From mediated to datafied recognition: The role of social media news feeds

1 Recognition on social media platforms: A complex issue

The works of Taylor (1994), Fraser (1995), and Honneth (1995) gave a new impetus to recognition theory in the fields of political sociology and philosophy. Despite its significance for understanding the relational essence of human life, the concept has only recently been explored by media and communications academics. The works investigating the processes of recognition in media environments use different terms to refer to the phenomenon, such as “mediated recognition” (Lorenzana, 2016; Maia, 2014; Muscat, 2019), “mediatized recognition” (Cottle, 2007), “cultural recognition” (Malik, 2014), or simply “recognition” (Couldry, 2010; Edwards, 2017; Nærland, 2017; Nikunen, 2018). In common, they all see the media as an institution that acts upon recognition processes necessary for the development of a more democratic and free society. Newspapers, television, and social media platforms, according to their stance, can produce spaces for deliberation and social representation that interfere positively (or negatively, when failures of recognition occur) in the formation of subjectivities capable of self-realization.

This article, however, proposes that the search for recognition on social media platforms should not be confined to the perspectives employed by research on traditional mass media. Social media operate in the social sphere of the news feed, which has an impact on how the individual makes the world intelligible (Bucher, 2018). In this sense, platforms like Twitter, Facebook, Tik Tok, and Instagram, responsible for a considerable portion of today’s sociability, modify the very principles structuring the idea of recognition. Publishing, liking, sharing, and commenting are just some examples of social media practices implicated in contemporary dynamics of recognition. What is distinctive in these new dynamics though is that they imply a type of sociability marked by a process of datafication, that is, a process in which people’s connection with others is put “in quantified form so that it can be tabulated and analyzed” (Mayer-Schönberger and Cukier, 2013, p. 165).

Social practices are captured and translated into data, which are then not only valued but also rewarded by algorithms according to their capacity for generating more data and, ultimately, financial gains (Couldry and Mejias, 2019). The news feed is particularly important in this function. It is there, in the news feed, where the logics driving the algorithms become concrete to the users of these platforms. For Duffy and Pooley (2019), these practices restrict expressive possibilities that do not fit into models of individualized self-promotion, thus compromising unconditional forms of recognition. In this sense, the present article draws on current discussions on data-driven platforms to argue that works on mediated recognition, traditionally based on theories of representation and deliberation, should make an effort to incorporate the materiality of such platforms in their analysis of social media practices.

Maia (2014) and Lorenzana (2016), for example, use Honneth’s theory of recognition to argue that social media can amplify the voices of groups historically silenced that struggle to be recognized by society in their legal and symbolic dimensions. Despite their valuable insights, these analyses do not explore whether such platforms follow rationalities that could pose significant challenges to the very idea of recognition and its correlated production of a collective sense of solidarity and equality.

Thus, the objectives of the present article are, firstly, to offer a brief review of works investigating recognition processes in media environments, with a special focus on social media platforms. It argues that the analysis of recognition in such datafied spaces should not lose sight of their working logics, which are responsible for the organization of media practices around the creation of economic value for the companies. This is followed by an examination of the news feed as a type of social space, which is particular to social media, where these logics are manifested in the decisions of the platforms about what and when to show to (and hide from) their users. Works analyzing dynamics of recognition on social media should, to some extent, take into consideration the way data are collected by the platforms, how they are translated into metrics, processed by the algorithms and, finally, how they help the platforms in rewarding particular forms of media practices. Ultimately, processes of recognition on social media cannot be disassociated from a type of power unique to datafied platforms, which directly interferes in how individuals and social groups are valued in society.

In short, this paper proposes a theoretical reflection on processes of recognition in social media platforms. It does not offer an empirical exploration of the theme. Instead, it sheds light on some key aspects of the data-driven nature of social media, connected to the quantification and processing of the social by algorithms, which are unexplored by works analyzing dynamics of mediated recognition. Towards the end, the article suggests that these dynamics of quantification and data processing manifested in the news feeds should be taken into consideration in future works drawing on recognition theory to analyze practices of sociability on social media platforms.

