The Coining of Convivial Public Space: Homelessness, Outreach Work, and Interaction Order

This article engages with the “convivial turn” in writings about the city and offers a reorientation of sorts. Beginning with encounters, rather than particular spaces, we make the case that conviviality and its limits are realised in practices. Rather than starting in set piece urban spaces designed to foster conviviality we start out on the move, with frontline street‐based care and outreach workers in Cardiff, Wales, and Manhattan, New York City, as they seek out and meet up with those sleeping on city streets. This provides a view of an improvised conviviality that makes the most of whatever the material affordances of a given city space happen to provide. Our research points to how these encounters necessarily take place in marginal settings and times due to the sorts of exclusions that can be built into contemporary city spaces that can at the same time be welcoming to the public, but hostile toward those most in need and vulnerably located in the centre of things. In this sense, we approach conviviality as a fragile interactional accomplishment and, in doing so, see questions of conviviality and conflict as less of a big‐picture paradox of togetherness and distance, hope and hate in urban life, and more of a dynamic relation of co‐presence and visibility. Public space, and indeed public life, might then be reconsidered not as a location but, rather, an active, shifting accomplishment, variously coloured by the politics of seeing and being seen.


Introduction
This article considers conflict and conviviality, two ways in which urban life is commonly represented: The city as the prototypical site of variance, difficulty, and ill-will on the one hand, or a site of co-presence and shared encounter on the other. We do not look to make the case for either one of these representations unconnectedly or vis-à-vis the other, but rather to explore how they might braid together on occasion and complicate or compromise one another. Our intent is to offer an engagement with the "convivial turn" in writings about public spaces and the status of the convivial as an uncomplicatedly good or even "groovy" thing (Wise & Noble, 2016) in urban planning. We are not, of course, critical of the idea of or hope for conviviality but we do aim to contribute to discussions thereof by considering-from a perspective that emphasises interactional practices over space and materiality and design-how conviviality might get done, practically, and how the limits of conviviality are produced in situ.
Our contribution does not begin in notably convivial settings, but rather, we set out and about the city with workers whose very job it is to seek out encounters marked by a convivial spirit; to meet up with "people who are different, but without the idea to [produce] a homogenous group" (Fincher & Iveson, 2008, p. 154). We provide some ethnographic details of such encounters between street-based outreach workers and their unhoused, rough-sleeping clients to demonstrate how conviviality can be found in overlooked settings and spaces that, certainly, were not designed for such "comings together" (Goffman, 1963). Outreach encounters necessarily take place in a range of settings; the very purpose and practice of outreach work is to meet vulnerable clients wherever they are, physically and socially (Hall & Smith, 2017). As such, these homeless encounters are coloured by a kind of improvised conviviality, their participants meet on common ground, socially and materially, relying on any material affordances at hand in producing a setting in which severe need can be met with care and kindness. Before arriving at these encounters, we begin with a consideration of some of the literature on conviviality in relation to interactionist treatments of co-presence in public space. In the context of enthusiasm for the possibilities of designing convivial settings, we discuss the practical management of interactions such that provide for co-presence and for conviviality to occur. In addition to some well-known observations on interaction in public space provided by Erving Goffman, the discussion is grounded in an alternative perspective on categories and category relevancy provided by Harvey Sacks.
Here we treat categories not as a fixed schema, nor a label that moves with the individual, but as practical, local, and situated practices for organising social scenes and participation in public space. At the same time, we consider how planned settings in which co-presence and conviviality might be found have local limitations and even inherent contradictions; welcoming and attractive on the one hand, exclusory and hostile on the other. This we suggest can be understood as both an interaction order matter, as well as something that is built into set piece convivial settings. We think that our examples of outreach workers and their clients coming together, convivially, and making do with the material resources of whatever setting in which they meet, offers further insight into this relation. Equally, our examples show that conviviality, wheresoever it is found, can be a fleeting and fragile accomplishment. More generally, our examples point to a politics of visibility in urban life that finds encounters between outreach workers and their vulnerable charges also appearing out of place, unwanted, and unwelcome, no matter how convivial they might be internally. We posit that such fleeting comings together at the margins of urban space demonstrate how convivial settings can be coined as a matter of necessity, and more broadly, that the very essence of public space (and of public life) is occasioned in and through the moments of these practices.

