Social practice research in practice. Some methodological challenges in applying practice-based approach to the urban research

ABSTRACT

Praxeological turn in social research resulted in many examples of empirical studies using the concept of social practice. They are mostly case studies of a single or clearly defined practices, such as cooking, dancing or energy consumption. There is a lack of studies employing this concept as a framework for research on complex, dynamic and empirically hard to reach socio-spatial phenomena, such as urban housing estates. Drawing from the experience of a 4-year research project in six housing estates in three Polish cities we present and debate methodological challenges in applying theories of practice to urban research. We discuss methodological implications of representational and heuristic models of explanation – which are being applied in practice-based research. We highlight the ambiguity of the notion of social practice in the existing literature and its troublesome ‘double nature’ – implying that social practices are both implicit and observable phenomena –which, although alluring in the stage of conceptualisation, poses challenges in the stage of data collection, analysis and interpretation.

KEYWORDS:

• Neighbourhood practices
• housing estate
• theories of practice
• practice theory methodologies
• representational and heuristic models of explanation

Introduction

Practice-based approach has become an increasingly popular and influential, however not unified, strand in contemporary social research. Its attraction stems in particular from ‘its capacity to resonate with the contemporary experience that our world is increasingly in flux and interconnected, a world where social entities appear as the result of ongoing work and complex machinations’ (D. Nicolini, 2012, p. 2). The ‘practical turn’ is visible in that ‘phenomena such as knowledge, meaning, human activity, science power, language, social institutions and human transformation [have become to be seen as] occur[ing] within and [being] aspects or components of the field of practices’ (T. R. Schatzki, 2001, p. 2).

But what is a practice? What does the notion of practice that has been increasingly used by contemporary social scholars, often in an ambiguous way without precise articulation, denote? T. R. Schatzki (2002, p. 73) defines practice as ‘a set of doings and sayings. Because these doings and sayings almost always constitute further actions in the contexts in which they are performed, the set of actions that composes a practice is broader than its doings and sayings alone’. One of the basic assumptions within a practice-based approach is the distinction between observable, explicit and directly performed actions (doings and sayings) and implicit, non-observable aspect of practices (i.a. practical understandings and teleoaffective structures). For T. R. Schatzki (2002) both aspects are equally constitutive of practices, only analytically distinguished as two sides of the same phenomenon. For Bueger (2014, p. 386) it is not the performative, observable and articulated but implicit, unobservable and not articulated level of practice of primarily importance and even ontologically prior. Nevertheless, the character of the relationship between both aspects remains controversial. As Galvin and Sunnika–Blank (2016, p. 65) argue it is because the nature of this relationship remains unidentified. They, following Harré (2009), consider a practice theory as an exemplification of a heuristic approach in which the models are built for better recognition of those phenomena and mechanisms which cannot be captured directly, through observation. Thus, a practice is a heuristic device that may help us to understand why people behave the way they do. However, it does not mean that what has been observed (doings and sayings/performances) equates with what has been introduced as a heuristic tool for its explanation. One methodological consequence of this assumption is that if we aim at capturing an implicit aspect of practice, we must follow its observable features. These features are: regularity, habitual nature, repetitiveness, recursivity (Bourdieu, 1990; Giddens, 1984; T. R. Schatzki, 2002; Reckwitz, 2002). We can easily notice these features in the everyday practices of cooking and eating (House, 2019; Twine, 2018), using a laundry (Mylan and Southerton (2018), ballet dancing (Müller, 2017) or even practices of parental involvement (Freeman, 2004) or walking for pleasure with sticks, that is Nordic Walking (Shove & Pantzar, 2005). It is not difficult to see them in the institutional orders (Jarzabkowski, 2004; Mol, 2002; D. Nicolini, 2012; Orlikowski, 2002). There are, however, many research areas, where regularity and routine are empirically hard to grasp. Regularity, repetition and recursivity can, for instance, be manifested in various, short and long, intervals. Moreover, even if we can capture regularity of an action, its habitual nature, in many cases, is still the object of our interpretation.

We encountered such difficulties in our research on the mechanisms of establishment and reproduction of urban territorial communities at the level of the housing estate.¹ We were interested in identifying practices in the dynamic, amorphous urban field, which – even though constitutes a focal point of urban studies – is still misunderstood. Urban housing estate, seen as a constellation of social practices, is an example of a research area which poses challenges when subjected to praxeologization.

The main focus of the discussion presented in this paper is thus a methodological question: how to empirically grasp practices which observable dimension – at first sight – does not take a form suggested by practice-based approach theoreticians: regular, repetitive, instinctive and recursive. In what follows we consider methodological challenges posed by the ‘double nature’ of practices and describe strategies for exploring the implicit aspect of practices that we have developed within the project.

