Memory and Everyday Borderwork: Understanding Border Temporalities

ABSTRACT The field of border studies has traditionally paid little attention to questions of temporality, leading to criticisms over its presentism and lack of historical reflexivity. A number of recent publications have brought temporal questions more centrally into border research, examining the changing and historically contingent nature of borders. This article intervenes in this body of scholarship, using memory as a means of studying the past and present of borders. Bringing border studies scholarship into a more systematic conversation with memory studies, the article shows how memories of the past play an important part in the symbolic construction of borders, and that processes of remembering are central to how citizens produce borders in everyday life. The focus on memory and everyday borderwork allows to go beyond linear and uniform conceptions of time that have shaped the writing on border temporality. It draws attention to how time is ordered and interpreted in non-linear and multiple ways and how these temporal orderings confirm, extend or question the meanings of borders. The usefulness of studying memory in everyday borderwork is exemplified through an analysis of memory narratives in the Russian-Estonian borderland, based on extensive fieldwork and the analysis of 58 narrative life-story interviews.


Introduction
In a well-cited article published in 2010, sociologist and border scholar O'Dowd (2010) put forward a critique of the state of the field of border studies. He argued that much writing on borders didn't demonstrate sufficient historical awareness, leading, among other issues, to scholars' 'incapacity to recognise the "past in the present" as in the various historical deposits of state formation processes' as well as to 'exaggerated claims of novelty of contemporary border change' (O'Dowd 2010(O'Dowd , 1031. O'Dowd was not the first and certainly not the last scholar to criticise the lack of historical sensitivity and more generally the limited engagement with temporality in border research. Already in 1995, Alvarez put out an early warning about exploring borderlands in an 'ahistorical' way (1995,462), while more recently Ballinger has noted that 'much work remains to be done on the diverse temporalities experienced by those displaced/emplaced persons who live, draw meaning from and give meaning to borders ' (2012, 390). Also Horsti writes in a recent publication that 'the flattening out of temporality and context (in border research) reinforces presentism, "an unintended and unremarked privileging of contemporary concerns and dispositions" (Inglis 2013, 100)' (Horsti 2019, 2).
The processual turn in border studies -the study of bordering processes rather than borders as fixed and stable objects -has created awareness for the changing and unstable nature of borders. Moving away from static depictions of borders, scholars have increasingly conceptualised them as discourses and practices that configure understandings of place and communities. The border is more of a verb, a practice and a relation rather than a noun or object, as van Houtum has emphasised in a much-cited phrasing (van Houtum 2010). Even though the acknowledgement of borders' malleability over time is a crucial part in such processual conceptualisations of borders (see also Parker and Vaughan-Williams 2009), there are relatively few explicit attempts to conceptualise temporality within border research. As Hurd, Donnan, and Leutloff-Grandits (2017, 3) argue, 'it is not so much that time has been ignored in border studies, it is rather that, where it does feature, it is less privileged analytically'.
A number of recent publications have brought temporal questions more centrally into border research, examining the changing and historically contingent nature of borders. This article intervenes in this body of scholarship by using memory as a means of studying the past and present of borders. While there is a significant and growing body of scholarship on borders and memories (for an overview see Zhurzhenko 2011; for recent examples see Stokłosa 2019; Klabjan 2018), discussions on border memory have largely taken place independently from conceptualisations of border temporalities. Furthermore, much of the writing on border memories has been empirically oriented and case study based. It regularly fails to engage with conceptual questions or does so in a limited way, referring to one field of study only -either border studies or memory studies -rather than drawing on the strengths that scholarship in both fields has to offer.
Conceptualising memory and everyday borderwork, the article brings border studies scholarship into a more systematic conversation with memory studies and shows how memory is a fruitful way of conceptualising border temporality. Memories of the past -for example memories of population replacements, competing ideologies, and past spatial orders -play an important part in the construction of borders and inform ways of constructing identities and evaluating the present in a context of border change. From the perspective of memory, past events and developments are relevant not as objective historical occurrences but as reconstructed accounts of the past that act in and on the present and serve to legitimise, shape or undermine borders. Focusing on border temporalities from this perspective doesn't mean to replace the focus on the spatial with one on the temporal but to look how time is used in the making of spatial orders.