Accomplishing Conviviality and Its Limits
A body of contemporary writings on urban life have aimed to identify and describe what it is about particular social settings that can give rise to conditions of conviviality. These settings range from the mundane-encounters in a café (Jones et al., 2015;Laurier & Philo, 2006) or on public transport (Wilson, 2011)-to more spectacu-lar planned spaces such as urban squares (Bates, 2018;Ganji & Rishbeth, 2020) and public ice rinks .
Much of the contemporary thinking around conviviality sets out from the work of Paul Gilroy. Gilroy (2004) encourages us to think about conviviality (along with melancholy) in such a way that moves away from troubles associated with the retention of essentialist notions of ethnicity, race, and nation in multiculturalist and communitarian thinking. Conviviality offers a means of revitalising a sense of the possibilities and modalities for a public life lived together. Gilroy's (2004) notion of conviviality overcomes the centrality of the autonomous individual as well as the constraints of the community by emphasising a connectedness of the human condition obscured by dominant categorisations of race, class, gender, and so on (Neal et al., 2019;Nowicka & Vertovec, 2013). Conviviality proper is defined by Gilroy (2004, p. 27) as: A social pattern in which different metropolitan groups dwell in close proximity but where their racial, linguistic and religious particularities do not-as the logic of ethnic absolutism suggests they must-add up to discontinuities of experience or insuperable problems of communication.
Along these same lines, much of the hope of planning for convivial urban dwelling stems from observing situations in which differences of ethnicity and race, along with other forms of categorial division, become "unremarkable" (Nowicka & Vertovec, 2013). This is not to suggest, of course, that racial identity can easily "disappear.'' Any hope for conviviality in society more generally runs up against the recognition that those who are viewed as "out of place" will continue to be treated with suspicion, discourtesy, and in far too many instances (lethal) violence (e.g., Anderson, 2022;Rawls & Duck, 2020). People do, however, routinely share space; perhaps not by communicating directly, but by acting in such a way that enables and supports co-presence and a continuity of experience as members of the setting (see Horgan, 2020, on civil inattention and strangership). It is in this sense that we suggest that pursuing urban conviviality and, indeed, a practically oriented sociology of hope (Plummer, 2013), might well begin by attending to the organisation of the scenes in which seemingly deep social divisions are made irrelevant or, at least, can be temporarily sidelined. In this sense, mundane situations are instructive as to the wider possibilities of public space. Speaking of the service line, for instance, Goffman (1983, p. 14) observes that the "first-come, first-served'' rule: ...produces a temporal ordering that totally blocks the influence of such differential social statuses and relationships as the candidates bring with them to the service situations-attributes which are of massive significance outside of the situation. (Here is the quintessential case of "local determinism" as a blocking device).
We recognise that conviviality properly relates to more than standing in line, but we also want to recognise that anything approximating conviviality cannot be achieved without the sorts of situated standing rules, commitments, and obligations identified in Goffman's writings on public space. Indeed, these practices undergird what Horgan et al. (2020, p. 147) have in mind when they discuss the "affordances of sociability" which, combined, produce what they call "soft infrastructures of sociability." Before we arrive at our empirical examples which demonstrate something of this relationship, we think there are additional insights to be gained from an engagement with the work of Harvey Sacks (1995) and insights from membership categorisation analysis (Fitzgerald & Housley, 2015;Hester & Eglin, 1997;Smith, in press).
