‘My home is my people’ homemaking among rough sleepers in Leipzig, Germany

Abstract By combining an analysis of the unhoused response with an ethnographic study of the lived experiences of unhoused people in Leipzig, Germany, this article offers insights into how rough sleepers understand home. The focus is on men and women who exited unhoused shelters and returned to the streets; on their experiences in shelters and their practices of homemaking on the streets. Whereas legal and policy frameworks often equate home with house, unhoused people locate home in relationships, affects, routines and in time. In Leipzig shared shelters that are separated by sex are used to move as many people off the streets as possible. But to unhoused people home is first and foremost their relationships. Policy and practice misunderstand that unhoused people exit shelters because they cannot live intimate, personal and family life there, not because they refuse assistance. To be effective, services must overcome the dissonance between their perceptions of home (infrastructural, spatial) and those of their target groups (social).

Notwithstanding efforts to show that place is just one highly specific aspect of home, physical structures/infrastructure and social and emotional factors have been married to understand home. Because 'place', as Clifford Geertz shows, must be locatable, it 'demands exactness' (1996, p. 259), home is often understood as a place that allows social relations to be lived and emotional states accessed. Peter Saunders and Peter Williams build on Anthony Giddens' (1984) theory of structuration to define home as 'simultaneously and indivisibly a spatial and a social unit of interaction' (Saunders & Williams, 1988, p. 82). The affects attached to home arise from a conflation between a certain building and the emotional lived experiences there, e.g. the security of a childhood home, the freedom attached to one's first apartment or the bonding in the shared flat with one's romantic partner. The physical structure becomes foundational for the social relations and emotions connected to home (see Cieraad, 2018). They can blend together so much that they fit Saunders' and Williams' definition of home as a '"socio-spatial system" that represents the fusion of the physical unit or house and the social unit or household (Mallett, 2004, p. 68 paraphrasing Saunders & Williams, 1988'. This fusion is terminologically substantiated. In many English-speaking countries, people without housing are called homeless and in Germany, where the terms obdachlos or wohnungslos [without shelter] are used, home and house haveas Joseph Rykwert shows in his analysis of Heimweh [homesickness]-become near homonyms (1991, p. 53). Etymologically, home [Heim] and house [Haus], refer both to a place and to the social relations in it. Hence, those without housing are understood to be without a physical dwelling and without a social/emotional home [ohne Zuhause].
Law and policy consolidate the association between home and house. In English case law this happened as early as the 17th century when a Judge pronounced 'the house of everyman is to him as his castle and fortress as well as his defense against injury and violence, as for his repose' (Rykwert, 1991, p. 53). The separation between private and public spheres upon which contemporary Western societies so firmly rest is still drawn by the walls of one's home and makes housing conditional to secure one's most fundamental needs in life without intrusion or surveillance. Socio-legally, housing is not only necessary for decent living standards but intimately linked to safety, belonging and wellbeing. The walls of one's own home came to symbolise the protective shield against the harms and vulnerabilities of this world behind which one can live safely without intrusion or surveillance (Bowlby et al., 1997, p. 344). Despite criticism that these walls also foster (gendered) oppression and violence, their symbolism remains firm. Unsurprisingly, demarcation practices-the extent to which people lock windows and doors, build fences, decorate or shut themselves in-have been analysed as an indicator of how secure communities feel (see Cieraad, 2018;Comaroff & Comaroff, 2016).
Moreover, home and house as interwoven categories have become essential for a person's social capital, status and respect. The notion of the 'walls of one's home' are an 'outcome of a cultural, societal, and historical development toward the progressive separation of two domains, represented in the separate … domains of consumption and production as the respective spaces of living and working (Cieraad, 2018, p. 1)'. Not only do home and work exist in social, spatial and legal opposition, they also become mutually dependent. Those without work have difficulties finding housing and vice versa (e.g. H€ ojdestrand, 2009) and home ownership enhances one's status (Dupuis & Thorns, 1998;Madigan et al., 1990). The house as a physical unit erects a boundary between leisure and work, self and world, family and public. Concurrently, this boundary becomes conditional for enjoying family life, finding and maintaining work and participating fully in society. What happens then when one has no housing, no walls to create a neat separation between private and public spheres? In a country where home has become imbued with housing, how, if at all, can people without housing create home and enjoy the privileges that come with the idea of home; how can they feel at home?