The paper consists of four basic parts. Part one presents the motivations that led us to apply theories of practice to neighbourhood research, the aims of the project and the outline of the methodological strategy. In the second part, we introduce the basic theoretical assumptions of practice-based approach, and we confront them with representational and heuristic models of scientific explanation. The assertion that theories of practice are heuristic devices has serious methodological implications for the research process. Thus, in the third part, we reconstruct the challenges we have encountered and our ways of dealing with them along the process of conceptualisation, operationalisation, fieldwork and, specifically, data analysis and interpretation. In the conclusions, we reflect on the impact of the applied approach on our way of theorizing neighbourhood as a dynamic complex of practices and express the need of further development of research procedures consistent with theoretical assumptions of practice-based approach.

Praxeologising neighbourhood

Traditional conceptual categories such as neighbourly bonds, local identities and local social capital have proved to be not only analytically useless in relation to the complexity and dynamics of everyday life on the estates and beyond, but also suspicious as they create a vision of local integration and community. We are not alone in these qualms. Objections to traditional and normative concepts are being raised by Van Kempen and Wissink (2014), Blokland (2017), and others, who claim a need for new approaches to research on neighbourhoods. Our own intuition and knowledge about what engages residents in everyday life grounded in previous research (A. Bukowski et al., 2007, 2014) suggested that we had to approach what is commonly called ‘neighbourhood community’ from a different perspective. We assumed there are specific mechanisms organising social life in urban territories, however their recognition requires a specific theoretical lens. Our research project (2016–2019), the results of which form the basis for this article, reflected a turn towards practices in our methods of studying urban housing estates. The main objective of this project was to develop a concept explaining the mechanisms of establishment and reproduction of urban territorial communities at the level of the estate. It is why we talk about research on neighbourhood practices and not on neighbourhood community understood as an integral entity. Furthermore, we aimed at reconstructing the mechanisms of social ordering on estate level through an examination of how particular practice, its elements and relations with other practices evolve, interfere or are in conflict.

The action plan sketched out at the application stage assumed conducting research at two housing estates in each of three cities, located in three historically different parts of Poland. The plan and methodology were built according to the principles of the case study research design, and because the ultimate goal was to discover the mechanisms at the highest level of universality, we sought the greatest possible diversity in choosing the cases (Seawright & Gerring, 2008). The selected cities of City1, City2 and City3 differ in many respects. The population numbers range from approximately 130,000 (City3) to almost a million (City1). Their functions in the country and regions are different: the city of (City3) is part of the Silesian agglomeration, mostly built after the Second World War, with the labour market still oriented towards industry; City1 is one of the most important academic, business and tourism centres in Poland; and City2, a metropolis on the regional scale, is an academic and tourism centre. In order to bolster the representativeness of the results for comparative research in each city, the chosen estates varied with respect to the date of establishment. Estates were selected after a reconnaissance and interviews with local experts. The first phase of empirical research was carried out as a pilot study in one of the estates in City1 built in the 1970 s and densified by development in the last 20 years. Estates in the other two cities were selected after the pilot phase.

Originally, we had planned to use mixed methods, hoping that we would obtain data from local institutions allowing us to analyse the demographic profiles of the housing estates, and that our team would gather qualitative material through interviews and observations. We soon had to change the strategy: the statistical offices did not have data separated for the areas we were interested in. Local experts (representatives of housing communities, cooperatives, parishes) agreed that a large percentage of the apartments was rented out (in some blocks one-third of the apartments), and data on this subject at the level of a building could only be estimated. However, it was not these difficulties (technical in the context of the following) that proved to be the biggest challenge.

Ontological and epistemological ambiguities of the practice-based approach

Different ways of understanding the concept of practices can be found in, among others, Warde (2005), T. R. Schatzki (2002), and Strengers (2010). D. Nicolini (2012, p. 214), synthesizing the state of the art in theories of practice, emphasizes that their diversity is characterized by ‘family resemblances’, which include the assumption that social reality consists of the connections between activities and meanings. These are materially mediated, and they cannot be understood without reference to a specific place and time.

As it was mentioned, a practice includes two general dimensions: observable activities, doings and sayings, and implicit ways of organizing them. Actions are understood here in a performative way – they are given meaning by social actors and are performed by means of bodily activities: what people do and say. What governs the activities and organizes them into a recognizable whole? Schatzki (2016; T. R. Schatzki, 2002) suggests that the organizational dimension of a practice consists of: 1) Practical understanding, an accepted operational definition of action, embodied know-how. Inexpressible through words, in the case of simple actions, it is marked by motor and perceptual abilities. In the case of complex actions, it means the abilities to perform, recognize, invoke and respond to a given action. Practical understanding, however, rarely defines the whole of what people do; 2) Rules of a given practice; these rules are explicit formulas, prescriptions, and instructions that command, direct or block-specific activities. To say that rules organize doings and sayings means that people who perform the same action take into account and refer to the same rules; 3) Teleo-affective structure – is a range of normative and hierarchically ordered goals, tasks and projects associated to various degrees with normative emotions and even moods. Normativity here refers to duty, and in the broader sense, to the acceptability of actions. 4) General understanding as a component that circumscribes the shared meaning of a given practice in the broadest sense and results in a sense of community. It unites different teleo-affective structures into one.