The first and foundational observation is that any individual can be described through a near-endless array of categories (Drew, 1978). Rather than starting from the sense that people "belong" to categories, Sacks, and the work that has followed, demonstrated just how category selections are made in any actual case through what has been called the "members' categorial apparatus" (Sacks, 1995). A key point for the current discussion is that categories and their organisation are local and occasioned, are yielded by the setting, and are operationalised by members in relation to observable activities. To return to the example of the service line, turn-generated categories (next to be served, end of the line, and so on) have more relevance than populationtype categories, unless, that is, there are good reasons for them to become relevant in and through the business of queuing (Watson, 2015). They also accomplish the materiality of the setting as socially salient. Population type categories (of gender, race, and social class too) are recognisable and available, of course, but the work of sharing a resource or a space reconfigures category relevancies in the course of the accomplishment of a given social setting. In a public square, categories relating to activities and spatial/mobile formations-skateboarder, cyclist, smoker, passer-through, a couple, a group, and so on (Lee & Watson, 1993;Smith, 2017)-are relevant categorisations bound up with the order of the space itself. We suggest that it is through this relationship that racial categories (or any other population type category) can be described as "unremarkable" (Nowicka & Vertovec, 2013), precisely because of a situationally reduced or side-lined practical relevancy (Coulter, 1996).
Attending to such interactional and categorisation practices can yield some insight into the production of public space, the possibilities of conviviality and its limits, and a wider interactionally-realised visibility politics of public space. Starting from matters of interaction order we now consider some of the potential contradictions and limitations of planning for conviviality.

Planning for Conviviality and the Limits of Tolerance
In terms of the planning of public spaces, the recurrent message-from Jane Jacobs, William Whyte, and their contemporaries through to the present day-is that inclusivity and even social transformation (Bredewold et al., 2020) can be achieved through a revised approach to the city's built-in equipment when that equipment is geared toward connection and openness, rather than division, privatisation, and securitisation (Jacobs, 1961;Minton, 2012;Stavrides, 2013). Across the contemporary critical urban planning literature, there is a growing recognition of the nuanced relationship between the built environment, social practices, and encounters and divisions, in an "emphasis on the spatial form and vital materiality of convivial places" (Bates, 2018, p. 985). Various spaces in which this vital materiality has fostered something of the public sense of easy togetherness found in Anderson's (2004) discussion of "cosmopolitan canopies" include settings such as cafes (Jones et al., 2015), swimming pools (Bates & Moles, 2022), and bowling alleys (Jackson, 2019). The materiality of the setting is matched by and supports a convivial sociality.
In keeping with the legacy of attention to the everyday life of public spaces, the vital materiality at the heart of many convivial spaces today is not so very grand at all, tending instead to small-scale interventions. "Edge" and "threshold" surfaces encourage and enable people to stop for a minute and engage in the pleasures of observing the activities of strangers (Bates, 2018;Ganji & Rishbeth, 2020) thus establishing the sorts of temporary and shifting co-presence that can foster conviviality. Benches and other forms of seating, open landscaped areas, and sculptures and fountains produce attractive and physically comfortable spaces in which people might stay for a while to "linger, sit, eat, drink, and converse" (Shaftoe, 2008, pp. 60-61). Even though such encounters along these edges might not seem so very significant, the necessary sharing of a space that is not observably owned by anybody provides for something like the temporary blocking of the relevancy of social status and identities described above: first come, first served (Goffman, 1983). Indeed, spaces which require the sharing of an open resource are likely best set to foster diversity and conviviality. An example can be drawn from one of the authors' personal observations. A bar in a city in the UK made food-produced from goods that were otherwise set to be thrown away-freely available to anyone and everyone who entered. This was coupled with a clear sense that no one was obliged to purchase a drink to be there, sit for a while, and help themselves to that day's offering. This accessibility plus the sharing of the space, the food, and an expectation of tolerance made for an uncommon inclusiveness and diversity. No one group could "claim" the bar as theirs over any other: a group of what appeared to be homeless people shared a table next to a group of smartly dressed people on a night out. At the same time, there were rules in place, tacit and explicit; the whole scene was presided over by the landlord who would remind people of the equality of all patrons in and through the equal application of the ground rules: "Hey, everyone! No double dipping the hummus, okay?" The sharing of resources is at the heart of the production of public space. Yet, at the same time, there is a contradiction, recognised in the literature, surrounding the accessibility of resources, designing for inclusivity, and the apparent requirement for the management of the space by agents (not unlike the landlord above) "with the special job of keeping 'order'" (Goffman, 1963, p. 210). Producing spaces which are open and necessarily shared by whomsoever is there at any one moment is one thing; producing and maintaining a shared and lived tolerance toward the wide range of activities and potential differences between the population of the setting is another. Often such tolerances rely upon the maintenance and management of the social landscape along with adjustments to its physical counterpart (Ganji & Rishbeth, 2020). Designated agents (park rangers are a case in point, see Ablitt, 2021) as well as the "eyes on the street" provided by members of the public play a part (Jacobs, 1961); as does, of course, the built-in equipment of any given setting.
