Platform or infrastructure or both at once? Detangling the two concept's knotty cross-articulations

Introduction

What is the meaning of infrastructure and platform in today’s digital landscape? How are these two concepts related to each other? The standard information practice when puzzling over a term or concept is to consult an encyclopaedia. This practice survives, of course, but books are no longer the dominant medium. Reach for the obvious digital version, which in Sweden is the Swedish National Encyclopaedia at www.ne.se, and the question follows us. Is it a platform or infrastructure? Perhaps it is both. Its entries for infrastructure and platform, translated using Google Translate, another platform-cum-infrastructure, are as follows:

a system of constructions and their operation, which forms the basis for the supply and the precondition for the production to function: i.e. roads and railways, airports and ports, power plants, electricity distribution and other energy supply devices, water and sewage systems, telecommunications and education. (NE, s.v. 'infrastruktur', 2022)

flat structure for passengers getting on and off at the railway station. (NE, s.v. ‘plattform’, 2022)

These definitions give little insight into the digital realm, however, and an encyclopaedia definition will not suffice when writing an academic paper. Instead, we need to engage with the scholarly community, and this entails finding related research. A well-established information practice here is to seek assistance from a university library, which like all libraries is a knowledge infrastructure—‘robust networks of people, artifacts, and institutions that generate, share, and maintain specific knowledge about the human and natural worlds’ (Edwards et al., 2013, p. 5)—while related research can also be looked up using Google Scholar. As scholars, though, we can also get recommendations from colleagues. This happens informally in various scenarios such as coffee breaks, but more interesting to us here is how this information is mediated digitally. At 05:53 on 29 December 2021, one author of this paper received an email with the heading ‘[Professor N or M] recommended this article’. The email had been automatically generated by ResearchGate (2022), a platform started in 2008 to address concerns about the way science is created and shared. This one email amounted to the platform using established infrastructures (email and research communities) to communicate to users based on other users’ data and actions.

Both platforms and infrastructures, then, are entrenched in scholars’ information practices, and it is difficult to define where one ends and the other starts. Libraries have been discussed as both platform and infrastructure. Bowker et al. (2019, p. 1) describe classifications in libraries as ‘thinking infrastructures’, which ‘structure attention, shape decision-making, and guide cognition’. Libraries also demonstrate platform qualities, as ‘by becoming a platform the library can better fulfill the abiding mission it set itself: to be a civic institution essential to democracy’ (Weinberger, 2012, p.36). To cross- articulate infrastructure and platform is to understand the concepts relative to each other, thus addressing the concerns relevant to library and information studies.

Digital media, in a broadest sense, is part of everyday life, facilitating day-to-day tasks, communication, and access to information. Infrastructures are often described as ‘invisible’ until they break down. We tend to be oblivious of the vital role of electricity until there is a blackout. And when Facebook went down on 2 October 2021, it became apparent that the platform had acquired such infrastructural properties. The Economic Times (Facebook…, 2021) reported ‘In Mexico, politicians were cut off from their constituents. In Turkey, shopkeepers could not sell their wares. And in Colombia, a nonprofit organisation that uses WhatsApp to connect victims of gender-based violence to lifesaving services found its work impaired.’

In recent years, scholars in media and communication studies have investigated the entrenchment of digital media using the two concepts of platform and infrastructure in conjunction, stressing the need for further theoretical insights into the dynamics between them (Constantinides et al., 2018). Information scholars have drawn on this theoretical framework to explain phenomena in the contemporary digital landscape, such as online search engines (Haider and Sundin, 2019). Some have highlighted the need for conceptual frameworks to better understand the conditions social media impose on our production, use and sharing of information (for example, Acker, 2020; Haider et al., 2012); others, that cross- articulation is the key to understanding the contemporary digital landscape (for example, Plantin et al., 2018). However, cross-articulations diverge. The relation between platform and infrastructure is defined differently depending on the setting and scope: the concepts have been discussed in terms of the ‘infrastructuralization of platforms’, the ‘platformization of infrastructures’ (Plantin et al., 2018), and ‘platform-as-infrastructure’ (Helmond et al., 2019). So how are infrastructure and platform related to each other in these cross-articulations? What properties do they emphasise, and most important, how do the concepts’ cross-articulations differ?