The article focuses in particular on the role of memory in everyday borderwork as part of a conceptualisation of border temporalities. Recent scholarship in border studies have emphasised that bordering functions, including the policing of borders, are increasingly handed over to citizens. More generally the making of borders can be seen as partly dependent on everyday practices which play an important part in legitimising borders and making them stick (Doevenspeck 2011;Johnson and Jones 2011;Rumford 2006;Yuval-Davis, Wemyss, and Cassidy 2018). I contend that an examination of memory in everyday borderwork is therefore an important area in the study of border temporalities: it allows not only achieve a fuller understanding of the forces that constitute and contest borders but also helps to account for the alignments and discrepancies of border temporalities that are usually being overlooked in existing conceptualisations of border temporalities.
The article is structured as follows: I will first discuss memory as one way of conceptualising border temporalities comparing it to other, chronological and genealogical approaches that have been developed by border scholars. I argue that memory allows us to capture the temporal complexity of borders in three key ways: firstly, it enables us to capture multiple relations between the past and the present beyond the linear conception of time that has dominated scholarship on border temporality. Secondly, it emphasises the existence of multiple and intersecting temporalities at different scales. Thirdly, a memory perspective also allows us to look at how memory itself is constitutive of borders, and spatial imaginaries that inform them, rather than limiting the study of time to the tracing of historical transformations of borders. These three issues will be discussed in more detail in separate sections. The final section of the article illustrates the usefulness of this approach through the empirical example of how memory is used in everyday borderwork in the Russian-Estonian borderland, based on extensive ethnographic fieldwork including 58 life-story interviews conducted in the border cities of Narva and Ivangorod between August 2011 and February 2012 and during a return trip in June 2016.

Conceptualising Border Temporality
Existing scholarship on borders has usually prioritised spatial over temporal questions. Among others, concerns over the location of borders -their proliferation and spatial stretching, locations both onshore and offshore (Mountz 2010) and increasing ubiquity in everyday spaces (Balibar 2002;Johnson and Jones 2011;Yuval-Davis, Wemyss, and Cassidy 2018) have been the subject of much writing on borders. As part of the ubiquity of borders, scholars have also noted the multiplication of actors involved in the making of borders, including the role of 'non-traditional actors' (Johnson and Jones 2011;Rumford 2006;Yuval-Davis, Wemyss, and Cassidy 2018) and have more generally explored how bordering processes work to categorise and classify, to mark spaces as ours and to construct some bodies as legitimate and others out of place (Green 2013).
In contrast to these spatially oriented questions, a number of approaches have brought temporality more centrally into border research, providing us with different ways of understanding borders in and through time. This section provides an overview of common approaches to temporality in border scholarship, that as I show, have usually shown interest in tracing changes over time and have conceptualised time in a linear and uniform way. This is however not the only way of conceiving border temporality. As Little (2015) argues, we need to pay attention to what he calls the complex temporality of borders: 'Complex temporality goes beyond the widely accepted notion that borders change over the course of time to focus on the nature and implications of that change across different bordering practices. (. . .) Instead of imagining political change to borders in a linear fashion, complex temporality highlights the disorderly manner and the uneven tempo in which change takes place in the real world.' I will first discuss established perspectives of conceptualising border temporalities before I show how memory studies can be one way of approaching this complex temporality.
The first broad perspective on border temporalities, which is the one most commonly found in scholarship, examines the chronological development of borders over time and proposes developmental taxonomies and, what has been called, biographies of borders to conceptualise border change. This perspective contends that what borders signify as well as their material, legal and technical infrastructures change significantly over time and aims to examine and in some cases, systematise these changes. Baud and van Schendel (1997) for example have proposed a life cycle model for the study of border change, which distinguishes between different stages of border development. They outline five ideal types or stages that a border goes through, from its initial delimitation to its hardening and maturing into a firm social reality and finally decline and disappearance. Drawing on debates on the biography of objects and places, Nick Megoran's concept of border biographies also examines border change over time and examines how borders 'appear, reappear and change, and disappear or become less significant in different ways and in different spatial and discursive sites over time' (Megoran 2012, 468). His biographical approach overcomes some of the determinist, biological undertones of the life cycle model, as it is more open to change into different directions rather than assuming a particular cycle of development.