The policy after which people become unhoused The above-mentioned questions are pertinent across Germany where the estimated number of unhoused people has tripled in the past decade and continues to increase (Bundesarbeitsgemeinschaft Wohnungslosenhilfe e.V, 2019). Across much of the geopolitical West these numbers are rising. Yet, approaches differ. Some countries see a housing issue. Based on the premise that housing is a fundamental right and the secure foundation from which social issues (be they social, administrative legal etc.) can best be addressed, they often employ a housing first model (see Tsemberis 2011) which means that people are offered independent housing unconditionally and as quickly as possible, and then receive social support in their new accommodation (see e.g. Finland, Norway). Other countries, including Germany, sees a social problem that should be processed and solved. Recognising the importance of having a home, but prioritising infrastructural over social factors, Germany attempts to assist people to get off the streets according to a 'staircase' model. The unhoused move through different stages of temporary and usually shared, supervised accommodation where one stage conditions the next. Unhoused persons are entitled to accommodation during day and night to protect from harm and weather. The police and regulatory law of the federal states tasks municipalities with its provision which differs widely across Germany.
In Leipzig day centers and temporary night shelters share responsibilities. The latter include the 'R€ ucke' for men, the 'Scharnhorststraße' for women, the 'Chopi' for active drug users or the 'Quecke' for wet alcoholics. 2 These shelters attempt to 'offer dignified and decent emergency shelter and so fulfil one of the tasks of the municipal order according to the Police Law in Saxony (City Leipzig, 2015)'. They operate similarly. For instance, the R€ ucke offers 50 beds in single and shared rooms and the Scharnhorststraße 24 beds in double or shared rooms. Showers, toilets and Television are communal and there are refrigerators with lockable compartments. Laundry costs 1,00 e. The shelters are open 24/7 on weekends and from 4.00 pm to 08.00am on weekdays. Day centres, though in different locations, are open when shelters are closed. Additionally, some temporary shared apartments are available for unhoused persons while they search for independent housing. This should serve as the next step on the 'staircase' before transitioning to independent accommodation. During these stages, unhoused people get a specific status within the welfare regime and are confronted with various assistance systems that are shaped by the social and labour activation policy paradigm (see e.g. Daigneault, 2014;Kenworthy, 2010). Germany placatively illustrates this 'active society' (Daigneault, 2014, p. 4) ideal with the popular motto 'F€ ordern und Fordern' (promote and demand) which essentially means''no rights without responsibilities' (Giddens, 1998, p. 65) and sums up the socio-political, labor market, integration and educational policy concept of the activating state. Benefits are not granted unconditionally. Instead, they form a contract of 'reciprocity' between state and individuals. Shelters cost 5,00 e/night and entry may be rejected. Newcomers may stay the first night for free but must then apply for sponsorship from Social Services. Unhoused people receive support only if they formally update their status, get their affairs in order and actively look for an apartment and work. For those who are unable to do so independently or at all, SGB XII foresees additional assisting services to overcome social difficulties or at least prevent further deterioration. Social support is thus an aid to get unhoused people's 'lives back on track, with an apartment as the ultimate reward' (Henley, 2019). However, this approach often fails, as the demands put upon unhoused people within this 'staircase' model are too high to be met.

Approaching home through fieldwork
This article builds on sixteen months of fieldwork in Leipzig, during which I accompanied 27 rough sleepers through their everyday lives and spoke to over 300 unhoused people. The research includes people affected by the four main living situations which the European Typology of Homelessness and Housing Exclusion (ETHOS) recognises: rooflessness, houselessness, insecure housing and inadequate housing (Amore et al., 2011;Busch-Geertsema, 2010;FEANTSA, 2005). Here, I focus mainly on data gathered with people who currently live rough (rooflessness), i.e., people who live on the streets, in public places, in cramped or abandoned buildings, in parks or under bridges. Specifically, I focus on people who have previously spent time in night shelters and who exited them even though they could have stayed there.
My collaborators are between 14 and 70 years old and from diverse walks of life: singles, couples, pregnant women, people with addiction issues (esp. alcohol, heroin, methamphetamine), people who beg, deposit bottle collectors, 3 (petty) criminals, sex workers, and former blue-and white-collar workers. About 70% are from Germany, some are EU migrants and few are from sub-Saharan Africa or the Middle East.I approached home as a situated category. I focused on people's 'everyday practices' (de Certeau, 1984), on how they speak about home, practice home, act, interact and position themselves in a social world which shapes them and which they help to shape. To highlight and problematise the hardship that can arise from the terminological, legal and practical fusion of home and house, this article uses the term unhoused in lieu of homeless which was proposed by OSL an aid organisation in Seattle to lessen negative stereotypes which view those without shelter as 'less than others' (OSL, 2014). By using the term in an academic context, I use a language which refers to the fact that my research collaborators are without a house but make no comment about their social or emotional connection to a home, to belonging somewhere. In so doing, I respect that for my research collaborators 4 home and house are not synonyms and that those without a house are not necessarily without a home. At the same time, I show the harm that can result from this fusion and offer a pathway to move beyond it.