Critics of the theory of practice formulate reservations about the ontological and epistemological coherence of this approach. Galvin and Sunnika-Blank (2016, pp. 65–66) make a reference to Harré’s (2009) argumentation that ‘there are two kinds of explanatory models scientists use to make sense of their data: representational and heuristic models’. A representational model is used in natural (but also social) sciences and it ‘attempts to explain what is actually there’. This model assumes that the principles that rule an observable reality reflect or represent unobservable, abstract rules, introduced to the explanatory model (Galvin and Sunnika-Blank 2016: 66). The relationships (for instance: causality) between both dimensions (observable and unobservable) can be verified or falsified in many ways: mathematically, logically or experimentally. If verified, we can talk about the real (observable) phenomena using the language of the model, because these phenomena ‘represent’ the abstract, explanatory categories and principles that rule the natural world.

In heuristic model the relationship between the observable (doings and sayings) and the unobservable – but what, as we assume, is affecting the observable – is not defined. For this reason, we introduce heuristic devices: categories such as habitus, institution, social bonds or structure, constructed by researchers by analogy to the other well-known areas or as a metaphor of other realms to help to understand social phenomena. However, we do not know the nature of the interdependences between the observable phenomena and the abstract categories. The same considerations apply to the notion of practice which is one of the heuristic devices used within practice-based approach (Reckwitz, 2002; Galvin and Sunnika-Blank 2016), understood as a set of sensitizing concepts (D. Nicolini, 2009), or as a lens through which to look at the social life (Warde, 2014). Consequently, sayings, doings, performances and other observable elements cannot be equated with social practices – social practices need to be abstracted from their observable traces registered in the research material. The family of theories of practice is currently being intensively developed and applied to empirical research, with regards to, among others, telemedicine (D. Nicolini, 2012), production of knowledge and innovation (Knorr Cetina, 2009; Shove & Pantzar, 2005) sustainable practices and energy consumption (Shove & Spurling, 2013). In the field of urban studies, there are also studies which indicate the considerable potential of the approach, and the need for further development (Blokland, 2017; Van Kempen & Wissink, 2014; Maller et al., 2016; Hall, Smith 2015).

We have noticed two benefits in applying the concept of social practice to research on housing estates. First, the possibility of developing urban research through the theoretical-methodological linking of elements that in the classical theories of urban sociology were reduced to the mezo scale (e.g. neighbourhoods that can be filtered out from the macro and global phenomena) or whose agency was questioned, e.g. by recognizing the objectivity of the city in the face of macroeconomic processes. Second, the experiences and conclusions from research in urban estates contribute, in turn, to the development of the practice-based approach itself, by trying it out within a new, previously untested empirical domain.

Conceptualizing and operationalizing the notion of practice in research on neighbourhood practices

Practice as a unit of research and analysis. Problems at the intersection of theory and methodology

The starting point for most of the research carried out in the field of social practice theory is a predefined social practice, embedded in colloquial experience, easily identified as a subject of interest by both researchers and research participants. When examining the practices of shopping, cooking, maintaining personal hygiene, managing a household, etc., from the very beginning we are dealing with an identified object of cognition, whose intuitive understanding is shared by the researchers and the practitioners. The study of housing estates understood as complexes of social practices does not have a similarly comfortable and firm point of departure. The identification of what constitutes practices of living on the housing estate was a research task, not a point of departure. In proceeding with the study, we did not have a solid opinion on what were the practices of living on the estates which were to be the target subject of the research, i.e. practices that have an effect of differentiating and boundary-making within the estates. So we started with a working list of a variety of activities that we expected to find during the field research. It soon became apparent that the matter was much more complicated.

We have adopted an analytical orientation that D. Nicolini (2017, pp. 26–27) describes as ‘situational’ and which means ‘the analysis of the concerted accomplishment of orderly scenes of action’. Myriads of activities are performed in every housing estate, and therein lies the problem. There is family and social life: people do shopping, take care of everyday things at the post office or in administration, travel by car, foot, and bicycle, take children and dogs for walks, play sports, and go to church. These activities are made up of even more detailed physical actions: opening and closing doors, lifting barriers at vehicle exits, sitting on benches, getting into elevators, walking, running, talking, etc. To discover the mechanisms of organizing communities, we had to determine, first of all, which practices are performed by the observable actions of the residents, and then, which of these practices require a greater degree of orientation towards other practitioners in the course of their performance. The requirement of ‘performative orchestration’ is considered an important criterion for identifying those practices which are ‘community-creative’ or, in other words, structuring the territorial community, organizing the activities of a significant percentage of the population and potentially creating the aforementioned ‘groups of common cause’.