Other scholars working on border temporalities have shown a similar interest in border developments but have put forward a different conceptualisation of time. For them time is not linear and orderly, progressing from past to present but the past is plural and contradictory and can persist in the present, shaping the forms and meanings of borders. O'Dowd (2010) proposed to combine border studies with historical sociology inspired by the longue durée studies of Michael Mann and Charles Tilley to examine the overlaps and continuities between empires and the contemporary nation-state system. Sarah Green's conceptualisation of borders as tidemarks (2009,2012) conveys the dynamic and historical contingent character of borders while at the same time recognising their repetitive and predictable aspects. Drawing on Derrida's notion of trace, her concept aims to capture the material and immaterial traces of borders, how they pattern the landscape and leave behind layers of movement and emotion. According to her, 'tidemarks are traces of movement, which can be repetitive or suddenly change, may generate longterm effects or disappear the next day, but nevertheless continue to mark, or make, a difference that makes a difference' (Green 2012, 585). This includes asking questions about the persistence and change of different classificatory logics that constitute borders.
Furthermore, the concept of borderscapes also takes into account how contemporary forms of bordering can bear traces of the past, for example when examining how the legacies of colonialism are traced in contemporary understandings of borders (Brambilla 2015(Brambilla , 27, 2014Rajaram and Grundy-Warr 2007). As Brambilla outlines, 'borderscapes allow a multi-sited approach to borders that is not only spatial but also temporal by encouraging a "genealogical" perspective on borders based on a new ontological standpoint on them capable of taking into account both their spatiality and temporality as well as the mutual implications between these two dimensions' (Brambilla 2015, 26-27). Despite the differences in these approaches, these perspectives on studying border change over time add substance to debates on the processual character of borders and challenge the essentialisation of what a border is through historically informed analysis.
Conceptualising border temporalities through the lens of memory studies provides a different, complementary perspective to the approaches outlined above: it examines how reconstructed accounts of the past shape borders and the spatial imaginaries. Studying border temporalities means to consider time not only as such but also through the temporal narratives that make it understandable. As Ricoeur (1984) shows memory narratives allow people to order time and appropriate it from the perspective of their own lived experience. Through processes of remembering the past is selectively brought into the present and is used to make sense of borders and one's location in relation to them (see Hurd, Donnan, and Leutloff-Grandits 2017, 2). Border memories can be located at different spatial scales; they are produced in media discourses, political speeches, museums and literature that have traditionally been the focus of much scholarship on borders and memory (for a discussion see Zhurzhenko 2011) but can also be found in everyday talk, narratives and commemorative practices (see Berdahl 1999;Meinhof 2002;Mihelj 2014;Smith 2013) -a perspective that is particularly relevant for the present article.
In the past years studies of borders have shifted their attention to the multiple meanings of borders and ways of making and experiencing them, going beyond the state as the main actor in the making of borders. Scholars have pointed out that 'bordering practices are less and less in the exclusive domain of the state and its agents' (Johnson and Jones 2011, 62) and that a disintegration of the state and the border has taken place. This emphasis on the multiplication of actors and particularly the role of citizens as everyday borderworkers has however not been reflected in research on border temporalities. The approaches outlined earlier have often privileged diachronic and linear conceptions of time, conceiving border temporality as largely uniform historical time without considering how time is experienced and made sense of at different levels including those of everyday live. While more recently authors have proposed the study of rhythms of emplacement and displacement (Ballinger 2012) and the experience of time by migrants held at borders (Andersson 2014), there are still relatively few attempts at bringing everyday or personal temporalities into the conversation on border temporalities. Using a memory studies perspective is one way of doing so as it allows us to look at how time is perceived, ordered and interpreted at different spatial scales and how these temporal orderings confirm, extend or question the meanings of borders. Memory scholarship is based on the general assumption of multitemporality (Macdonald 2013, 52-56), the existence of a multiplicity of temporal orderings and forms of telling the past, and therefore lends itself well to the study of the 'complex temporality' of borders described earlier.
The following sections will develop the conceptualisation of memory in borderwork, bringing border scholarship into a closer conversation with memory studies. I will firstly discuss how memory goes beyond linear conceptions of time by making sense of the past in relation to the present. Secondly, I focus on how memory at the level of the everyday needs to be conceived in relation to other scales of remembering which it can draw on, use as a resource within a memory narrative as well as actively contesting them. Thirdly, I will focus more specifically on the effects of memory narratives and the spatial imaginaries they support.

