The hopeful labour of begging – Homeless people’s struggles for a better life in Paris

Both in public imagination and scholarship, begging is often associated with passivity, waiting, inactivity and as such defined as the opposite of ‘real work’. During two years of fieldwork with people sleeping rough in Paris, I observed my informants to the contrary describe begging as their occupation. In this paper, I present an ethnographically grounded argument that the people experiencing homelessness I accompanied over the course of two years experienced begging as labour which involved constructing and supporting emotional narratives of neediness, deservingness and personal connections in scripts, stories and hustles. My interlocutors often understood this labour to be a hopeful one, one that structured their day (and thoughts) but also one that enabled their projets de vies, their struggles into the future. In the wake of overwhelming suffering and structural violence, I observed begging not only a short-term fix for consumption but also as keeping the future open – the more long-term future of eventually leaving the street behind.

row of cars on Rue Lafayette and knocked on the driver's window. The man looked up at him and shook his head. Franc¸ois wasn't too persistent; somehow, he knew -from years of experience perhaps -whom to make money from. Confidence was key. The next car had its window open. Franc¸ois addressed the man inside: 'T'as une petite pi ece ou une cigarette, chef?' [Do you have change or a cigarette, boss?] 'Je n'fume pas. Mais, tiens'. [I don't smoke but take this.] [hands over a 50c coin] 'Merci, chef'. [Thanks, boss.] The donor nodded at him as Franc¸ois slowly walked to the next car, a smile on his face. Before addressing the next taxi driver, he looked at me: 'Not too hard, is it?'. The row of about 20 cars brought in e1.5 and two cigarettes, all for less than 10 minutes of work.
Francois, a Frenchman in his late 50s I met early during my two years of fieldwork among homeless people in Paris, made begging look like an easy-undertaking, but he at the same time called it his 'work'. In fact, begging as a practice of 'asking money for nothing ' (McIntosh and Erskine, 2000) is as old as human society (Ribton-Turner, 1972). In both the public imagination and scholarly analysis, however, it is usually seen one-sidedly: it is associated with inactivity (O'Neill, 2017a), a passive time pass (McIntosh and Erskine, 1999), something that marginalised people are (often structurally) forced to engage in to make a meagre living or simply a symbol of societal decay (Swanson, 2010: 2). It is widely seen by the public as something that needs to be banned, curbed or at least questioned, often based on its (assumed) links to alcohol and drug consumption (Amster, 2008;Barnett, 2016). The law often reflects this latter view classifying begging as somewhere between a nuisance to be kept out of public space or a (criminal) deviancy (Hermer, 2019) -depending on the geographical context and historical period. In urban studies and geography, a specific focus has been on concrete extensions of such 'deviancing' into the realm of the (built) environment by both private and public actors. Anti-homelessness or more specifically anti-begging architecture -what Petty (2016) calls 'hostile architecture', from spikes to 'hostile' benches and metal sheets have been critiqued based on their reproduction of spatial exclusion and security scapes (Amster, 2003;Lenhard, 2020;Maguire and Setha, 2019).
In social science, begging has mostly been observed as a side effect of homelessness or poverty and (economic) marginalisation more broadly (with the exception of religious begging, e.g. Laidlaw, 2002). Groups affected in this way would include migrants and refugees, Roma (Friberg, 2020;Ruggiu, 2016), street children (Hecht, 1998), certain specifically marginalised people with disability (Devlieger, 2018) or drug addiction (Bourgois and Schonberg, 2009) and also the group I worked with, people experiencing homelessness. For all of these groups, begging would be described as part of a wider spectrum of 'street-level economic activity' (Dean, 1999), 'street level informal economic activity' (Adriaenssens and Hendrickx, 2011) or an 'informal income opportunity' (Hart, 1973) and as such as a secondary, unreal kind of occupation with very little positive potential. 2 In the studies that do analyse begging among homeless people, the scope has so far been similarly limited; begging has mostly featured marginally as part of for instance the survival strategies of people with drug addiction (Bourgois and Schonberg, 2009) or a simple time pass (O'Neill, 2017b). In the cases where it featured more centrally, the analysis is mostly focused on the 'begging encounter' and its consequences as seen through a lens of gift giving. The work of both Erskine (1999, 2000) and Hall (2005) writing about the UK describe begging as a kind of ambiguous gift exchange. While the gift itself can be stigmatising and wounding (see Hall, 2005), there is a moral confusion for the giver about whom to give. Which 'beggar' is genuine (see McIntosh and Erskine, 1999)?