The second problem was the ontological contradiction between the notion of practice and the concept of ‘housing estate’ as a socio-spatial unit. Social practices are ‘ordered across space and time’ (Giddens, 1984, p. 2), and they ‘migrate’ and extend in space (Hui, 2013). This theoretical assumption poses a challenge for studying every practice – consumption, travel, knowledge production and innovation, etc. – but in the case of research on territorial communities, it poses additional specific problems. Many social practices are located in or focus on the urban housing estate, but they are not limited to the estate: people shop in local shops, but also in large shopping malls and via the Internet; many children attend kindergartens and schools outside of the housing estate, and may take additional courses in other districts, etc. The praxeological analysis of the estate therefore required of us to make difficult and arbitrary decisions about the operational delimitation of specific practices in terms of their ability to organize life within the housing estate understood as an urban unit. From interviews with residents of the estates came reports on different ways of engaging in social activities outside the housing estate, but two things were of critical importance to us. First, the specificity of the practice performance in the direct materiality of the estate, in the relationships among people, animals and things on the estate. For example, an excess of cars in relation to the available space is common in large Polish cities, but these urban practices of mobility and car mooring are not treated as context (and another variable), but as a wider picture of car parking in an estate. The consequence of citywide car driving practice is the socio-spatial reorganization of the estate and the formation of, as we call it, the ’parkinghood’ (Smagacz-Poziemska et al., 2018). Secondly, the diversity of everyday life on the estate creates a high probability that interference, feedback, and conflict will occur among the practitioners, or rather their performances in time and space. Practices can be mutually reinforcing, such as supervision and parking, or monitoring and maintaining cleanliness, as we observed. Conflicts may arise, for example, when practices are enclosed and strengthened materially, and their expansion takes place at the expense of the space of other practices. Parking practices where pavements, passages or lawns are annexed create conflict with practices of caring for greenery or walking (by elderly people, parents with small children, or teenagers). It was, therefore, necessary to look for ‘intersections’ of practices located on the estate, which D. Nicolini (2017, pp. 26–27) defines as conflict-sensitive orientation: ‘the inquiry into the co-evolution, conflict and interference of two or more practices’.

Eliciting practices through interviews

Before designing the research tools, at the stage of initial operationalization, based on our own (research team members’) experiences of living in various housing estates, we prepared a list of a dozen or so practices. We asked ourselves: what do people do on the estate, in the place of residence? What do they engage in? When do they come into contact? The working list of ideas included, among others, managing the common space, fencing the settlements (which is common in Polish cities); quarrels and conflicts; parking and getting around the estate by car; religious activities; civic activities; walking dogs; recreation; and tending to greenery. This initial list was meant to sensitize us to discovering practices while conducting interviews and observations planned as the two main research techniques.

The interview scenario was not structured according to this list: the questions were open-ended, and the researchers were only instructed on how to expand on them and probe further when needed. The interview began with two groups of questions: about moving into a given estate and general opinions about this place (‘How do you find living here?’). The third part of the questionnaire concerned other residents (’Who lives here on the estate?’). The fourth part was about basic, everyday life on the estate (‘Please describe your usual day’, ’Who do you have contact with daily? In what situations?’ etc.). The fifth included questions about unique, distinctive situations, including those leading to conflict and annoyance. The last part was defined as the ‘institutional characteristics of the housing estate’, and included questions about formal institutions on the estate and their activities, as well as local rules (as they are followed and broken).

After transcribing several dozen interviews and their preliminary analysis in terms of the accuracy of the questions, we came to several conclusions. First, that during the interview, the structure of the tool was more often a hindrance than a help: it stiffened the language of questions, and limited our ability to show a flexible, natural interest in what the respondent was saying. Second, that conducting the interview requires a good knowledge about the material already collected by other researchers in the same housing unit, because from the point of view of the project’s goals the most important were those activities, events, situations and emotions which were repeated in subsequent conversations and told from different perspectives. Third, that conducting an interview in terms of social practices requires a special skill: authentic wonderment that gives the obvious and the ordinary the status of unobvious, interesting and important matters. During the preliminary analysis of the interviews, we saw that we have not always managed to pick up on seemingly insignificant remarks given by the respondent, that we have not always avoided the error of asking about opinions and not about actions (this happened especially when the respondent was taciturn, or responded casually), and that there were cases of implying our own ‘expert’ categories and interpretations.

From the perspective of 150 conducted, fully transcribed and analysed interviews, we can see that the key challenge in conducting the conversations was to stay open to the individual, direct experiences of a particular respondent, while at the same time recognising them as shared with other residents, as experiences known to us from other interviews. Acquiring information through interviews about the sense of intelligibility of the discussed experiences was very difficult and yielded the best results when the minimum two conditions were met: the researcher played the role of a complete layman who wants to learn how to live on the housing estate, and the respondent praxeologized the story of his or her life on the estate. In a interviewing practice, the most efficient tactic was admitting to not understand the interviewee’s statements, particularly when they referred to understandings, intentions, feelings or rules.

