Conventional approaches in urban design tend to impose control over public spaces to manage their use (Carmona, 2010; 2014; Madanipour, 2010). Conversely, more flexible approaches allow freedom of choice to accommodate different uses (Whyte, 2001; Gehl, 2013; Mehta, 2014). This flexibility extends to the co-production of spaces and users and their interaction (Lefebvre, 2012). Based on the latter approaches, recent studies have explored the potential of urban interventions to transform spaces into spontaneous, open arenas with new use possibilities and different ways of engaging users (Cohen et al., 2014; Hunter et al., 2015). Some studies have focused on interventions that are temporary in nature (Lehtovuori, 2011; Almousa, 2015; Wagner, 2016; Madanipour, 2017; Rossini, 2019; Paukaeva et al., 2021; Skytt-Larsen et al., 2022), modify spaces continuously, and allow different emerging uses.

These aspects have caught the attention of relational researchers (Pinder, 2008; Tornaghi, 2015; Manfredini et al., 2017; Wohl, 2017), with some adopting actor–network theory (ANT) (Law, 2009; Latour, 2012; 2018; Law et al., 2012; Kim, 2017; Sharif, 2020a; 2020b; Yaneva, 2022). These relational approaches, including ANT, view space as a collaborative achievement between existing urban elements, new elements, and users (Hotakainen & Oikarinen, 2019). A consideration of the temporality of new elements facilitates an improved understanding of unfinished, ambiguous, and continuously upgraded spaces in relation to produced uses and engagements. However, although these studies explore urban interventions and their potential to produce spaces that are continuously altered to generate new uses, they rarely address the variety and continuity of the interactions between interventions and their users.

This research utilizes actor–network theory (ANT) and deploys its potential to view the socio-technical world through networks and produced relational effects to investigate the reality of interactions between space interventions and their users. ANT allows for a specific focus on the variety and continuity of interactions by understanding these relations and unpacking their diverse and dynamic nature, which has rarely been utilized in previous ANT-related research. The research applies ANT to provide a nuanced perspective on the potential emergent interactions between introduced urban interventions and their users. It explores how these short-lived alterations in public spaces induce different uses. In particular, this study examines how temporary arrangements, introduced by designers, result in different and continuous relational patterns (rearrangements of relations that are built/strengthened, cut/weakened, or fluctuate between interventions and users). Through considering their temporality, the research aims at accounting for the spatiotemporal interactions occurring between interventions and their users through their presence and absence in space and time (i.e., before the interventions were installed, during their existence, and after their removal), which stimulate diverse and continuous arrangements and changes in relations. In this study, ANT is used to empirically investigate a temporary intervention in Abdali Boulevard in Amman, Jordan, to understand the different interactions revealed in terms of stabilized relational patterns. This investigation is likely to be useful for designers in developing new ways of designing spaces through offering interventions that engage the public and understanding the nature of such engagement when (i) taking into account the multiple, complex, and even unpredictable intervention-user interactions and (ii) considering the different capabilities of co-functioning of spaces when new interventions are introduced and new uses are induced.

Relational approaches in urban studies view the socio-technical world based on the dynamic relations between space and its use (Tornaghi, 2015; Tornaghi & Knierbein, 2015; Wohl, 2017). Accordingly, spaces are viewed not as final products with pre-arranged configurations but as processes of arrangement between spatial elements and users (Hotakainen & Oikarinen, 2019). These approaches explore new ways to challenge and transform spaces into dynamic, unpredictable arenas that offer diverse possibilities of use and engagement. In particular, relational approaches are interested in disruptive encounters in the form of interventions to upgrade, modify, and activate spaces to create different uses (Altrock & Huning, 2015). These interventions disrupt the order of the urban space, infuse new qualities that enrich user experience, and produce a creative space that becomes livable rather than lived (Hotakainen & Oikarinen, 2019). Consequently, such interventions become points of creativity, change, and resistance, producing different spaces and expanding the emerging experiences of their users.

