Below the Radar: Private Groups,Locked Platforms,and Ephemeral Content—Introduction to the Special Issue

Abstract

This special issue of Social Media + Society originates from the first AoIR Flashpoint Symposium, entitled “Below the Radar: Private Groups, Locked Platforms and Ephemeral Content.” The aim of this conference was to investigate platform-driven changes and emerging practices of everyday-life content production occurring “below the radar” of internet research, or outside of previous standards of data visibility and accessibility on which most internet studies have been based over the last decade. In the current context, online spaces seem to be heading toward more circumscribed and unsteady forms of publicness, which contrast with the platform affordances upon which the theorization of networked publics has been built. Private groups, locked platforms, and ephemeral contents are some of the challenges that require the development of new perspectives and research tools capable of adapting to this shifting environment. In this introduction, we will illustrate how the theme of “below the radar” has evolved since the initial call thanks to the confrontation with the researchers who participated in the conference, and this special issue, and we will introduce the nine articles that make up the collection. These articles, which combine different research disciplines and techniques, provide a map of some of the most urgent theoretical, ethical, and methodological issues concerning the current transformations of the visibility regimes of online social action.

Introduction

In 2008, danah boyd suggested that permanence, replicability, scalability, and searchability were the four significant properties in configuring networked publics (boyd 2008). Said factors have long been essential tools through which scholars have thought about the affordances of online social networks. These properties have been subsequently expanded, criticized, and discussed by numerous contributors (Bucher & Helmond, 2018; Costa, 2018; Treem & Leonardi, 2013) and the author herself (boyd, 2014). Nevertheless, this idea of describing the social media-centered internet as a whole through aspects of its publicness has become a founding model for a great deal of work in internet studies over the last 15 years. Similarly, it has fostered the assumption that social media provide a more accessible and concentrated version of human action than offline research contexts (Giglietto et al., 2012).

About a decade after this first phase of social media research, these properties still apply to much of the content produced and disseminated online. However, we are also witnessing a multiplication of digital spaces and online behaviors that move toward more situated, unstable, and unpredictable forms of publicness.

The growing use of group chats in messaging apps shows a new geography of unsearchable “small conversations” (Boccia Artieri, 2017), which many have recently tried to define as “dark social media” (Swart et al. 2018), “meso-newspace” (Kligler-Vilenchik & Tenenboim, 2020), or “crypto-publics” (Johns 2020). The shift from permanence as a default condition for content to the multiple temporalities of stories and video streams has raised the issue of a possible ephemeral turn in digital media (Haber, 2019). Recent platform lockdowns have led some authors to talk about a “post-API age” (Freelon, 2018; Perriam et al., 2020) or “APIcalipse” (Bruns 2019) and have made social media contents more difficult to scale and replicate, especially for researchers (Walker et al., 2019).

The attempt to define the general connotations of social media’s affordances and practices—even if only by focusing on their mainstream dimension—is becoming a less and less practicable research endeavor. This is due to the increasing plurality of online contexts and the fact that various obfuscation tactics have become common among users’ repertoire of actions, next to the most studied practices of visibility. The assimilation—by protest movements, extremist groups but also by average users—of methods to escape censorship and being banned (Rojas-Galeano, 2017), along with the widespread use of VPNs, the algorithmic opacity of companies and institutions, and the use of memetic idiolects as forms of cultural cryptography, indicate that online social action takes increasingly place below the radar of research.

This observation gave rise to the theme of the first AoIR Flashpoint Symposium, entitled “Below the Radar: Private Groups, Locked Platforms and Ephemeral Content,” which was held in Italy at the University of Urbino “Carlo Bo” on 24 June 2019. With the Flashpoint Symposia, the AoIR wanted to create opportunities for intensive study of specific themes and geographical areas to complement the extensive approach of the annual conference. The aim of the first event—from which this special issue of Social Media + Society originates—was to investigate those platform-driven changes and everyday practices that occur “below the radar.” By this, we meant changes and practices outside the previous standards of data visibility and accessibility on which most of the internet studies have been based over the last decade and beyond those platform affordances upon which the theorization of networked publics has been built.

