Making the city through participatory video: implications for urban geography

ABSTRACT Participatory video-making represents a significant opportunity within urban geography. Reflecting on two participatory video-making processes in Seville, Spain, this article examines, firstly, the epistemological opportunities contained within a participatory video-making process. Specifically, it reflects on the potentials and limitations of participatory video as a mode of urban geographical inquiry. Secondly, this article draws on critical urban theory to examine the dialectical relations between the practice of participatory video-making and the social production of urban space. In doing so, the article sets out a potential approach toward constructing a critical ontology of urban geographical research, which elucidates the relations between research, methodology, subject, and place.


Introduction
The nature of geographical knowledge is a primary concern for geographers (Agnew & Livingstone, 2011;Castree, 2006;Dewsbury & Naylor, 2002;Döring & Schnellenbach, 2006;Jazeel, 2014). For some scholars this issue is epistemological -how do geographers know things? For others, particularly in the field of urban geography, it is a political concern (Jazeel & McFarlane, 2010;McFarlane, 2011). Critical to both of these perspectives is the issue of methodology.
In the past three decades human geography has embraced participatory, discursive, and other previously marginalized approaches to qualitative ethnographic inquiry (Caretta & Riaño, 2016;DeLyser & Sui, 2014;Lees, 2003). However, the significance of these approaches for geography as a discipline is seldom the focus of research. It should be. Methodology is more than a question of evidence and knowledge; it is an issue that goes to the heart of the relationship between research, practice, and reality. It is, in other words, an ontological concern.
In the past two decades, participatory video-making has emerged as a distinct methodology within geographical research (Mistry & Berardi, 2012;Pain, 2004). In the simplest terms, participatory video-making is a process through which a group of people make a video about an issue that is important to them. But participatory video can take on a multitude of forms and be used for a diversity of ends. It is useful, then, not to think of it as a specific methodology or set of methods, but as a process of collective, visual, narrative inquiry.
Recent scholarship has recognized the potentials of video-making to contest hegemonic systems of knowledge production; masculine, Western, Northern, academic, etc. Makamba et al. (2019), for example, set out the potentials of participatory videomaking to contribute to the epistemic sovereignty of rural communities in Zimbabwe. Drawing on Fricker's (2013Fricker's ( , 2015 important contributions to the notions of epistemic justice and epistemic capabilities, Boni and Velasco (2020) have also drawn attention to the political significance of the capacities developed through a participatory videomaking process for challenging systems of epistemic marginalization and oppression.
This scholarship speaks to a wider effort to understand how knowledge produced though participatory video relates to the broader political economy of knowledges. However, much of this critical scholarship engages with the sum of the knowledge produced through participatory video-making rather than disaggregating and critically unpacking the diversity of ways of knowing and learning that can emerge throughout the process. Each of these different ways of learning, as this article will explore, represents distinct opportunities and challenges within urban geographical research. Moreover, there has been no attempt to unpack and examine the multiple relations between participatory video-making as methodology and the socio-spatial context in which it takes place, which represents an under-theorized and under-examined set of changemaking potentials.
The aim of this article, then, is to unpack the specific epistemological opportunities contained within participatory video-making processes as well as the ontological relations between participatory video-making and urban space as a site of inquiry. In doing so I reflect on the significance of participatory video-making as an approach to geographical research and the questions that participatory video-making raises regarding geographical research methods. I reflect on how we might begin to construct a critical ontology of geographical research methods, which engages with the relationships between urban geographical knowledge, methodology, and "the urban" as a sociospatial construct. To these ends, the article addresses three interrelated questions: to what extent does participatory video making represent a significant epistemological contribution to urban geographical inquiry; how might we conceptualize the relationships between participatory video-making and the social production of space; and finally, what issues does participatory video-making raise in regard to the relationship between urban geographical knowledge and methodology?
This article draws on critical urban theory, specifically the works of Henri Lefebvre, to reflect on two experiences of using participatory video making in Seville, Spain. Lefebvre's conception of space, set out in The Production of Space (1991Space ( [1974), has informed a wide range of urban geographical and sociological research (Castells, 1983;Eizenberg, 2012;Gottdiener, 1985Gottdiener, , 1993Moles, 2008;Sheppard, 2002). His celebrated triad of space represents an ontological intervention that urges us to reconsider the nature of urban space as an object (Pierce & Martin, 2015); as the dynamic and contested product of both material and social processes. In this way, Lefebvre's conception of space is active, dynamic, relational and subjective, emphasizing the dialectical relationships between the way space is rationalized and planned, how it is perceived and used, and how it is experienced through emotion, memory, and imagination. For these reasons, Lefebvre's conception of space provides a useful framework for examining participatory video-making as a socio-spatial practice, which can be approached in terms of its dynamic ontological relations with the site of inquiry.
In the following section I position this inquiry in terms of critical literature on participatory video-making as well as Lefebvre's critical approach to urban space that is useful for understanding the relations between research methodology and subject. In the third section I introduce the geographical and research context in which two participatory video processes took place as well as the combination of methods on which this article draws to analyze these processes. In the fourth section I examine the distinct stages of participatory video-making in terms of their epistemological potentials, as well as their ontological relations with the city. In the final section I reflect on the significance of participatory video making as geographical research method and reflect on implications of a critical ontology of participatory video-making for urban geographical inquiry more broadly.