Properly open, accessible, and comfortable public spaces are just that and are so for whomsoever happens to be there at any time. That openness is, surely, the very grounds of convivial public space. Yet, for all that we can celebrate the insights provided by Whyte (1980) and others in fostering comings together and conviviality in public space, it must be recognised that these recommendations were also at the heart of the "domestication" of areas of the city, such as Bryant Park, Manhattan (Zukin, 1995) that once provided a space, a sanctuary of sorts, for those whose lives were caught up in street homelessness, drug use, and the displacement caused by the de-institutionalisation moment. Indeed, Bates (2018, p. 987) writes of the same process in the regeneration of London's General Gordon Square as involving the "transmutat[ion of]…a derelict space, frequented only by street drinkers and cat-sized rodents, into a vibrant hub of multicultural life," and goes on to note that a: ...series of public drinking bans have been enforced in an attempt to move the street drinkers out of the square. These bans are intended to make the square a safer and more appealing place to other residents, but they also exclude those people who are already marginalised from other places and may be most in need of access to public space.
Much of this contradiction is embodied by the humble urban bench; a necessity, and valued resource. Yet the concern with the potentially disruptive spectre of unwelcome users is reflected, materially, in the form of benches, hostile to anything but constrained individualised and temporary sitting (Bates et al., 2017). There is a growing consensus that aggressive architecture is bad (Chellew, 2016;Smith & Walters, 2018) and that drinking and begging bans might be a necessary aspect of managing public spaces which, nonetheless, have an exclusory effect (Bates, 2018;Ganji & Rishbeth, 2020); but much of the discussion still finds the street homeless at the edge of things, or problematically visible, or simply not considered at all. One significant proposition in this context is found in the suggestion that a city that is excessively planned-too smooth, with all the wrinkles ironed outlacks those spaces to which marginalised groups might otherwise go to "get on with their own lives out the way of others." (Shaftoe, 2008, pp. 26-27). Accordingly, the same argument runs that certain areas of the city should be left unplanned, allowed to run down a little, to become leftover spaces. Such "slack spaces" provide a necessary public resource to a whole range of groups who might find the open and planned city does not contain the built-in equipment suited to a whole range of needs (Ablitt, 2020;Cuyvers, 2006;Hall & Smith, 2017;Worpole & Knox, 2007). The very inclusion of this mention of "slack" locales in a text directed to the production of convivial urban spaces points directly to the relation at the heart of our argument. Slack spaces, then, provide a quintessential sort of public "free" space.
The key point we are making here is that the potential of conviviality and convivial space has its limits. Limits that are, from our view, produced in and through interaction order and specific practical resources. As there are good reasons to be hopeful and to support the push for the kind of inclusive urban realm in which different and potentially divisive identity categories might become unremarkable, it remains the case that there are invariably those who, under their activities and appearance, are liable to become treated as very much remarkable, as "out of place," ahead of any other consideration for their needs or reasons or rights to be present. Indeed, in addition to recognising the practices that sustain the possibility of co-presence, Goffman understood public spaces and social settings as sites that are policed in relation to situated rules of conduct and expectations and obligations of self-presentation: "To be awkward or unkempt, to talk or move wrongly, is to be a dangerous giant, a destroyer of worlds" (Goffman, 1961, p. 72). Individuals and groups can thus be found "out of place" by breaches-assumed and actual-of locally established norms of conduct. This sort of mundane exclusion sits at and defines the boundary of public tolerance and can itself stem from little more than appearances.