2 Struggle for mediated recognition

As human beings we evolve our sense of self through our interaction with other people. Without being properly recognized by others, the development of our capacity for self-realization and self-determination is undermined. Since the nineteenth century, philosophers and academics from different fields have explored the concept of recognition and its implication for the formation of personal identities and for the transformation of society. This discussion gained new vitality in the mid-1990s, when debates connected to identity politics and social movements started to address the issue of recognition and, perhaps more importantly, the question of failure or misrecognition, and what it means for the constitution of subjectivities. Charles Taylor (1994) proposed, in the introduction of his Politics of Recognition, that “our identity is partly shaped by recognition, or its absence” (p. 25). His work on multiculturalism, identity, and recognition, for instance, remains one of the most influential references in the field, despite the insightful critiques it received by the likes of Nancy Fraser, who defends the idea that cultural change should be associated with demands for economic and material change as well (Fraser, 1995).

Axel Honneth, also a central figure in the debate, is known for his normative effort to describe the dimensions involved in identity formation and, in relation to this article, is particularly relevant because of his account of the conditions at stake for the production of social-esteem. For him, esteem recognition is characterized by relationships founded on principles of symmetry that promotes the individualization and autonomy of subjects who perceive themselves as similar to each other while maintaining an interest in the particularities of others (Honneth, 1995, p. 128). Full recognition, for Honneth, involves processes occurring in the affective, legal, and social dimensions. It is mostly to the latter, however, that media and communication theorists have given especial attention when analyzing the role played by the media in the struggles of marginalized groups for voice and positive representation.

This is broadly the case of the works on mediated recognition analyzed here, which were discussed or used as key references in two recent international events on the theme. Recognition theory started to be used by media and communication studies in a more or less consistent way from the 2000s onwards, but only in 2018 did it become the subject of a dedicated discussion among researchers in the field. In that year, the ICA conference, held in the city of Prague, hosted a thematic panel called Mediated recognition: Agency, paradoxes and struggles for visibility. This was followed, in the year after, by an ICA pre-conference, that took place in the city of Washington, DC, entitled Mediated recognition: Identity, justice and activism. Although it is not the objective here to discuss the particular studies presented on those occasions, the current article draws on their main references to mediated recognition to form the starting point of its arguments.

Using the notion of political listening, Muscat (2019) offers a representative example in her research on news audiences in the city of Sydney, Australia. She shows how mainstream media discourses are key in establishing racialized notions of otherness, usually through negative representations and practices of exclusion. While some participants in her research confronted mainstream mediations of racism, they contradictorily expressed Islamophobic fears, thus reflecting existing values articulated by these same media. For Muscat, the results are evidence of the key role played by public service and community media for facilitating political listening and, ultimately, processes of mediated recognition.

Malik (2014) sees similar challenges presented by media representation in her analyses of the European context. In her view, the approach of European public service broadcasters to minority groups overlaps two imbricated principles: “multiculturalism on the one hand and equal citizenship on the other” (Malik, 2014, p. 38). Again, she proposes that unbiased representation in public services is a central strategy for ongoing struggles aimed at group recognition and social justice.