Researching sensitive matters attached to home requires building trust. This is rarely possible via surveys, questionnaires or structured interviews. Similarly, abstract theories developed without those they speak about often remain divorced from the issues at the heart of the lives they purportedly render intelligible. Becoming immersed in people's lives is the condition for gaining some understanding about the reasons behind rough sleepers' rejection of shelters in a society where home, family, emotional and sexual life are inextricably tied to having a roof over one's head. I therefore conducted participant observation which allows for long-term relationship building and for research collaborators to share what they deem important in their own time and way. My research was consistently openly communicated and all collaborators decided whether and to what extent they wanted to participate (informed consent). Some identifying characteristics and biographic details have been omitted to protect identities and many chose their own alias (anonymity, confidentiality).
Rough sleepers lead mobile lives. My research collaborators were constantly on the move. The journeys between the places where they slept, ate, showered, met people, tried to make money etc. spun networks across the city. These daily routes were complemented by seasonal movements. Many also circulated between prison and street and chose imprisonment as an alternative to paying fines. I therefore applied Henrik Vigh's 'rhizomatic fieldwork' (2006, p. 18) which entails letting the interconnected webs spun by the movements of research collaborators define the research site(s). In this decentralised set-up, I added 'go-along interviews' to observe research participants' 'spatial practices in situ' while speaking about their experiences and interpretations (Kusenbach, 2003, p. 463). I am also in close contact with criminal justice personnel and service providers who engage with my research collaborators. Hence, this article attempts to 'parry the prevalent public images of homelessness with an account drawn from the homeless themselves' (Desjarlais, 1997(Desjarlais, , 1997 while simultaneously unravelling the broader cultural, historical, social and legal forces that weigh on notions of home.

State accommodation is no home
It was horror. I did not know where Tim is. I could not remember when I slept anywhere without him. I moved in with him straight from my parents' house. He is my home! By separating us, they took our house and our relationship, our home. In the shelter, I was lying awake the entire night surrounded by strangers with a pit in my stomach. The women in my room made strange noises. I felt so unsafe, lost and disturbed that I decided it would be better to sleep on the pavement with Tim than in this strange place without him. Better to be together than to be separated if we don't know whether we will ever find an apartment again. This is how Lena (54) describes her and her partner Tim's (55) experience. When the couple was evicted from the apartment they had shared for 29 years, they could not imagine living apart. However, when they turned to social services for assistance, they were directed to the night shelters and informed they needed to separate, and Lena must sleep in the shelter for women and Tim in the shelter for men. After one night, they checked out of the shelters. They have been sleeping rough for almost two years now. Lena's and Tim's story is not unique. Only one couple I followed agreed to separate to stay in the shelter. Singles too have voluntarily exited temporary accommodations because, as they variously put it, the shelters' structure made it impossible to 'be at home', 'feel at home' or 'create home'. This resonates with work on the UK private renting sector, where 'the struggle to continually assemble, de-assemble and re-assemble a sense of home drastically reduces … tenants" wellbeing through stress, anxiety, depression and alienation' (Soiata & McKee, 2019, p. 148).
In the makeshift spaces of temporary shelters which are often, of necessity, shared, users are unable to enjoy privacy and intimacy, rights which they associate with home and are unwilling to give up. The temporary and emergency shelters are divided by sex and often by target group, e.g. people struggling with addiction or minors, which means they separate beneficiaries who must go through this process independently to possibly be reunited in independent accommodation later on. However, many rough sleepers reject putting their relationships on hold for a temporary roof over their head and an uncertain and distant future. Instead, they stay together even if this means exiting the 'staircase' and diminishing the possibility of finding independent shelter. Hence, users exit shelters because they cannot live intimacy there not because they refuse assistance-a widely misunderstood practice.
Additionally, the shelter management collides with users' notions of home. Shelters are temporary and users should find other accommodation as quickly as possible. There is a high turnover and rooms must frequently be shared with new people. Shelters are assigned communities largely between strangers which makes some user feel uncomfortable (for France see Blanchard & Bruneteaux 2019). Despite frequent theft, only few storage spaces exist. Most must keep their valuables on them. Moreover, the accommodation can usually not be decorated or personalised.