We would suggest it is better to proceed by detangling the knotty issue of diverging definitions, shedding light on what is at stake in the cross-articulations, and thus enabling nuanced conceptualisations of the role of digital media in everyday life. Cross-articulations can assist in developing conceptual tools and frameworks to understand digital media’s information structures and the associated activities and practices involved in producing, utilising, sharing and retrieving information. This paper, therefore, provides a survey of the literature on recent cross-articulations of infrastructure and platform to gauge how the concepts are cross-articulated by different authors and given new content in the light of contemporary information technologies. By focusing less on pinning down definitions and more on examining various cross-articulations, the paper contributes an information perspective to the debate about infrastructure versus platform, while confirming the relevance of the concepts’ cross-articulations to library and information studies.

Theoretical background

The concepts of infrastructure and platform both have dual origins. Infrastructure has been theorised through a historiographical lens, which emphasises large-scale structures such as the railway system (Goddard, 1994) or electricity generation and distribution (Hughes, 1983). It has also been discussed in terms of how ‘individuals and communities meet infrastructure’ (Bowker & Star, 1999, p. 33), applying methods such as ethnography to the study of infrastructures as social phenomena, which inevitably stresses the relational perspective—the dynamics between local user practices and formal or institutional structures. In information studies, a body of infrastructure research exists for topics such as information systems (Star, 1999; Star & Ruhleder, 1996), standards (Bowker & Star, 1999), the web (Sandvig, 2013), information literacy (Haider & Sundin, 2019), knowledge in work settings (Huvila, 2009), education (Sundin & Carlsson, 2016), libraries (Centerwall & Nolin, 2019), and research structures (Anderson, 2013; Borgman et al., 2019) among others. Infrastructure can be understood through a practice-discursive lens (for example, Juneström, 2019), where the interactions between local practices, standards and systems generate a transparent infrastructure sustained over a more extended period in a specific setting, and thus is routinised as part of everyday life (Haider & Sundin, 2019, p. 4). Due to its naturalised form, infrastructure is rarely apparent until it breaks down, rendering the everyday tasks or practices it supports suddenly difficult or impossible to perform (Star & Bowker, 2002).

Platform conceptualisations, meanwhile, address the relations between the technical artefact’s layers of hardware and software and the socio-cultural perceptions associated with it (Bogost & Montfort, 2008, 2009), emphasising the material dimensions where the underlying mechanisms, properties, and design processes matter most (for example, Friedhoff, 2014). In media studies, the platform concept has been developed in pace with the study of game consoles (Montfort & Bogost, 2009) and general computing (Maher, 2012), while extending towards other technical artefacts such as social media with enquiries into the platform’s political-economic dimensions (for example, Gillespie, 2010) and user data generation processes (for example, Gerlitz & Helmond, 2013). The platform concept stresses characteristics such as programmability and modularity, which are what enables any platform to adapt and evolve (for example, McKelvey, 2011). These characteristics enable connectivity, where the platform expands by connecting to other technical instances such as applications, websites and other platforms via its application programming interface (Helmond, 2015; see, for example, Van Dijck, 2013). The concept’s focus and scope have been criticised for ignoring the user perspective. Apperley and Parikka (2018) point out that the material analysed in platform studies often originates from official sources and prioritises publicly available materials from the platform company or media, but disregards user practices and the technical artefact’s role in everyday life (p. 354). Consequently, the platform concept can assist in making sense of a platform as a technical artefact and business enterprise, even as its dynamics with various actors, such as the user, are frequently overlooked.

Broadly speaking, these are the understandings and assumptions inherent in the concepts of infrastructure and platform, highlighting each concept’s unique properties and diverging applications. Similarly diverging operations are to be found in the cross-articulations of infrastructure and platform, to which we will now turn.

