Borders that relate: Conceptualizing boundaries in relational space

Introduction

It is not only from the perspective of numerous theories of globalization that a world without borders seems increasingly to be becoming a reality; theorists also often write about space as if the question of borders is inconsequential when space is viewed as relationally constituted. Jan Aart Scholte (1997) detects three dominant perspectives in the globalization debate: ‘The first identifies globalization as an increase of cross-border relations. The second treats globalization as an increase of open-border relations. The third regards globalization as an increase of trans-border relations’ (Scholte, 1997: 430). In globalization theory the increasing significance of a ‘space of flows’ (Castells, 1996) goes hand in hand with a tendency in spatial theory that Jeff Malpas critiques: ‘space appears as a swirl of flows, networks, and trajectories, as a chaotic ordering that locates and dislocates, and as an effect of social process that is itself spatially dispersed and distributed’ (Malpas, 2012: 228). As distinct from mathematics and the natural sciences, which at the beginning of the twentieth century dealt with the relativity of time and space, in the social sciences and humanities it was in the second half of the twentieth century that space could no longer be treated as it so often appears in everyday life, namely as a container of pure materiality that stands in contrast to the world we describe as social (Harvey, 1973; Löw, 2016). Although Georg Simmel had already published a ground-breaking text on space and spatial ordering in 1908, it was only in the 1970s that the permeation of the social realm by spatial forms was addressed in French sociology, for example by Pierre Bourdieu (1977 [1972]), Henri Lefebvre (1991 [1974]) and Jean Rémy (1975), and in Anglo-American geography by David Harvey (1973). It was through this process that space was introduced as a relational arrangement. The insight that space has to be understood through social practices as a relationing of objects, places and (groups of) people, leading to new forms of human and cultural geography, as well as the gradual development of a sociology of space. It is perhaps not surprising that the question of borders was left relatively under-explored during this development, widely regarded as a paradigmatic change of theory: the spatial turn. Nigel Thrift, for example, writes: ‘I would claim, the “spatial turn” has proved to be a move of extraordinary consequence because it questions categories like “material”, “life” and “intelligence” through an emphasis on the unremitting materiality of a world where there are no pre-existing objects’ (Thrift, 2006: 139).

Boundaries seem to challenge this conceptual advance. Doreen Massey, for example, finds that a systematic distinction between space and place is rather dubious (Massey, 2005: 68) and argues that a place is a ‘particular constellation of social relations’ instead ‘of thinking of places as areas with boundaries around’ (Massey, 1994: 154). And as the ‘root of any approach to space’, Thrift asserts ‘there is no such thing as a boundary’ (Thrift, 2006: 140). However, the theory of space is not so much characterized by the fact that boundaries, borders and frontiers are explicitly declared to be irrelevant objects (on this point see Schroer, 2006: 23, 68) as by the fact that they simply receive little attention. Just as the sociology of borders has developed largely apart from the sociology of space (for a critical summary see Eigmüller, 2006), the sociology of space has largely neglected to include the phenomena of bordering in its analyses. The insight that spaces and places develop from relations seems as though it might be threatened by borders or perhaps to be theoretically incompatible with it. Empirically, however, it seems that in certain contexts, at certain times, for certain goods or people spaces can become containers by virtue of borders (Harvey, 2008). Refugees from Syria, for example, who in 2016 are stranded in Turkey or Greece because escape routes are ruptured by border closures, are certainly able to confirm this. If social scientific space research does not want to simply ignore these phenomena, then it is challenged to describe the constitution of such empirically demonstrable containers on the basis of a relational concept of space. We conjecture that without a concept of border this task cannot be solved. ¹

The fact that it was possible to develop a theory of space largely neglecting borders is also a result in literature that grants little distinction between space and spatial metaphor. Theories that are used to develop a heuristic for understanding various hierarchies, or to explore simultaneous positioning in terms of power politics, or for imagining dynamic flows, are undoubtedly insightful and important, but they remain metaphorical as long as no information can or need be given as to ‘where in the world’ this is happening. If space is deployed as a metaphor for purposes of reasoning, for example to argue against a logic of progress and for the simultaneity of different developments, it can indeed be irrelevant or distracting to focus on boundaries. However, an analysis of spatial phenomena in everyday life and on-the-ground experience, by contrast, often requires a precise concept of boundaries and borders.