In my article, I want to pick up a different thread from a very small group of scholars taking begging seriously as work (or labour) from the perspective of our informants. Like Kassah's (2008) description of people with mobility difficulties begging in Ghana, Swanson's (2010) observations among indigenous women and children in Ecuador and Lankenau's (1999) analysis of panhandling in Washington D.C. I appreciate the complexity and possible negative connotation of begging, but want us to also understand its productive, positive possibility. While Lankenau's focus is more on how begging and the relationships springing from it contribute to the status enhancement of the panhandlers, Kassah focuses on how begging-as-labour increases the feeling of self-worth. Swanson, on the other hand, describes how begging enables her female and underage indigenous informants to enter the urban space (coming from the rural highlands of Ecuador) and even more so to work towards a variety of goals from 'societal status' to access to education. For my informants, begging was similarly important not only as a means of making money but a way of structuring and ordering their daily life, and thus contributes to making a home on the street. Begging was part of my informants' survival practices but also identity shaping and futureopening. As such I will describe begging as part of a practice of hope (Pedersen, 2012).
For my informants such as Francois begging consisted of a developed set of skills and practices he learnt over the years as a Gitan, a French traveller. For others, such as German ex-soldier Carl or Congolese-German Pascal, begging was initially harder and more challenging -both physically and emotionally -while over time they became similarly more skilful, reflective and thought through. Taking as a starting point my informants various exclusions (from for instance the formal system of work or welfare), I follow their understanding of begging as a source of hope, of keeping the future open. Begging was on the one hand part of what Stettinger (2003) in her analysis of homeless newspaper vendors in France calls 'survie', a necessity that kept them alive; the practice of begging itself was also more than an economic necessity, however: it kept my informants busy (and particularly mentally occupied), engaged them in an active and structured occupation, exhausted them and over time helped them to keep the future open. It provided structure, routine and hope and was part of what Stettinger calls 'travail'. All this is encapsulated in my understanding of labour (see below) going beyond a binary understanding of begging. Describing begging as labour is hence an important act valuing my informants' activity and struggle to be engaged in the world. In this, I follow a recent impulse to include in our focus not only suffering (e.g. Scheper-Hugh, 1993) but also what Robbins (2013) calls an anthropology of the good. I am complementing a predominantly critical view of begging through its relation to structural inequality or practices of policing (through people or urban architecture) and as part of a secondary, informal economy; instead I am asking what good does begging do for my roughsleeping informants? Using the Arendtian analytic of labour explicated below allows me to encapsulate both, the hard and at times devastating 'work' and effort and the possibility for productive and positive outcomes. Bracketing off my informants' structural exclusion, as many of my informants are forced to do on a day to day basis, I am asking: how do people sleeping rough deal with it, survive and think ahead? The answer I will go into detail below is by engaging in the concrete labour of begging (Harvey and Krohn-Hansen, 2018: 20).

Theorizing the labour of begging
Adding to the definition(s) of labour in the introduction to this special issue, I am following Arendt's differentiation between labour and work. She defines labour following ancient traditions as a process which produces 'vital necessities' (Arendt, 1998;Harvey and Krohn-Hansen, 2018: 7) and assures individual survival (Harvey and Krohn-Hansen, 2018: 47). Unlike work, 3 the product of which is always something material (Harvey and Krohn-Hansen, 2018: 86, 93), labour leaves nothing behind, is in this sense unproductive. Labour is never-ending and repetitive (Harvey and Krohn-Hansen, 2018: 102) and as such cyclical, because the need to consume doesn't stop. Labourers are, according to Arendt, bound up by the necessity of daily survival (Harvey and Krohn-Hansen, 2018: 83). She likens the labourer to the 'menial servant' (Harvey and Krohn-Hansen, 2018: 93), what in antiquity was the way of life of the slave (Harvey and Krohn-Hansen, 2018: 12). But unlike her ancient predecessors, she doesn't completely understand labour as something one needs to rid oneself of. She believes, to the contrary, that 'the perfect elimination of the pain and effort of labour would not only rob biological life of its most natural pleasures but deprive the specifically human life of its very liveliness and vitality' . Labour -and the balance between pain and repetition but also its direct link to the need to consume (Harvey and Krohn-Hansen: 134) -is an essential part of the human condition. I will further illuminate this understanding of labour -not in the sense of part of a market society, of participating in pre-defined labour relations and structures -as part of the human condition, driven by daily survival for my begging informants.