Beyond Linear Time: memory as Past-Presencing
As Barbie Zelizer writes, 'collective memory is predicated upon a dissociation between the act of remembering and the linear sequencing of time' (Zelizer 1995, 222). The process of remembering is about recollecting the past and producing it in the present. A focus on memory thus challenges the assumption that time moves only in one direction, from the past to the present, and that it is this movement that is of primary interest in the study of border temporality. Already genealogical approaches to borders have departed from such linear conceptions of time, studying how the past as a trace or historical sediment can survive over time and structure the present. Research on memory is largely interested in the process of past presencing as an active reconstruction by the remembering subject: through processes of remembering, past events and experiences are recast from a present-day standpoint.
That remembering is an active process shaped by the present was already emphasised in early sociological conceptualisations of collective memory (Halbwachs 1992), and it is a common place among memory scholars today to state that remembering involves telling stories as much about the past as about the past's relations to the present (see Keightley and Pickering 2012;Misztal 2003;Ricoeur 1984). For the study of border temporalities, this means that border-related memories do not just give us insights into past bordered lives but are of interest because they tell us something about the border today, how it is lived and evaluated from the present. Daphne Berdahl's ethnography (1999) of the Eastern German border village of Kella during the fall of the Iron Curtain is an example of how memories are used by borderland inhabitants to come to terms with border change. She shows how the opening of the Iron Curtain allowed Eastern German citizens to re-establish contacts with their West German neighbours, a process through which economic asymmetries and uneven power relations between East and West became apparent. Through processes of remembering and reconsidering the past, the villagers of Kella reoriented themselves in the present. Rather than telling a story from a constrained past to present liberation and reunification, they now used the past to critically engage with the present and to contrast themselves from their neighbours, an act of borderwork.
When studying border memories such as the ones of the villagers of Kella, it is important to pay attention to the selectivity of memory (which events and periods are remembered and which ones are forgotten?) as well as to the process of emplotment. Emplotment gives significance to the past and makes it understandable by 'allow(ing) us to construct a significant network or configuration of relationships' (Somers 1994, 617, author's emphasis). There are multiple stories that can be told about any past event and period, and remembering involves putting it into an order, a particular narrative form, rearranging, reinterpreting and recontextualising the past in the present (Keightley and Pickering 2012, 37). Examining border memories one can distinguish between different plotlines, as for example progress narratives, decline narratives and zigzag narratives (Zerubavel 2003) which are mobilised with different effects, as will be discussed in more detail below. Studying emplotment is important as it draws attention to the concrete ways in which the past is used to stabilise or undermine borders -how exactly, in what form, memory is used in borderwork.

Tracing Memory's Interscalar Relations
The previous section showed us how memory research allows us to go beyond linear conceptions of time by studying use of the past in the present. This section explores complex border temporalities by examining multiple and intersecting temporalities at different scales. Studying memory in everyday borderwork involves studying how the everyday relates to different scales of remembering, including the local, national and transnational. Everyday processes of remembering regularly make reference to national events, histories and symbols and can adapt and reinterpret them as they are incorporated into a memory narrative. Everyday memory narratives can also provide alternative frames focusing on localised events and frameworks or making reference to the transnational scale, transcending the scale of the nation-state and thus mobilising alternative spatial imaginaries. A multi-or interscalar perspective allows us to examine these processes in more detail.
Multi-or interscalarity has only been a relatively recent concern in memory studies. Memory scholarship has often assumed that memory in society is homogeneous -for example when studying national memory, generalisations about the memory held by the wider population are made on the basis of the analysis of particular memory texts (for a critique see Bell 2003;Kansteiner 2002;Rigney 2018). Kansteiner for example criticises that memory studies scholarship has been shaped by a 'desire for cultural homogeneity' (Kansteiner 2002, 193), assuming that people who have knowledge of a particular past 'have similar perceptions and form a stable interpretative community' (Ibid.). Another strand of memory scholarship goes beyond such homogeneous accounts by limiting the study of memory-making to one particular scale only, with political scientists studying memory politics, social psychologists studying personal memory and cultural studies the mediated representation of the past in museums, films and other media products (see Keightley and Pickering 2012 for a critical discussion). This has led to a particular scalar vision of memory as social phenomena taking place at neatly divided levels. If a relationship is assumed, it is usually one of stacked scales and hierarchical relations between the transnational, national and local.