Many of my research collaborators carried items with them which symbolised home (see also McCarthy, 2020) and which were located in time. Some were attached to memories of home and served as mnemonic devices that keep their home close even if it has been lost or is far away, such as the wedding ring of a late wife, a daughter's sock, a photograph, a driving licence, some jewellery, an item of clothing (see also Kaplan & Stowell, 2004). Some were aspirational items of a future home e.g. a cloth to sew a pillow cushion for a baby's crib, a picture frame to capture future family memories or a business card from a University or employer for their future daughter. These items were often inconspicuous. Their real value and meaning could only be understood if the owners decided to reveal or display them. Yet, losing these items would have further severed their memories or aspirations of home and was an unbearable thought which drove them out of the shelters.
Bj€ orn: never again. I slept there one night and they stole my bag. My mother's picture was in there. Last memory I had. All my life died that night.
Kai (overhearing our conversation from another bench): Sorry man. See this card (shows business card). It's the last person who believes that I can make it. Keeps me going. I lose this I'll never build home. So no to R€ ucke. Can I sleep where you are?" Concerns over personal safety played a role too. Paule (47), a former teacher who became unhoused after his wife divorced him and he lost his job, elucidated: The shelter is a place that prevents you from dying. It is unsafe. You sleep with complete strangers, some are loud, drunk or fighting. If you close your eyes maybe you will wake up without your shoes. You don't take your clothes off; you sleep with them and you don't take anything there. Now the first rule of home is that you feel safe. You don't have to worry because you are protected. But the people you love cannot be in the shelter with you. Only strangers. And you are not alone for one second. You cannot retreat. Everything is shared. Bedroom shared, showers shared, everything. Even in prison you can decorate your cell, put pictures up, make it yours. But shelters, they can never be home. No. The street is much better. On the street you can decide where you go and with whom, and you can create the places however you want them. You decide who sleeps there with you and how. You decide when you wake up, when you go to sleep. Yes, home is much more on the streets than in such shelters.
Sentiments of feeling policed, controlled and judged resonated in most users' stories. Many felt that the shelters are designed to 'make you feel uncomfortable so that you look for something new'. Although entry and exit are more fluid, these shelters share procedural features associated with total institutions (Goffman, 1999) in that they operated according to 'a single rational plan of action promoted by a unified supervisory staff; the regimented regulation of daily activities, usually carried out in the immediate company of others (Desjarlais, 1997, p. 35)'. By contrast unhoused people attached personal autonomy, the agency to design their lives and live in a selfdirected manner to the notion of home. Being able to 'create a home' suggests the ability to decide what home constitutes and with whom a home should be built. This agency is stripped off the residents of such shelters who must follow a rigid set of rules in a depersonalised manner (see also Watts et al., 2018, for Scotland). Karl (53) who worked as a contractor until he was let go after prolonged illness said thatthey [the shelter's staff] have many expectations. They look down on you-this person who is a nobody, a loser, scum on society-and they expect you to get out of there quickly. They are everywhere watching you, judging you and waiting for you to find work, find money and make things happen. If you don't manage, it is like you disappoint them. Home to me is where you are respected and loved for who you are not just for what you do. In the shelter, they don't care who you are. They don't care that I worked 15 years of my life until my back prevented me from continuing. All they see is a man without a house, money … You stop becoming a person and you turn into a problem that they must solve. Now tell me, if you would describe home, would you say home is where people think of you as a problem and a burden?
By contrasting the circumstances in shelters with their perceptions of home, Lena, Paule and Karl teach us about the discrepancies between the priorities of the city and service providers and that of unhoused people. The Municipality, who is unable to offer independent housing immediately, is using such shelters to move a maximum number of people off the streets. By contrast, to unhoused people home is first and foremost connected to their relationships. While they need and want a roof over their head, they are often unwilling to trade this for being with the people they love. Mark (59), a former truck driver who became unhoused after his wife passed away, summarised this in the following way: 'A house is a certain dry place to sleep. But apart from that it is just four walls, a roof and a floor. A house is not a home. A home comes with the people in it and with feelings of home, with safety, independence and happiness'. Kim Hopper (2014) agrees and says 'one key issue is to recognize that the opposite of homelessness is not a homeless shelter. Rather, it is a home, which is currently lacking in treatments of homeless people'.
The distinctions rough sleepers draw between the forms of life available to them in shelters and home reminds of Georgio Agamben's (1998) concept of bare life. These shelters may preserve the biological dimension of life, zo e, in that they provide a roof over a person's head thereby preventing them from dying on the streets. However, they are reduced to offering necessities to preserve the sheer biological fact of life, of survival, but imply no guarantees about bios, the quality of life. Indeed, as Credlandet al. (2004) show, while life is preserved, temporary shelters adversely affect physical and mental health. Neither the possibilities or potentialities of life are considered, nor do shelters provide spaces to live needs, desires and rights such as relationships, family-or personal life. Users are expected to return to mainstream society, but for the time they are in the shelter, they are in a 'liminal period' (Turner 1964) between street and independent housing. They have no private sphere, no autonomy over their living circumstances or social relations, and some of their most fundamental rights remain unprotected.