Participatory video-making and the nature of urban space
The notion of "participation" within filmmaking emerged in the 1960s through the struggles of feminist filmmakers, radical film collectives, and film activists in the global South (Mistry & Berardi, 2012). However, it was only consolidated as a distinct approach to filmmaking in the 1990s with the increasing availability and affordability video cameras and recording technology, primarily in the context of international development; participatory video-making has been used extensively, for example, for monitoring and evaluation of multi-lateral development programmes in combination with other participatory methods such as Most Significant Change (Lunch, 2007). In this sense, participatory video can be thought of as a method that "fundamentally blurs the distinction between research and community development" (Low et al., 2012, pp. 53-54) Scholarly literature on participatory video falls broadly into two areas. The first celebrates participatory video-making for its empowering and transformative potentials (Kindon, 2003;Sundar Harris, 2008;White, 2003). The second is the more recent body of critical literature on participatory video-making in which scholars have increasingly reflected on dynamics of power and participation in participatory video-making (Kindon, 2016;Milne, 2016;Milne et al., 2012;Shaw, 2012).
Overall, what has emerged is a more nuanced understanding of the potentials of participatory video-making to create multiple spaces for critical engagement and activelearning that challenge academic dominance in knowledge production. As Kindon (2003, p. 144) argues, "The knowledges produced [through participatory video-making] are both for and by the participants, which challenges dominant representations and goes some way to breaking down usually hierarchical researcher/researched relationships." Yet scholars have emphasized that participatory video-making cannot exercise issues of power, identity, and the "dangers of participation" from knowledge production (Walsh, 2012), nor Western realist tendencies, which permeate participatory videomaking (Kindon, 2016). Walsh (2016), for example, argues that participatory videomakers must recognize the "liberal, technocratic presumptions" underpinning the practice, in order to realize more collective and socially just outcomes. Elsewhere, Rogers (2016) warns that participatory video can lead to "individualistic and deficit discourses" that marginalize participants. Participatory video-making is also susceptible to critiques that have been made of other participatory research methods (Cooke & Kothari, 2001) and other participatory visual methodologies such as photovoice (Shankar, 2016).
It is beyond the scope of this article to respond to these specific critiques. However, it is important to note that the critical discourse surrounding participatory video has focused either on the politics and dynamics of the process -specifically regarding the nature of participation (Milne, 2016) -or on the relationship between audio-visual and non-audio-visual research outputs (Luttrell & Chalfen, 2010), rather than on the relationships between the participatory video-making and the site in which the process takes place. Moreover, the body of critical literature has neither examined the specific constellation of epistemological potentials contained within a participatory video process, nor their significance for urban geographical inquiry.
There are a number of reasons why human geographers might choose to use a participatory video methodology; some practical, some political. Practically, participatory video-making combines a number of processes that closely resemble qualitative research methods that are frequently employed by urban geographers such as semistructured qualitative interviews, focus group discussions, participatory workshops, participant observation, and ethnography. As such, a participatory video process can be thought of as a framework through which to structure and augment qualitative fieldwork.
In order to understand the political potentials of participatory video, it is useful to position the process within a broader framework of Participatory Action Research (PAR); an approach that emphasizes the need to produce knowledge specifically oriented toward social transformation (Kindon et al., 2007) and which consciously contests traditional hierarchies of knowledge production (Fine et al., 2007). Through a PAR approach, geographical research can be candidly normative, biased, and political; rendered meaningful through tangible, real-world change. Beyond the capacity to produce knowledge, participatory video-making can also be an opportunity to build capacities and amplify the voices of marginalized groups (Lunch & Lunch, 2006).
Moreover, it is useful to turn to feminist critical theory in order to understand the potentials of participatory video-making to challenge forms of marginalization and oppression. Laura Mulvey's seminal essay, "Visual Pleasure and Narrative Cinema" (1975), for example, draws on psychoanalytic theory to show how the male gaze dominates filmmaking and cinema. In this way, the act of filmmaking can either contest or entrench gendered systems of domination. It is significant then, that participatory videomaking has been characterized as "a feminist practice of looking" for the way that it can destabilize power relations and elevate diverse perspectives (Kindon, 2003). Feminist theorists, such as Mulvey, point to the issue that is at the center of this article: the relations between object and observer; between representation and reality; and between geographer and subject. In doing so they urge us to recognize the way that research, like film, both reveals and produces reality.
Lefebvre's theory of space offers one framework through which we might conceptualize the relationships between participatory video-makings epistemic potentials and its ontological relationship with the urban environment. Within Lefebvre's spatial ontology, space is continuously remade through the coexistence of, and dialectical relations between, three "moments" of space, which he termed spatial practice, representation of space, and spaces of representation (Lefebvre, 1991), and which are commonly referred to as perceived space, conceived space, and lived space.
Perceived space refers to the material space that constitutes both built and natural environments; it is the material arena in which day-to-day actions occur. Conceived space refers to the psycho-social construction of space -the way space is rationalized -the knowledge this entails and the ways that space is represented. Lived space combines elements of both perceived and conceived space; it is each individual's experience with space in everyday life; the ways that people connect to space through, amongst other things, memory, expectation, and imagination. Lived space is a constituent of social relations and thus, social life (Purcell, 2002).
In this way, Lefebvre proposes that urban space is not the disinterested setting in which social life takes place, rather, space continuously shapes and is shaped by social life. As Soja (1996, p. 46), drawing on Lefebvre, writes, "Social reality is not just coincidentally spatial, existing 'in' space, it is presuppositionally and ontologically spatial. There is no unspatialized social reality. There are no aspatial social processes." Lefebvre's spatial ontology draws on the philosophy of Nietzsche, Hegel, and Heidegger, amongst others, and underpins Lefebvre's predominantly historical materialist interpretation of urbanization. It is also the foundation of his political philosophy, often referred to in terms of the right to the city (Lefebvre, 1996); that urban space is continuously (re)produced by all urban inhabitants therefore all urban inhabitants have a right to participate in its governance and management.