Inhabitant1:

Nothing annoys me here [in this estate]. Only … […] Well, they could organise the parking space here … That’s the only … it bugs me … They were making a road here, they could expand it a little bit. No. They do nothing for the young so that they would have a piece of the car parking space. All they made, they made out … for other housing blocks, over there. Nearby the health centre they made, but here, inside our estate … barely, starting with the arcade, in this direction …

Researcher:

But why did you tell me that they do nothing for the young? I don’t understand it …

I1:

They don’t do a favour for the young, that he doesn’t have to stand in the mud up to the hubs. That he could park the car on the tiles.

R:

So here is the rule, that first inhabitants have got some specific parking slots?

Inhabitant2:

No, it doesn’t work that! It doesn’t work that! It’s who gets there first. It’s not that they have got any …

I1:

But there are … there is a space that could be a parking space, you know? And they don’t do that … If you look right here …

R:

So I misunderstood … Because you said that you hadn’t any problem, as you have the fixed parking place … [all persons approach the balcony to look at the courtyard – the researcher takes the pictures]

I2:

well, I always have a fixed place, but it is not that it’s mine, but I park there, and what is going when somebody is parking there … ? Then, when I come back home, and my place is occupied, then I am parking anywhere, and when that somebody is driving away I am going back and moving my car. How this would be possible otherwise?

[married couple, 60y, long-time residents; TS_3]

The position of the ‘estate-alien’ or ‘trainee-inhabitant’ adopted by the researcher helped the interviewee (practitioner) to use natural language, describe and explain everyday activities and situations. For example, instead of just saying ’we manage to clean the stairwell ourselves’, one explained how the ‘cleaning’ was done, saying that the neighbours clean up every Saturday, that they are mainly women, and that they sweep and clean the floor, but there is a neighbour who always polishes as well. And that until recently there had been an unspoken competition over who could clean better, but now the neighbours are older and are considering hiring a cleaning person. Parts of the interviews were conducted outdoors, during walks conducted along respondents’ daily routes or during their daily activities, such as caring for children at the playground or shopping at the local store. However, we were unable to systematically stay on the estate and participate in everyday life. The scale of research conducted in the field of social practice theory with full use of observation techniques in three cities, in six different locations, would have required a much larger and fully available research team and several times more resources at our disposal; thus, observations were only complementing interviews.

Such an approach may raise certain methodological concerns among those practice scholars who express reservations about researching social practices through interviewing practitioners. For some consider interviews as providing ‘primarily ex-post rationalizations of an individual actor’s behaviour which are worthless as data sources’ (Bueger, 2014, p. 400). Those who question the value of using interviews in practice-based research often refer to Pierre Bourdieu’s version of practice theory, and his claim that “once practices are established, they are unlikely to be subject to much examination thereafter” (Bourdieu, 1990, p. 73). Thus, they argue, information about the object which is not subject to people’s reflection cannot be obtained by asking them about it. In Russel Hitchings’ opinion, however, this is a faulty interpretation: „Indeed, though Bourdieu views practices as ‘opaque’ (Bourdieu, 1990, p. 12) to their practitioners, he later advocates a discursive form of ‘disposition analysis’ (Hitchings, 2012: 62)”. ‘There is scope for reflexivity in Bourdieu’s model’, Hitchings (2012, p. 62) continues and suggests that practitioners’ self-reflection enables us to reconstruct the implicit aspects of practices. Davide Nicolini not only grants practitioners the ability to reflect upon their actions, but also admits that they ‘have and use quite precise words to describe the details of practices’ (D. Nicolini, 2017, p. 29). He thus deems studying practices through verbal accounts possible (although not desirable; the preferred method is a participant observation).

Central to our considerations is the question about effective ways of eliciting practice-centred verbal accounts. Drawing on our own experiences with interviewing practitioners about their practices as well as other successful interview-based practice-oriented studies (Bueger, 2014; Hitchings, 2012; D. Nicolini, 2012) we would like to suggest three strategies for facilitating practitioners’ self-reflection (as a methodological ‘way in’ to the implicit aspects of practices) in interview situations:

1. asking interviewees to provide detailed instructions on how to do things, e.g. how to park a car on the estate, and in this way to reveal their practical knowledge in a form of ‘detailed first-hand descriptions of bodily movements, activities, utterances, or handlings of things’ (D. Nicolini, 2012, p. 401)

2. questioning the ‘unquestioned’ ways of doing things by presenting and discussing alternatives with interviewees (Hitchings 2012), e.g. how car parking is being arranged on other estates, in order to explicate motivations, beliefs and desires that ‘tacitly’ accompany habits and routines

3. enquiring interviewees about ‘inappropriate’ performances, e.g. instances of inappropriately parked cars, in order to reveal implicit standards, norms and rules.