The following article therefore focuses on the question as to how the phenomena of borders can be integrated into relational concepts of space. To this end, we begin by articulating the distinction between space and spatial metaphor in order to more precisely mark the realm covered by a sociology of space. Thereafter, we outline our proposal for a conceptual framework of space (for details, see Löw, 2016; Weidenhaus, 2015) so as to define borders within this framework as a possible relational arrangement of spaces. In this context, borders must not be conceived merely as walls or fences; rather, borders place at least two spaces in specific relation to one another. Lastly, we would like to test the applicability of our concept of space and border by using an empirical example, the dynamics of national border controls in the context of the refugee movements in the last quarter of 2015. We conclude with a brief summary of our arguments and findings.

Space and metaphors of space

Almost all social phenomena can be described with the help of spatial metaphors. Family structures, communication networks, the structure of organizations and class theoretical analyses of society can be presented through a heuristic of spatial metaphors. We assume that a sociology of space is appropriate for research when the ‘where in the world’ – the places at which the goods and living beings of space constitution are located – is in fact relevant. It is only in this case that we are concerned with spaces rather than spatial metaphors, and it is only here that the appropriate object of a sociology of space can be found. The use of two examples clears up the difference between space and spatial metaphors.

Space changes when the elements constitutive of space (social goods or living beings) change their place. When in contrast the hierarchy of a family structure is depicted using spatial metaphors (for example when we claim the mother or father is at the top or head of the family), the spatial metaphoric structure is obviously unchanged when the mother moves physically in space, not losing this ‘position’ when going downstairs to the cellar. In the spatial metaphor, her position is unchanged, whereas in the constitution of space, she has moved lower. Alternatively, a socio-spatial perspective on family in modern western societies might grant the constitution of the ‘home’ a central status, but would seek to ask whether there is an actual spatial example that anchors the hierarchic order in the spatial order (for example, if the mother has the largest room on the top floor for her study, whereas others have no such room of their own). Objects relevant for socio-spatial research are foremost those realms in which a ‘where in the world’ can be differentiated (Werlen, 1993), where physical space is undeniably relevant.

The virtual and material worlds of the Internet provide a good example to illustrate the distinction between space and spatial metaphor in its full complexity. Spatial metaphors are an excellent way to describe the Internet and its uses. A chat room or text-based portal for chatting in an online gaming community, for example, is a user interface of a communication program. In the first place, it is irrelevant where the chat room is located (the physical location of the server on which the program is running). But more importantly, the relative spatial positions of the communication partners are of limited consequential significance. If chat rooms and portals necessitated relative spatial positions, bodies would be displaced, and anyone engaging in a chat room would have to first look for a place, but of course this is not the case. In the chat room it is irrelevant who is seated to the left and to the right. For the functioning of the communication space both the ‘where in the world’ and the relative positions of the participants are generally insignificant. Hence, the room in ‘chat room’ is a spatial metaphor. To understand social processes in a chat room, a communications analysis is more useful for most enquiries than is an analysis of space. In short, the Internet is overall not a space, but rather a communication system that embeds spatial metaphors.

Nonetheless, there are research questions concerning the Internet for which expertise in the sociology of space is useful or even indispensable. There are two main perspectives from which the space of the Internet is relevant: first, when virtual worlds are simulated in the Internet in such a manner that the relative positions of the elements are in fact significant (generally in games), and second, when questions are posed so that the location of the material infrastructure of the network comes into focus. ²

On the first case: within many video games, it is essential what and who is located where. Avatars move along different paths through virtual worlds and continuously have to pay attention to their position in relation to their surroundings. Thus, relative positions are simulated in virtual spaces. ³ It is of interest to the sociology of space that, for example, these virtual worlds are in the first place free of spatial restriction found in the real world. Gravity is in the first place irrelevant to mobility or to the architecture of simulated buildings, just as every element can in principle be transferred immediately to any place in space. In fact, however, the programmers generally make an effort to program the most exact possible depiction of the real world; they have highly complex physics engines running permanently in the background to undermine the potential boundlessness of the virtual worlds in order to bring virtual spaces as close to our experience of the physics of the space of the offline world. Subsequently, the games often provide the opportunity to transgress these boundaries, which were so arduously and artificially drawn. In the course of a game, a player’s avatars might learn to fly, to be ‘beamed’ or to look through walls. It is precisely the transgression of boundaries that seems more compelling than boundlessness.