Shelters, says Robert Desjarlais are 'a novel hybrid of the almshouses and asylums of an earlier era (1997, p. 35). They carry 'the vestiges of an earlier culture of incarceration and psychiatric care, though … in a more deinstitutionalized present, itself founded on a network of institutions … less entrenched in a particular locale than concerned with activities and procedures (ibid.)'. The social and labour activation paradigm, places residents on a ladder that must be climbed. Due to the rules and expectations of the shelter and the broader subject position residents must take within society as 'unhoused people', residents become the bodies which feed a decentralised total institution-encompassing shelters and daycentres, aid organisations like Caritas or Diakonie, job centers, social ministries etc. Because completing the ladder is difficult, this liminal stage, that is characterised by social exclusion, may become (near) permanent. The ways in which research collaborators explained losing sense of time, perspective, meaning and direction in the shelter indicates a taking over of what Desjarlais termed 'struggling along' over 'experience ' (1997, pp. 19-22), a more selfdirected, reflexive, temporal engagement with the world. The constitution of subjects within the shelter does not allow for love and was therefore rejected by many. Such shelters cannot provide a home and, instead of giving up on 'feeling at home' altogether, many unhoused people attempt to create home on the streets. Rather than staying in the shelter and submitting to the directions, obligations and limitations of the decentralised total institution, by exiting such shelters unhoused people attempt to (re-) take control of their lives. They are now 'experiencing' on the streets which entails suffering but also agency (see Desjarlais, 1997, pp. 19-22). This is not to dialectically oppose control in shelters with freedom on the streets. Indeed, the agency of people sleeping rough is under threat due to the heightened surveillance and regulation of public space in many countries (see e.g. Johnson et al., 2018). Leipzig too is increasingly policed, but, as of yet, streets do not equal the panopticon of shelters and, while shelters are rigidly regulated, surviving on the streets necessitates being part of a social matrix which is in flux and thus requires constant relationship building, positioning and re-positioning.

Social and affective dimensions of homemaking on the streets
Homemaking on the streets concentrates on social and affective dimensions. Similar to Joseph Rykwert's (1991, p. 54) assertion that 'home does not require any building even if a house always does' to many research collaborators, home and house are not synonyms. When asked where home is and what home means, they rarely name a physical structure or geographical location. Instead, they refer to people and reiterate: 'Home is my love'; … where I lived with my wife'; … where I raised my children,' ' … my parent's place because of my mum's smell' … Hence, memories of home are also rooted in family and loved ones. Family events such as separation, divorce, death of a loved one or eviction by household members play a prominent role in rendering people unhoused. Becoming unhoused is often connected to losing a relationship and thereafter losing housing. Some research collaborators hide their unhoused status from friends and family. Unhoused people's relationships to housed kin and friends are often complicated or severed. Whereas memories of home when they were still housed centre on a few key figures e.g. mother, partner, children best friend, today home is spread out between various social networks, some interdependent, some unconnected, others opposed. Unhoused people's homes are embedded in manifold specific relations with friends, allies, supporters, opponents or competitors Dynamic 'families of choice' tend to form and fictive kin receive titles like 'stepson', 'street father', 'friend-sisters', 'baby brother' etc. These relationships form and dissolve very quickly. Newcomers are usually received with open arms, but also let go without further ado if they move on, disappear or pass away. In some sense rough sleeping communities are also assortments of strangers. However they have not been grouped together but choose whether and for how long they want to stick together. 'In the end' explained Fin (36), who became unhoused after losing his leg and his jobthe street makes you form families of choice. We are communities. We exclude no one, we make room for new people. The street accepts you. There is no way down. You already hit the pavement so you might as well be who you are. We don't try to change one another; we try the best to support one another. If you go, you go. If you come back, the street is where you left it, you've come home.