Strategy of the neighbourhood practices identification

It was a challenge to prepare the material for analysis and interpretation in accordance with the premises of the practice-based approach. This task took about half a year, much longer than planned, for two reasons. First, we had to agree on the conceptual model of social practice. We considered using the model proposed by Shove, Pantzar and Watson (2012), combining the elements of materials, competences and meanings. The simplicity of this model was appealing, as was the fact that it had already been employed in research on innovations (Shove & Pantzar, 2005) and on the materiality of neighbourhoods in health practices (Maller et al., 2016). We decided, however, that we would use Schatzki’s concept, because it is more sensitive to the analysis of the relationships between individual and mutually oriented performances and the general patterns organizing the ‘communities of practice’; it is also more comprehensive, as it takes into account both emotions and motivations (teleo-affective structures) as well as rules of operation as elements organizing individual performances into practice as a whole.

Following T. R. Schatzki (2002), we have identified two general levels of practices as units of analysis: A) observable actions along with their corporeal-material correlates and B) organization of actions (implicit aspect of the practice as a whole: norms, beliefs, rules, intentions and practical knowledge) and material dimensions, including infrastructure, resources, tools, as well as location and time of actions. We had to decide whether we could at this stage analyse interviews in terms of specific practices (i.e. when we would arbitrarily decide that a particular statement is an indicator of one practice or another), or whether we should apply a bottom-up strategy, first identifying statements about activities, as they were named by the respondents, and only then combine the activities into broader sets of activities, and on this basis create a list of everyday practices of living on a housing estate.

Second reason of our schedule extension was that the idea of analysing interviews at the level of singular actions turned out to be highly controversial in our research team. Some team members were willing to try it, justifying the effort as it would allow us to avoid pre-interpretation (because, e.g. each researcher conducted only part of the interviews, and each of us showed interest in slightly different aspects of neighbourhood life) and verify the initial list of everyday practices of living on a housing estate, created before the interviews began. Some were very critical of this idea arguing that, first, going down to a level so detailed and behavioural during the analysis carries the risk of losing the wider sense of practice as an assemblage of activities, beliefs, knowledge and rules; second, that this major effort to identify all activities would be disproportionate to the expected results: the de facto salient point is not how the respondent refers to the activity, but what larger whole the activity is a part of.

The greatest difficulty was in choosing an effective strategy of identification (reconstruction of social practices based on the transcribed interviews were only activities we mentioned) of all the wholes that can be defined as ‘implicit knowledge’, indicating the existence of social practices. This is not a problem when the starting point is a predetermined practice, such as in research on, e.g. cooking or Nordic Walking. The quintessence of the whole, the ‘implicit knowledge’ is then known, and it gives meaning to other elements, including singular activities. In case where we do not know the WHAT being investigated, we lack this point of reference. If ‘a doing or saying belongs to a given practice if it expresses components of that practice’s organization’ (T. R. Schatzki, 1996, p. 103–104) and the organization of the practice is unknown, then we may get caught in a vicious circle of interpretation. The dilemma, therefore, was how to identify what to investigate at all. Was it to inspect what people are doing? But then, how to answer the question: what is the supra-individual sense of the activity (i.e. to what kind of practice does the activity belong), if at the same time we do not have an interpretive framework in the form of implicit knowledge, a practice as a whole?

As a compromise, we decided on an exploratory analysis of 20 interviews carried out in City1, on two different housing estates, with residents diversified in terms of socio-demographic characteristics. Five members of the team proceeded to code selected transcripts (in QDA Miner, a qualitative data analysis software) according to the following scheme, developed on the basis of Schatzki’s model of a social practice:

1. Category: Actions (as in Schatzki’s model of social practice)

1. Code: Activities: Doings and Sayings

1.1.1. name of activity [inductive coding]

1.2.Code: Location (no match in Schatzki’s model)

1.2.1. name of the location of the activity [inductive coding]

1.3.Code: Temporal Dimension (no match in Schatzki’s model)

1.3.1. name of the activity timeline (inductive coding)

1.4.Code: Practitioner (that enacts practice)

1.4.1. the name of the category of inhabitants’ and non-inhabitants [inductive coding]

2.Category: Organisation of Activities

2.1.Code: Rules (as in Schatzki’s model)

2.2.Code: Norms (in Schatzki’s model: the element of teleoaffective structure)

2.3.Code: Intentions (in Schatzki’s model: the element of teleoaffective structure)

2.4.Code: General Understandings (as in Schatzki’s model)

2.5.Code: Knowledge (no match in Schatzki’s model)

2.6.Code: Categorisations [how do interviewers categorise people in the estate] (no match in Schatzki’s model)

3.Category: Other

3.1.Code: Boundaries of The Housing Estate (no match in Schatzki’s model)