In the second case, the ‘where’ of the material infrastructure of the Internet becomes relevant. In Germany, for example, the user who wants to call up a YouTube video is often faced with the message that this content is not available in this country. The cause is a relatively restrictive, territorially limited copyright law. But if the user manages to conceal their own IP address or to use one that gives the provider the misapprehension that they are at some other location in the world by drawing on an IP address abroad, then the video can be viewed. With respect to the investigation of Internet crime, the spatial perspective is unavoidable, traced along the distribution of network resources and the real locations of servers in bordered jurisdictions. Similarly, questions of access to the Internet and thus to options in the social world can benefit from a spatial perspective when researchers analyse the spatial distribution of wireless and wired access to the Internet.

Thus, a sociological study of the Internet is not always dependent on a perspective grounded in the theory of space. As long as only communication is in focus, expertise in other fields suffices, even if spatial metaphors are frequently used and useful. In certain perspectives, however, the spatiality of the Internet does indeed become relevant far beyond a metaphorical use of space. In these cases, the Internet becomes an object of study in terms of sociology of space.

In summary, this articulation of space and metaphor, through the example of the Internet shows how the objects of spatial sociological research can be usefully analysed in terms of ‘where in the world’ and that the relative positions of the elements constitutive of space are sociologically significant. In case of a change of the relative positions, space itself is changed; if not, we are not going to find benefits in spatial theory, but perhaps we might still find researchers wielding spatial metaphors. Moving forward, the objects of our following pages are spaces, not spatial metaphors.

Space as a relational arrangement …

Relational concepts of space have become increasingly influential in spatial theory (Hubbard and Kitchin, 2011: 3; Malpas, 2012). Spaces are the ordering forms of proximity; we understand spaces as the result of operations linking physical objects. Questioning a stark division between human bodies and objects, we define spaces as relational arrangements of social goods and living beings in places. To perceive something as space, two things are required: first, an operation of placement of objects in places (spacing), and second, operations of conceptual synthesis and making sense of the relational significance of this spacing in this space. Spaces can always be grasped by way of the question ‘Where?’ Both the operation of synthesis and spacings are, of course, often institutionalized, that is, they are socially pre-structured and shaped by conventions. It is not simply the case that everybody spans their own individual space; but rather, the ways spaces are perceived, experienced and remembered, and how people place themselves and their objects is shaped and pre-structured by institutions. Where and how what is stored or built is a social process; this also applies to when objects are experienced as near or far, associated or unconnected and so on, and accordingly synthesized.

With this relational concept it is possible to ask about various arrangements that form spaces, for example as a network, region, topology, scale or territory (Jessop et al., 2008; Mol and Law, 1994). Spaces can be distinguished according to symbolic and material dimensions, they require different placements and are based on varying operations of synthesis, but they are all spatial formations that have an ordering social effect. Space emerges from the subject’s operation of synthesis and placement (in relation to institutionalized structures that shape, determine or inflect what is possible). But since subjects themselves are in relations to other subjects, space is not the product of a two-pole relation between subject and objects; and, as Hubert Knoblauch (2016) emphasizes, it is not only the product of relations between subjects or between objects, rather spaces emerge from triadic relations. Spacing takes place not only according to convention and subjective judgement, but also reciprocally by virtue of subjects dynamically related to each other in the course of action (Christmann, 2015: 98; Knoblauch, 2016). Accordingly, surfaces, layers, boundaries and openings emerge from the relation of the objectified formations of human and other corporeal bodies, producing specific social meanings, both shared and individually inflected. The production of space must in this sense always be conceived as bodily. The place constitutes the third factor in the triadic relation between synthesizing subject and the object(s) or subject(s) in spacing. The place is a point, but also a relational location. Its meaning refers to other places and to other dynamic subject positions.