With these families of choice memories, circumstances and aspirations can be shared, one can re-invent oneself or drop dramaturgy altogether. Many share life histories and stories with each other that conjoin places, times and people, thereby creating a sense of home between past experiences, current listeners and potential avenues. Hence, these families of choice provide antidotes to homesickness for lost homes, for those who never existed and those who could someday exist (see also Blunt & Dowling, 2006). Whereas men and boys tend to find their social home within these dynamic networks, women and girls often find a partner on the street, a 'husband' thereby taking the role not of child or parent, but of spouse. My research collaborators seek aspirational kin across and within generations through diverse forms of guardianship filiation and partnership. Male friendships are another way to reconstruct forms of semi-structured domesticity that suits unhoused person's individual needs and produces a sense of home. Moreover, 'street-level bureaucrats' (Lipsky 1980), police, probation officers, social service personnel, street workers, case workers, social workers etc. become such fixpoints in the lives of many unhoused people that they take the role of guardians. Dennis (34) who became unhoused after turning 18 and losing his place in the foster home, reiterated this in the following way: I never knew my father and my mother did not want me. I was pushed from flat share to flat share until they kicked me out after I turned 18. Then the street became my bed. But home to me is mostly my probation officer because whatever I do, he must take care of me. And also, the woman in the jobcentre, because she must always provide. Like a real mother would do. Nowadays mothers can decide against having you. But these state employees, they must care for you. They are family. They are home. So, in some way you can never really be homeless. You can just be on the streets, but so long you go and see them they must care for you.
Dennis shows how in a life marked by the absence of reliable care, structure and belonging, it is street level bureaucrats who he depends on, knowing that they are bound not by affection but by law to support him. Both his probation officers and his case workers change frequently, but this is of no consequence to Dennis, who is part of a generation of marginal people who rely more on the role and the institution behind a person, than on the specific interpersonal relationship. Through the continuums of care that result from prolonged dependence on state support, institutions, through their representatives, become permanent figures and primary caretakers in unhoused people's lives with wide ranging implications for notions of family, responsibility, social connections and care (see also Thelen & Alber 2018). The inclusion of care takers, fellow unhoused people and animals into 'fictive' or 'aspirational kin' speaks to an attempt to belong and to create continuity and normalcy amidst disruptions and uncertainty. This tactic which has predominantly been recognised among refugees and migrants who must quickly forge meaningful social relations (Diggins, 2017;McGovern, 2012), applies to unhoused people for they too have been uprooted from their previous living circumstances or never had security at all and must find their feet in unpredictable and insecure circumstances. In many ways, to the people I worked with home resembles the 'embryonic community' proposed by Mary Douglas (1991) in her analysis of the idea of home; a community which is based on strategic solidarity. My research collaborators show that Joseph Rykwert's assertion that in most languages and cultures, home is 'at the centrifugal heart … the family' (1991, p. 54) holds true, but here this family is one of choice. It may be a biological family, one's spouse or partner, people who share similar fates or street level bureaucrats who are employed to provide, oversee, protect and punish.

Re-inventing privacy in the public sphere
Despite a focus on social relations, unhoused people's need for the privacy that allows intimacy to be lived remains. Many unhoused people spend large proportions of their days in public spaces where they have little to no control over who crosses their paths and when and how they can be alone. Being static in a steady stream of people as is common for places like main stations, inner-city parks or popular shopping streets can feel both isolating and overwhelming. Tova H€ ojdestrand, who conducted research with unhoused people in post-socialist Russia, declared 'there is no privacy in this never ending flow of people and events, no option to distance oneself from the here and now, be it physically, socially or mentally ' (2009, p. 14). Even under a bridge, in a forest or an abandoned backyard, one never knows when and if another person might pass by.
Creating home on the streets thus requires establishing a barrier between public and private. People who must do without independent housing, attempt to reinvent privacy by establishing makeshift separations between the street and the various places they frequent.
Just like rooms in a house serve different purposes, many unhoused people divide their activities between different localities across the city to separate the place where they sleep from where they try to earn money, meet friends or use services. To protect their relationship(s) -their home-and to create some form of privacy, unhoused people engage in spatial and infrastructural creations of shelter in the public sphere. When sleeping, most research collaborators go to great length to cover themselves from public view. Ede, an unhoused man in his sixties told me once thatyou will never see a rough sleeper's face when he lies down. Never. Why? Because sleeping is something very intimate. Only people you trust will see you sleeping. We don't have that luxury. If we don't have a sleeping bag, we will pull our clothes over our heads or turn to the wall or something. So at least all you will see is a body and not a specific person. They think we do that to blend out the light or the weather. But really, we just create minimal privacy.
