In the last two decades, urban geographers and historians have radically reconstituted their object of investigation. Recognizing that infrastructural patchworks, short-lived structures, momentary exchanges, and sensations hold the key to understanding political agency, urbanists have departed from prioritizing built form’s permanent inscription to focus on the impermanent traces of material culture. Yet few have considered the cumulative impact of a large number of “temporary” interventions operating within the “permanent” infrastructure of the city. What would it mean to describe, “map” as it were, the everyday and the contingent aspects of urban experience that involve multiple temporalities beyond the usual longer arc of urban studies?
In 2018, a group of us launched a research project titled Mapping Ephemerality that brings together GIS mapping tools, architectural analysis, ethnography, and urban history to explain and visualize the relation between the ephemeral/temporary and the perennial/permanent aspects of city-making. The first research investigation is a focused reading of an emergent set of spatial relations initiated by the festival of Durgapuja in Kolkata.
Hundreds of temporary pavilions—pandals—are built in Kolkata to facilitate the autumnal celebration of Durgapuja, worship of the goddess Durga. The temporary pandals work as urban “infill” within the perennial urban fabric, and are dismantled at the conclusion of the event. The most elaborate of these structures take as long as three months to construct, and entail year-long planning and fund-raising, whereas more modest interventions take as little as three days to a week to assemble. In terms of design they are strikingly innovative, and can put any architecture biennale to shame (Figures 1, 2 and 3).
Undertaken by community organizations and clubs in each para (neighborhood/locality), at core the festivities are place-based events that engage multiple publics and attract thousands of visitors. The short duration of this five-day event brings to the fore social relations and spatial affordances that might otherwise remain unnoticed, and generates new claims to space and recognition that would not be otherwise possible. This includes the capacity of a dense urban fabric to absorb the stress of road closure, infrastructural overload, and additional footfalls by stretching its existing infrastructure. Or it may be a lower-middle-class neighborhood claiming equivalence with its upper-class brethren through inventive design schemes made possible by the agency of clubs, celebrities, and political parties.
This project asks questions about location, emplacement, and distribution: What is the pattern of the pandals at the scale of the para, and at the larger scales of the ward (unit of municipal governance) and the city? How do these structures attach to, extend, modify the existing physical and social infrastructure? How is the territory of a puja pandal defined? What role does the seemingly permanent and temporary play in marking the territory of a pandal and a para? How does this align with the ward boundaries that are defined by the municipality as opposed to para limits that are fuzzy and changeable?
Last fall, along with thirteen research assistants from Jadavpur University, Kolkata, we documented the location, form, and access conditions of the pavilions across all 144 city wards using an ArcGIS platform with an app, Collector. Each pandal was given a unique identification number and three types of data were collected for 1,983 pandals—location, emplacement, and access—using the three Collector features of points, areas, and lines. A color-coded distinction was made among pandals on streets (in orange), green islands (in purple), parks (in green), and vacant lots (in blue). These were augmented with photographs and field notes, embedded in the Collector data. In addition, we conducted focused investigation of fourteen individual sites that involved interviews with people in the neighborhood, and measured drawings of the site showing plans, street edge conditions, and urban context.
The city-wide data on pandal location shows a discernible pattern of distribution and dispersal (Figures 4 and 5). Not surprisingly perhaps, there is a direct correlation between the sites on which these pandals are constructed and the road infrastructure. Most of the pavilions in the older parts of the city in the north are on streets. The same pattern holds for the established southern wards of the city, although these contain more pavilions in parks and open green spaces. Most of the pandals in the southern suburbs, incorporated recently into the metropolitan limits, are on “vacant lots”: an array of unbuilt and found spaces, as well as forecourts of community halls, schools, and temples.