Thus, we would like to maintain a distinction between space and place on the conceptual level: we understand places as the products of the placement of elements constitutive of space. However, this is not meant to classify social entities of the empirical world in a place–space pattern. For example, whether a marketplace is interpreted as a space or a place in a study is dependent on the perspective. If the study is focused on the space ‘city centre’, then the marketplace can be understood as a place in this space. If, however, the marketplace itself is of exclusive interest, then it should be conceived as a space that for its part consists of elements constitutive of space (sales stalls, fountains, people and so on) at certain places.

… with borders

If the constitution of spaces is understood as the synthesis of certain elements, specifically goods and living beings at places with the relative positions of these elements taken into consideration (Löw, 2016), then spaces appear in the first place as networks into which any other goods and beings can at all be ‘smuggled’ and go unnoticed as long as they are not part of the spatial synthesis. The advantage of such a conception is that it easy to imagine how such networks overlap with each other, how certain elements are a part of several spatial constitutions and thus how several spaces can emerge at one place. Precisely this is the strength of relational conceptions of space as they were developed by Thrift (2006) and Castells (1996).

However, spatial constitutions can be empirically observed that have a highly exclusive character, at least with respect to specific elements. At a place where Spanish law applies, French law does not apply; when this exclusivity is challenged, conflict results. Exclusive spatial constitutions of this kind, such as nation-states, have one essential feature: they are bounded by borders. Hence, the question arises as to how boundaries and borders are developed and how they can be integrated into the framework of a relational conception of space without abandoning the advantages of the ‘spatial turn’ and returning to the container model of space (see Fuller and Löw, this issue). The challenge is to think of space as relational arrangements with borders.

The first decisive point is to interpret borders themselves as relations (Weidenhaus, 2015: 46), but not as relations between individual elements constitutive of space, but rather as relations between spaces themselves. Whenever a border is in focus, at least two spaces are constituted and set in relation to each other by this spatial boundary. It is possible for this relation to be highly asymmetrical because the second space can be conceived as ‘outside’ or as ‘surroundings’. It is often the case that only the space within a set of borders is of interest, while the external space can remain blurry, indefinite or ignored. Nonetheless, this second space plays a constitutive part in the definition of the first. The empirical example outlined below will show that the production of an indefinite and unspecific external space can be the goal of the constitution of borders. To be specific: in an indeterminate exterior there is no need to deal with complexity, whether spatial, communicative, organizational or moral.

As a first step in the conceptualization of borders, we can thus maintain that in the framework of a constitution of spaces in which corresponding but distinct elements are linked with each other, a commonality is also constructed between these elements and separated from what is different (other spaces). The operation of synthesis thus also contains an operation of differentiation. Nevertheless, different emphases on synthesis and differentiation can be observed in various space constitutions. The more significant the synthesis of complementary elements in the constitution of two or more spaces, the more these spaces appear to be networks. By contrast, boundary constructions become more significant and a territorial space emerges to the extent that the differentiation from the outside by means of the production of interior homogeneity, or at least commonality, is deemed more important. ⁴ In the synthesis of nation-states, for example, differentiation traditionally plays an important part (language, history, culture are often appealed to as internal common ground). For this reason, it seems unimaginable to think of nation- states without clear borders. The spaces of modern metropolises, by contrast, are synthesized on the basis of various complementary elements (Simmel, 1950 [1903]); boundaries and borders seem of less importance (Held, 2005).

The operation of differentiation thus emphasizes commonalty within and difference from the outside; everything within the boundaries has a common feature (for example, being German and not French or Swiss), which thus has a homogenizing effect. If, however, the differentiation from other spaces plays a subordinate part, the differences between the elements constitutive of space remain. Such ensembles can be constituted and perceived as spaces without clear demarcation of boundaries. Thus, the construction of borders is more important to the extent that the determination of the relationship to other spaces is more important in the constitution of space.