The coding scheme reflected Schatzki’s conceptualisation and his distinction of two aspects of practices: activities and components organizing activities (in the scheme: rules, norms, intentions, beliefs, and knowledge). We complemented the original model with a location and time horizon of activities, information on agents and membership categorizations of residents of the housing estate. Computer-aided analysis of the content of the transcribed interviews using the prepared scheme consisted in the analytical decomposition of the narratives into elementary particles (individual components of a practice) and their reassembly according to the intelligibility of a practice as a certain whole. The translation of Schatzki’s formal, nuanced model into a coding scheme that would play the role of an effective tool for identifying content relevant to researchers has proved problematic. Although the coding scheme was not very complex, and each of the coding categories had an operational definition and examples of correctly assigned fragments, using it caused difficulties for the researchers. Particularly troublesome was the meticulous process of coding, caused by the nature of statements packed with single names of a myriad of activities performed by the residents of the housing estates, e.g.:

There are the hills right over here, between our housing estate and XY estate, and there is a lot going on there. There are the hills for the sleds in the park right there, you can run there, play ball, there are people that jog or do Nordic walking. Well, actually, that’s it. KRK_KR_M_07

Coding:

ACTIVITIES/Sled ridingACTVITIES/RunningACTIVITIES/Playing ballACTIVITIES/JoggingACTIVITIES/Nordic walking

Following Schatzki’s suggestion that ‘a significant clue to which practices constitute people’s lives is the vocabulary they use to classify their activities’ (T. R. Schatzki, 1996, p. 104), we have tried to hew closely to the language of the respondents. It was also difficult to identify fragments of interviews that unambiguously relate to distinct organizational elements of practices (rules, norms, intentions, beliefs, and knowledge): when articulated in interviews, the differences between them turned out to be extremely subtle, often manifested in the form of the statements rather than in their content, forcing us to spend time interpreting short fragments of text, often individual sentences, e.g.:

I always close the windows [in the corridor] when I’m back, but within 15 minutes the windows are opened by somebody else. Well, I’d have to go down and close them every time. And then you get mad, and you think that when you catch the guy red-handed, then you will break his legs. I’m done! How many times? The board of housing cooperative decide in these things, but it’s not working, because people are somehow undereducated. KRK_KR_M_07

The coding resulted in a list of over 450 terms referring to the activities on the housing estates. From this emerged a picture of the estates as highly complicated systems of everyday activities performed in specific, sometimes implicit and sometimes directly negotiated ways: using common property (e.g. corridors, waste receptacles, playgrounds, parking spaces), keeping things clean and tidy, walking dogs and walks with other dog owners, going for walks and using the playground with children, intervening when rules are violated (e.g. noise, improper parking, damage to property), including taking actions resulting in the installation of fences or monitoring (cameras). Also talked about, albeit less often, were activities regarding the cultivation of greenery around buildings, participation in community meetings, and organizing and participating in celebrations (such as receiving pilgrims during the World Catholic Youth Days, groups watching football championships).

Further analysis was limited to fragments of statements related to activities. At this stage of work with the research material, information about the other components of the practices (rules, norms, intentions, beliefs, and knowledge) extracted from the interviews has not been synthesized. Concentrating on activities, although cognitively valuable in many respects, turned out to be insufficient to reconstruct the practices as wholes, but it enabled the extraction of internally coherent sets of activities focused on specific ’themes’ of neighbourhood life, which we felt could lead to the identification of social practices.

The identified activities were categorized and a list of a dozen or so ‘themes’ was created, those which most often appeared in interviews. To sum up, the first stage of the analysis consisted in a systematic exploration of a sample of research material aimed at identifying information on activities, which together with the components organizing the activities form a set of everyday practices of living on a housing estate. The aim of this stage was to create a catalogue of social practices performed in the studied estates, which would allow verification of our initial intuitions.

Due to the difficulties we had encountered at the stage of exploratory analysis, before proceeding to the analysis of the entire research material, the coding scheme was simplified:

1. Category: Actions

1.1.Code: Activities: Doings and Sayings

2. Category: Organisation of Activities

2.1. Code: Explicit Knowledge (rules, norms, guidelines, standards)

2.2. Code: Teleoaffective Structure (intentions, desires, emotions, feelings)

2.3. Code: Practical Knowledge (competences, skills, understandings, know-how)

2.4. Code: Materiality (infrastructure, resources, materials, tools)

2.5. Code: Location

2.6. Code: Temporal Dimension (time of day, season, rhythm)

At this point we changed the analytic strategy: five sets of activities were selected and named as the practices of: care, delivery, supervision, parking, maintaining neighbourhood clean. One hundred and fifty interviews were then analysed regarding these sets (practices). The criteria for their selection were: 1) sheer repetition of the mentions of the activity in the interviews with various respondents. 2) importance of the ‘theme’ of neighbourhood life to the respondents, indicated by, among others, their emotional involvement when speaking about it. 3) commonality, which concerned whether a specific ’theme’ appeared in the interviews conducted in all the examined estates. 4) internal complexity: we were looking mainly for those sets of activities that would allow us to reveal all the components of the organisation of the practices, i.e. rules, intentions, competences and the material dimension. The general coding scheme was ‘personalised’ (questions oriented on the particular practice were formulated).