Some studies have paid particular attention to temporary interventions (Hotakainen & Oikarinen, 2019), where time is a dynamic and crucial variable that provides a temporal frame for the produced interactions (Purpura, 2016; Madanipour, 2017). Through these interventions that interrupt the space for a short period of time, the space transcends the expression of fixed, predictable content to reveal transformative, unforeseen elements that can respond to different issues and upcoming events (Purpura, 2016), stimulating different uses and activating new, previously unlived experiences. In this way, temporary interventions signify a type of spatiotemporal production and present an essential component of urban development processes (Hotakainen & Oikarinen, 2019). Although such interventions are ephemeral, the emergent interactions leave a lasting effect that connects space with its users in different and unexpected ways.

Studies on temporary interventions address the aspect of ephemerality through introducing temporary events or changing the space to host short-term use (Wohl, 2017). However, the present study addresses these aspects by focusing on the establishment of flexible (changeable or removable) structures that introduce different events and produce changes in use. Previous similar studies have explored adaptable temporary structures that are assembled and disassembled (Tardiveau & Mallo, 2014; Hotakainen & Oikarinen, 2019) or continuously transformed (Almousa, 2015), potentially upgrading spaces and generating different responses from their users.

Among these studies, and based on similar relational approaches, ANT provides a way to understand the socio-technical world through networks that produce relational effects between humans and non-humans (Latour, 2012; 2018; Akrich, 2023). ANT particularly involves exploring new ways of transforming spaces into dynamic and uncertain entities to extend their possibilities of use and engagement. Urban studies utilizing ANT (Kim, 2017; Sharif, 2020a) use concepts such as inscription/prescription (Akrich, 2023) to refer to the initial creation of spatial relations by designers, and translation (Law, 2009; Latour, 2012) to show how these relations are variously and constantly amended by users. Time and space are folded into and affect these relations (Latour, 2012). Consequently, interactions in urban spaces are distributed through entanglements and rearrangements of relations that are built/strengthened, cut/weakened, or fluctuate (Yaneva, 2022) between human and non-human entities through similar or different time–space frames (Murdoch, 1998)). Through these interactions, relations oscillate between different forms of presence and absence of actors in time and/or space. In some cases, actors might be physically present in time and space or present in one dimension and absent in the other, while still producing specific relational effects (Murdoch, 1998; Latour, 1999). ANT differs from other relational approaches in that it does not consider any reality outside the relational effect (Harman, 2009), which makes it valuable for empirical investigations, as it specifically follows the relations and emerging translations wherever they may lead (Yaneva, 2022). However, as these relations continuously change to create alternative arrangements, they result in variable and continuous translations (Law, 2009; Latour, 2012; Law et al., 2012). Accordingly, translations can only be traced through their stabilized moments, when they appear through enduring and recurring forms or repeated relations with continual/regular production of interactions (Law, 2009; Law et al., 2012).

Although relational approaches have been used to explore these urban interventions, the adoption of ANT to study them is rare (Boonstra & Specht, 2012; Harboe, 2012; Costa, 2018), despite its potential in urban studies to investigate the interactions between space and users by following relations and continuous translations (Murdoch, 2006; Kim, 2017; Sharif, 2020a; 2020b). Among the few researchers utilizing ANT for these explorations, Boonstra and Specht (2012) utilized the example of Heemraadpark in Singeldingen, the Netherlands, in 2008, where a poorly developed and almost abandoned park was turned into a dynamic attraction through deploying a temporary kiosk made out of a spring-roll cart. This stimulated a string of activities in the park lasting 6 weeks. Costa (2018) described a temporary community garden structure at an abandoned space in Maria de Feira, Portugal. This flexible structure, comprising ceramic pots and wooden piles, assisted in place activation and public engagement by encouraging planting and maintenance activities and associated events. Harboe (2012) explored a temporary structure installed for 2 weeks in a deserted area of northeast London. Made of freely constructed steel bars and scaffolded roofing, this structure comprised easily modified activity spaces and components that enlivened an otherwise abandoned space. Harboe (2012) also explored an intervention constructed at a dumping site at the edge of Parc de la Villette in Paris. This flexible, scaffolded structure could be unfolded, extended, or altered in response to daily changes, to promote residents’ interaction during events hosted for 5 weeks. In each of these cases, the introduced interventions, while present, helped users produce different networks in space through building relations with the intervention, the surrounding space, and other users. The interventions also had long-lasting effects, even after their removal (Boonstra & Specht, 2012). These examples provide valuable insights into how urban interventions may upgrade the possibilities of space and generate user responses vital for encouraging interactions. However, few empirical investigations of the reality of these interactions, especially in terms of their variability and continuity, exist.