Boundaries can be of quite different types according to what is related by them. Whereas national borders relate realms of application of law to each other, landscape boundaries will relate geological formations or climatic conditions to each other. At this point, the different ways of constructing spaces can be described as ideal types by means of a thought experiment. In the constitution of a space, a geologist could pay attention to common features on the surface of the earth. She might examine the rock at various places. As long as she finds granite, for example, she could synthesize these places to a space: a 200-million-year-old low mountain range of volcanic origin. If she suddenly encountered slate (developed 50 million years ago from the sediments of a marsh landscape), she would differentiate two spaces and construct a landscape boundary at this place that she would mark in a geological surface map. Networks are constructed in a completely different way: in the determination of the space of the global financial market, one would search for the different elements that are necessary for this market. A banking house at this place, a computer centre at another and the top manager in his home office could be synthesized to a space that need not conceptualize an outer border. Spaces thus need not be enclosed in border constructions. Borders can, however, be constructed to determine the relation between various spaces.

This example reveals a further important aspect of borders: they specify. Spaces constituted as territories with borders only relate particular aspects of the world to each other. Rock formations, the weather, certain commodity flows in the European Union and even the mobility of certain people among the Schengen countries remain unaffected by the national border between Germany and France. This leads to a permanent overlapping of different spaces. Even if social reality were to consist only of territorial spaces (which is not the case), the world would not simply be divided into territories next to each other; rather, even territorial spaces enclose and overlap each other. Borders are thus relevant only to certain objects or people in certain contexts. Borders are rarely, if ever, completely closed to all objects and persons, but are better understood as a membrane allowing different permeations.

Against this background, it can be shown that borders organize circulation and in certain cases it is borders that produce circulation (Dittgen, 2003). Some areas of Mexico can only be spatially understood when taking into consideration various cross-border flows of people, commodities and money, such as remittances sent, for instance, by those living and working in the United States (Pries, 2010). In 2006, for example, 660,000 people and 12,400 trucks traversed the border crossings between Mexico and the United States each day (Migration Policy Institute, 2006). In addition, there is also illegal cross-border movement, which is difficult to estimate (Andreas, 2009). As producers of circulations, the borders of Great Britain and Panama institutionalize different levels of taxation, resulting in a capital flow towards Panama. From this perspective, borders can be interpreted as generators and organizers of circulation. Nonetheless, we would like to argue that such an interpretation misses the essence of the border because it skips an important initial step in analysis. Borders arrange spaces in relation to each other by controlling the circulation of certain things, goods or living beings in certain contexts for certain times, even prohibiting certain circulations and seeking insofar as possible to separate the spaces. A border construction is always predicated on separation. This function of separation can seem so obvious that the social sciences sometimes lose sight of it; or the connecting effect of the border is regarded as coeval with the separating function in the framework of much work in border studies. The statement that ‘they [borders] connect as well as divide’ (Mezzadra and Neilson, 2013: 4) is deemed a truism in this context. But in the first place, the borders of Great Britain and Panama divide the realms of application of the taxation law of these two states. A company domiciled in London cannot submit its tax declaration in Panama City. For this simple reason, it ‘has to’ set up a shell company in Panama and then transfer its profits there. The division between taxation realms constructed in terms of territory and lacking overlap is thus the cause of the circulation. Hence, what circulation is generated or organized by borders can only be recognized when we understand what has been divided under what conditions by a border. An analysis of the interruption of circulation, as intended via a border construction, should always be at the beginning of a study of borders. If it cannot be stated what is supposed to be divided by a border under what conditions, there is no border. Circulation, by contrast, also takes place without borders. It is important to note that the division by borders exists only as a claim. Whether such a division can be empirically maintained or what circulations are developed to undermine it must be studied in the framework of an empirical analysis of the border being investigated. It is nonetheless valid that every border construction begins with a claim to differentiation and division from other space constitutions, a claim to control and, if indicated, interruption of circulation. The specific links and circulations that can be generated by borders are effects of this division.