By referring to particular sets in a particular way (by naming them ’a practice of care’, ‘of supervision’, ‘of parking, etc.), we have put forward a thesis about the sense and relationship between the activities that we had grouped together. The names given to practices did not appear in the interviews, they were the results of our interpretation, second-order abstractions (more about the results of the research: A Bukowski, Nóżka, Smagacz-Poziemska et al., 2018, 2018; Smagacz-Poziemska et al., 2018).

Concluding, the data analysis in the project is carried out in five stages: 1) Initial coding scheme development (operationalization of the Schatzki’s notion of social practice); 2) Exploratory analysis of 20 interviews focused on the identification of the activities; 3) Categorisation of the activities leading to the identification of 5 coherent sets of activities – practices of care, delivery, supervision, parking, maintaining neighbourhood clean; 4) Final coding scheme development (operationalization of the 5 identified social practices); 5) (An on-going task) Analysis of 150 interviews with regard to the 5 identified social practices.

As the aim of the project was the reconstruction of the mechanisms of establishment and reproduction of urban territorial communities at the level of the estate, we are now in the stage of analysis of selected, defined and operationalised practices. The next, fifth step will be focused on the analysis and interpretation of the co-evolution and interference of two or more practices (D. Nicolini, 2017, pp. 26–27).

Conclusions

Our experience sheds light on the vast potential of theories of practice in urban research, but also on the need to further develop research procedures consistent with ontological assumptions of practice-based approach. Even if the heuristic categories do not have the complete real representations in doings and sayings, it is possible to explore the traces of the implicit structures in the world of performances. Observable and unobservable elements of practices make a complex arrangement (nexuses, regimes), that require a specific set of techniques of data collection and analysis. What is essential for conducting an effective and successful practice-oriented study is the conceptualization and operationalization of the sensitising concepts used throughout the research process: conceptualizing and operationalizing the notion of social practice, as well as designing and using adequate research tools to generate data, and developing standards and strategies for practice-sensitive analysis of the research material.

Our work described in this article, which consists in breaking down interviews into ‘elementary particles’ in the form of basic actions that are performed in the estates, and then in an attempt to thematically link them to larger entities (bundles, threads, themes), has been the necessary, exploratory stage leading us to the identification of social practices responsible for structuring the territorial communities and further analysis of the relationships between two and more practices in accordance with the conflict-sensitive orientation (D. Nicolini, 2017). Our work contributes to the development of a practice-based approach to empirical research by demonstrating how the notion of practice can be used as an analytical category, applied to the research material at the stage of analysis, and not only as a sensitizing concept at the stage of fieldwork. In our opinion the potential of practice-based approach – in relation to urban research – lies in new perspectives of seeing the neighbourhood realms and recognising the mechanisms not fully captured through the lens of classical concepts of, for instance, neighbourhood bonds or a place attachment.

Acknowledgments

The authors would like to thank Marta Klekotko, Ewa Kopczyńska and Seweryn Rudnicki for their comments on the draft of this article.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The article is an output of the research project ’Differences and Boundaries in the Process of Creating Neighbourhood Communities in Large Cities. A Socio-spatial Study’, funded by the Polish National Science Centre, under grant no. UMO-2014/15/B/HS6/01949. Research team: Marta Smagacz-Poziemska (PI), Andrzej Bukowski, Marcjanna Nóżka, Karol Kurnicki, Natalia Martini and Krzysztof Bierwiaczonek.

Notes on contributors

Marta Smagacz-Poziemska

Marta Smagacz-Poziemska - a professor at the Institute of Sociology of the Jagiellonian University. Her main research interests involve urban neighbourhoods and local communities, public space, urban sustainability. Author and co-editor of publications on social exclusion, urban revitalisation and urban everyday life. Member and principal investigator of several national and international research projects; coordinator of the 37. Research Network (Urban Sociology) of the European Sociological Association, member of the Scientific Advisory Board of the JPI Urban Europe.

Andrzej Bukowski

Andrzej Bukowski – a professor at the Institute of Sociology of the Jagiellonian University. His main research interests include regionalism, local and regional development, democracy and civil society as well as social theory. A participant and co-ordinator of over a dozen scientific research projects funded by Polish and international like: State Committee for Scientific Research, National Science Centre, European Union and World Bank among others. Currently researching the institutional factors of innovation and innovativeness as well as the collaboration of the social sciences and the economy in broad terms.

Natalia Martini

Natalia Martini - Doctoral Student at the Institute of Sociology, Jagiellonian University, Kraków, Poland. Principal Investigator of the project ‘City experienced in the homeless situation. A socio-spatial study’ funded by the Polish National Science Centre under grant no. 2016/23/N/HS6/00810. Sociologist with a background in cultural studies, who has a strong interest in urban everyday life, as well as creative, transdisciplinary methodological approaches to urban phenomena. Her work cuts across sociology and human geography and favors critical and activist approaches to scholarship.

Notes

1. See the Acknowledgments.