The research aims at applying ANT to unpack the potential emergent interactions between introduced urban interventions and their users by accounting for the different and continuous relational patterns occurring between interventions and their users through their presence and absence in space and time. Specifically, an ANT-inspired ethnography was used to empirically investigate a temporary intervention in Abdali Boulevard in Amman, Jordan, to understand the different interactions and relations traceable through stabilized moments of endurance and occurrence. This investigation examined the possibilities of co-functioning that depend on both the scrutinized elements in space and the responding users as well as the diverse possibilities of associations between spaces and users.

The case study used in this research comprised a temporary intervention developed in the Abdali project, a mega-urban generation project and mixed-use development located at the center of Amman developed as its new downtown, between 2015 and 2020 (Abdali website, 1 January 2021). Within the project, Abdali Boulevard is an east–west pedestrian spine considered the main public space in the project’s first phase (Figure 1). It follows the site’s contours, combining terraces allocated at three levels. It also reflects a certain design quality through its outdoor spaces, landscape features, and furnishings surrounded by buildings that include ground-floor restaurants and retail spaces and upper-floor residential and office spaces (Zalloom, 2015).

According to direct communication with the Abdali project management, temporary interventions have been introduced in various forms and changeable modes to enhance the quality of the urban space, suit different events and situations (four monthly signature events and 30 daily or weekly events annually), and expand user experiences. One commonly used temporary intervention has been an arch-like structure decorated with hanging flowers, typically used in spring, which was installed and uninstalled in various forms and several times between 2015 and 2020 (Figure 2). As the intervention introduces a different type of spectacle to the boulevard, it encourages users to experience it while passing by or standing beside it, observing and taking photos. The significance of this case is that it represents a highly repetitive intervention that allows direct and intimate interactions, unlike other, less approachable interventions (such as those that are distant or fenced off). Studying the former type opens specific possibilities, although studying the latter type in future research would allow for different horizons to be investigated.

This study utilized an ANT-inspired ethnography, which differs from traditional ethnography, as it allows for in-depth investigation through developing sensitivity and receptivity toward the emergence and change of socio-technical relations between actors involved in complex and dynamic processes and folded into similar or different time–space frames (Murdoch, 1998; Latour, 2012; Kärrholm, 2016; Sharif, 2016; 2019; Yaneva, 2022). This allows the researcher to actively scrutinize and refine understandings of and from the research process to enable effective and pertinent reflexive translations of the research project (Sheehan, 2011).