Whilst plans for convivial spaces are indicative of a move away from revanchist models of the city (see, Lawton, 2018;Smith, 2005)-in both a political and aesthetic sense-the contradictions and tensions sketched above appear to produce some continuing uncertainty as to what to do about those unwelcome characters whose sheer visible presence is conceptually inadmissible as a part of any conventionally imagined convivial scene. Wherever one stands concerning the production, management, and potential of convivial city space (and its contradictions), the limits of conviviality seem unavoidably bound up with the visually available relationship between observed activities and their doers and situated notions of who "belongs" in any given setting. Returning to the activities of "lingering" and "drinking" mentioned in the discussion of tolerance and slack spaces above (Shaftoe, 2008), these activities are irrevocably tied to the visually available categories of their doers. "Drinking" is a category-neutral activity in the sense that anyone can do it, yet, of course, "street drinking" is treated entirely differently depending on just who is seen to be doing it. So too with the contrast expressed in the action-categorisations "loitering" and "lingering," even though, descriptively, the activity might be the same (making use of a bench in a park, to have a drink and a chat, and maybe a bite to eat with friends). Eldridge (2010), for example, traces something of this category-action relation when considering differential treatments of and tolerances for urination in public spaces in the night-time economy, adding an important temporal dimension.
The lingering of a mix of certain groups produces a convivial setting; however, the lingering of others can generate unease and conflict relating to the right to be in the "open" setting in the first instance. The wrong type of presence can quickly become a problem that needs managing and likely threatens the success of any planned convivial space until managed (i.e., removed). In this way, the presence of "street drinkers" might, ironically enough, have a unifying effect, categorically speaking. Street drinkers are likely to be reported by multiple disapproving agents, and then discouraged or removed by another class of agents. In more extreme cases-one of which we have observed directly in Cardiff and are sure has happened in any number of cities-the benches that were providing for the sociality of street drinkers are themselves removed. This, of course, has the consequence of individuals and groups of "unwanted" urban citizens being dispersed and meeting up elsewhere. And knowing only too well that they are likely to be so-to be removed, that is-they may look to assemble elsewhere, to begin with, on their own terms and away from whatever set piece convivial spaces the city has offered up, perhaps gathering in those "slack" spaces, away from eyes on the street that might take offence. The dispersal of homeless individuals is a regular and routine occurrence that finds them in a more or less constant state of mobility (Hall & Smith, 2014). Their movements trace something of the limits of tolerance, and patterns of cruelty and kindness, in the city centre. Their movements also produce a challenge for those workersvery briefly introduced at the outset of the articlewhose job it is to seek out encounters with these individuals in order to-attempt to-open up a relationship of some sort, to-attempt to-tend to some immediate and long-standing needs, and, primarily, to-attempt to-be there. We want to suggest that these sorts of contacts in these sorts of spaces point to a primal sort of public space; encounters with others, coloured by a togetherness (if only physically, for a while), the sharing of resources (even if just space or shelter), and the sort of improvised and self-managed character that is so very hard to build into city life more generally. Of course, these gatherings themselves produce a spectacle of sorts. Comings together of groups of people have that quality. That is, after all, one of the pleasures of convivial spaces. These encounters and gatherings, however, are not often celebrated. As we go on to describe below, the potential and hope for convivial urbanism that might be observed in these settings is tempered by the same politics of visibility and tolerance of difference that equally sets the tone and pace of the city as "vibrant," "diverse," "accessible,'' and so on.

Homeless Outreach
In this section, we turn to two brief descriptive examples drawn from ethnographic research undertaken with teams of outreach workers-municipal and charitabletasked to seek out therapeutic and supportive encounters with individuals experiencing street homelessness wherever such encounters can be accomplished. The workers are required to be experts in encounteringattuning to, searching for, and finding-those who are "different" (Hall & Smith, 2017). Their daily business, much of it conducted well outside of a 9-5 hour shift, is accomplished across the city, on the move as frontline workers (Smith & Hall, 2016). These two ethnographic examples are based on fieldwork undertaken in Manhattan, New York City (Williams, 2022) and Cardiff, Wales (Hall & Smith, 2017) respectively, although we wish to minimise the importance of their geographic locations here. What is noticeable is that the methods and settings of these outreach work encounters are similar on both sides of the Atlantic, and indeed the emergent convivial practices that the occasion will be recognisable across cultures and continents.