The central theme of the symposium was defined by three key terms: “private,” to indicate the set of online spaces and processes that redefine the idea of publicness of online action, “locked,” to designate the limitations in the accessibility of data obtainable from platforms, and “ephemeral,” to indicate the reconfiguration of the expected persistence of online data. The three event panels and one poster session took place around these three nodes. At the opening and closing of the panels, we had the privilege of hosting two keynote speakers, Crystal Abidin and Rebekah Tromble, who made an essential contribution in clarifying and extending the stakes of the original theme.

The range of themes, approaches, contexts, and theoretical perspectives went beyond our wildest expectations. In 19 articles, we touched on a wide variety of national contexts—China, Singapore, UK, Germany, Italy, Netherlands, Israel, Russia, US, Sweden—, platforms—WhatsApp, Telegram, WeChat, Facebook, Twitter, Instagram, 4chan, Spotify, Gab—and research methodologies—ethnography in digital and physical places, qualitative interviews, computational methods, theoretical speculation, research-action, cultural criticism.

The debate among the participants in the conference and, subsequently, the articles that make up this special issue have transformed what was mainly intended as a theme to convey papers into a “sensitizing concept,” capable of raising researchers’ awareness toward a precise research perspective. There is no ambition here—nor is there any need to feed the already extensive list of “turns” and “studies.” However, we think that the “below the radar” perspective is necessary to point out that the construction of (a) the invisibility of behaviors, (b) the inaccessibility of data, (c) the misunderstandings of languages, (d) the temporariness of contents is a constitutive element of the digital environment, instead of a “privation,” an “obstacle,” or a “denial” of the condition of publicness experienced as the state of normality within social media.

Even more so, the discussion between researchers has shown how partial our initial approach was. While our first question was “what are the current blind spots in internet research?,” it soon emerged that the first blind spot is to think of these problems from a research-centric perspective. That is, when we talk about “below the radar,” we must also consider how the line between what is observable and what is non-observable involves a wide variety of “radars” in addition to those used by researchers, namely the “radars” of users, institutions, companies, movements, platforms, and AI systems. This is a problem that can be framed, in the manner of the cybernetic theory (Luhmann, 2012; von Foerster, 2003), as a leap from a “first-order observation” to a “second-order observation” that moves away from the pretension of occupying an essentialist point of view whereby data simply “exists” for a privileged audience of researchers and starts observing the multiplicity of actors who observe data or, rather, build data through their observation operations.

Second, a need has emerged to understand what we mean when we talk about “private,” “locked,” and “ephemeral” as properties of the current digital environment. It is clear, if not apparent, that even these characteristics can present themselves in opposite ways depending on the observer. A conversation is private until we suspect the presence of intruders (or that it can be shared through screenshots to others); the data of a platform can be locked for certain users but not for other interest groups; an Instagram story is ephemeral for its audience but not necessarily for its creator. We therefore hold that we can identify a material dimension of these features where they are observed as affordance by the observer: the encryption that makes a conversation private, the impossibility of copying data that blocks the extraction of data from an online space, the tendency toward deterioration or self-erasure that makes content temporary. However, there is also a deontological dimension in which these constraints depend on adherence to an observation ethic. A group may remain private, even though we have access to it, because we choose to uphold the pact established with the members; we may choose not to use accessible data because we do not support the ethical conduct of the company that provides it; we may decide not to document an event when asked by the performers not to, even if we have the technical possibility to do so. Finally, we can distinguish a relational dimension when these constraints depend on the position of the observer in the network of actors. Relational proximity to the members of a private group increases the possibility of viewing the group’s contents; knowledge of data gatekeepers influences the possibility of using them; a condition of proximity/trust can allow us to view contents designed to be ephemeral.