Lefebvre's spatial ontology offers a number of ways of understanding the relationship between participatory video-making and urban space that we can unpack in terms of the multiple, dialectical relations between moments of space. As the following section will demonstrate, each element of participatory video-making -the training, planning, shooting, editing, and screening -not only constitutes a distinct epistemological opportunity for geographical research, but also a distinct ontological intervention regarding the nature of urban space.

Context and methodology
The following account draws on my experience of participatory video-making in Seville, Spain between 2016-2017. I was engaged in two participatory video-making processes with urban gardeners in the city as part of a wider doctoral research project that examined the politics of self-organization and the social production of space in urban community gardens.
I first made contact with gardeners at five sites across the city during a seed exchange visit organized by Garden Organic, UK, in June 2015. On a follow up visit to the city, I reinitiated contact with gardeners from three of the sites and arranged visits. During these and subsequent visits it was made clear that gardeners from one site were not prepared to be part of a research process, gardeners from another were willing to be spoken to as part of a research process, and the third group of gardeners were keen to be more active participants in a participatory research process.
Following weeks of discussions with this third group, based at Huerto del Rey Moro, introduced below, we co-developed a participatory video-making programme that would contribute both to my research and to their shared aims of the gardeners: to learn to produce their own media and to produce their own films about the garden. Participants were not paid for their time in training and during filmmaking. However, all travel expenses were covered for workshops and shooting across different locations, and meals were provided. The first process took place over six weeks in 2016. Four urban gardeners -three female, one male -made a short film exploring the themes of communication and transformation within and between two urban gardens in the city, Huerto del Rey Moro and Miraflores Sur.
Huerto del Rey Moro is an occupied, community-managed garden in the Macarena district of Seville's Old Town (Casco Antiguo). The garden occupies approximately 2000 m2 between Calle Sol and Calle Enladrillada on land that was formerly the private orchard of a large house. The space was hidden from public view until it was "rediscovered" in 2002 by local residents and gradually transformed into an urban garden. Today the site combines garden areas with large communal spaces. The garden area is used to grow a combination of vegetables, aromatics, and medicinal plants according to a mixture of organic and permaculture principles. The communal areas are used by local adults and children and visitors to the garden on a daily basis. The site also hosts workshops, festivals, and public events throughout the year. In the first participatory video process, the participants were all regular users of Huerto del Rey Moro, with limited knowledge of Miraflores Sur.
Miraflores Sur is a community-managed and local government-owned growing space within Parque de Miraflores in the north of Seville. It is the largest urban garden in the city. The former flood plain that now constitutes the park was designated a "Green Zone" in the 1960s, meaning that it could not be developed. However, the site became a dumpsite for construction debris and was almost entirely inaccessible to the local population. In 1983, local residents established the Comité Pro-Parque Educativo Miraflores (Miraflores Park Educational Association) with the aim of reclaiming the land as a public park, which eventually received the support of the city hall. Miraflores Sur was established within the park in 1991, comprising 36,400 m2 (approximately 4.2% of the total park area). The gardens are divided up into 162 individual plots, of approximately 60 m2 and 10 school gardens of approximately 150 m2. Today, the gardeners are predominantly retired residents from the local area; retired men hold the majority of plots.
The second participatory video process took place over eight weeks in 2017 with an urban permaculture collective, La Boldina. Eight urban gardeners -three female, five male -made a short film about the collective, which had emerged from Huerto del Rey Moro in early 2017 and was working in sites across the city. The group was started by a small number of gardeners with a strong interest in permaculture but has grown to include a diverse group of 30-40 people that cultivate sites across Seville, as well as outside the city. Beyond urban gardening, La Boldina is involved in public workshops, advocacy, lectures, and performance art to promote permaculture principles and practice. During the second participatory video-making process, the group was working in Huerto del Rey Moro, in a school garden, in an occupied house, in Parque de Alamillo gardens, on land made available around a local radio station, and on a small farm outside of the city. La Boldina's other activities, such as street theater, take place predominantly in Macarena within Seville's Old Town.
In each participatory video process, I considered "the participants" to be those that were involved in every element of video-making, and who had agency to significantly shape the video output. However, in each process a number of additional participants also took part in one or more activities. The first process involved four additional participants; the second process involved twelve.
The first process resulted in a short film, "Jardin Interior: Garden Inside", which was made publicly available through a number of websites including Resource Center on Urban Agriculture and Food Security (RUAF); City Farmer; the Huerto del Rey Moro website; Agroecology Now; and the Canadian Food Studies journal website. Additionally, the film has been screened at a number of academic conferences including the American Association of Geographers (AAG) 2017 Annual Conference and screened in part at the Royal Geographical Society/Institute of British Geographers (RGS/IBG) Annual conference in 2017. The second participatory video process resulted in a film, "La Boldina", which the participants decided not to make public, due in part, to the sensitive and precarious nature of many of the filming locations.
In addition to participatory video-making, the research employed a combination of qualitative research methods including semi-structured qualitative interviews, ethnographic observation, and auto-ethnography. I used semi-structured qualitative interviews to deepen my understanding of issues that arose through the participatory video processes. The interviews were informed by an adapted form of narrative inquiry, "[focusing on] an interest in life experiences by those who live them" (Chase, 2011, p. 421). This was a pragmatic, applied approach to qualitative interviews that focused on the form and content of narratives as constructed and communicated by participants. Across the two participatory video processes I conducted 36 interviews with gardeners, including the participants, as well as local residents, local academics, and members of civil society organizations.
