII (Osoba 2008;Chromý et al. 2009;Šerý 2014). In connection with this recently introduced conceptual framework, many researches deal with individual partial shapes of region and their imprints in people's consciousness. These comprise in particular names, symbols and logos (Šifta, Chromý 2014;Semian 2016a;Semian et al. 2016), mediated regional images (Kučerová et al. 2016), regional institutions (Kašková, Chromý 2014;Chromý et al. 2014), or people's perception of historical borders (Siwek, Bogdová 2007;Šerý, Šimáček 2012). It must be said that many of these works incorporate Paasi's concept uncritically. They fail to further elaborate or critically discuss it in view of more recent regional conceptualizations. I can claim with a bit of a hyperbole that numerous works keep scrutinizing regions as a mere category which serves as a context for the study of various phenomena, namely "regional identity", i.e. an articulated emotional relationship to a region. The region is either poorly defined or reduced to a specific unit in these researches (e.g. Nikischer 2013;Štefánková, Drbohlav 2014;Bucher, Ištoková 2015;Ryšavý 2015).

Societal production of region
It is clear, the further theorization of the concept of region is necessary. From above mentioned, one can conclude many approaches to the concept of region have very similar bases building on the understanding of the region as a process, as a construct in flux. The region thus defined is usually addressed within three mutually interdependent dimensions: (1) given: the way "regional assets" are present and organized; (2) made: the way "regional assets" are produced and mediated by various regional actors having different goals; and (3) perceived: the way inhabitants perceive "regional assets" and establish bonds to a region. Here, the term "regional assets" is used to encompass all tangible and intangible, environmental, natural, cultural, political, economic and historic features and components of any particular region. It is also clear that an institutionalized region is something more than a mere sum of regional features (Nay 1997). Regional identity is thus one of the key phenomena in the study of regions. "Obviously, regions can be identified and characterized by the use of regional identities" (Riukulehto 2013: 45). This implies that not only regional actors, but also regional community are crucial for the research of regions.
One can find the link to Lefebvre's constructivist framework of the societal production of space (Lefebvre 1976;1991;cf. Brenner, Elden 2009) which can be adapted to the societal production of region. Thus, region can be approached as a product of the interaction between three distinct layers: (1) Practice of region -the way people interact with their surrounding environment in their everyday life. This interaction reproduces the region as a relation and lived space. The everyday practices are influenced by the existing regional representations and they also produce information which affects human perception.
(2) (Re)presentation of region -the way a region is mediated by various groups of actors with various goals and through various media. This means that information is produced and mediated the way someone would like others to perceive the region. This layer involves generating mediated images, but also various constrains that can be imposed on everyday life depending on the level of institutionalization.
(3) Idea of region -the way people perceive a region as a spatial entity. Their perception is influenced by many aspects, both outer and inner, wherein the quality and quantity of information are very important. Such perception also has an influence on and is influenced by everyday activity of each person. This interaction is reproducible in time and space as all the three layers undergo perpetual changes. All the three components are in a constant interaction and mutually influence one another during the process of reproduction. And so, region is a social construct that is reproduced through, de facto, constant teetering at the interface of these layers. Although it can exist lacking one or even two of these layers, it becomes fully developed (institutionalized) only through the blending of all three of them. A fully developed region is more resistant to disappearance. Furthermore, for analytical purposes it is possible to interlace this framework with the three Paasi's shapes of region: territorial, symbolic and institutional. The fourth shape, termed as "regional embeddedness", is de facto substituted by the interaction of those three layers which gives the identity to the region (Fig. 1).
The proposed concept integrates functional, normative and perceived regionalization in a complementary way. The partial components have been widely discussed, yet their mutual relation has been at the margin of regional scholars' interest. The study of this relation is necessary in order to understand the process of the formation and reproduction of region as well as to explore the general idea of any particular region in question.

Conclusion
There is a plentiful complex of regional conceptualization in contemporary geographical literature. Nevertheless, many of these concepts explain the process of regional formation and reproduction only to a limited extent. Any complex approach to region must incorporate three levels of region: "given" (practice of region), "made" (representation of region) and "perceived" (idea of region). It is possible to explore the general idea of region by way of studying these levels. The change of the society's value orientation comes along with the general changes of society. Thus, the meaning, the importance and the gist of the need of "moorage" are also changing and this change brings into question human relation to place and region (the sense of belonging). Therefore, the third, perceived dimension has been gaining both strength and importance.
With the use of the inhabitants' regional consciousness and their sense of belonging it becomes possible to introduce some order and clarity into the plurality of regional manifestations. Thus, the complementarity of images of regions gets synthesized through people's perception in the collective consciousness. It is important to point out that such synthesized region will not be universal but shared, representing the most common image of a produced, mediated and perceived region. Moreover, regions delimited according to the general idea appear to be widely accepted by the public, and thus, they have a potential to become a unifier Fig. 1 Societal production of region. Source: author's own framework based on Lefebvre 1976;1991;Paasi 1986. that builds the community, ties it together and stimulate the inhabitants' willingness to collaborate in regional development and social life. Or, at least it represents a good trademark suitable for the regional promotion often addressed by regional actors.
Region is truly an intricate concept. Thus, this paper should be seen as a contribution to the never-ending discussion on the conceptualization of region and not as an attempt to come up with a universal answer. Such a universal conceptualization is neither desirable nor possible. Nevertheless, a better understanding of the concept of region is still more necessary and relevant especially due to the increasing pressure on the applicability of regional research (regional development, tourism management, etc.). Therefore, the discussion on region should focus more on the understanding of the process of formation and reproduction of region instead of a mere delimitation and evaluation of regions. Further, closer insight into the interrelation of above mentioned three levels of region seems to be crucial from the view of the contemporary state of art.

Aknowledgements
This paper was supported by the Czech Science Foundation under project No.: P410/12/G113 "Historical Geography Research Centre". It builds upon materials published in author's dissertation thesis (Semian 2016b). Further, author would like to thank to Pavel Chromý for the supervision and Iveta Čtveráková for the help with the graphics of figure 1.