Cross-articulations found

Cross-articulation is defined as a concept or context which is read or understood relative to another. While articulation concerns ‘the breaking of wholes into elements’ and ‘circumscribing an object from a bigger whole and distinguishing different parts within the constitution of the object’ (Rünkla, 2015, p. 173), cross-articulation intersects to develop new forms of categorisations and distinctions (Blacksell, 2015, p. 84). The present survey of the scholarly literature was designed to explore recent conceptualisations, the sample criteria being intentionally broad to capture variations in the cross- articulations, but limited to papers published in the last 10 years that treated the concepts of platform and infrastructure relative to each other. In the resultant corpus we identified four ways of cross- articulating the two concepts: (i) platform as infrastructure; (ii) infrastructure as platform; (iii) dual processes of infrastructuralisation and platformisation; and (iv) dynamics between platforms and users. The first three emphasise process and the fourth practice.

Platform as infrastructure

The first cross-articulation found was the platform as infrastructure: a process where the platforms acquire infrastructural qualities such as scale, become indispensable, and yet retain certain platform characteristics. Plantin and Punathambekar (2019) exemplify infrastructuralisation with companies such as Google, Facebook, and Amazon, which have transitioned into ‘vital infrastructures in the world at large’ (p. 164). They argue the two concepts must both be considered, because ‘an integrated approach helps us grasp how platform-level features (such as participation, programmability and modularity) and those of infrastructures (such as scale ubiquity and temporality) together constitute hybrid digital artifacts’ (p. 171). Infrastructuralisation also advances qualities such as invisibility, where the platform becomes entrenched in everyday life to the point where we take it for granted. Pujadas and Curto-Millet (2019) situate infrastructuralisation in the sharing economy, arguing that the tendency of platform owners to frame platforms as ‘neutral, clearly bounded, digital infrastructures’ (p. 276), merely acting as mediators between different actors (such as academics) downplays the platform owners’ power. Uber is taken as an example of how platforms work to be absent while being the nexus of activity. This positioning is a political act since it reduces the platform owners’ role to one of matchmaker, when in fact their algorithms do far more, playing ‘an agentic role, that is, they impact on decision-making’ (p. 283).

Another aspect of infrastructuralisation is how platforms transition into infrastructures by integrating with external resources and existing infrastructures. Helmond et al. (2019) discuss infrastructuralisation by tracing the evolution of Facebook from a social networking site to a ‘platform-as-infrastructure’. Their point is that platforms allow third-party developers, such as marketing agencies, to ‘build “on top of” a platform’s core infrastructure’ (p. 127), complementing the platforms’ own products and services. In this way, though not being infrastructural at launch, social networks have ‘gained infrastructural properties over time by accumulating external dependencies through computational and organisational platform integrations’ (p. 141). According to Plantin (2018), Google Maps is another example of a platform transitioning to infrastructure. Cartography is traditionally an essential service, and while it remains so, ‘the rise of the Web has promoted an alternative configuration for mapping: cartography as platform’ (p. 490). Google Maps has not replaced traditional infrastructures because it relies on them for several of its components; however, by latching onto existing infrastructures, platforms such as Google Maps have ‘reached a scale and social status that was previously attained only by knowledge infrastructures’ (ibid).

Knowledge infrastructures are ‘robust networks of people, artifacts, and institutions that generate, share, and maintain specific knowledge’ (Edwards et al., 2013, p. 5), which leaves platforms to latch onto existing knowledge infrastructures’ resources, embedding them into the platform to scale up to infrastructurality. Plantin et al. (2018a) discuss infrastructuralisation relative to knowledge infrastructure research to see how web-based platforms alter scholarly communication infrastructures, and find that platforms are not simply taking over, but can become ‘bridging entities’ with the ability to ‘re-integrate scattered actors and institutions of the scholarly infrastructure’ (p. 5). However, they note that platforms also pose risks as they ‘amplify the existing tendency of infrastructures to splinter, with their more profitable functions gradually pruned off and privatised’ (p. 12). Data-sharing platforms enter a field traditionally organised along infrastructure lines, because of the de-integration of existing academic infrastructure. With the development of the internet, technologies and cultures emerged to challenge the existing, black-boxed, scholarly infrastructure by rendering it visible, which opened the way for platforms to assume an essential role in data sharing.