Tracing such mobile work reveals a shifting distribution of mostly rogue locations-rear alleys, disused or neglected lots, vacant office frontages, fire-escape stairwells, or residual corners of the cityscape-at or near which outreach workers set up shop for not much more than 10 or 20 minutes, providing hot food, information and advice, health and wellness checks to known and potential clients from the back of the team's van (loaded with thermos flasks, leaflets, donated clothing, needleexchange kits) before moving on again to continue an exploratory roving patrol. The locations of this work are, as Popovski and Young (2022, p. 2) describe, "mundane and unremarkable" locations-everyday locations that are available for adaptive use and offer the potential for "subversion" (Amin & Thrift, 2002). In the accounts that follow, the attention is not necessarily on the illegitimate or subversive use of places (as in Popovski & Young, 2022), but on the convivial practices which happen in everyday, mundane, and unremarkable locations. Practices which do not represent the intended or designed purpose of those places, but which produce a kind of conviviality all the same.
There is much more to say here, but the relevant point is that these workers are in the business of actively seeking out encounters with difference, and in doing so they practice their own sort of conviviality, sharing in the production of a shifting and unique collection of (sometimes) convivial spaces which are themselves temporary, fleeting, and vulnerable; sometimes tolerated, and sometimes not. The two examples comprise fieldnote selections from two separate but complementary ethnographic studies of homeless outreach practices in Manhattan and Cardiff. Both studies involved the researchers becoming part of the outreach team that they were researching, and actively doing the job of outreach work in their respective city remits. The Manhattan study was carried out for one year, while the Cardiff study was an extended seven-year engagement. Both accounts that follow were jotted down in situ as shorthand "scratch notes" or prompts and later written up by the participant researchers as full fieldnote vignettes at the end of their respective shifts. They both closely attend the multimodal practical action of the outreach encounter in terms of capturing the interaction as it plays out. These examples provide a foil to a more conventionally imagined urban conviviality, and they can perhaps shed light on the convivial possibilities of such comings together.

Bobby's Radio
An urban fly-over and the area beneath it resembling a parking lot, a wide paved space with white markings; to one side a series of construction sites, to the other buildings and a street intersection leading either into town or onto a slipway and back to the multi-lane highway. The space is inexact, lacking an obvious and intended use: taxis wait, trucks pull in and turn around, water pools on ground here and there. The outreach team park their van directly beneath the centre of the elevated highway. There are two reasons for this: one, in heavy weather cars driving along the highway spray rainwater across the barriers and down-if you were to one or other side you would be soaked through; two, there is a low wall here, separating two parts of the space which can be-and isusefully re-purposed as a seat by those already waiting for the team to arrive. This is the first stop along an evening's longer route, begun at seven o'clock. It is not a busy stop, usually only five regular clients will gather here-two of them labourers from the nearby construction sites, the others are known to the team as homeless clients. They sit spaced out along the wall, waiting for the outreach van to arrive. The two labourers wait together and when the van arrives, they queue together. Whilst serving food and supplies, Fran (one of the volunteers) will speak to them in basic Spanish, explaining what is on offer that evening; they respond in simple English. This brief moment of practicing second languages is a staple of the interactions between the labourers and Fran. The labourers are polite and quiet, smile a lot and, once they have received a portion of food, they leave together.
One of the others waiting at the stop is a man, Bobby, a long-term client of the team, who has been in and out of the homeless shelter system for years. He pulls a small shopping trolley with him, and usually waits at the back of the queue for the others to be served, allowing himself more time for conversation with the team. On occasion he will ask for particular items, things the team likely already have in the van, such as clean socks, underwear, or items of warm clothing. One evening, Fran asks if there was anything else they could get for him, to make his days a little more comfortable. Bobby asks for a battery powered radio to listen to and mount on the shopping trolley as he walks through the city. A couple of days later the team give him a radio.