Throughout each period of fieldwork, I also maintained an (auto)ethnographic account of the video process in the form of a research diary in which I recorded ethnographic observations as well as reflections on my position within the research and participatory video-making processes. The process of (auto)ethnography enabled me to track the development of specific ideas and themes, many of which changed in nature and prominence throughout the research process. The following reflection on the participatory video-making processes draws on findings from each of these methods.

Participatory video-making in Seville's urban gardens
This section is divided into five sub-sections, each framed around a stage of the participatory video-making process: training, planning, shooting, editing, and screening. In practice, the stages often overlapped. However, for the purposes of this article it useful to disaggregate the elements in order to unpack their distinct epistemological potentials and relations to urban space.
In each subsection I outline the process that I went through with participants in the first and second video-process, I then reflect on how I was learning as an external researcher and how the participants were learning, and the nature of knowledge being produced in each stage. Finally, I reflect on the ontological significance of these forms of learning and knowledge production in terms of the production of urban space.

Training
In both the first and second participatory video processes, technical training was very brief; limited to one afternoon in in the first process with four gardeners, and a threehour session with eleven gardeners in the second. Training was basic, tactile, and collective. In each process, I sat the group in a circle around the bags containing the camera equipment. I asked one member to unpack the bags, another to set up the camera, another to set up the microphone, and another to set up the tripod. During this exercise, I did not touch the camera equipment and encouraged the group to help one another to set up the equipment for filming.
We then conducted a series of short exercises to encourage the group to become familiar with using the equipment and seeing themselves on film. For example, we recorded short interviews of each other talking about our favorite places in the city. Each participant took a turn with the camera, with the audio recording equipment, and being in front of the camera. We immediately watched back these short interviews back using a laptop in the first video-making process and a projector in the second. The groups discussed the composition, exposure, and sound quality in each recording; the only rule being that they could only comment on the work of the people behind the camera and not in front of it.
Outside of the training workshops, I also set specific exercises for the group throughout the processes, designed to deepen their critical engagement with filmmaking. For example, during the first participatory process I asked the gardeners to capture five moving images of no more than 10-seconds each that "tell the story of the garden". To achieve this, the group negotiated how to plan a shoot sharing one camera and decided what might illustrate the "feel" of the garden. These shots were screened with the group followed by critical discussion.
The aim of the training exercises was not to teach the group specific ways to compose shots or to train the group to a high technical standard. Rather, the aim was to de-mystify the equipment and support mutual learning and a culture of experimentation within the group. Learning, in this sense, was collective, horizontal, and often spontaneous; shaped by the dynamics of the group that were made visible through the act of training with the cameras.
Through the training process, gardeners began to think about issues in the garden and community in audio-visual terms as well as in terms of narrative. This meant experimenting with using visual images and sound as ways of communicating "the feel" of the garden to an imagined audience that had never visited the site; how might they recreate the feeling of entering and spending many hours in such a dynamic the space within a small number of shots? The linear nature of film storytelling also encouraged the participants to think about how they might communicate a narrative of the garden; should they begin by explaining that this is a community space or that it is a garden, why or why not? In raising these issues through training, the aim way not to resolve, but rather to introduce them as frameworks for discussion throughout the video-making process.
As a researcher, the key learning opportunity was to observe these dynamics and better understand how I might better manage issues of power and voice through the participatory video process. One of my initial impressions was that despite the garden being managed as a horizontal space, the epistemic authority of some gardeners, for example, those with many years of experience growing food, meant that some individuals were given disproportionate influence in the design and management of the space, which causes some tension within monthly assemblies. This sense of hierarchy was present also in the filmmaking process, whereby the most active participants would often defer to those less-active participants with greater gardening knowledge, regardless of whether this knowledge was expressly relevant for the topic under discussion.
Additionally, the discussion of the city, neighborhood, and accompanying shots presented a useful early opportunity for scoping themes and issues that were likely to emerge through the research process. In the first process, one question that emerged clearly was whether urban community gardens were defined by their natural environment or in terms of the ways that the communities used the spaces.
Lefebvre's spatial triad reveals a number of simultaneous dialectical relations between the training process and the gardens in which we were situated. The ways that the group appropriated the space as a "classroom" -the ways we organized and shaped the space for trainings, screening, and discussion -impacted the materiality of the gardens, both for the group and for other users of the space. Moreover, the act of "telling the story of the space" evokes Lefebvre's conceived space through the collective negotiation of how the garden could be represented on film; the rationalization of the space and the negotiation of an audio-visual proxy. This exercise also raised important questions for the group regarding the politics of representation; who can decide how a project is represented? One gardener said, for example, "I feel I don't have a right" (male gardener, Huerto del Rey Moro, April 2016, first participatory video process). Training and discussion about the politics and ethics of video-making continued throughout the video-making process.

Planning
In order to plan the themes of the films we held a series of participatory workshops. In the first process we conducted two workshops. The first followed an adapted form of a problem tree exercise, which maps challenges from cause to effect. The aim of the exercise was to think through the interrelations between different "problems" and issues toward identifying a central challenge or set of challenges -the trunk of the tree -that is both a central cause and consequences of other issues. I asked the gardeners to think about "problems" at different levels: facing the garden, facing the gardeners, and facing the wider city. These challenges were written on post-it notes and discussed at length before they began to map the causality and relationships between them. In this instance, the issue of communication formed the "trunk" of the tree.
A second workshop held three days later was designed to explore the issue of communication more deeply as a potential theme for a short film. First, the gardeners developed a mind map around the theme of communication. As we proceeded, the discussions progressed from communication as being "about the message", to communication being about identities, motivations, and relations between those communicating. As part of this workshop the gardeners also proposed a number of questions that they hoped might be addressed through the video-research process and we developed conversational questions for interviews across both sites.