Borrowing from Arendt in her conceptualisation, I describe begging as a process of labour. 4 My informants begged to survive; they didn't produce anything of lasting value or importance beyond their immediate ability to consume in a repetitive circuit. As such, begging is categorically unproductive -unlike the work of shelter-making, which is at least temporarily about creating a material home (Lenhard, 2020). The public space and its infrastructure figures prominently in what I will theorise as the labour of hope (Pedersen, 2012; see also Zigon, 2005). At times, the labouring practices of homeless people consist in making themselves visible -by portraying neediness and deservingness or by using connections to regulars -to make money from passers-by. At other times, it involves becoming invisible and blending in to gain access to the right benches, the right location. Begging at all times crossed the line between physical effort -walking around, sitting on the pavement, monotonously repeating the same sentence and the same narrative -and emotional labourovercoming shame and embarrassment, making up narratives (what Summerson, 2011 calls 'scripts'), supporting them through their appearance, creating a network of regulars -in order to portray being needy and deserving. *** I developed deep trust relationships (friendships) with many of the informants featured in this article as well as my wider work which allows me to produce a particularly rich contribution from the bottom up. 8 Tracing my informants' 'mobility paths' (Wolch and Rowe, 1992), begging featured most prominently as their 'occupation'.
The structure of this article will now follow the process of begging as it cyclically repeated itself again and again for my informants: featuring detailed ethnographic descriptions from my fieldwork I describe different parts of the begging labour from finding a good 'begging spot' to choosing the right script, hustle or narrative. I dive into what kinds of labour (emotional and physical) and what kinds of strategies (evoking neediness and/or deservingness, using personal connections) these practices involve before closing with an -also empirically grounded -discussion of the hopeful quality of this labour for my informants.
Choosing the begging spot -Finding a good workplace Carl, who had been a solider in the German special forces for over eight years -travelling the world before being injured and traumatised in a grenade attack in Afghanistan -introduced an important dimension of the begging process to me. He had come to Paris only months earlier escaping the trauma that, as he explained to me in many conversations, was haunting him so pressingly in Germany. Like most of the other people featured in this paper, he spent his time around the train stations in the North of the city and made his money from begging. 9 While we will see how crucial self-presentation through language and also clothing was, the first decision concerned the location, the begging spot: what made one location better suited than another? For the longest time, the Gare de l'Est was Carl's preferred venue not only due to the arrival of the ICEs from Germany carrying a lot of potential revenue in the form of German tourists into Paris, but because of the train station's architecture: The big entrance -200, 250m with the taxis -where you can walk along in about 10 minutes and in exactly the same interval the people [you just asked] have already disappeared into the train station again. [. . .] Most are there for the length of one cigarette. [. . .] My strategy is to not stand on one spot but to walk around. You can make 100-150 people in 10 minutes or something like that.
Doing round after round without asking the same people twice, Carl was easily able to make his 10 or 20 Euro at the station in the course of a morning or afternoon. Enough people were streaming in and out of any of the two neighbouring stations to produce a large enough crowd of potential donors. While Carl preferred the closed circuit of the Gare de l'Est, others such as my Bulgarian informant Barut chose to walk from West to East following the long axis of the Gare du Nord. My informants consciously thought about their route, testing new ones in an effort to maximise the outcome.
A second factor which made certain locations preferential was the availability of givers, in particular regular givers, ideally with their purses already in their hands. Carl developed a special connection to a second venue, closer to the Gare du Nord. Every day around the same time -at about 6 pm when people left the offices in the area and were on their way homehe placed himself right next to a bakery just south of a park next to the church Saint Vincennes de Paul. It was a busy street leading down to the Grands Boulevards where people were passing through en masse. While not all of the people he asked knew him, a decent number became regulars over time. They would sometimes see him every day and get to know him little by little with advantageous outcomes: After some days, some people got to know me. And then they gave me money without me even asking them. Then it's easy. [. . .] You talk to them a little -like at the bakery. Short conversations.
The bakery and its customers provided two favourable components for Carl's begging work: it was a place frequented by both a changing set of people over short time cycles and returning cohorts over longer times. People would only come once a day, but they came every day -or at least several days a week. This made it easy for Carl to become acquainted with the regulars, who in turn were more likely to give him money (Garnier-Muller, 2000: 175;Prolongeau, 1993: 163ff;Lenhard, 2014: 98ff). This meant less effort in the sense of asking people and walking around and explaining himself, with greater income thanks to the personal connections developed over time. Additionally, people came to the bakery with the intention of spending money, and Carl would catch them either just about to open their purses to buy something or when coming out of the bakery with change in their hands. 10 The barrier of having to make an extra effort when asking for a donation was thus minimalized.
Others amongst my informants banked on similar locations where regularity was paired with a preparedness to spend money. A group of Punjabis I got to know with a leader called Sabal were often outside of either bank branches or supermarkets, and so was Franc¸ois (whom I introduced in the opening vignette); a changing group of people was camped close to a tobacco shop inside the Gare du Nord; another one around Natasha -whom I will introduce in the next section -always sat outside the busy fast-food restaurant Quick on the other side of the street from the station's main entrance.