Moreover, an approach to borders informed by spatial theory makes it possible to conceive borders as spaces without mixing borders and space on the conceptual level. When borders are linked to functions of control and separation that occupy space, the border itself becomes a space (see e.g. Lebuhn, 2013). This takes place, for example, in transit zones when goods or people are temporarily detained at the border (see the empirical example below). It is self-evident that the analysis of such border areas can gain from the perspective of spatial theory. Nonetheless, we would like to maintain an analytical division between space and border allowing very different phenomena to come into sight. On the one hand, not every border benefits from being analysed as a space. There seems to be gain in analysing the masonry boundary between two rental dwellings, that is, a wall with the function of a border, as a space. On the other hand, border areas such as transit zones themselves have their own bounded borders; the analysis of these borders would result in conceptual messiness if each border were conceptually equated with space. The study of the border area focuses on the goods and living beings in their relational arrangement by means of which this space is constituted. The exploration of the border of the border area, by contrast, brings the spatial relations between the border area and the spaces surrounding it into focus. It can then be asked who and what enters and leaves the border area in what direction and under what conditions. Border and space therefore do not coincide on the conceptual level even though border spaces emerge as empirically relevant in some cases.

In summary, it can be maintained that borders are to be understood as relations between spaces. Border constructions always involve the constitution of at least two spaces. When in the framework of the synthesis of space the differentiation from other spaces plays a significant part, this difference can be spatialized through the construction of a border; this tends to result in territorial constitutions of space. Borders relate spaces each in its own specific way. According to what differential markings they involve, they divide defined goods or people in defined contexts at defined times. For these specific object realms are constructed with the claim to make a division, to stipulate ‘so far and no farther’ – they are intended to control and interrupt circulation. As a consequence, borders organize and generate circulation at the same time, thus having a linking effect. Linkages created by borders are thus logically subordinate to a border’s role in dividing. The links created by borders can only be understood in the context of their dividing and controlling function. Someone who wants to make a spatial separation will construct borders; someone who only wants to connect can do without them.

Against the background of this conceptualization of border constructions in the framework of relational spatial theory, we would like to briefly present a case study that focuses more precisely on some of the social differentiation processes in the construction of national borders. This is not a comprehensive analysis of the border regime under discussion, but rather an exemplary reconstruction of specific differentiations and interruptions of circulation that were carried out in the last quarter of 2015 on the borders along the ‘Balkan route’.

Case study: Border constructions in and around Europe

On 3 December 2015, the United Nations Refugee Agency (UNHCR) withdrew its personnel from Indomeni, a Greek village on the border to Macedonia. Within the borders of the European Union, the United Nations no longer saw any possibility of guaranteeing the safety of its personnel, who were supposed to provide the thousands of refugees from Syria, Afghanistan, Iraq, Morocco, Eritrea and Bangladesh with blankets, water and food. This step was preceded by the introduction of border controls in Germany and Austria and by the partial closure of the borders of Croatia, Slovenia, Serbia, Hungary and finally Macedonia (except for Syrians and Afghanis), so that for many people in Indomeni there was nowhere to go, no way to proceed. After three days of protest against the closure of the border to Macedonia, violence escalated under the pressure of hunger and cold (see Süddeutsche Zeitung, 4 December 2015, section ‘Bayern’: 6). Border constructions in and around the European Union, the Schengen states and the neighbouring countries are decisive for people’s chances to live and survive. In the last quarter of 2015, these border constructions became increasingly dynamic. Under the influence of large-scale flight, above all from Syria and Afghanistan, where civil wars radically endanger the lives of civilians, the practices determining who is and who is not permitted to cross which border changed almost weekly.

A media analysis of two leading German newspapers in this period can provide some insights into these dynamics. The primary goal is to decode, at least in part, the discursive constructions of the semantics of borders in this context and the rhetoric legitimating their closure. To this end, a corpus was complied of over 60 articles published in the last three months of 2015 in the newspapers Frankfurter Allgemeine Zeitung (FAZ) and Süddeutsche Zeitung (SZ) with the word ‘Grenze’ (border) in the title.

At the beginning of 2015, mobility was still possible without border controls in large parts of the European Union (among the ‘Schengen states’ [Figure 1]). At the end of 2015 and in the beginning of 2016, controls had been reintroduced at most border crossings and some borders were closed for people of certain nationalities by means of fences and walls (Figure 2).