In the second participatory video process, we held only one thematic workshop with the primary participants in which we repeated the problem tree exercise. However, the primary thematic development took place incidentally at one of La Boldina's weekly meetings when we arranged a screening of the film, Jardin Interior. We used this as an opportunity to discuss the changes in the gardens and their organization over past year and develop ideas about the scope of the film and their vision for urban agriculture in Seville.
The workshops in the first participatory video process created a critical space that allowed for focused discussion of complex themes. Such spaces are not unique to a participatory video-making process, however the combination of workshops and filming exercises allowed for a productive dialogue between distinct critical processes. For example, the participants reflected continuously on how they might illustrate social challenges on film and many of the challenges that were raised reflected footage that was captured through the training processes.
Through the process of planning, the gardeners thought in different ways about both the challenges and the benefits of the gardens. What images, for example, should be used to represent the impacts of the garden in the neighborhood? These questions encouraged the gardeners to think more critically about what these impacts really are. Through these processes, the gardeners undertook a process of collective critical reflection that was characterized by the free association of ideas and images.
As a researcher, trying to understand the politics of self-organization in urban community gardens, the planning element represented an incredibly rich opportunity to understand how individuals and the group articulated issues and how they chose to represent them audio-visually. In some instances, these representations were literal: if we want to communicate that this is a space for all generations, then it is important to show old and young people together in the garden. But frequently the representations of issues and concepts became more metaphorical. How, for example, might they represent ideas of citizenship?
In addition to the uses of space that emerged in the training element of the process, the planning element speaks specifically to the conceived and lived moments of space. Through the problem tree exercise, the group debated the ways that the garden was rationalized, including how it was perceived by them and by others, how it was managed and organized spatially, and how these issues could be represented on film.
In asking the group to think about challenges facing the city, they also began to think about broader issues through the space of the garden and its relations to the wider neighborhood. These issues often brought in elements of memory, experience, and emotion from the gardeners. For example, concerns about sustainability and water management at the municipal level were frequently expressed in terms of analogy to the garden space. In Lefebvrian terms, the rationalization of the garden on film through memory was in dialectical relation with the space itself; simultaneously reflecting and reproducing the character of the space; by rationalizing the space as "green", "public", and "self-managed" amongst other characteristics, the group affirmed these qualities in the garden. Moreover, the act of deliberating over how these qualities might be represented on film further contributes to this dialectical relation.

Shooting
In the first process, following the workshops, the film was shot over three consecutive weeks. Much of the filming involved the gardeners traveling between Huerto del Rey Moro and Miraflores Sur to conduct and film interviews with gardeners, academics, and other users of the space, based on interview questions collective developed in the planning stage. Additionally, filming often occurred spontaneously, including events and scenes that may only have had a tangential connection to the themes developed in the planning stage. The recording equipment was also available to the group to take away and use to record events and further interviews in the gardens, or simply to experiment with filmmaking.
The second film was shot over a period of four weeks. Compared to the first process, the shooting stage was far more sporadic, spontaneous, and opportunistic. In part this reflected the availability of the gardeners, and in part it reflected the fact that a coherent narrative was not agreed in the planning stage. Additionally, at the suggestion of the gardeners, I took a more active role in filming, for example, by conducting interviews using questions that they had developed. Filming took place in five locations in which the group of gardeners were working including private, public, institutional, and occupied space, inside and outside of the city.
The power dynamics within the group of the participants observed in the training stage were largely mitigated in the shooting processes as the actual footage captured reflected a great degree of opportunism in addition to following the plan collectively developed; split decisions to begin recording were often made by members of the group who were available at the time. This meant that those members of the groups with more free time were given disproportionate agency to influence the filming processes. But it is important also to recognize that those with greater free time were also more likely to be those in part time or precarious employment. And therefore, within the wider dynamics of the gardens and the neighborhoods, the most active participants were frequently those that could be considered more vulnerable members of the community.
Shooting the film offered a number of epistemological opportunities both for me and for the gardeners. The first was a continuation of the process that began in the planning stage that I termed, learning through space, whereby the research questions and research subject, across both processes, were continuously reshaped and renegotiated through the images that the gardens were capturing; the form of the garden and its composition on camera became the means through which the group discussed, for example, issues of cooperation, solidarity, and sustainability. The images that had been a way of representing abstract ideas in the planning stage, became the primary mode of inquiry, to which the gardeners would continuously attach new meanings and significance. A large number of qualitative interviews were filmed by gardeners in both processes; however, their inquiry was shaped by the imagery and the representations of the gardens on film to a greater extent than the content of the interviews. This process was incredibly dynamic, given that multiple perspectives were captured as the camera is passed around the group, shots were set up, and "the gaze" of different participants -women, men, local, visitors, young, old, etc -was revealed and captured.
The second epistemic opportunity relates to what I term, displacement; the process of dislocating one's position in relation to everyday practices and familiar contexts through the camera. As one of the primary participants in the first process explained: This has been an opportunity that you gave me. I took it immediately because it was another way to learn about the garden. I was really conscious of that in the beginning. It was a really conscious thing, to be on the camera and work with the camera, it was like putting yourself in another perspective. And at the same time working in the garden, and changing from one to another during the day, for me it is accelerating the process of knowing the place (Interview, June 2016, Huerto del Rey Moro).
This process of displacement is not unique to participatory video-making, but here it takes on a particular dynamic and collective character as participants share recording equipment and switch continuously between perceiving space through and outside of the camera.