As I will argue below, procuring money through begging is perceivable as labour in Arendt's sense of the word. The decision for the best location, the begging spot, prepared this labour in an important fashion: it was the first step towards successfully practicing begging by finding the best place to work. Where to stand, and when, was something that my informants learnt from experience and something that was often thought through and reflected upon. The begging encounter with its stories, narratives and the presentation of self, was crucially enabled by this first step. In the following, I will describe the different kinds of labours involved in begging; how did my informants display and balance neediness and deservingness, the two most important attributes in the process of begging? What role did the network of regulars, often built on personal connection, play? How did my informants' skills in balancing these three axes improve?

Emotional and physical labour -Scripts and hustles
Carl: It is also dependent on the weather, the time of the day. It is easier to ask at night because you are less present. Working when the sun is shining is much harder for me. While Carl was concerned about both the exhaustion which comes with walking around and constantly asking people for money, and the embarrassment of asking itself, Pascal was focused only on the second aspect. Pascal was a Congolese German in his 20s who had left Germany escaping from legal prosecution. Still in Germany, he had completed an apprenticeship as an electrician and was hence used to hard physical work every day. The emotional drain of begging, which at the time was his only way of earning any kind of money, 11 was the biggest hurdle to overcome for him. Pascal was a newcomer to the world of begging and the street in general. His fear was that begging marked him as dependent and ultimately as poor and homeless. He didn't want people to know that he was on the street. He didn't want people to put him into this category. Overcoming shame and embarrassment and accepting the categorisation of 'homeless' required psychological labour, often enabled by marijuana or alcohol -things which then had to be concealed again using chewing gum. Pascal felt inhibited: Pascal was aware of his own inferior position when asking people for money. The risk of categorisation and the implied social hierarchy that came with 'outing himself' as homeless worried him. He wanted to fight against what according to Simmel (1908) was the quintessential feature of being poor: 'What makes one poor is not the lack of means. The poor person [. . .] is the individual who receives assistance' (Simmel, 1908: 140). A sense of belonging outside of the group of homeless people, of the ones who were assisted was important both for his feeling of self-worth (see also Lankenau, 1999), but in a second step also as a tactic. Below, I will further elaborate how Pascal and others managed to keep up a fac¸ade through the use of on the one hand clothes (Smiley and Middlemass, 2016) while on the other hand I will show how words, stories and narratives were the most important part of this labour. Pascal used the potential ambiguity of his own standing in relation to the group of donors as a way of hustling.
A clever hustle -part of the 'art of deception [. . .] changing the rule of the game and misdirecting the audience in order to "get over" [to get their money]' (Williams and Milton, 2015:5) -is key to how successful you are begging. The goal is to convince people to give to you without being too forceful and intrusive. In fact, linguistically speaking, mendicit e (begging) is in its origin related to mensonge ('lie'), further strengthening the connection between hustling as deception (lying, making up facts, displaying part truths) and the labour of begging. Referring back to Pascal, he learnt over time through trial and error which part of himself to present in which situation while at the same time attempting to be true to himself as much as possible. He was learning how to hustle people without necessarily revealing his homelessness by forging personal connections (as a German or a Congolese for instance; see below).
Even more suitable in the context where narrative is the most important part of the begging encounter is the related concept of the 'script'. Summerson (2006Summerson ( , 2011 explains how, in the context of US drug treatment programmes, language is used by patients as a way of getting what they want -often a certain prescription drug -rather than what they would necessarily need. Users in the outpatient programme in the American Midwest would engage in what was called 'flipping the script': 'clients' linguistic interactions with therapists were commonly characterized by carefully constructed, institutionally astute, and strategic performances rather than simple acts of self-reference' (Summerson, 2011: 196). In other words: they told the doctors what they wanted to hear in a verbal performance, mimicking a certain kind of -in her case -recovering client without giving away the actual inner state, and without being honest (Summerson, 2011: 188ff, 213). Over time, people learn which ways of speaking -which scripts or hustles -work and which don't, as they engage with more and more institutions and individuals. These scripts -narrated and performed presentations of certain aspects of the self, as I will further unfold below -are part of the emotional labour which begging involves. Hochschild (1983;Wharton, 2009) categorises emotional labour as one which is face-to-face (or voice-to-voice) and which aims at producing 'an emotional state in another person -gratitude or fear, for example' (Wharton, 2009: 147). While his study focuses on Delta-airline trained air attendants, doctors, lawyers and salesman, at least part of my informants' labour fits into this category. Emotional labour is one which goes beyond 'suppress[ing] feelings of frustration, anger or fear' (Wharton, 2009: 154) and is about 'the production of a state of mind in others' (Wharton, 2009: 156). While there is no employer managing a staff's emotional state in the case of my informants (something important in Hochschild's study), I find his categorisation useful when paired with 'scripts'. My informants engaged in both repressive emotional labour -suppressing emotions of shame and embarrassment -and expressive emotional labour, by displaying worthiness and neediness and evoking sympathy in the givers (Hoang, 2010). As we will see further below, learnt 'scripts' -both in their core form of a spoken narrative and supported through appearance and clothing -are crafted as a tool of this second kind of emotional labour. They are used to solicit money from people by evoking sympathy.