The pandals, irrespective of size and elaboration, are products of a shared building practice attuned to the specificities of the urban locale. The vast majority of these pavilions are savvy interpretations of the surrounding buildings and sidewalk infrastructure. Bamboo poles are tied to columns and railings of adjacent buildings for structural stability; verandahs and stoops (ro’ak) become in-built platforms to house the deity; and the width, length and volume of the street set the parameters for the form of the pavilion (Figures 6 and 7). In this the pandals share morphological and temporal characteristics with the myriad ephemeral structures and “encroachments of the ordinary” that define urban public space in cities in much of the world: food carts, hawker stalls, billboards, street games, and gatherings.
There is much to learn from all this. The pattern and mode of construction of pandals in public space, constituting a stupendous bottom-up mobilization of city life and urban community, helps us understand not simply the geography of the festival but also a whole set of allied relations that depend on canny recognition of the relation between the ephemeral/temporary and the perennial/permanent. The very potential of the ephemeral, however, also poses challenges to fieldwork in terms of duration and representation. We become aware of those aspects of urban materiality that resist representation: of what can and can’t be mapped.
 See for example, Ash Amin and Nigel Thrift, Cities: Reimagining the Urban (New York: Wiley, 2002); AbdouMaliq Simone, For a City Yet to Come: Changing African Lives in Four African Cities (Duke, 2004); Christine Bonnemaison and Christine Macy, eds., Festival Architecture (London: Routledge, 2008); James Holston, Insurgent Citizenship: Disjunctions of Modernity and Democracy in Brazil (Princeton: Princeton University Press, 2009); Jeffrey Hou, Guerilla Urbanism and the Remaking of Contemporary Cities (London: Routledge, 2010); Swati Chattopadhyay, Unlearning the City: Infrastructure in a New Optical Field (Minneapolis: University of Minnesota Press, 2012); Saskia Sassen, “Complex and Incomplete: Spaces for Tactical Urbanism,” in Pedro Gadhano, ed., Uneven Growth (New York: MOMA, 2014), 40-47; Andrew Herscher, Displacements: Architecture and Refugee (New York: Sternberg Press, 2017); Arijit Sen, “Discarding Corb’s Shoes: Marginal Voices and Local Histories from the Urban Edge,” in Manu Sobti, ed., Chandigarh Rethink (San Francisco: ORA Editions, 2017), 65-75; AbdouMalique Simone and Edgar Pieterse, New Urban Worlds: Inhabiting Dissonant Times (New York: Wiley, 2017); Amit Rai, Jugaad Time: Ecologies of Everyday Hacking in India (Durham: Duke University Press, 2019).
 For documentation of temporary infrastructure constructed for a religious festival under the aegis of the state, see, Rahul Mehrotra and Felipe Vera, eds., Kumbh Mela: Mapping the Ephemeral Megacity (Ostfildern: Hatje Cantz Verlag, 2015). The geography, infrastructure, and pavilion designs of Durgapuja in Kolkata are fundamentally different from that of Kumbh Mela.
 Swati Chattopadhyay is the PI for this project conducted by the Calcutta/Kolkata Co-Lab consisting of Renee Chow, Arijit Sen, Tania Sengupta, and Jeremy White.
 The para (neighborhood) does not have an administrative/jurisdictional boundary, but constitutes a conventional understanding of locality that is defined by a sense of community, use, and affect. One ward typically contains several paras. For more on the para see Swati Chattopadhyay, Representing Calcutta: Modernity, Nationalism, and the Colonial Uncanny (London: Routledge, 2005).
 The group of research assistants in Kolkata consisted of Ankana Das, Pradipta Das, Nazmul Hoque, M.D. Mahasweta, Meghna Roy, Sounita Mukherjee, Saptadipa Nandy, Suchi Priyadarshani, Aninda Roy Ahmed, Aparajita Sarkar, Shreyoni Singharoy, Aaheli Sen, Shalini Sengupta. At UC Santa Barbara Mallorie Chase, Somak Mukherjee, Adriana Ocasio, and Thomas Crimmel served as research assistants.
 We documented only those pandals that occupied public space and were accessible to the public. This did not include the celebrations held within apartment complexes or in individual households.
 Asef Bayat, “Un-civil society: the politics of the ‘informal people’,” Third World Quarterly 18, no. 1 (1997): 53-72.