Both these approaches to memory are unable to capture the multiplicity of border temporalities -while the first one sees temporality as uniform and provides an overly homogenous picture of temporality, the latter restricts itself to one scale of analysis and thus cannot provide a broader picture of how temporalities at different scales work together -or against each otherto shape borders. A multi-or interscalar approach provides us with a perspective that can overcome these shortcomings: it involves firstly, a differentiation between different levels of memory, and secondly, analyses how these overlap and interact with each other (Bekus 2019;de Cesari and Rigney 2014). According to de Cesari and Rigney, studying memory across multiple scales means to examine the 'multivocality that is brought to play in the interlocking social fields connecting the "local", the "national", and the "global"' (de Cesari and Rigney 2014, 3). This also means to pay attention to dissensus and differentiation as well as power relations at play in the production of memory (de Cesari and Rigney 2014, 3). Memory-making then is a contested field in which multiple actors participate in the interpretation of the past. While these actors are equipped with different resources and capacities to give voice to their interpretations and institutionalise them, they all participate in the construction of a plurality of temporalities existing within societies.
Keightley, Pickering, and Bisht (2019) have suggested using the concept of interscalarity to explore the relations between different scales of remembering. Similar to de Cesari and Rigney, they see different scales as 'mutually interactive even as they are influenced and mediated by differential sources of power and authority, endorsement and legitimation' (Keightley, Pickering, and Bisht 2019, 28). Individual memories are never simply based on individual experience alone but incorporate shared or second-hand knowledge, 'shuttle-(ing) back and forth between our own experience and that of others (. . .), and in doing so move(s) across and between individual, collective and cultural scales' (Keightley, Pickering, and Bisht 2019, 28). For a study of memory in everyday borderwork this means to keep a relational focus on different scales of remembering and to examine if interscalar relations -as for example in references to the national collective -are made and in what ways. It also means to pay attention to tensions and/or lack of connections between everyday remembering and particular spatial scales such as the national.

Remembering as Borderwork
While the previous section already looked at the relations between everyday and other scales of remembering, this section looks more closely at the effects of processes of remembering and the particular spatial imaginaries that are mobilised through everyday memory. As de Cesari notes, memory works as a 'border device' (de Cesari 2017, 21), producing 'spatial imaginaries of containmentsocial imaginaries of bounded communities-cum-territories or bundles of territory-community-culture' (de Cesari 2017, 18-19). These spatial imaginaries help to naturalise borders and exclude others from the cultural community. While memories are immaterial, they have real material effects, feeding back into physical borders and legitimising their fortification and surveillance (de Cesari 2018). But how exactly is memory involved in the making of borders?
Based on the earlier reflections on emplotment, it is important to consider that memory doesn't come with any guarantees in how it might be actualised and therefore past events can be mobilised to inform multiple spatial imaginaries. This assumption is useful in thinking about borders and temporalities because it shows the malleability of memory in the making of borders as well as the need to empirically analyse for what purpose memories are mobilised. Rather than assuming that memory either generally works to support territorialisation or helps to undermine it as it is often assumed in writing on transcultural memory, I follow Jansen and Loefving's suggestion that we should 'approach the key concepts of sedentarist and placeless paradigms -including territorialisation and deterritorialisation, and emplacement and displacementas empirical issues to be investigated rather than as philosophical assessments about what characterises our age' (Jansen and Loefving 2009, 5).
Memory studies have traditionally not paid much attention to how memory is used to constitute borders. When Maurice Halbwachs (1992) first proposed a conceptualisation of collective memory, mnemonic communities were conceived as fixed and pre-existing. As a pupil of Émile Durkheim, Halbwachs was interested in how group membership -belonging to a family, religious group or also a nation -provided the social frames of remembering and shaped what is considered worth remembering and what is forgotten. According to Halbwachs, memory reconstructs an image of the past and re-organises it in regard to social frames provided by groups and thus stabilises communities. Like Halbwachs, more recent memory scholarship has often taken memory's borders for granted and has focused on the operation of memory within distinct containers; studying memory often involved the study of allegedly stable and clearly demarcated groups and cultures, with 'the most popular social unit being the nation-state, which was then swiftly seen as isomorphic with national culture and a national cultural memory' (Erll 2011, 6). Erll proposes to replace this focus on fixed memory containers with one emphasising memory's fluidity, a move that has been increasingly popular among memory scholars who have started to examine memory's transborder movements as part of the scholarship on transnational or transcultural memory.