Third, it became increasingly clear how these seemingly unprecedented challenges evoked some very old problems in the study of media and society. The reduced accessibility and hyper-contextuality of new digital spaces resonate with the resistance to observation that physical environments and tight-knit social groups have always opposed. The approach to private contexts, for example, is one of the main concerns underlying the ethnographic method. As suggested by several essays in this collection, this research tradition can provide a rich contribution to the ethics of approaching such spaces, even when there is no direct application of the ethnographic method. The question of locked platforms, instead, reminds us of Foucault’s lesson on the power/knowledge binomial (Foucault, 1980), namely the way the production of knowledge is always structured by the power relationship between scholars and apparatuses that archive knowledge. The growing asymmetry between platforms and research institutions in the description of the social world implies the expanding of the question of power over the epistemological one. In other words, the central problem in the production of data passes from being mainly a question of the relationship between data and reality to being a question of the power relationships between researchers and platforms. The difficulty in studying the ephemeral has long been at the center of many research traditions: it is a pivotal issue in the field of performance studies (Auslander, 1999; Phelan, 1993; Taylor, 2007), which has developed a vast theoretical and methodological repertoire on the implications of documenting what leaves no persistent traces (Reason, 2006; Sant, 2017); ephemerality is also an inescapable condition in the study of oral cultures, thus we find folklore studies (Blank, 2012; de Seta, 2020; Fernback, 2003; Ruggles & Silverman, 2009) can offer an important toolbox to approach the transience of new digital content; the ephemeral status of many media para-texts is also the focus of extensive debate in film and TV studies (Grainge, 2011; Pesce & Noto, 2016); moreover, as Jonathan Gray (2016) recalls, studying television before the diffusion of the VCR meant indeed studying an ephemeral medium. In this sense, the issues raised by the below-the-radar internet are also opportunities for internet studies to further amplify their interdisciplinary vocation.

All the contributions in this special issue, though diverse in themes and approaches, participate in making the ways we build the observability of digital phenomena deeper and more complex. They not only identify some crucial cores that currently remain under the scope of internet research, but also help us to understand the fallacious opposition between visibility and invisibility of behavior, accessibility and non-accessibility of data, ephemerality and persistence of digital content.

For the presentation of the articles, we have decided to provide a double reading path. The order of the articles in the journal reflects the sequence in which they were presented at the conference, with the contributions of the two keynote speakers—Crystal Abidin and Rebekah Tromble—to open and close the issue. However, they can also be grouped according to the focus of their contribution. We therefore distinguish theoretical perspectives on the visibilities and invisibilities of data and individuals, methodologies to circumvent the black boxes of online and offline digital environments, and researches presenting findings on how publics employ and are affected by the manifold gradients of privateness, ephemerality, and inaccessibility of online spaces.

Theoretical Perspectives: Thinking about Data Visibilities and Invisibilities

In the article that opens the special issue “From ‘networked publics’ to ‘refracted publics’: A companion framework for researching ‘below the radar’ studies,” Crystal Abidin lays the foundation for a conceptualization of how online publics elude or take advantage of digital visibility regimes. In the article, Abidin extends the reflection of her keynote speech, presenting the characteristics of what she defines as “refracted publics,” that is, the set of “circumvention strategies” that groups and individuals develop in response to the ubiquity of human, algorithmic and corporate scrutiny in today’s social media landscape. The analysis stems from decades of research experience in the field of influencer cultures. As the author specifies, however, it also draws on the reflexive nature of her own experience as a minority person, for which it was necessary to develop a peculiar sensibility about the visibility demands of different groups of individuals, and on strategies to notice, support or transgress these “radars.” The article then goes on to present the framework of refracted publics that the author builds within a dialectical relation with the foundational theorization of danah boyd’s networked publics. Abidin then defines the four conditions—transience, discoverability, decidability, and silosociality—and the three dynamics—impactful audiences, weaponized contexts, and alternating the public/private—observable in refracted publics, specifying how these characteristics develop from the context of 2010s internet culture. This context is characterized by perpetual content saturation, hyper-competitive attention economies, gamified and datafied metric cultures, information distrust, and infodemic fatigue. The following part of the article provides an overview of the field of influencer culture studies from which the refracted publics framework derives. In the last part of the article, the author analyses six strategies that emerge from the observation of refracted publics. These strategies provide a first repertoire of actions by which groups and individuals build, negotiate, or elude their own image with respect to the multiple gazes of algorithmic cultures.