I observed how the expressive 'scripts' focused particularly on three axes that were explicitly balance against each other: neediness, deservingness and personal connection (ad hoc, or in networks over time). In short, I will demonstrate ethnographically how potential givers are more likely to give when they understand that you need their personal help; most, however, want to make sure that you are deserving of their help (for instance won't spend the money on drugs or alcohol), and they are more likely to engage with you when they feel a personal connection. Ideally, all three come together, but at times a missing element -looking deserving -was made up for, for instance, by a stronger personal connection. I will work through these three axes -what I see as the core of the labour of begging through scripts and hustles -by focusing on physical presentation, but mainly language and narrative in the following section. I will describe how my informants balanced the three axes based on their own (and the passers-by's) subject position (as having a particular nationality (Pascal), as speaking a particular language (Natasha) as being unhealthy (Sabal)).

First axis: Neediness
It was early on in my fieldwork in 2014 when I took a late-night tour around the back of the Gare du Nord -a site which was, at that point, still unfamiliar to me. It was around midnight; the weather was bad, and it was little wet out. I had just come from observing people running around in the station and decided, on the way home, to walk down the road towards Rue Lafayette. On the corner of the Caisse d'Epargne bank, two men in their late thirties were sat on the floor. One of them spoke only English, the other had relatively good French. They were both originally from India, but had already spent some time in Paris. Within a minute or so, the one who spoke English had directed my attention to his foot. Even in the dark of the night, I was able to see how bad the skin looked. Not only was it black from the dirt of the street -he was wearing neither shoes nor socks -but it was also weirdly wrinkled, almost as if broken into little pieces. He explained to me how much it itched, and that it hurt when he scratched it. Finally, they asked me: could they have some money to get into a room for the night? The other man was constantly gleeming at me out of his dark eyes, his head bent down. He was supporting his friend's demand by adding another, non-verbal layer to it. I almost couldn't bear facing the two; I was appalled to see them sit there like that, in desperate need of shelter, medical care, help. I felt that I had to help. I gave some money to the two men and walked back home.
This was the first time I met Sabal and Bouti, who became some of my main informants, and it was one of the only times I was convinced to give money to yet unknown informants. 12 What was different in the situation I described? I perceived the two Indians as acutely in need of my help on that night -because of their illness, their apparent neediness -a feeling which was aggravated by the visibility of their bad health, their lack of hygiene, the weather, the state of their clothes. It was also the fact that the two didn't really seem to speak French, that they were and looked foreign.
Studies in other contexts -for instance Moeschen's (2008) overview of feigned disability among people who beg in America in the twentieth century, and Schak's (1988) ethnography of disabled people who beg in contemporary China -show even more extreme forms of displayed neediness as an important part of the begging encounter. Looking both at historical material and film, Moeschen describes how disability and impairment were at times deceitfully performed by people who beg, adding another dimension to the display of neediness to solicit gifts. Lankenau (1999), in his study of panhandlers in Washington D.C., observed how his informants 'manipulate [d] signs and symbols to demonstrate [. . .] need' (p. 290) by adapting their dress code and shaping their public persona (Lankenau, 1999: 305) to earn what Clarke (1997) calls 'sympathy credits'. The aim was always to appear in such a way that one's sympathy margin was high: 'Panhandlers that do not look impoverished may unwittingly drain their sympathy margin and receive fewer contributions'. (Lankenau, 1999: 307). The emotional (and physical) labour of appearing needy is not necessarily enough to solicit gifts; however, people also needed to be perceived as 'deserving' of a contribution (or what Lankenau, 1999: 309 calls 'respectable') which takes us to the second axis of the begging narrative.

Second axis: Deservingness
At the opposite end of the spectrum, I observed how, in the begging encounter, neediness is balanced off with what I call deservingness. People are more likely to give if they think you are not only needy but also deserving of their gift. A common hindrance, for instance, is the perception that homeless people will spend donated money on drugs or alcohol (McIntosh and Erskine, 2000). While this was true for many of my informants, 13 too, it was something that was hidden in the begging labour (the hiding being in itself a kind of labour) in order to balance neediness with deservingness. The people I accompanied over two years perceived their substance use as a problem for the narrative of deservingness (mirroring public perception and also the more general 'welfare conditionality' (see for instance Johnsen et al., 2014).