Rather than taking borders for granted or questioning them altogether, looking at the role of memory in everyday borderwork requires an analysis of how memories are productive for the making of borders (de Cesari and Rigney 2014; Pfoser and Keightley 2019; Rigney 2018). As de Cesari and Rigney (2014, 9) write, 'the dynamics of remembrance are (. . .) intimately bound up with community-making since narratives about events belonging to "our world" continually reproduce, redraw or challenges the lines between "them" and "us". And while cultural remembrance helps thus to create bonds, it is a two-edged sword whose power can also be deployed to discriminate against groups'. Analysing how memories are used in borderwork involves a close examination of how the past is related to the present and with what effect: How does everyday remembering relate to territorial borders? Which events, among the many that can be remembered, are selected, how are they interpreted and used to shape, legitimise and imagine present and future spatial orders? Can we observe in everyday remembering the banal nationalism (Billig 1995) imbuing our everyday lives, and does it justify the reinforcement of territorial borders? Or does everyday memory provide us with different spatial imaginaries that can cross borders and classificatory systems based on the distinction between a nationally defined us and them?
Scholarship shows us that memories can be used for different purposes -for example memories of past violence can be used to legitimise hard borders through historical analogies suggesting past violence will repeat itself in the future or the construction of a continued threat emanating from a neighbouring country or from the figure of international terrorists (Innes and Steele 2014). Memory work can however also offer possibilities for mourning, reconciliation and care and can provide the basis for an imagination of different presents and futures (Till 2012;Zhurzhenko 2011). As remembering involves interpretive acts articulated from a changing present-day standpoint it also allows for reconfiguration and different ways of connection-making rather than simply bringing together pre-established communities.

Memory and Everyday Borderwork in the Russian-Estonian Borderland
This section analyses the role of memory in everyday borderwork on the example of the Russian-Estonian border. The Russian-Estonian border underwent significant changes from an unmarked border between two Soviet republics to a gradually fortified border between two nation-states (from 1991 onwards) and EU external frontier (from 2004 onwards). The analysis focuses on how Estonians and Russian-speakers living in the Estonian border town of Narva use their memories to make sense of this border change and engage in borderwork. I show how different ethnic groups use different emplotment strategies to bring the past into the present, engage in different relations to the scale of the nationstate and use their memories to put forward different sociospatial imaginaries. In particular, we find competing interpretations of the socialist past -a time of suffering and insecurity or a more ordinary time of making careers, families and building everyday relations -that construct the border of the nation-state in opposed ways, supporting a hardening of the territorial border as well as constructing an alternative spatial imaginary based on localised and transnational sociospatial relations rooted in the past. Overall, I show the significance of memory for borderwork as well as how multiple and contradictory temporalities work together to produce the border as a contested space.
The section is based on several months of ethnographic fieldwork in the borderland, which were conducted between August 2011 and February 2012 and during a return trip in June 2016. During this time I conducted 58 narrative life-story interviews with different ethnic groups as well as doing participant observations of social gatherings (see Pfoser 2014; Pfoser 2015 for a more detail description). Participants were contacted through voluntary associations, educational institutions as well as earlier personal contacts.
They were asked to tell their life-story in a free narrative form, followed by follow-up questions and more specific questions about their experiences of the border. While Russian-speakers (ethnic Russians as well as members of other ethnic groups such as Belorussians, Ukrainians and others who share a common language) constitute the large majority (over 90%) of borderland inhabitants, I purposefully sought to include Estonians living in the borderland to capture perspectives of the national majority. As an Austrian national conducting research at a British university, I had the advantage of being an outsider to both groups and was not seen as partial to the often competing interpretations of the past put forward by participants. Following Josselson (2007), I sought to develop an empathic and respectful relationship to participants, listening with attentiveness, trying to understand the reasons and motivations behind people's narratives, while remaining sensitive to silences and exclusion. The interviews were analysed using narrative analysis, focusing on the identification of key narrative patterns (Chase 2005;Riessman 2001) aimed at generating a broad perspective into the dynamics of remembering and bordering among different groups. The analysis below identifies key memory narratives that were articulated in the interviews in relation to the border.