Infrastructuralisation also connects to the mobilisation of new and existing data where these resources are folding into the platform structures. Langlois and Elmer (2019) find platforms are becoming infrastructures helped by the intense and extensive datafication of everyday life. Social media corporations transition from enclosed platforms that see individual users as valuable sources to impersonal infrastructures that aggregate and mobilise large sets of data, created intentionally and unintentionally from users, non-users, humans and non-humans (p. 238). The platforms move beyond individuation towards ‘connecting heterogeneous systems and networks’ (p. 239), embedding themselves in everyday life by taking control of material aspects of life rather than specific online spaces (p. 241). Hunger (2018) examines platforms as infrastructure by mapping how platforms and database infrastructures generate value. Platforms are understood as database-driven, and the emerging mode of production and data extraction depends on database infrastructure. According to Hunger, electronic databases can be narrowly defined as ‘infrastructures for the structured storage of information’ (p. 53). Databases do not exist independently, but consist of different ‘infrastructural stratifications’ (ibid). When searching, buying things, or recording movement (for example, using training apps), users voluntarily and involuntarily produce transactional metadata that platforms can harvest. From the perspective of the platforms, we become ‘dividuals’, a scattered form of subjectivity, instead of individuals, because each platform has a limited perception only of who we are. For example, ResearchGate and Facebook have different perceptions of us as (in)dividuals. In other words, fragmented dividual parts of agn individual are what the platform databases store and use to produce value in a new capitalist regime where ‘human labor embodied in data in microscopic doses adds valueg’ (p. 62). Microscopic doses of human labour entail recommending a paper through ResearchGate, sharing a news article on Facebook, or endorsing someone’s skills on LinkedIn. Computational capital, according to Hunger, is ‘the disposition over data and computing infrastructure’ (p. 60), and the production and extraction of data depend on database infrastructure. Platforms become infrastructures by adopting similar qualities such as invisibility as they become embedded in our daily routines. At the same time, infrastructuralisation also entails platforms scaling up by means of extensive, intensive data production and aggregation, connecting the platforms to existing infrastructures.

Infrastructure as platform

The second cross-articulation found was platform and infrastructure discussed in terms of infrastructures adopting a platform logic and being platformised. Helmond (2015) defines platformisation as ‘the rise of the platform as the dominant infrastructural and economic model of the social web and the consequences of the expansion of social media platforms into other spaces online’ (p. 5), and pointing out that social media platforms strive to extend into the rest of the web and make external web data ‘platform ready’ (p. 4) by applying programming interfaces, reformatting the web according to social media logic. Platform infrastructures reconfigure external data to fit the platform’s agenda. Application programming interfaces are, according to Helmond, ‘part of the material infrastructure of social media platforms’ (p. 4). In a similar vein, Nieborg and Poell (2018) highlight platformisation using examples from the cultural industry such as news and games, where ‘platformisation can be defined as the penetration of economic, governmental and infrastructural extensions of digital platforms into the web and app ecosystems’ (p. 4276). Infrastructures are platformised by adapting to a platform logic of how cultural content should be produced and distributed in order to participate in the platforms’ extensive collecting of user data. One example from the field of archive, library and museum studies is K-samsök or Swedish Open Cultural Heritage (SOCH), a web service that allows cultural heritage institutions to make their data available to a broader audience. SOCH is described as ‘a junction box between the institutions’ databases and actors who want to use the information in their own applications (for example, on websites or in mobile solutions)’ (Riksantikvarieämbetet 2020).