The radio proves to change the atmosphere at the stop. Bobby plays music and tunes in to news broadcasts and traffic updates. Whenever the team arrives, those waiting (the five regulars, at least) are sitting closely together along the wall, listening to the radio, discussing the news, the game, and sometimes singing along to a song. The team joins in whilst food is served, with everyone present involved in the conversations. "What's on the radio?" becomes a regular talking point. Before leaving, Bobby lets the team know about the traffic updates for their likely route ahead saying, "You don't want to keep anyone waiting.'' For a time, the stop continues like this, the team providing fresh batteries to keep the radio going. However, Bobby's attendance at the stop becomes less regular; he finds a place (to be and sleep) further away and can't always arrive on time. He begins to find other means of getting by. Without Bobby and his radio, the playful atmosphere subsides a little. Added to which a new face appears, waiting at the stop. An ill-tempered Norwegian man who consistently complains about the food or the quality of the supplies. His negativity grates on the team's, and the other clients', patience. The clients now sit apart on the wall, the established regulars avoiding the newcomer. Fran still talks to the labourers in Spanish, and everyone present would still make small talk about events, sports, music, and the traffic. But Bobby's radio is missed, and the new arrival too often interrupts easy conversations. Things have changed.

Dylan the Vicar
The outreach team arrives on Church Street. Vinny and crew have made this street-and, specifically, the modern church and its open porch area-a temporary home. As such, it has become a regular stop on the early morning patrol, and the team have skipped another less likely spot in getting there.
As well as Vinny, Dylan and Bob are there, sleeping bags lined up together under the limited shelter of the church porch. Dylan-an imposing character, but a gentle giant, really-is in one; he opens his eyes, and, seeing Jeff, says, "Oh no, it's you, the bloody Womble." Jeff responds, deadpan: "Do you want any breakfast or not? Come on. Get up!" Getting up to his feet, and his full six foot three, Dylan stretches. He's got hold of a black round-collared shirt from somewhere. "This lot keep calling me a bloody vicar. All I need is a dog collar and I'm sorted," he says. "Bless me father for I have sinned," he adds, putting his hands together in front of him in mock prayer. "Yeah, I know that" says Jeff. Dylan starts moaning about the sausages. Again. Says he dropped his on the road yesterday and "it bloody bounced! Even the seagulls wouldn't touch it." There's laughter at the review. Bob is "out of bed" now and seems to be OK, if a little out of it. Unsteady on his feet. Bleary eyed. "God, I was off my head last night. Didn't even get in my sleeping bag.'' Bob pipes up and says: "Show Rob that thing on your phone, Dyl." Dylan's reluctant-perhaps because of the presence of a woman, Charlie the outreach worker-but after a little encouragement, he ends up showing the group a pornographic cartoon version of the Wizard of Oz. Charlie's come over to have a look too, mockingly rolling her eyes and tutting. As they're watching, Gary turns up with a "new face." The new guy is wearing jeans, boots, and a red hiking jacket. He's probably in his early thirties although it's hard to say for sure. He leans into the group and says, about the video, "that looks like something a paedophile would watch" and laughs. Gary says, "Don't say that round here" and new face replies "Why, what's he going to do? Beat me up?" It isn't quite clear who he's referring to, but big Dylan seems the obvious recipient. The admonishment/threat is left to drop, and the team go back to sorting out the food for the gathered recipients. Gary has his usual: sausage roll with a "cocktail" of brown and red sauce. The new guy is asked what he wants. He then enters into a somewhat over-the-top monologue of thanks about how important outreach work is, how it has been a real life-saver: "Literally. You don't appreciate how much we appreciate what you do." "No problem, man," says Rob.
Jeff, Charlie and Rob leave for the next stop, laughing about the encounter. They're discussing Davey's whereabouts, how he's dropped a load of weight and really hadn't been eating enough recently. Rob asks who the new guy was and is told he's just met the "infamous Chazza; a nasty piece of work, thinks he's a bit of lad. He's pulled a knife on an outreach worker in the past."

Discussion and Conclusion
What to make, then, of these two encounters in the context of a discussion of urban planning and conviviality and conflict? The first thing, the most obvious, is that these are loose comings together in improvised settings. These settings are not designed for this purpose. The ambiguous space under the flyover provides shelter. The low wall is repurposed as seating. The overhang of the church frontage offers a place to sleep out of the weather and, consequently, a place in which an encounter between outreach workers and their clients might take place. We have comings together in open settings. People arriving, some people already waiting, others turning up, joining in in different ways, still others getting on with things, some disrupting. Not really settings at all, but loose and temporary spaces; and yet there is assembly, organised by way of shared space and shared resource, and not only the staple outreach offer-Bobby's radio, two languages, a risqué cartoon, jokes and complaints, cigarettes on occasion, and companyat the margins.