Both participants' main narration and their responses to the follow-up questions revealed differently structured memories among Estonians and Russian-speakers. These (selective) memories played a key role in how participants related to the border and its gradual fortification since the break-up of the Soviet Union. One memory narrative, dominant among Estonians living in the borderland, focused on painful memories of the Soviet past and used them to justify the existence of a hard border regime in the present. The years of Soviet rule in Estonia (1940-41 and 1944-1991) when Estonia was incorporated into the Soviet Union were narrated by several participants as a time of state violence, insecurity and displacement, constructing not only the years of World War II and Stalinist terror but the whole period of Soviet rule as a time of suffering and extended displacement. This was only reversed by the restoration of Estonia's independence in 1991when relative order and security was reestablished. As one participant, 65 year old Sofia recounted: I was born in 1946. I still remember the pre-school time when we had to whisper when we spoke about everyday things. This was the Stalin time. Maybe somebody listens and gets something wrong. We spoke quieter than usual. And if somebody said something with a normal voice, it was like "Quiet!". This is what the Soviet power did. Estonians went through a lot. Some people less, some people more. (. . .) Every family was suffering in some way, no matter how, under this Soviet power (Interview with Sofia, b.1946, 26.10.2011 In this brief excerpt, Sofia narrates her experience of past insecurity as part of a larger story of national victimhood, shifting from insecurity experienced at personal and familial scales to the level of the nation as a whole, similar to the interrelations Koresaar (2004) observes in her analysis of Estonian autobiographical memories of Stalinism. Memories of past violence and insecurity were in the case of several participants directly used to justify a hard border regime, making connections between personal and nation-state scales. The restoration of national independence and the establishment of a border to Russia were experienced as a personal relief; they meant at least a partial return to (a nationally defined) 'normality'. At the same time, participants projected past violence onto the present and future to describe a continuing sense of insecurity deriving from Russia and the Russian-speaking minority inhabiting the borderland. Russia was not perceived as directly aggressive (the interviews were conducted before the annexation of Crimea and the war in Ukraine) and the local Russian population not a direct threat, however memories of the Soviet period were used to construct them as potentially dangerous and untrustworthy, constructing a continuity between past insecurity and future threat that one needed to be always prepared for.
While this emplotment strategy perpetuated the scale of the nation-state and symbolically legitimized a hard border regime, other narratives related to the border differently, drawing on the memories past mobilities, mundane experiences of conviviality and, less frequently, the idea of historical reconciliation. These memories mobilised alternative scales of meaning, grounded in locally and transnationally rooted spatial imaginaries. Despite being marginalised within national public memory discourses, these spatial imaginaries were locally influential in shaping relations to the border and could also be exploited by Russia as a way of destabilising the Estonian border regime (see Pfoser 2015). Particularly memories of Russian-speakers living in the borderland emphasised the past experience of an integrated local space between the border town of Narva and an its neighbour Ivangorod, formally part of the Russian Soviet Federative Socialist Republic. Participants also remembered mundane convivial relations based on work relations and friendships including those between Estonians and Russian-speakers. Beyond such localised spatial imaginaries, memory narratives made reference to other parts of the Soviet space or to the idea of the Soviet Union as a larger homeland. Particularly those Russian-speakers who had moved to the borderland from other parts of the Soviet Union to work in its large industrial complexes often narrated their lives as stories of social and geographical mobility, enabled by an integrated Soviet space and homeland which also allowed them to maintain cross-border kinship ties to other places in the Soviet Union uninhibited by border controls and documentary regimes. Participants' memories of past mobility and a local convivial culture were contrasted to the period after the break-up of the Soviet Union and were used to construct the border as violent disruption of and intrusion into local and transnational lifeworlds.
While memories were then largely structured by ethnically divergent historical experiences based on suffering and violence on one hand and historical mobilities on the other, in some Estonian narratives additional layers of memories of the late Soviet period -particularly of friendships and work relations crossing ethnic and national borders -complicated accounts of suffering and displacement. In some cases, participants were also able to transform experiences of violence into accounts of transnational reconciliation. This is similar to the reworking of difficult pasts in other border regions (Zhurzhenko 2011), even if in the Estonian case these voices are less heard and haven't been institutionalised yet. When asked about the border, Laura, an ethnic Estonian of 74 years, for example remembered her memorial tips to Russia, visiting among other places the Levashovo Memorial Cemetery, a cemetery of the victims of Soviet repressions in St Petersburg. She said that she didn't hold any anger in relation to the repressions her family had experienced and constructed Russia as part of a shared historical and memorial space which she visits to remember the dead.