The third distinct opportunity that emerged through the shooting process was the idea of trans-local learning (Allen et al., 2018;McFarlane, 2011); forging connections between places to support learning across places and, emerging from this, the process of learning one place through another. The process of trans-local learning took on a different character in each of the video processes. In the first, gardeners traveled frequently between two gardens which enabled the participants not only to learn about a new space, but also to think about their own garden differently. As one participant explained: I have been visiting Miraflores with you. That was a huge opportunity to know . . . It taught me more about this place, visiting Miraflores. It taught me really a lot about this place. How it works, with the neighbourhood, with those involved, and how the people here are working. It's been amazing. Was it last week we visited? We talked to the people -it taught me a lot of things. Even if they were only brief conversations, they throw a lot of light to this area and this garden (Interview, May 2016, Huerto del Rey Moro).
In the second process, the act of video-making across multiple sites supported a process of collective, comparative inquiry, whereby participants generated and discussed critically commonalities and differences between the sites, as well as the ways in which different spaces and projects had been impacted by processes and events, such as policy changes, that occurred at the municipal level.
The final learning opportunity, which was particularly useful to me as a researcher, emerged from discrepancies between the ways that the gardeners had articulated issues in the planning stage and how these issues were represented on camera. I found that certain challenges identified in the planning stage, such as insecurity of tenure, which were often discussed by the participants and the wider gardeners were entirely absent from the shooting process. This epistemological pluralism was particularly rich in terms of research insight; tracing how some ideas emerged only in discussion and others only in the context of shooting.
In ontological terms, there were myriad and complex relations between the act of shooting the film and the urban space in which we were working. The act of videomaking has an immediate impact upon the organization of space through the ways that the film subjects (those appearing on camera), the filmmakers, and the camera equipment itself shape and delineate space into "on-screen" and "off-screen". This reflects both the way that space is rationalized and experienced by the filmmakers, as well as the ways that bodies moving through material space fundamentally alter the character of that space. Moreover, the collective labor of the filmmakers, the discussions and selforganization, such as the sharing of roles, in particular the control of the camera, across the group generates a further dynamic level to the production of space through participatory video-making. Furthermore, the ways that the group chose to include or exclude specific angles of the gardens in the film contributed to the ways that they, individually and collectively, understand and experience that space. The shooting process evoked, in particular, Merrifield's (2011, p. 475) description of Lefebvre's spatial philosophy: "It's not in space that people act: people become space by acting. Nothing is scenic anymore . . . participants' own bodies become the major scenic element, the spatial form as well as the spatial content."

Editing
Each process contained a participatory video editing stage. In both processes we watched back all of the footage captured over a period of three days in order to construct a "paper timeline" -using post-it notes and other materials -to replicate the way editing software is used to edit films.
In the first process, the primary participants reviewed all of the footage captured during the filming period (approximately seven hours in total). The group initially selected clips for inclusion by drawing the scene, noting the timecode and filepathway, and writing short description of each clip on a post-it note. There was no consensus as to a precise message in the film. As such it retained a "messiness", reflected in the non-linear storytelling, and inconsistent exploration of central ideas; the aim, the participants agreed, was to produce an "honest" film, rather than a promotional one.
The post-its were initially grouped thematically before we began to lay out specific sections of the film shot by shot. I then reproduced the timeline using Adobe Premier Pro video editing software and showed the resulting film in a small screening with the primary participants. At this point, they gave feedback and I made their suggested changes to the film. This process was repeated three times before we had a final product. The subtitles were added at this point, written collectively by the author, one of the participants, and a UK visitor to the garden who was working in Spain.
In the second video-making process we also developed a paper timeline, this time with twelve co-editors. We watched approximately eight hours of recordings across three halfday workshops. This time I reproduced the timeline with the assistance of one of the primary participants who wanted to learn how to use editing software. We arranged one screening of the film with the group who discussed changes and suggested edits. I then returned to the UK and we continued to discuss the draft versions of the video, shared privately between us online.
In both editing processes the sessions were long and occasionally fraught, with different members of the group entering the process with different ideas and expectations at different times. Occasionally a member of the group would want to start a discussion about that they felt was important, for example, whether an image should or should not be included because of what it implied about the group's philosophy. Sometimes these points were taken up for discussion, which could involve a room-full of people in intense debate about what would ultimately amount to a few seconds of footage in the final film. Other times these discussions were avoided, not because the issue was not important, but because there was only so much energy for debate across three days; people soon became exhausted.
In this sense the output film represents what could be agreed by a diverse group of people, rather than an output that reflected accurately the collective position regarding the gardens. This was further compounded by the differential engagement of the group with the editing process. The second output film's credits list twelve editors that were involved in every aspect of the editing process, however twenty people were involved in some capacity. This was a challenge for me as facilitator, managing the expectations of individuals that were only able to attend part of the editing process. The complexity of the editing processes is reflected in the messiness of the final output films.
Across the two video processes, the participatory editing workshops were not only about the assemblage of material, but also an opportunity to reflect on how we have represented the gardens and the gardeners, how we want to represent them, and how different themes have been explored implicitly. In both cases, the themes of the film changed substantially during the editing process from the plans developed in earlier workshops.
In the first film, the theme of self-management (autogestion) that had been proposed in the planning stage, became far less central. At the same time, the gardeners noticed the amount of material captured that related to the reuse and rehabilitation of waste materials, which became a prominent theme within the output film. In the second editing process, central themes such as water management were abandoned in favor of an overview of some of the group's projects and statements on the group's philosophy.