Data were collected through a number of methods to allow effective engagement with the intervention and its users and the different interaction modes between them. The fieldwork was preceded by a desk-based study of documents and websites for sourcing relevant data on the Abdali project and its boulevard design and use from 2015 to 2020. This was followed by semi-structured interviews with five management members (30 min each), conducted between January 2019 and March 2019. The interview schedule was flexible, targeting the main design intentions in terms of involving users and simulating their interactions. Furthermore, the research involved semi-structured interviews with 35 boulevard users (10–15 min each). The users were directly approached during site visits. They included 21 visitors and 14 employees, which distributed the sample into less and more frequent users. The sample also comprised 19 females and 16 males, with ages ranging between 15 and 75 years. They were asked about their experience and engagement with the structure, regardless of whether it was in situ. The interviews were aimed at exploring and understanding how interventions, introduced by designers intended to engage users, and how the actual engagement of the users presented different and continuous interactions and relational patterns occurring between interventions and their users through their presence and absence in space and time, which might not be initially anticipated by the designers. The research also involved fifteen participant observations of the temporary intervention (2–3 h each, two times a week, on different days each week to ensure a representative distribution between and within weekdays and weekends) conducted between March 2019 and October 2020 to capture the reality of interactions between the structure and users during its existence and before/after its installation/removal. While the observations allowed the author to observe and follow the actions and interactions of users, the interviews helped in understanding these interactions and how users explained and justified them. Both interviews with users and observations were carried out on weekdays and weekends in the afternoon when people were often anticipated to be outside (after finishing work, school, or chores on weekdays, on their free time on weekends, and when the weather was pleasant for being outside).

The management members were identified and accessed through colleagues in companies in the area and they were recruited by email. The users were recruited during the site visits, which was mainly affected by users’ availability in the site. The number of interviews and site observations was defined based on reaching saturation–the state when the emerging data confirms, or mostly repeats, the previously collected data while accounting for the many possible variations (Glaser & Strauss, 2010). The researcher provided a statement of the study purpose, duration, and procedures to all participants and obtained informed consent from them. Furthermore, the researcher maintained participants’ confidentiality and ensured their anonymity through employing pseudonyms for users and with general referral to management in relation to data from managers, in addition to the use of blurred photographs.

The interviews were documented through either notes or tape recordings and observations were recorded through notes, sketches, and photography. The collected data were analyzed thematically using the qualitative data analysis software Nvivo 10. Data analysis was aimed at (i) identifying codes that relate to the different interactions and emerging relations between the intervention and its users, (ii) grouping the emerging codes to identify potential themes concerning the types of emerging intervention-user relations during the presence and absence of the intervention, (iii) defining and finalizing themes, and (iv) presenting the results in a coherent and logical sequence. The interviews and site observations presented various, continuous, and sometimes unpredictable interactions and relational patterns occurring between interventions and their users through their presence and absence in space and time, where users attended to, engaged with, overlooked, bypassed, disturbed the intervention, or floated between different modes of interaction.

The process initially included exploring different interactions featured through stabilized relational patterns (rearrangements of relations that are built/strengthened, cut/weakened, or fluctuate between interventions and users), while the intervention was either present or absent (i.e., before the intervention was installed, during its existence, and after its removal). This included the following three analytical forms:

a) The co-presence of actors and their mutual existence in time and space;

Intervention–user interactions can stabilize different relations. Figure 6 provides a diagram that summarizes the results showing the different types of intervention-user interactions in cases of presence and absence in space and/or time, illustrated through different relations.

While related studies by Boonstra and Specht (2012), Costa (2018), and Harboe (2012) considered how relations are built in space due to introduced interventions, this study focused on different types of relations (Murdoch, 2006; Kim, 2017; Sharif, 2020a; 2020b), which were described in terms of building/cutting, strengthening/weakening, and fluctuating during the presence and absence of the intervention. It is challenging, although of critical importance, to understand more fully the nature of these relations, the types of actors they connect or disconnect, the degrees of strengths and weaknesses, and the timespan through which they stabilize.

Considering Abdali Boulevard’s intervention, building/strengthening relations between the intervention and users can happen through people who may attend to the structure because it catches their attention or others who may want to engage in the interactions happening around or inside it. Both interactions entail building or strengthening relations but differ in terms of the actors involved, the degree of such building/strengthening, and the time during which the stabilizations last. Attending to, for example, entails building or strengthening relations with the structure and possibly with users. Engaging involves more actors and entails stronger and more durable relations. The engaging user not only builds strong relations with the structure, space, and users but also helps other users build relations with the structure. Other cases of building/strengthening could connect people distant in time and space because people are interested in thinking about, talking about, or checking intervention-related sources while the intervention is there before its installation or after its removal.