These tactics of creating minimal privacy become pertinent when attempting to create togetherness. Unhoused people use large sleeping bags and blankets that can cover two people completely or tents that are either hidden in bushes within parks or erected on the grass at night and taken down and hidden in the mornings. Most of the people I followed sleep and create temporary spaces to enjoy privacy and intimacy in that way. Kay (19) who ran away from his parents' house said that-If we crawl under the blanket or into the sleeping bags, we can switch our clothes without someone seeing. We can then pull them over our faces to be anonymous. Some are lucky, they even have tents and stuff. But most of us, we fuck behind the bush in the sleeping bag if we are so lucky.
Others move into stairways, abandoned buildings or under balconies. Sara (24), who is four months pregnant and who sleeps under the stairs of a public building in Leipzig explained her choice in the following way: Since I met my boyfriend, we need to be at home. We want to have sex and we cannot really do that if we sleep out in the open on the street or if we stay in an abandoned building with many, many others. So, we moved under the building where there is a small hidden space between stairs and ground. Because nobody sees us there, we can even leave some of our things there and with my man being there, it is not just a place to sleep, but something like a home. There we are alone. I got pregnant there actually (laughs). Maybe we will soon be a real family in our own private home under the building.
As illustrated by Sara's pregnancy, such practices also enable bodies who dwell on the streets to build future homes in the social sense. Therefore, home is tied to privacy and to the ability to be together away from the glances and the intervention of others (see also Dowling 2012).
'Home starts by bringing some space under control (1991, p. 289)' wrote Mary Douglas. The boundary work and demarcation practices of unhoused people are practices of seeking control, routine and regularity. Night shelters take away control by treating home as something which can be given and taken away again. By contrast, unhoused people' create home through 'individual strategies of control defended respectively in the name of the home as a public good' (Douglas, 1991, p. 306). Although they cannot change the state's strategy or response to them, by moving tactically through the city, they prioritise their own needs and attempt to tacitly widen the holes in the web of state strategies. Unhoused people resist the state's strategy by exiting shelters and shape it (to the extent possible) by exerting control over public spaces.

Home as routines
Establishing control plays an important role in the making of routines that enable a sense of belonging and of having a place in the social world. Both in their memories of home and in their attempts to create home today, unhoused people referenced routines as the texture that weave the fabric of home. The ability to structure one's life and the need to do so to balance various obligations often constitute taken for granted and ignored parts of housed people' lives. To unhoused people however, they are essential markers of a sense of home through which they become integral parts of a well-oiled social matrix.
In their contemporary attempts at homemaking, many unhoused people engage in establishing and polishing such routines and boundaries. Far from being marked by idleness, the life of an unhoused person in Leipzig today is characterised by manifold activities (see Lenhard, 2019 for Paris). While housed people have a fixed location to sleep, eat, shower, undertake bureaucratic tasks or meet people an unhoused person often travels vast distances and has to synchronise his or her schedule with the opening hours and the movements of social services, social workers, friends, peers etc. This requires meticulous planning and almost constant movement. Additionally, many of these tasks still cost money which must be made first, e.g. by collecting deposit bottles, begging or assisting people with tasks. According to the proverb 'eine Hand w€ ascht die andere [one hand washes the other]' many rough sleepers invest much time and energy into doing favours which create debts that can be cashed in later. The binding obligations and relations of compulsion they create, are a key tactic of unhoused people to earn, stay connected and needed (see Schneider, 2018). Mark describes his routines in the following way: When I had an apartment, everything was in one place. Now, it is all over the place. I must get up before the police catches me outside. The Bahnhofsmission, 5 where I get my coffee and the soup kitchen, where I get lunch are in different parts of the city, so I have to move. If I want to eat something else, I either have to go withdraw money or collect deposit bottles. I must always remember which toilets are accessible … when and where I can wash myself. If I want to do laundry or buy something, I collect deposit bottles first. The people I meet have different routines and many do not have phones. If we want to meet, we must align schedules way before. Everything must be planned. There is no day where I do not at least go to three places. Most of my income I get from organising things for them and then getting them to do me favours. Without constant organising, home would fall apart, and I would be all alone.
Mark's way of creating home through routines relates to Veena Das et al.'s observation that 'the modalities of the domestic, in the sense of forms of doing that connect possibility with actuality, the subjunctive with the indicative' (Das et al., 2008, p. 352) are enacted over diverse spaces thereby including them into a web of domesticities. The lives of unhoused people span a multitude of places. The roadmap of their movements is drawn by the people, services and institutions connected to a certain place, not the place itself.