In this sense, many of my informants thought about bodily and clothing hygiene as something that benefitted both their health and their begging work. Hygiene -shaving one's beard, showering regularly, washing and changing clothes -is part of one's presentability. For Camilla, a young Eastern-European woman doing her rounds at the Gare du Nord, asking people for money for a train fare, looking appropriate was the most important part of her spiel. Camilla was only a fleeting acquaintance of mine and I mostly saw her at the needle exchange North of the Gare du Nord where I volunteered. She was a heavy injecting drug user in her mid-20s. During the summer of 2015, for several days in a row I observed how she interacted with people and her line of argument was based on the fact that people took her to be 'one of them', passing as not being homeless. Her usual story was a variation of the following: Excuse me, I am really sorry to bother you, but I don't know what to do anymore. I tried to reach my family but nobody is picking up the phone, and somebody has stolen my wallet, so I can't take out any more money or go to the bank. I don't live far away, and I only need 5.30 Euro to pay for my train there. Would you be able to help me? I would be so grateful to you. I am really sorry to bother you.
The story, however, was only one part of her presentation which was dependent on her looking 'as if' she had actually just lost her wallet on the way home. Whenever I saw her, her hair was usually combed; her clothes were in a good shape, particularly her shoes -the item of clothing most under pressure on the street. She always carried around a handbag -small, in comparison to the bigger bags that many of the other people living on the street would have with them containing all the necessities of life. She fitted in very well with the general crowd of tourists and commuters at the Gare du Nord.
Camilla engaged in what Gonyea and Melekis (2016) in their study of homeless women in Boston described as 'passing' (see also Goffman, 1990): some of their informants used a certain way of presenting themselves to pass as what she calls a 'normal' person rather, than a user of a homeless shelter, both in relation to members of the (potentially giving) public and professionals, such as health-care employers. While the focus of the above study (see Donley and Jackson, 2014) is on reducing visibility as a homeless person in order to decrease stigma (Donley and Jackson, 2014: 47), my analysis sees 'passing' or 'blending in' (Donley and Jackson, 2014;Hodgetts et al., 2010) as part of the labour of begging. Camilla needed to be visible to begin with -she was trying to get the attention of passers-by to ask them for money -but not as a homeless person. While there are moments where invisibility can be the aim -usually to procure access to semi-public infrastructure such as shops -being invisible for Camilla is about not being perceived of as unworthy, or as homeless, but instead of deserving of help.
Carl further elaborated on deservingness, and how he made it easier for people to judge him adequately as a deserving person. He would try to display certain parts of his identityhis orderliness -but hide others, such as his alcohol consumption: You need to be able to make contact with people before you actually ask them for money. [. . .] I don't like it when people see me with beer. I take a quick break to drink -10-15 minutes.
While he himself likes to choose the people he asks, he also wants to make it easier for them to see that he is orderly, well-dressed and not aggressive, dirty or drunk. He wants to appear worthy of donations, and as if he won't spend his money on alcohol (despite the fact that he, like Pascal, did like to drink whilst begging). He was hiding his alcohol as he learnt that it would decrease his deservingness, and, as a result, his income.
The important puzzle which all of my informants had to deal with was how, in the end, to balance neediness and deservingness. Both are part of what Goffman (1959) calls 'impression management' and Snow and Anderson (1993) term 'identity work': my informants engaged in a constant effort to 'anticipate, project, define, interpret, assess, accept, resist and modify images of self' (Dietz et al., 1994: 60). Rather than acting in accordance with their own perception of the self, my informants were trying to appease the expectation of potential givers (Erickson, 1995). Appearing 'too needy' -scruffy, unwashed, with ripped clothes -could put off potential givers. Appearing 'too deserving' might in turn raise questions about their neediness. I found that elaborate narratives, such as Camilla's, were one way of addressing this question. Another way was to overcome these initial, firstcontact considerations in the mind of the giver, and to build up a network of personal connections which was the third important axis of my informants' narratives.