In the article “Where Have All the Data Gone? A Critical Reflection on Academic Digital Research in the Post-API Age,” Rebekah Tromble challenges the narrative of a paradigm shift in internet studies that supposedly happened after the Cambridge Analytica scandal. Tromble argues that this event has, if anything, highlighted the ethical, scientific, and power issues that the relationships among researchers, data, and platforms have always brought with them. The article, which expands on the theme of Tromble’s keynote speech, offers an epistemological and deontological reflection on the data-based digital research that calls for a rethinking of quality standards in the collection and use of platform data. The article first disputes the idea that the “post-API age” should be understood as a situation that aims at disadvantaging academic research. It points out that, from the beginning, application programming interfaces (APIs) have been black boxes whose preferred users have never been researchers. Second, she questions the exceptionalist perspective with which academic data research has often been framed. Tromble argues that if we reflect more honestly on research practice, we will see how it has sometimes taken advantage of the opacity of certain privacy and data-construction standards. Third, attention is drawn to another fallacious narrative, that of the “golden age of data,” stressing how, instead, data biases have always concerned the research carried out with APIs. This point in particular is supported by a previous study on Twitter’s API by the author, which shows the high degree of time sensitivity in its data-retrieval process. In conclusion, the article invites us to imagine and put into practice a series of requests that go beyond the demand for “more data,” showing that we should instead be asking for high-quality data that can encourage more ethical research with a greater impact on society.

Next to issues related to data-access limitations, the question of how we conceptualize what we can observe of individuals’ online behavior depends on how we understand the visibility and invisibility of data. The article by Christina Neumayer, Luca Rossi, and David M. Struthers provides an accurate reflection of how platforms, users, analysts, and researchers construct the boundary between visible and non-visible data. In “Invisible Data: A Framework for Understanding Visibility Processes in Social Media Data,” the three authors challenge the dichotomous opposition between data visibility and invisibility. These should not be understood as object properties but as processes of visibilization and invisibilization. In this respect, the article presents the notion of “quasi-visible” data as an intermediary state of the continuum in which social media data moves. Drawing on a heterogeneous set of examples and traditions that interweave the study of social movements, social network analysis, and the epistemology of history, the authors describe four dimensions along which the processes of visibility and invisibility move. The first dimension is that of “people and intentionality,” namely the way people create data in a particular social context; the second is that of the “technologies and tools” employed to store data on social media platforms. Platforms then define also a third dimension of visibility determining the conditions of “accessibility and form” up to the fourth dimension of “meaning and imaginaries,” with which researchers and analysts order and interpret data. At the end of the article, the reader is invited to reflect on how each level of complexity raises fundamentally important ethical, epistemological, and methodological questions. The authors call for a more reflective approach in the choice of data representation regimes, the way we deal with data bias, and the way we decide to bring out what was meant to be invisible in the users’ intentions.

Methodologies: Accessing the Black Box

Three of this collection’s articles focus on indicating problems and methodological solutions to address the locked, private, and ephemeral components of the current evolution of digital platforms.

In their important contribution “Walking Through Twitter: Sampling a Language-Based Follow Network of Influential Twitter Accounts,” Felix Victor Münch, Ben Thies, Cornelius Puschmann, and Axel Bruns propose and demonstrate the effectiveness of a sampling method to retrieve the following networks of influential accounts on Twitter. The need for new Twitter following network mining techniques is decisive because of the limitations on data accessibility via the platform’s standard API, which restrict the possibilities for independent research on the Twittersphere. The methods currently available for collecting large amounts of data from Twitter impose a high entry barrier in terms of economic resources (access via premium and enterprise APIs) or tend to represent only the emerged part of active communicative actions on Twitter (i.e., mentions, replies, re-tweets), leaving aside the description of the following networks, which can also be based on silent-listening activities. The method presented here, built on the rank degree method, offers a low-cost alternative (in terms of time and budget) that enables a wider audience of independent researchers to conduct studies on large-scale following networks. The authors test this method by successfully presenting a representation of the top 10% of German-speaking Twittersphere’s most influential accounts. The possibility to carry out studies on the overall structure of the global Twittersphere via the cost-free API standard of Twitter that this method offers is of paramount importance. This is because the “locked” condition of platform data not only acts as a “quantitative” deprivation but as a constraint on the choice of the possible research topics, making the study of fragmented portions of Twittersphere prevail and leaving aside the emerging qualities observable at the level of the overall structure.