In terms of epistemological opportunities, the editing process presented a rich collective discussion whereby the themes identified in the planning stage and the images captured through the shooting stage, were transformed into a narrative. The narrative did not necessarily mean "the message". Rather it was an attempt to give collective meaning to the process that we had been through so far. In this sense, the act of narrative creation represented a distinct opportunity to reconsider the issues that the group had anticipated being in the film, and the issues that had actually emerged through the shooting process. Like the other stages of the video processes, participatory editing was also horizontal, visual, deliberative, and critical; the narrative that emerged in both instances represented an opportunity for me as a researcher to learn about the issues, and for the participants to learn about the perspectives and priorities of the others.
The ontological relation between the processes of participatory editing and the production of the space can also be understood in terms of narrative creation. In creating a narrative regarding a space, the participants are generating representations of that space, thus contributing to its social (re)production. For example, each person that forms the idea (or learns) that a particular space is safe or dangerous, quiet or busy, public or private, contributes to the ontological character of that space. Lefebvre (1991) understood the three elements of space to have a continuous, dialectical relationship. Therefore, the ontology of the space (and the lived experience of the space) cannot be distinguished from the knowledge produced through and entailed within representations of that space; each contributes to the ontology of the other.

Screening
The final stage of each process was a screening. In the case of the first film, we screened the output film first on a laptop in Huerto del Rey Moro, before traveling as a group to Miraflores Sur, to watch it on a projector with some of the gardeners involved, as well as the Heads of the Gardeners' and Park Associations and academics from a local university. This screening was followed by a discussion between the gardeners, from different the sites and of different generations, about the future of urban gardening in the city. For example, I asked the gardeners to discuss how they would use a small plot of land in each other's site. The purpose of the discussion was not to develop a roadmap for collaboration, but, with the aid of the film's audio-visuals, to use concrete examples of action to explore differing visions for community-managed spaces in the city. This discussion was particularly rich and brought together a large number of the themes discussed though the entire process. One gardener from Huerto del Rey Moro noted after the discussion that she could see the growth in the group at this point: I am a very sensitive person, and this is another transformational work that I am going through in this place. And trying not to feel asphyxia with all the information that comes along. The process with the workshops for me was really intense because I was seeing all this information and all these intuitions that I have been feeling in Sevilla this month all condensed . . . It was an intense process but also beautiful. I felt so much growth in the group in that discussion (interview with female gardener, Huerto del Rey Moro, June 2016).
For the second film, "La Boldina", I missed the opportunity for a substantial discussion with the group at the screening, as the film was not completed before I left the city. However, we were able to organize a screening of the draft film at a local community center for twenty members of La Boldina. I felt that at this point, it was clear that second film had raised more questions than it answered for the group. In the first video process, the gardeners had sought to explore specific issues across two sites, and I believe, developed a better understanding of both gardens through the process. In the second film, the process had highlighted disparities in thinking within the group, and different visions within La Boldina for urban agriculture in the city, which the process had not been able to reconcile. However, at the screening many of the gardeners noted that above all, the film captured the collective energy of the group.
The first film was shared widely online and is available through a number of different websites. This in turn generated interest and contributions to the discussions that we had begun in Seville. We received correspondence from individuals interested both in urban agriculture and participatory video-making from Europe, Africa, Latin America, and North America. While the second film was not put online, La Boldina went on to produce a number of further videos that are widely available online.
The screenings represented a new opportunity for learning in that they made the conversations public; suddenly individuals that had not been involved in the development or production of the film were invited to comment on the films and the issues they raised. This form of public examination was welcomed by some participants, but not by all. The discussions that followed the screening of the film were useful to me as a researcher, as the points made by the audience were frequently supported and made concrete by examples, both spoken and visual, from the film. In a similar way to the shooting and planning stages, the visual elements of the film provided a means through which the group generated new insights about their own projects.
In terms of the ontological significance of the screening, this was the point at which the conceptualization and representations of the gardens entered into public dialogue with existing conceptions of the space. By raising questions in the first film about the importance of reusing and repurposing materials from across the city, the garden became a site in which conversations about these issues are welcomed and promoted. The dissemination of the video online also contributed to the virtual presence of the gardens and iterated their importance as green public spaces.

Reflections on participatory video-making in geographical research
Overall, the participatory video process was extremely rich and productive. It allowed me to develop close relationships with the groups of gardeners that would not have been possible through an interview-based methodology. These relationships allowed me to better understand the micro-politics of the urban gardens and complex the relationships between the gardens, the gardeners, and their neighborhoods.
The video processes and outputs were particularly useful for exploring the lived experience and diverse perceptions of a variety of spaces across the city. The output films preserved the messiness of the participatory processes, which in turn reflected the diversity of perspectives across the groups of participants. In this sense, participatory video-making was a particularly apt and productive mode of inquiry into inherently dynamic and contested spaces such as urban community gardens, which are often defined by their contradictory politics (McClintock, 2014). Dissensus, then, was preserved throughout the processes and outputs, rather than flattening it out or explaining it away.
The power dynamics that existed within the groups of gardeners emerged at several points in the video-making processes, particularly in the planning and editing stages, when there was a pressure to reach collective agreement, however temporary, on a direction forward within a finite amount of time. However, these dynamics were mitigated to a great extent by the actual shooting process, which allowed for a diversity of inputs from all the participants, and incidentally privileged those who were less likely to be in full time employment. It is important to note also that the video-making processes occurred within the context of contested relations within the gardens (detailed in Yap, 2019) as such the processes may have contributed to a minor extent to the consolidation of factions within the gardens, even though this was not suggested by any of the participants.