Haider and Sundin (2019), in studying search engines, suggest that information infrastructures are platformised so the concept of infrastructure might not suffice ‘to fathom the role and function of these dominant search engines, where economic power, social control, and material structures work together to concentrate control over information’ (p. 58). They find the characteristics of platform and infrastructure cannot meaningfully be teased apart. The concept of the multisided platform is proposed to describe ‘companies whose business model is based on enabling interactions between at least two partners, for instance, users and advertisers, software providers and consumers’ (ibid). Platformisation is better understood as a process of reconfiguring existing infrastructures, alteringg how information is produced, distributed and circulated in ways aligned with a platform logic. Cultural heritage institutions are thus part of an existing infrastructure that, for example, SOCH both endorses and challenges. Developers can build e-services for different target groups using the platform’s open API, thus extending the reach of archives, libraries, and museums; but those with portals to the system are thus dependent on the platform owner, which in the case of SOCH is the Swedish National Heritage Board.

Infrastructuralisation and platformisation

The third form of cross-articulation to be found conceptualises platformisation and infrastructuralisation as occurring simultaneously or even entangled in each other. Constantinides et al. (2018) argue that the two processes are undertaken in such a way as infrastructures and platforms adopt each other’s properties and characteristics simultaneously. For example, a platform becomes more similar to infrastructure when cementing itself as something crucial. Google Docs, which the authors’ used to prepare this paper, is an example of a platform that many researchers rely on, especially when writing in collaboration, if it can be perceived as an infrastructure. Google Docs is not an official digital resource at our university, and were it to malfunction there is no guarantee any text could be recovered, and yet many of us assume it will not happen. Similarly, Helmond et al. (2019) suggest that infrastructuralisation and platformisation are described as ongoing processes in interplay:

While platformisation speaks to Facebook’s growing capabilities to mediate the interactions between multiple stakeholder groups and their diverging needs and interests, infrastructuralisation speaks to Facebook’s growing ubiquity by embedding itself in other markets and industries to render technical and business operations more widely and immediately available. (Helmond et al., 2019, p. 141)

Although each process foregrounds different aspects of economic and technological growth, the authors note that the processes are engaged simultaneously. Facebook enacts both platform and infrastructural qualities, integrating and enveloping existing infrastructures. De Kloet et al. (2019) hold that platformisation and infrastructuralisation are neither uniform nor unidirectional processes, but arise locally and simultaneously (p. 254). Taking China as their example, they note that private platforms and public infrastructures are merging, because ‘the boundaries between platforms and public infrastructures have become increasingly porous’ and ‘platforms are more and more a part of our everyday lives’ (p. 251). Similarly, Plantin and de Seta (2019) investigate the Chinese app WeChat to demonstrate how infrastructuralisation and platformisation proceed. They argue that platforms such as WeChat gain infrastructural properties such as scale and ubiquity wherever information and interactions are centralised and, enclosed inside the platform (p. 4). The platform replaces or even competes with vital services in everyday life economically, administratively, and socio-culturally, simply by ‘rising to an infrastructural standard’ (p. 8).

Infrastructuralisation and platformisation can also be understood in the dynamics between private and public. Plantin et al. (2018b) uses Open Web, Facebook, and Google as their empirical examples to investigate the technological developments which have supported the dual processes, while changing the dynamics between public and private:

digital technologies greatly assisted in this process, often making possible lower cost, more dynamic and more competitive alternatives to governmental or quasi-governmental monopoly infrastructures, in exchange for a transfer of wealth and responsibility to private enterprises. (Plantin et al., 2018, p. 306)