Esther Hammelburg’s article, titled “Being There Live: An Ethnographic Approach for Studying Social Media Use in Mediatised Live Events,” examines the articulation between the online and offline dimensions and between the digital and material ones in live events. In these settings, a deep exchange between physical co-presence and the digital conversations that make up the “eventsphere” is observed. The study of the eventsphere raises important ethical and methodological questions: the conditions of ephemerality and privacy of the live experience intersect ephemerality and privacy as an affordance of the digital spaces through which the event is prolonged. Therefore, Hammelburg reflects on how to study this kind of assemblage and deal with the ethical challenges concerning the need to get in touch with the elusive personal and emotional conversations surrounding the event. The author explains the fieldwork conducted with the audience of three very different types of live events—Dutch events Oerol, 3FM Serious Request, and Pride Amsterdam—in which she demonstrates the use of a composite patchwork of methodologies including participant observation, media diaries, short in-situ interviews (with 379 event-goers), and in-depth interviews. When analyzing this research experience, Hammelburg assesses the tactical meaning of “being there” for what concerns the study of the practice and the meanings that participants develop around the event. On one hand, being there makes it possible to reconstruct what is not traced in the online testimony of the event, for example, by being able to observe “what you choose not to post.” However, being there is also fundamentally important for its value of mediation with respect to the relational component of privateness: carrying out participatory ethnographic work on the ground changes your perspective on digital data sets, since you face the people whom you research.

Continuing on the theme of ethnographic investigation between online and offline and between digital spaces and materials, the article by Tiziano Bonini and Alessandro Gandini offers an interesting point of view on music streaming platform companies. The study entitled “The Field as a Black Box: Ethnographic Research in the Age of Platforms” provides a clear example of how the “locked” component of platforms not only concerns the limitation in accessibility to digital data, but also the way platforms’ corporate culture participates in the black-boxing of the human and algorithmic processes on which they are based. The article examines what the authors call “a failure” in their attempt to conduct ethnographic research in some of the leading companies in the platformization of music. As the authors recall, the institutional “black box” is a typical problem of production studies, which depends on working ideologies and therefore precedes the context of platformization. However, these obstacles find in the digital environment both positive reinforcements as greater ease in the opacification of technical gatekeeping processes, but also strategies to counter this opaqueness that researchers can employ. The article details the processes of approaching the field with which the two authors have attempted to carry out an “online platform production ethnography.” The results show that there are several “concentric” (cultural, algorithmic and relational) black boxes detecting recurrent patterns in the way corporate working ideologies have attempted to take their eyes off of the field (the “deflect and silence” protocol). Hence, Bonini and Gandini propose their own set of tactics to enrich the ethnography of digital cultural industries—rely on personal connections, multi-sited ethnography, focus on ex-workers, be undercover, digital methods for ethnography—defending the ability of the ethnographic methodology to open a crack in the “platform fortress.”

Both articles underline that it is unnecessary to invent a new ethnography for private online spaces or for digital/material articulations. Instead, it is necessary to continue the development of knowledge about the presence of the researcher and the access to a multi-situated field. Where some online spaces return to having that kind of recalcitrance to the datafication of physical spaces, this invites us to look at that wide repertoire of resources, methods, and experiences matured by the rich ethnographic tradition that has always been related to the problems of the privacy of spaces, the transience of action and meanings, and the power relationships involved in the construction, archiving, and access to knowledge.

Research: Online Spaces Affordances Beyond Publicness

In addition to theoretical and methodological perspectives, among the contributions of the special issues, there are also results of research carried out in “below the radar” spaces and practices. These three articles are important elements in research development on the affordances of private spaces, on the repertoires of actions with which users make themselves less visible, and on the discursive regimes that develop around these dynamics. These research experiences show how the definition of spaces as “private,” “locked,” or “ephemeral” is a dynamic process resulting from negotiations of meanings and clashes of perspectives.