The fact that La Boldina decided not to make their output video public, in part to protect various occupied spaces and houses, points to a wider issue regarding the production and dissemination of visual research outputs, namely that they have the potential either to jeopardize groups engaged in illegal or impermissible activities or to perpetuate the ways in which certain urban practices, behaviors, and activities are marginalized and made invisible in media and research. In part this reflects the challenge as to what types of knowledge can be represented in audio-visual terms, but it also reflects the wider set of challenges regarding the access and use of urban space by urban inhabitants. Participatory video-making doesn't enable us to overcome these challenges, however, the capacities built through the process can be used strategically to enhance the claims of mobilized groups whilst mitigating the immediate precarity of its members, as we have seen with La Boldina.
To some extent, video outputs can reveal in greater depth than photography the complicated social and material processes that combine to give urban gardens their distinctive character. However there remain important limitations. The selection of material, through shooting and editing, might give a good indication of what it is like to experience the garden, but it is inescapably a deeply partial process. In using a participatory approach to video-making, this partiality is rendered more collective and democratic. Nevertheless, a representation of the garden on film is an implicitly partial and prescriptive representation of how the space can or ought to be conceived. In this way, the very act of participatory video-making is the production of conceived space, both reflecting and contesting the spatial ontology of the garden.
It is also important to recognize what Soja (1996, p. 20) identified as the "tendency in post-modern critical urban studies to overprivilege the local -the body, the streetscape . . . the micro-worlds of every day life and intimate communities -at the expense of understanding the city-as-a-whole." This risk is especially important for methodologies such as participatory video making, in which the lived experience and engagement with highly localized spaces fundamentally determine the process and outputs in the ways described here. One way of addressing this issue is to bring a participatory video-making process into dialogue with other research methods that could include policy analysis, discourse analysis, ethnographic observation, qualitative interviews, and focus groups amongst other methods.

Epistemological potentials of participatory video-making
As this article has demonstrated, the main stages of each participatory video-making process -training, planning, shooting, editing, and screening -represent distinct epistemological opportunities both for researchers and participants. These activities offer distinct opportunities for engagement with often complex issues. Moreover, participatory video-making necessarily creates a series of collective, critical spaces which enable facilitators to construct conceptual narratives of the development of key ideas or arguments through these spaces, in a way that reflects the emergence of a narrative within the film.
The ways that certain ideas emerged or were articulated within each stage could be complementary, but it could also be contradictory. For example, participants may set out to represent a situation in one way when shooting the film, but this might be contradicted by the way the same situation in represented during and as a result of the participatory editing process. The dialogue between these distinct forms of engagement, and the opportunities to reflect on this dialogue, is one of the greatest opportunities of participatory video-making as a research methodology. Overall, through the creation of epistemologically diverse processes and multiple critical spaces, participatory video-making enables a diversity of learning pathways to emerge both for the participants and the researcher, which are incredibly valuable within the field of urban geography.
As outlined above, the change-making potentials of participatory video-making, as well as its limitations are well-documented through the vibrant critical participatory video-making discourse. What this article has attempted to demonstrate is that participatory video-making represents a uniquely dynamic approach to the production of urban geographical knowledge that is particularly relevant in terms of "learning the city" as a dynamic process of developing perceptions (McFarlane, 2011), and which can complement and augment existing approaches to ethnographic, participatory, and qualitative geographical inquiry.

Towards a critical ontology of participatory video-making in research: implications for urban geography
Drawing on the work of Henri Lefebvre, this article has attempted to outline one approach to understanding the reciprocal and dynamic relationships between participatory video-making and the production of space. Participatory video-making, in this sense, elucidates the relations between research practice, subject, people, and urban space.
Lefebvre's spatial ontology helps us to trace the numerous reciprocal relations between participatory video-making and the city; the myriad ways that urban space is itself produced through the ways groups organize to use the city, through the production of narrative, and through the conceptualization, representation, and experience of space, each of which occurs throughout a participatory video-making process. In this sense, participatory video-making is more than a way of understanding cities; it is a way of producing cities and producing urban social life.
As is well established in the field of anthropology, the process of research can fundamentally alter the subject of research. This truism exists also within the field of urban geography, exemplified by the critical turn toward issues of identity, positionality, and postcolonialism. However, the specific mechanisms through which this "alteration" occurs, remains under-theorized. However, the relationship between participatory videomaking and the production of urban space is only one example of the ways that we might begin to construct a critical ontology of urban research.
While this analysis has drawn on the work of Lefebvre, it has not engaged with the rich and established discourses regarding film theory and semiotics, for example, the work of Deleuze (1997), which speak directly to the issues raised here and which represent significant opportunities for further research into the ontological significance of participatory video-making in research.
What issues, then, does a critical ontology of participatory video-making raise for urban geography as a wider discipline? The first is that there are multiple ways of conceptualizing, and thus managing, the diverse and multi-directional relations between geographical research methodology and the subject of inquiry. The analysis of participatory video-making through the lens of Lefebvre offers just one opportunity for recognizing these complex ontological relations. This approach is not exhaustive, nor does it try to be. The challenge, then, for urban geographers, and urban geography as a discipline, is to recognize, conceptualize, and reflect upon the relations between research as urban practice and the production of urban realities. This represents a significant area for further scholarship.
The second is that participatory video-making makes visible the plurality ways that urban geographers produce urban reality through the collective narrative that is academic discourse. In this way we might begin to think about urban geography in terms of making cities as much as it is about understanding them. A participatory video-making process is not, by any means, a microcosm of urban discourse, however the ways that themes and issues emerge through dialogue between moving images, critical collective discussion, and lived experience -the ways that plural knowledges are produced, circulated, and contested -points us toward ways that urban geography as a discipline might productively engage with a plurality of voices, approaches, and epistemologies not only to better understand cities, but also to remake them.