Plantin et al. thus understand infrastructuralisation and platformisation as being engaged simultaneously, their point being that platform-based services obtain infrastructural qualities and become indispensable in everyday life. They underline that existing infrastructures adapt to platform logic because of adaptive characteristics such as modularity and programmability, which makes data generation possible in a manner similar to platforms. ResearchGate (2022) strives to offer ‘a place to share your work and connect with peers around the globe, traversing the borders and silos of science’, in contrast to existing infrastructures, all while their business model relies heavily on the same infrastructures. Gehl and McKelvey (2019) connect the private–public dynamic in infrastructuralisation and platformisation with darknet phenomena, taking the differences between the concepts to be one of standards: ‘infrastructures tend toward relations based on open standards, while platforms depend on proprietary application programming interfaces’ (p. 221). Infrastructures are defined as ‘modernist state interventions’ and platforms as a ‘postmodern pastiche of similar products arranged in a marketplace accessible by private interfaces’ (ibid). In terms of the darknet, infrastructures and platforms are noted as both ‘privatising and publicising in their relations’ (p. 227). Darknet utilises the public internet’s architecture and establishes private communication channels, while also forming networks of personal computers to enable sharing of bits and bandwidth (pp. 222–224).

Dynamics between platforms and users

The fourth category of cross-articulation emphasises the relations between different actors through a practice-discursive lens. In this category of cross-articulations, authors have primarily focused on how digital technology and media impacts users. Alaimo and Kallinikos (2019) highlight the impact of social media on our sociality as they ‘provide the backstage operations and technological facilities out of which new habits and modes of social relatedness emerge and diffuse across the social fabric’ (p. 289). In this perspective, sociality is being standardised or infrastructured through platforms, since they offer limited sets of social actions users can perform. Infrastructuring is understood as a practice undertaken by platforms to establish new forms of social interactions, generating ‘calculable and machine-readable data footprint out of user platform participation’ (p. 289). The platforms’ ability to alter behaviour is also highlighted by Bowker et al. (2019), with examples of platform components such as algorithms or ranking and rating systems that alter material and social infrastructures (p. 8). Platforms are described as ‘investment in form and organisation’ (p. 1), infrastructuring patterns of (visible) human interaction, which can be valued, traced and governed. Platforms such as Uber do more than merely matching or mediating; they ‘render visible, knowable and thinkable complex patterns of human interaction in and out of the market, in feedback loops of learning, reformatting and redoing’ (p. 1). Lambotte (2019) offers the different example of social media analytics, which work in similar ways because their tools are ‘objects to think of our digitally mediated interactions with’ (p. 308), enabling calculation and visualisation of actions in the form of data taking place on platforms. The platforms’ agentic dimension and technical components are understood as undertaking the practice of ‘infrastructuring’, moulding user activities and interactions into standardised forms of actions, thus making platform data generation and aggregation possible.

Another aspect of infrastructuring involves the dynamics between the local infrastructures and global standards. Mohan and Punathambekar (2019) define the practice of infrastructuring as connecting the local to the global, so the implementation of global platforms such as YouTube is possible in local infrastructures. The local and global standards mutually impact on one another: local conditions and resources adapt to global (Western) standards, while local infrastructure can also drive platform development. In YouTube’s case, local circumstances lead to the development of functions such as offline viewing, enabling users in places with unstable internet connections to access the platform. As noted, infrastructuring is linked to standardisation, which further relates to platform data generation and aggregation. However, differences can be noted in how the relations between platforms and users are portrayed, where the dynamic is considered a unilateral process (for example, platforms impacting users) or mutually constitutive (for example, platforms and users affecting one another).

Discussion and concluding remarks

The several forms of cross-articulations found in the corpus can be divided into the categories of process and practice. Cross-articulations of the two concepts can be understood as processes, namely infrastructuralisation and platformisation. The processes are discussed separately and as a dual process, where the two concepts adopt each others’ properties and characteristics. It is explained how platforms grow to become infrastructure, while infrastructure adapts to platform logic. Aspects involved in these processes include scale, invisibility, and rendering itself indispensable. For example, the infrastructuralisation of platforms equates to platforms scaling up to infrastructural standards and presenting themselves as neutral facilitators of connectivity (for example, Pujadas and Curto-Millet, 2019, p. 276).