Third axis: Personal connection
Pascal was very good in building on the third axis -personal connection based on certain commonalities with donors -and was clever in adapting his 'scripts' to do so. He had learnt how to craft the narrative he presented to potential donors in order to bring about the desired result. In the following example, he was very successful because he shared a country of birth, a language, an interest in marijuana and a similar age with a group of young German donors, and, later, a Congolese woman: It was a crazy day. There were Germans -living in Paris -young Germans who wanted to buy marijuana and I got it for them and they bought me stuff for e10. Not only is this experience an example of how scripts are consciously used and changed according to the situation, in order to build first of all temporary personal connections, it is also noticeable how Pascal's relationship to begging changed over time. He gains positive feelings when he is successful, when the work goes well. This success -firstly in the procurement of money and marijuana, secondly supposedly a night at the hotel -was not random, but created by the presentation of Pascal's story. Pascal in fact admits that he consciously changed his story from a needy homeless person into somebody who shares a country of origin with the second donor to forge a relationship with her based on his understanding of her identity (as a French-Congolese woman). Since he started begging, Pascal had developed important skills to increase success. He had become more attunedthrough what Summerson Carr calls 'metalinguistic labour' and experience in the field -as to which 'script', which part of his narrative, identity or life -being German or Congolese, for instance -was going to help him. In this instance, he quickly switched from a more direct narrative of a needy, young homeless man to building an ad hoc connection. The third axis -personal connection -was enough to make the encounter successful, and offset any potentially less convincing displays of neediness or deservingness. * * * For Natasha, begging also revolved around building personal connections with potential donors. Natasha is of Algerian descent but has lived in France all her life. In 2015, Natasha was in her late 60s. I first met her early in 2015 in front of Quick, the French equivalent of McDonalds, opposite the main entrance to the station; I came to learn it was her usual spot. While she slept rough elsewhere (I never found out where exactly), Natasha spent most of her time 'hanging out' at the train station. She always wore a woollen hat, even in the summer, from under which her dark brown eyes would look up at you when you talked to her. Her voice was deep and kind. It didn't quite seem to match her diminutive body, but perhaps years on the street had given her an inner gravity, a weightiness and depth that radiated out through her words. Over time, I observed that Natasha shared an important commonality with many of her donors: the Arabic language. This was especially true of many of the street merchants, delivery people and passers-by Natasha talked with, and the shared language created an immediate connection. Bringing these bonds to the fore judging the commonality of her own experiences as a French-Algerian woman with the passers-by, was her key for making money.
On the other hand, Natasha's way of connecting was more long-term and less situational, and included more of what we could call networking. She banked fully on her group of regulars which she had built up during over a decade in the Gare du Nord. She didn't make up a script of worthiness or neediness, but one of 'having something in common'.
But importantly, Natasha used connections she had created over years of begging on the streets and spun them further by giving away personal details. She talked about the past, how she grew up in Algeria before coming to Paris with her parents and seven siblings when she was still very young. She talked about her criminal career after she moved out from home age 17, leading her into prison as well as a disastrous marriage, which brought about three children who have all grown up and apart. Building connections through these details, Natasha was able to beg almost without moving, often without even asking people to give to her. She called people by their names, joked with them as they walked past and engaged in small-talk about the weather, her life and the police. Very rarely, and usually only towards the end of a conversation, would she ask for money directly. She waited for the moment in which the person was ready to walk away after having stopped to talk. She used this moment of insecurity and vulnerability in which the person was busy disentangling herself from the conversation to make her advance. It worked more often than not. Natasha was needy and deserving enough for people to give to her, but the key factor in her begging encounter was this personal contact and her network of regulars (Lankenau, 1999).
Investing in regulars as part of the labouring process -as with capital investments -paid out over time for people like Natasha. Her narrative in this sense had been spun over years and rested mostly on the third axis. Natasha's developed skill as a begging woman consisted less of a scrip of being needy or deserving but a long time and continuous investment (of time and personal details) into her personal connections with people that had something in common with her.

Labour of hope
During a trip organised by his day centre to a park just outside of Paris, Carl and I spoke about his plans, his next steps. He wanted to reconnect with his ex-girlfriend whom he had left behind with their son in Berlin. I am on the street. I can't host her or show me my home or anything. I can't even offer her a cup of tea right now [. . .] it is my son's birthday soon and I am saving up to buy him something, Lego, perhaps.
Carl was still sleeping rough at the time and earned all his money through begging; he was slowly engaging in the French social care system and on the way to being offered accommodation and possibly also some kind of minimal benefit. Begging, however, was his only way of earning money. It on the one hand kept him alive -of what Stettinger (2003) calls survie -but on the other a travail. Stettinger describes travail as a structured routine and project with an aim for the future in mind. Carl had a clear aim in mind: getting better and re-connecting with his family; this was part of what his assistant social would call his projet de vie. Working -or labouring -towards this was enabled by begging. Symbolically speaking, Carl saving begging money to buy Lego for his son makes this ambiguous nature of the practice -between present-day surviving and longer term planning -very clear.
Similar to Millar's (2008) trash-collecting informants in Rio de Janeiro, the unwaged labour my homeless informants in Paris engaged in was both a result of their situation (an unstable daily life, often involving suffering) and a refuge (Millar, 2008: 35). The labour is destabilized by life which demanded an irregular and flexible, rather than wage-producing post-Fordist, kind of occupation (Millar, 2008: 48). But it on the other hand also helped to stabilize life, keeping the future open, enabling connections -like Carl's present for his son. In this sense, begging went far beyond mere present-day survival. For many of my informants it was future-oriented -not necessarily in that they would all save money for a certain purpose, but in its concrete contribution to survival and hence the opportunity to keep a future perceivable. Begging was part of a set of practices -including shelter making or coordinating with the assistant social at a day centre -constituting and structuring the daily life of my informants. This daily life was the necessary basis from which a future is imaginable. People on the street are forced to spend most of their energy on being in what seems like a present-oriented state -begging, drinking, sleeping rough -while often this was the only option for them to keep the future, the longer term hope -for leaving the street, for instance -open while they were waiting to progress into temporary housing or other types of accommodation. It is in this sense that I following my informants categorise begging as travail, and as a form of ordering one's day and world (part of what Douglas, 1991 calls home-making) and creating a sense of self-worth (Kassah, 2008;Lankenau, 1999) as well as income.