OPEN IN VIEWER

Figure 1.

Schengen states.

OPEN IN VIEWER

Figure 2.

Refugee routes and border controls.

In the first place, there are two central functions specific to national borders: they territorialize laws and control mobility. The territorialization of law is associated with the state’s claim to enforce this law. The borders thus define the space within which a state deems that it can give force to its law. The control of mobility proves to be first and foremost the control of access to laws and rights (on entry) or the detention of people under certain legal conditions (on exit). The border controls within Europe and on the external frontiers of Europe almost exclusively serve the first goal. The point is to control access to rights within the European Union and its nation-states; what is discussed is almost exclusively the control or the prevention of entry. ⁵

It can thus be maintained that national borders reduce complexity by unequivocally defining the space within which a state takes measures to enforce its legal norms. This act of territorialization hugely increases the chances to give force to law. At the same time, the border ensures that complexity is retained inasmuch as a second space is constituted that from this state’s perspective remains legally exterior. From the state’s perspective it is often contingent or even irrelevant whether and which law has validity beyond the borders. The construction of borders and thus the territorialization of state spaces thus pragmatically increase the chances of endowing the law with force.

In the analysis of the newspaper material, however, one discursive mode regarding European borders stands out; it is often implicit, but significant. Not only can law be territorialized through borders, borders also territorialize morals. One of the most frequent argumentations for the closure of borders is that appropriate, humane accommodation cannot be provided in the receiving country (for example FAZ, 20 October 2015: 20) without consideration that the moral claim to minimal standards of human accommodation and provision does not end at the borders of a nation-state. The legal proceedings in Germany against those who assist in facilitating illegal border crossings (German: Fluchthelfer, here translated as escape agents) for refugees also bear witness to a territorialization of morals. In these proceedings, the judge’s moral assessments with respect to the wellbeing of the refugees (how safe are the refugees in the transport vehicle?) obviously play a decisive part (cf. FAZ, 20 October 2015: 4). However, a comprehensive moral assessment of the situation would require, on the one hand, that the refugees’ alternatives be taken into consideration (perhaps if escape agents prevented their deaths), and, on the other hand, that we ask whether the people assisted by the escape agents might, for example, have a claim to political asylum in Germany so that the agents enabled them to avail themselves of their right. In concrete terms this means: in order to judge in moral terms the escape agent’s decision to take another family into an overcrowded transport van or rubber boat, the threat to the family that might otherwise have to remain behind is also taken into consideration. Similarly, it would have to be considered whether the people who are transported by the escape agent have the status of civil war refugees according to the criteria of the United Nations, or whether they are entitled to asylum, but would not have been able to make a valid claim to this status without the help of the escape agent. However, nothing is reported about the legal proceedings of these questions within the courtroom, and journalists were not calling for these moral territorializations to be examined. While at the same time, the character of the actions of the escape agents is treated as morally objectionable, and this reprobation certainly contributes to their sentence. Let us again emphasize the distinction: within a national legal system it is legitimate to convict escape agents due to a violation of law. But it is a problem to judge their action morally when the moral claim is territorially restricted (if refuge in the receiving country is legally possible) or when it is assumed that a material assessment of a transport container or boat provides sufficient basis to assess the moral behaviour of escape agents. When moral questions end at national borders, the distorted image of the escape agent as a brutal exploiter, predominant in German media, becomes possible. During the entire period of the study, both of the media outlets under review used the negatively connoted terms Schleuser or Schlepper (human trafficker or smuggler), whereas those who facilitated escape from the former German Democratic Republic were habitually called Fluchthelfer (escape agents), even when they accepted hefty payment for their services. The territorialization of morals by drawing borders necessarily results in a double moral standard: one moral standard for inside, one for outside. If, however, it is assumed that basic norms (the right to life) can claim universalistic validity, then this form of double standard must itself be judged to be immoral. The extension of border policy from a territorialization of law to a territorialization of morals is itself morally precarious and profoundly consequential for those who are condemned. This transfer between law and moral seems to be more easily performed to the extent that the spatial constitution of outside remains unspecified and unclear. In this case, it is easier for media reporting to thrust aside the specificities of the conditions of what is outside the border. The exhibition Frontières in the Musée de L’Histoire in Paris in 2015–2016 demonstrates this impressively. A diachronic review of refugees in print media through the autumn of 2015 shows that at first individuals, families or small groups were in the photographer’s focus. Within a few weeks, the perspective changed, and masses of people are depicted from a wide-angled distance.