Recent scholarship of place recognises that attempts to 'theorise a notion of place in a world of flux-a world of displacement and reattachment'-must treat place as something fluid, elastic and processual (Ward, 2003, p. 93). Similarly, prison and confinement scholarship highlights the dynamic relationship or 'deadly symbiosis' (Wacquant, 2001) between prisons, and other places like city or household. Social relations, these scholars say, ebb and flow between these sites and simultaneously, becoming socially connected to some place can restrict access to other layers of society (see Da Cunha, 2008;Schneider, 2020;Waltorp & Jensen, 2019). Hence, while place(s) are geographically specific, social relations may transcend boundaries. They can render the rigid boundaries between specific places-like a prison cell and a bedroom in one's marital home-porous and create fluid and interconnected spaces which, together constitute a person's home. Concurrently, a person's social positioning can prevent access from certain places or change the texture of the place entirely. Since for unhoused people, place is attached to the social, and the social is spread over a multitude of locations, the physical aspects of home are not rooted in a singular, stable locale, but marked by fluid plurilocality in time and space. In terms of place, home constitutes 'the site of several domesticities' (Das et al., 2008, p. 351).

Discussion: unhoused people and the dynamics of home
Home is not simply where the key fits but where we can create home and feel at home, notions associated with loved ones, with agency, security and the freedom to design. Home is then intimately tied to a sense of belonging and to the ability to shape. Home is one's people and homemaking constitutes a social phenomenon that is connected to intimate relationships, friendships, to building a family or living familial relationships. To unhoused people homemaking means navigating a diverse social world at a specific moment in time. Practices of home-making centre on preserving and protecting these relationships. Home then can be the sum total of these relationships across time and space. Unsurprisingly, socio-political interventions and service provisions that fail to capture the vernacular resonances of home among unhoused people rarely succeed. Understanding why they fail-for instance by comprehending why so many people exit shelters-is the condition for improvement. Yet, while a growing body of scholarship disentangles the conflation of home and house (e.g. Das et al., 2008;Douglas, 1991) in numerous countries, law, policy and practice continue to champion infrastructural over social factors and to tie rights and protections to independent shelter. For instance, the staircase system used in Germany and many other countries, radically limits belonging and the ability to shape. In Leipzig, it creates a situation where a roof over one's head is conditioned, at least temporarily, on separation from loved ones. Leipzig's shelters privilege home as a place. They are run according to a strict set of rules, are separated by sex and offer places mainly to individuals. Here, the condition to separate to receive support forces a choice whereby place loses to social relations. Unhoused people must separate from their homes-their people and routines-for a roof over their heads; a trade-off many are unwilling to make. Instead, they decide to sleep rough together and attempt at creating makeshift, though often unlawful barriers between public and private spheres. So long emergency shelters do not provide viable pathways towards housing, affected people will ultimately prefer to navigate the dynamics of the streets together getting stuck on the staircase alone. For many unhoused people, homemaking so becomes independent from domestic dwellings but is still firmly tied to a separation between private and public spheres. This demonstrates that so long the aid system is premised on individualization and separation; so long we expect affected people to overcome all of their problems before we offer them the security of housing, while we push them from one volatile situation to the next, the aid system will strengthen the very mechanisms that lead to the loss of housing and will, inevitably, fail.
Disentangling home and house as a category of analysis and a category of practice could be a useful first step towards overcoming the disconnect between some service providers' and affected people's concepts of home (e.g. Mallett 2004). This requires re-examining the linguistic fusion of home and house and the basic premises on which housing responses are built. Using the term unhoused in lieu of homeless already popular in activist circles, could be one way to de-naturalize the conflation. Evidently, renaming something does not do away with the problem. It is however a helpful tactic to reflect on taken-for granted concepts which then allows for the development of context-specific, bottom-up strategies. Housing first concepts, where those affected first obtain an apartment and are then offered social support, is one model which overcomes the conflict between social and spatial dimensions of home. The roof over one's head so becomes a secure base from which a home can be maintained or formed. Shelters for couples, families, friends as well as animal friendly places which would allow privacy and intimacy to be lived without discouraging users from finding independent accommodation are another option. This could improve the success of the unhoused response which is fundamental at a time where the number of unhoused people across most of Europe is dramatically rising. Notes 1. People without tenancy-protected housing. See discussion of terminology below.
2. The shelter for women is operated by the Advent-Wohlfahrtswerk on behalf of the City of Leipzig. The R€ ucke is run directly by the City of Leipzig; the Quecke by the SLZ Suchtzentrum and the Chopi by the center for drug help of the City Hospital St. Georg Leipzig. 3. People who collect empty bottles or cans for drinks for which some money is refunded if they are returned to the store. 4. Throughout this article, the term research collaborators refers to the unhoused people I conducted research with and who helped me think through, analyse and present my data. 5. a Christian relief organisation with free access points at almost 100 railway stations in Germany aiding travellers and people in need.