In his study of urban Mongolia, Pedersen (2012) further describes the kind of ambiguous practice as 'work (or in my case: labour) of hope'. Describing people struggling ('muddling through') the lower end of society in Mongolia, Pedersen observes his informants 'practicing hope'. Their daily practices -meeting people, pursuing creditors, talking money out of people, convincing people to postpone the payback date for a debt -was not about reaching a goal, it was, in Stettinger's above sense, about surviving (survie) and continuing to live life (Pedersen, 2012: 11). Pedersen's informants often didn't have realizable goals in a far-off future; there were no goals that could be broken down into step-by-step action. In fact, they were often "hoping for the magical manifestation of 'profit'" (Pedersen, 2012: 4) while continuously doing things which were "active, intersubjective, and [. . .] social" (Pedersen, 2012: 11). The activity itself was to hope as it kept them busy and gave them something to do (Pedersen, 2012: 12) and as such was directly connected to keeping the future open.
My informants were similarly engaged in begging as a way of surviving and 'keeping the future open'. It was not part of a long-term strategic plan, but part of the daily necessity of continuing life. The future was not on their mind every time they were begging but begging was a necessary occupation to be able to imagine a future. It was both an activity focused on daily survival (Bird-David, 1990;Day et al., 1999) on the one hand, and a futureoriented, structured and aimed undertaking in the sense of keeping the body alive and making time for things to unfold, to work towards, to hope for (Pedersen, 2012;Zigon, 2005).

Conclusion
Thinking through material collected during two years of fieldwork with people sleeping rough in Paris I saw a lot of suffering and a lot of violence. But I also observed much active engagement, creative ordering and struggle. Following Robbins (2013) in his call for describing the good life I found the latter set of practices particularly curious in a setting that is often characterised by its passivity and suffering. Focusing in this paper particularly on the economic aspect of my informants' struggles on the street, I theorise begging as a labour of hope. My informants begged to survive; on the one hand, they didn't produce anything of lasting value or importance beyond their immediate ability to consume in a repetitive circuit. In this sense, begging is quite useless, categorically unproductive. On the other hand, however, Still, begging is a skill which my informants acquired and practiced and one that structured their day and routines and kept their future open, enabled their projets de vies. It involves significant amounts of labour in both physical and emotional ways once skilfully prepared (by choosing the right spot): the labour of building personal relationships and developing networks (Pascal, Natasha); of choosing a spot (Carl); the emotional labour of suppressing the shame and embarrassment to ask people for money (Pascal); the physical labour of standing, walking or sitting searching for potential givers; passing as 'one of us' by staying clean and consciously dressing up (Camilla); and, most importantly, the overall labour of cleverly constructing narratives and scripts managing the three axes (neediness, deservingness, personal connection). These narratives were built around balancing three main axes, of deservingness, neediness and personal connections; they were about covering up one's status as a homeless person and passing 'as if' or about being visible as a deserving person begging. My informants learnt how to balance the different axes in their begging scripts depending on their own and the passers-by position (of gender, age, religion and language ability for instance).
Despite its apparent unproductiveness, begging was a fundamental part of my informants' striving: as a labour of hope (Pedersen, 2012; see also Zigon, 2005), it was at times future-oriented, reflective and structured, a way of making money but also of ordering one's day (what Stettinger, 2003 calls travail). At other times, the labour itself was hope; less about the future, more about doing something in the present, keeping busy (Stettinger's survie). As a kind of labour among marginalised people such as my informants, it allows homeless people to struggle along on a daily basis (Desjarlais, 1994;Pedersen, 2012), keeping hope alive and the future imaginable. While it came with significant problems which required additional (emotional) labour to overcome, begging had both symbolically and materially a positive connotation for many of my informants. The labour of beggingmendecit e -was a key part of my informants' struggle for a not necessarily a good but at least better life keeping the bridge from the present into the future open.

Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/ or publication of this article.
12. All of the people I call my informants knew about my research and agreed to be part of it; only once that boundary had been established would I (rarely and irregularly) give money to them specifically to compensate them for time focused on questions I asked them. 13. Official statistics of the prevalence of drug and alcohol use among people who are homeless is mostly underestimating the extent; particularly having worked predominantly with people sleeping rough, more than 80% of my informants regular drank (most of them every day) and/or consumed illegal substances.