Socio-technical constraints that make digital environments more or less private can, for example, be used to construct different symbolic boundaries according to different user groups. In this sense, Daphna Yeshua-Katz and Ylva Hård af Segerstad investigate how the affordances of different digital discussion spaces relate to the boundary work of stigmatized online support groups. Their article “Catch 22: The Paradox of Social Media Affordances and Stigmatized Online Support Groups” makes a decisive contribution to the issue of online support groups for individuals suffering from stigma, as it observes the role that groups’ different degrees of anonymity and visibility play in today’s digital ecology from a comparative perspective. The research investigates four types of stigmatized communities—eating disorders, infertility, bereaved parents, and posttraumatic stress disorder—and four different discussion environments—blogs, forums, WhatsApp chats, and private groups on Facebook. Through 66 qualitative interviews with the members of the groups, the researchers observe the rhetorical construction between ingroups and outgroups by detecting two central processes: the monopolization of the networked public and the removal of those who are considered inauthentic members. Observing the way these processes relate to the affordances of the observed spaces, the researchers distinguish between high anonymity and high visibility spaces (such as forums and blogs) and low anonymity and low visibility spaces (private Facebook groups and WhatsApp chats). Consequently, they point out the following paradox: the same affordances of privateness that allow the construction of a space protected from intruders emerge as obstacles even for stigmatized individuals who need the group’s support. The study questions the widespread understanding of the concept “private” as “safer,” highlighting the plural readability of technical boundaries that can erect a problematic barrier for those in need.

While privacy supports the protection of vulnerable people, private social space can give rise to mechanisms of impunity for action threatening victims’ reputation, privacy, and safety. The article by Silvia Semenzin and Lucia Bainotti titled “The Use of Telegram for Non-Consensual Dissemination of Intimate Images: Gendered Affordances and the Construction of Masculinities” analyses the role of the Telegram messaging app affordances in the circulation of non-consensual intimate content (NCII). The circulation of erotic and intimate material in private and group chats is increasingly widespread. In Italy, numerous scandals have emerged regarding the exchange of intimate images spread non-consensually on Telegram, sometimes with underage victims. Semenzin and Bainotti’s study questions how Telegram affordances have influenced this phenomenon and investigates the relationship between these affordances and the construction of the hegemonic masculinity of the users. Starting from the concept of “gendered affordance” (Schwartz & Neff, 2019), the study focuses on how platform affordances not only define a range of uses but also a spectrum of preferential gendered subject-positions. Through an ethnographic content analysis, the research examines 50 groups and Telegram channels in which this kind of content is distributed. The results show how, on Telegram, the characteristics of anonymity, weak regulation, and sociability facilitate homosociality dynamics that reinforce the hegemonic male identity. In this context of misogynistic homosociality, a normalization of objectification, classification, and victim-blaming processes is observed. The authors point to the link between platform affordances and constructions of masculinity that feed the diffusion of NCII. The article, therefore, demonstrates how the study of affordances of new private digital spaces cannot ignore the pre-existing gendered cultural repertoires.

The affordances of networked (and refracted) publics should be observed starting from the imaginaries that guide their identification and interpretation. Tetyana Lokot’s article, entitled “Articulating Networked Citizenship on the Russian Internet: A Case for Competing Affordances,” draws the attention to the discursive component of visibility affordances. Analyzing the Russian context, Lokot underlines the role that different narratives on the visibility and accessibility of citizens’ social media data in the construction of different imaginaries of citizenship play. After reviewing the status of digital rights in Russia, her study observes the clash between two discourses on visibility: the one produced by the state through Roskomnadzor (the Russian state media and internet regulator), and that of the digital rights activism groups called Roskomsvoboda and the Internet Protection Society (OZI). Through a comparative narrative analysis of reports and documents by these subjects, the research shows how two competing narratives of networked citizenship emerge. While the Russian state frames the digital traces of the “dutiful networked citizen” in terms of visibility, permanence, and searchability, the activist discourse tends to favor the vision of a “self-actualizing networked citizen” that exercises their agency as discretion on the visibility and ephemerality of their data. Lokot’s article, therefore, raises an important question about the redefinition of social media affordances for activism. This transformation concerns both the search for safer spaces and channels of communication and the social imaginaries of visibility.

The purpose of this special issue was to expand the themes presented during the conference and to raise the problem of how the material and epistemological constraints produced by private, locked, and ephemeral conditions affect internet researchers. We believe that the articles presented here have succeeded in doing much more. Overall, they provide an overview of the transformations of digital publics across different visibility regimes and of the research practices needed to study them. We hope that this discussion will continue beyond this special issue, and that it will serve to raise the level of awareness on those research blind spots that we often conveniently contribute to nourishing.