There is only one article in our corpus of newspaper articles that focuses on the fate of an individual. It is about Carlito, who illegally crossed the border from Mexico to the United States (SZ, 30 December 2015, section ‘Munich’: 7). To describe the situation in Europe, by contrast, de-individualizing terms such as ‘influx of refugees’ or ‘torrents of refugees’ are predominant, as are pictures in which masses of people can be seen. The spaces beyond the border would become more tangible in the stories and faces of individuals, but this is precisely what is avoided. What is outside remains undifferentiated despite the fact that the debate is centred on the border, where more than one space is constituted.

The establishment of transit zones may well be the most rigorous constitution of an unspecific outside. In the last quarter of 2015, the demand for the creation of transit zones for refugees at the German border was one of the most widely discussed proposals of conservative parties and politicians (see e.g. FAZ, 13 October 2015: 1; 7 November 2015: 10; SZ, 12 October 2015, section ‘Munich’: 6). In this case, the border is not merely the relation between two spaces, but rather is itself constituted as a space. The artificial creation of such an exterritorial frontier space generates a definite blurring of applicable law together with complete control lying with the executive of the state. In the artificial no man’s land of the transit zone, power can be exercised while law is suspended (for this reason, American torture prisons such as Guantánamo are established, as are transit zones, as exterritorial spaces in which, however, the law of another state also does not apply). In this specific case, the point is to be able to deal with applications for asylum without the legal proceedings prescribed within German territory. ⁶

Overall, it is conspicuous in media discourse on borders that there is more at issue than a pragmatic structuring of space, which can be implemented by means of border constructions; beyond this, national borders are charged with meaning to an extent far beyond their functional purpose. Beyond the territorialization of morals mentioned above, borders are endowed with an intrinsic value that obscures the genuine motives for border closures. Someone who claims that the border has to provide protection from immigration (e.g. FAZ, 15 October 2015: 1) seems not to be required to explain what exactly has to be protected behind the border.

In conclusion, it can be argued that borders can readily be used to spatialize social phenomena in an almost exclusively territorial form. Precisely this is the danger associated with their construction. Once a border has been institutionalized, it becomes difficult to ask whether it is reasonable to spatialize a certain phenomenon (for example, a more universal and less bordered set of morals) and whether the territorial form generated by border constructions remains appropriate, valid or justifiable.

Conclusion

The point of departure for the present article is the fact that in a relational sociology of space, borders are given little conceptual consideration because it is difficult to harmonize them with a relational understanding of space as an arrangement of objects, places and (groups of) people. Compensating for the lack of such a concept of borders in theorizing of relational space over the past decades, the centrality of spatial metaphors seems to suggest other spatial modes might be cited. But while these metaphors can be theoretically stimulating inasmuch as they make it possible to conceive simultaneity beyond borders, they mix up the sociological analysis of ‘where in the world’ with depictions of social order in terms of an imaging of space, without concrete spaces.

In this article, spaces are understood as the result of operations of synthesis between physical objects. Borders are then conceived as relations between spaces. Borders set spaces in relation to each other and close them to each other (or they organize circulation after a previous closure). Space is spanned when different elements are set in relation to each other in such a manner that they are structurally experienced as belonging together and localizable. The consequence is that at the same time, other spaces are separated. We argue that the operation of synthesis inherently includes an operation of differentiation.

What is of central importance is that different relative emphases on synthesis and differentiation are observable in different constitutions of space. The more central the synthesis of complementary elements of a space constitution is, the more these spaces appear to be networks. By contrast, boundary constructions become more significant and a territorial space emerges to the extent that the differentiation from the outside by means of the production of inner commonality is deemed to be more relevant. It is a matter of empirical analysis when spaces are perceived as unbounded and open or differentiated and closed.