We might also see that material features are perhaps not, in themselves, so very vital, but become enrolled in encounters. Thinking about those material features as "producing" or even "encouraging" convivialitymarginal or otherwise-misses, in a broader sense, how convivial settings are coined as a matter of necessity. If you want access to a resource that is open to whoever turns up, then you have to play the game; just as you must if you want to join a queue. People will quite often be just "making do," and it is the "making do" that makes for a convivial setting. Just ask anyone who has huddled with others in an outside smoking area on a winter's night. Conviviality only needs a very few props (Peattie, 1998, p. 248), but also, certainly, in these instances, some degree of need, whether that be simply a place to sit, or some food to eat, or a jacket to keep off the cold.
At the same time, things are never quite comfortable. This is inhabiting a threshold space (Stavrides, 2001), not simply sitting in it and watching the world go by, and conflict is not so far away. New faces can bring tensions. There is an inside and an outside. The specifics here get done through categorisations of clients, clients with names, clients without, regulars; dangerous categories too-not applied to an individual, necessarily, but of which even a mentioned can signal trouble. The point is that as much as interactional practices produce convivial spaces, they can destroy them too-in a moment. As much as conviviality is about rubbing along and making do, it requires management, and that management can play out in different ways, scales, and temporalities.
For all that we can point to the internal dynamics of conviviality and conflict in our descriptions, these encounters play out in a city where appearances do matter. Being tucked away under a flyover is one thing (the truckers and taxi drivers might not mind so very much). Sleeping in the entrance to a church might matter a little more and certainly so over time with daily outreach visits extending for several weeks-visits that came to be seen as legitimising a continued (and unwelcome) homeless occupancy. Eventually, the covered entrance was renovated out of existence. A wall-mounted camera and movement-sensitive lighting were installed, shortly followed by an encompassing glazed facade pushed flush to the street and eliminating the porch as an available space altogether. Regardless of what we might have to say about encounters and interactions between outreach workers and clients in this one location, the encounters themselves were judged in conflict with the surroundings. Convivial encounters produced in a space for which that space was not designed were designed out. It is in the event of the exclusionary redesign of the space that the "population type" category relevancies of the incumbents, and the incumbent practices, are brought to the fore and aggressively so. These people and practices are found to be out of place in the church porch, to the extent that the porch is redesigned entirely to exclude them. Perhaps more pertinently, the materiality of the space is made acutely enrollable by two opposing situated practices (much like the previously mentioned anti-homeless benches) equally coloured by the politics of visibility.
To finish, we draw attention back to the fact that we have not made much of either of the two cities, as cities, in which each of the above (sequence of) encounters between street-level workers and their clients took place-deliberately so; our point being that it does not matter so very much at all. New York and Cardiff are different sites and cities, markedly so. And, yes, urban planning and homeless policies can vary significantly from one city to the next. But the kinds of careful and spontaneous conviviality, the encounters themselves, that occur between workers seeking out the difference (and damage) and clients making do at the very edge of things-the edge of injustice, of acceptance, of recognition-in each of the two cities referenced are in fact markedly similar. Whatever the differences at the "grand scale," and in city-wide approaches to dealing with the "problem" of homelessness, these encounters look very much alike on either side of the Atlantic. The talk, the care, the management of touch, of information, of closeness, and distance. Of visibility. If you know how to do outreach work in one place, you already know a lot about how to do it in another. When attending to interaction, in situ, in actual existing convivial encounters, there is something stable and consistent and essentially human at the heart of these interactions.
If conviviality is an ethic of openness and care (Fincher & Iveson, 2015), as instanced and accomplished in practice, then that kind of openness and care, combined with an awareness of potential conflict, might be a better place to start than with the material design of convivial locations. This is to propose, in closing, that attending to outreach workers' practices-and to the practices of others like them operating in threshold spacesindeed, attending to practices full stop, instead of attending to spaces, might get us a little closer to the potential of "the convivial," wherever it might be found.