Platformisation, meanwhile, is the reconfiguring of traditional, existing infrastructures according to platform logic in order to stay relevant. For example, Langlois and Elmer (2019) concentrate on platformisation’s alterations according to information flows, changing traditional media and information infrastructures such as news and how it is produced, distributed and circulated. Together, these processes alter the dynamic between public and private, which in cases such as the Chinese app WeChat intersects governmental and business interests (Plantin and de Seta, 2019, p.3). Other examples include proprietary enclosed structures merging with existing public infrastructures such as Google Maps latching onto cartographic infrastructural resources (Plantin, 2018). Ultimately, any debate about the public–private dynamic revolves around a perspective on the distribution of power, or as Edwards et al. (2013) stress, ‘All infrastructures embed social norms, relationships, and ways of thinking, acting, and working. As a corollary, when they change, authority, influence, and power are redistributed’ (p. 23). Lastly, a dimension of power is identified among the corpus as the platforms, gaining momentum as they scale up to infrastructural forms, connect to the increasingly intensive datafication of everyday life and the extensive tracking of our daily actions.

Another strand of cross-articulation is evident through a practice-discursive lens. For example, platforms practise infrastructuring on their users, shaping fundamental aspects of human life involving our sociality (Alaimo and Kallinikos, 2019) and thinking (Bowker et al., 2019; Lambotte, 2019). Further, platforms can standardise our actions and activities, rendering it possible to extract platform data. For instance, Alaimo and Kallinikos (2019) points out how platforms like Facebook reduce our sociality to a specific set of interactions such as the like button. Meanwhile, understanding platforms and infrastructures through a practice-discursive lens allows for insights into how platformisation and infrastructuralisation are sustained by various actors involved, such as social media companies, government bodies, third-party actors in advertising and marketing, and, last but not least, users. For example, according to Plantin (2018), the public maintains Google Maps as a cartographic infrastructure, because Google has immense power to shape and circumscribe the forms and level of participation. The relation between platform and users is discussed here in a unilateral form to focus on the platforms’ impact on users. Mohan and Punathambekar (2019) stand out among the cross-articulations, because they discuss the interplay between local infrastructures and global platforms where they are mutually constitutive. They make a compelling argument that platforms also have to adapt to certain circumstances on a local level, touching on a key tension between global platforms and local structures.

We would agree with Plantin et al. (2019) that cross-articulations of platform and infrastructure can assist in grasping the contemporary digital landscape. Although cross-articulations have primarily been discussed by media and communications scholars, we find they are also relevant to library and information studies as they offer a conceptual entry point into contemporary information infrastructures and practices, aligned with the increasingly intensive, extensive datafication of everyday life. Platformisation and infrastructuralisation are not straightforward processes, especially as their technological and political levels become ever more entangled with the micro- and macro-historical levels. These dimensions have to be acknowledged when investigating a specific, local subject such as a public library. Further, we note that user perspectives are often overlooked, because cross-articulations focus on the platforms’ impact on existing socio-cultural structures. For future conceptualisations, we would suggest a stronger emphasis on the relational, namely the dynamics between the platforms and users. We need to include a user, practice perspective when investigating the question of power, if only to shed light on how users negotiate contemporary platforms and infrastructures.

Acknowledgements

We would like to thank the paper’s anonymous reviewers for their helpful comments and suggestions.

About the authors

Shirley Chan is a doctoral student in library and information studies at the Division of Archives, Libraries and Museums and Digital Cultures at Lund University, Sweden. Her doctoral research concerns the preservation challenges of digital traces produced between online fan communities and social media platforms, exploring the dynamics between contextuality and ephemerality. She can be contacted at Division of ALM and Digital Cultures, Box 192, 221 00 Lund, Sweden, shirley.chan@kultur.lu.se
Ann-Sofie Klareld is a senior lecturer at the Division of Archives, Libraries, and Museums and Digital Cultures at Lund University, Sweden. She has a PhD in Archives and Information Science from Mid Sweden University. She publishes on digital archiving in the public sector, and on contemporary family history research. She can be contacted at Division of ALM and Digital Cultures, Box 192, 221 00 Lund, Sweden, ann-sofie.klareld@kultur.lu.se

How to cite this paper