When studying borders and bordering processes, it is important to attend to such memories and historical experiences as they play a key role in how people make sense of borders and relate to them -as spaces to be inhabited, as places of everyday life but also of clashing political visions, fear and insecurity etc. As Karolina Follis shows on the example of the Polish-Ukrainian border, when the EU border was constructed in the region, it was done on top of historical memories and on-going debates on historical reconciliation and neighbourly relations between the two states. She writes, 'quite independently from the Western European debates concerning security, asylum, and the desirable scope and nature of immigration to the EU, the Polish-Ukrainian frontier has been a site where the new European border was contested for its exclusionary tendencies, but sometimes also affirmed as a much-needed fence dividing Europe from a post-Soviet unknown. These sentiments have been animated by a different set of historical experiences from those informing the actions of EU leaders, experts, and technocrats who are responsible for the regime's construction' (Follis 2012, 172-173). In the case of Narva, local memories of past violence were in several cases aligned with the bordering effort at the national -and also EU -scales, making connections between personal experiences and the nation-state scales. At the same time, alternative spatial imaginaries based on a different understanding of the socialist past as a time of mobility and conviviality were also put forward and were used by participants to critique the current border regime and to imagine more open and porous borders that would ease the continuing of connections and historically formed conviviality. Only few participants, all of them Russianspeakers, used their memories of the Soviet past to openly question the national border and the legitimacy of the nation-state, also showing how the border that in 1991 appeared as new and unfamiliar has gradually become normalised, even if it continued to be critiqued by participants. The study of memory then adds temporal depth and ambiguity to the study of borders, providing a complex picture of how locals relate to borders and participate in their making, based on different historical experiences.

Conclusion
The emphasis on spatial over temporal questions in border research seems obvious, given that researching borders is about examining processes of ordering and classification and the organisation of relations between here and elsewhere. However, the prioritisation of the spatial often leaves us with a flat picture of borders, oblivious to the histories that have shaped them and continue to act on and in the present. If one of the key aims of border studies is to denaturalise borders -to show that they are socially constructed and the result of both material and symbolic practices -considering their temporality is one important way to do so. Studying border temporalities allows us to go beyond the presentism that shapes much current border research. It allows us to examine the historical trajectories of border production, to show that borders are historically contingent and that for example the present-day relations between states, the perception of refugees and migrants or the construction of insecurity are temporary phenomena with particular histories. It also sheds light on how temporality is actively and selectively used in the making of borders, by remembering particular pasts and the forgetting or deemphasising of others. Memory -in the form of temporal narratives that allow us to make sense of time -is an important resource for the construction of borders and can be used to legitimise their fortification and surveillance as well as to question and undermine them.
This article has argued that among the different approaches to border temporality, memory provides us with a lens that is particularly suitable to study borders' multiple and complex temporalities: memories are multidirectional, plural, contradictory and generative. They allow us to account for the fact that temporality is not linear -that the past is not over but acts in and on the present -and that temporalities are multiple, overlapping and contradictory rather than singular. In his conceptualization of complex temporality, Little points out the uneven tempo through which border change takes place (Little 2015, 431) -when borders change, new temporal narratives are mobilised to make them stick, at the same time alternative memories can persist, or be newly created as a response to border change, and be used to contest the new spatial order.
I contend that an examination of memory in everyday borderwork is an important area in the study of border temporalities as it allows not only achieve a fuller understanding of the forces that constitute and contest borders but also helps to account for the alignments and discrepancies of border temporalities that are usually being overlooked in existing conceptualisations of border temporalities. As I show on the example of the Russian-Estonian borderland, a study of memory and everyday borderwork helps to uncover the different temporalities that exists simultaneously based on different historical experience and that work together to construct the border as a contested political space. In recent years, border studies have emphasised multiplicity of border-making actors, it is time to also see this multiplicity in terms of temporal orders that are put forward by borderworkers and to develop appropriate conceptualisations. A deeper engagement with the memory studies scholarship that this article has undertaken is one way of doing so.

Funding
This work was supported by the Economic and Social Research Council [ES/R011680/1].