Similarly, cutting/weakening relations, in the case of Abdali Boulevard’s intervention, can occur through interactions by users accustomed to these changes or those focused on specific tasks (overlooking), those who prefer to navigate an interruption-free environment (bypassing), or those who play or act in a disorderly manner (disrupting). However, it is challenging to identify the types of actors and the extent and duration of such cutting/weakening relations. Overlooking, for example, constitutes weakening relations with the structure because users steer their focus in a different direction, whereas bypassing is a more extreme case of weakening/cutting relations, wherein users are annoyed by interactions occurring around the structure. Disrupting is the most extreme, as users contribute to cutting other people’s relations with the intervention. Other cases of cutting/weakening could disconnect people while distant in time and/or space.

Finally, fluctuating relations can occur in the mentioned cases through shifts and oscillations in relations between the intervention and other structures, surroundings, events, and other users through the same or different times and spaces. The challenge in understanding the nature of relations is greater in this case because varying timespans are also concerned in which relations are formulated/reformulated and stabilized/destabilized, which range from a few minutes to much longer durations. Understanding the various types of relations as well as their nature entails diversifying and extending possible intervention–user interactions and expanding the involved networks in endless and hard-to-capture ways (see also Latour, 2012; Yaneva, 2022).

This study used ANT to provide a nuanced perspective on the potential emergent interactions and relational patterns between introduced urban interventions and their users during the intervention’s presence and absence. This was achieved by following the relations as they were built/strengthened, cut/weakened, or fluctuated with these interactions through the cases before the interventions were installed, during their existence, and after their removal. The study showed how the emerging relational patterns were indeed various and ongoing. The nature of these relations, the types of actors they connect or disconnect, the degrees of strengthening and weakening, and the timespan through which they stabilize all worked to further extend this diversity and continuity.

The findings illustrate how the interactions and changes in the relations between the intervention and users can contribute to changing, upgrading, and elevating spaces (Boonstra & Specht, 2012; Costa, 2018). Through these interactions, and as the installations are foregrounded, backgrounded, or slip in between, and as users engage with them in different ways, spaces are manipulated accordingly. Interactions that entail shifts and oscillations in intervention–user relations affect spaces. Furthermore, space is transformed with these interactions and changes in relations when the intervention is present. It transforms again when it is not yet installed or after its removal, as users approach the space differently. Spaces are involved in continuous and long-lasting effects despite, or because of, the temporality of the structures and their corresponding usage.

This understanding can be useful for designers in developing new ways of designing spaces through offering interventions that engage the public. Designers should consider the following:

a) Spaces comprising multiple, complex, and even unpredictable interactions should not be underestimated or pluralized, where each designer-initiated change in relations could present different translations. As relations change and fluctuate, they could transform such translations differently.

This understanding shows how spaces can be reproduced to counter stabilities and create change, emergence, and processuality. Spaces can also counter certainty by introducing ambiguity, spontaneity, and improvisation (Sennett, 2009). Spaces can cater to different capabilities of co-functioning and give rise to an infinite number of associations between current and new elements and users.

This study’s exploration of a temporary intervention in Abdali Boulevard contributes toward an enhanced understanding of micro-level interactions occurring within multiple space–time frames that make up spaces and allow for different and ongoing ways of engagement. This investigation followed the relations concerned and provided insights into their differing nature in a specific setting. However, one limitation of this study was that, given relations are various and fragile, it was difficult to fully identify and elaborate on them using only one practical example. Furthermore, these relations are specific to the example and could differ in other cases or even between different interventions within the same case. This limitation creates an opportunity because it shows the particularity of each presented case and prompts consideration of different cases in future research, which is likely to extend the discussion on interactions, relations, and their effects on spaces and ways of engaging their users.