Couldry (2010) articulates voice and recognition to rethink the role of media in democracy theory. In Why voice matters, he suggests that the individuals’ capacity to realize their abilities as human beings and to be widely recognized before the rest of society is connected to a more democratic way of life. Any notion of social justice must, therefore, be linked to the plurality of voices that can be heard, whether through alternative media or not. For Cottle (2007), television journalism should engage in these processes of giving voice to the voiceless and rehabilitating the representation of “former others”. Mediatized recognition, following his perspective, can be achieved through deliberation and the public display of differences, which results in increased political respect for asylum seekers, refugees, and other minority groups. With a similar perspective, Edwards (2017) draws attention to the role to be played not only by mass media, like television and newspapers, but also by public relations. She reports the results of a study on YouthVoice, a UK charity that helps young people to speak out about issues that matter to them (Edwards, 2017). YouthVoice offers the skills and material support to marginalized individuals opening up “localized spaces of appearance”, prompting responsive listening from different segments of society. While her research demonstrates the importance of this type of initiative, it also draws attention to tensions arising from constant (re)framing of identities involved in the process (Edwards, 2017, p. 14).

Rousiley Maia’s Recognition and the media (2014) is perhaps one of the best examples of work which extensively explores the relationship between media and recognition. In the book, Maia and her collaborating researchers present Honneth’s sociological program in detail in order to lay out the theoretical foundations for her research on representation of minority and stigmatized groups in different media. The articulation between power and the media is discussed, mainly, in relation to the construction of discourses and representations of disadvantaged and minority groups on television series and newspapers articles. Hence, the power of media organizations, especially mass media, is read from their capacity to reflect and produce cultural and political meanings, a stance overtly inspired by the Frankfurt School tradition (Maia, 2014, p. 4). The authors are particularly interested in understanding how minority groups seek to exert pressure to change production practices of media organizations in order to promote a more inclusive society. According to Maia and her collaborators, disadvantaged groups portrayed in the cases studied do manage, on occasions, to successfully challenge the decisions related to the creation and circulation of social representations taken by traditional mass media agents.

The chapters in her book investigating the struggles for recognition in the online context, however, pay little attention to the power exercised by digital platforms, such as Facebook or YouTube. The latter, for example, is basically portrayed as an infrastructure that allows various social actors to become content producers. The platform is presented as a space where free conversation and deliberation take place, even if, sometimes, in a conflictive and disrespectful fashion (Maia, 2014, p. 161). But such description leaves out a key type of governance performed by its algorithms in relation to what is published and what is made visible, which reconfigures the distribution of what can be seen, heard, and thought (Rancière, 2009). YouTube’s materiality is therefore treated in a soft manner. Moreover, this perspective also ignores the development of algorithm imaginaries (Bucher, 2016), which have a direct impact on how people feel and relate to the workings of the platforms, ultimately interfering in the building of practices connected to identity formation. In this sense, the problem of media power in Maia’s case studies suddenly loses some traction when it comes to analyzing struggles for recognition occurring in social media platforms. To put it differently, the structuring economic logics of these platforms, which are acting on the horizon of subjective possibilities of its users, are simply ignored.

This issue can also be perceived in the ethnography conducted by Lorenzana (2016) with Filipino transnational citizens living in India. In his work, Lorenzana analyzes daily practices on Facebook by Filipino migrants employed as strategies to convey esteem and social recognition. He shows how the publication of photos and texts displaying professional achievements and affirming social connections plays a fundamental role in the identity formation of these individuals. This is particularly evident when members of the migrant community convey their skills and contributions to the Indian society on social media platforms (Lorenzana, 2016, p. 4). Although Lorenzana acknowledges that Facebook predisposes users to self-representation, he does not advance the proposal to investigate how this occurs and what are the implications of this finding.

As seen in the debates above, discussions on deliberation theory and media representation are the structuring frames of debates articulating social recognition and traditional mass media, such as TV and newspapers. Nevertheless, these perspectives appear to show their limitations when it comes to analyzing recognition processes in digital platforms. The works of Maia and Lorenzana, for example, share a generally positive perspective on the possibilities offered by YouTube and Facebook for individuals to express their voices. If, on the one hand, the case studies reported by the researchers successfully demonstrate how such platforms provide the space and symbolic resources for recognition to take place, they on the other hand, do not reveal what the conditions are and the context in which the production of these resources take place. As Couldry and Mejias (2019) argue, social media companies are responsible for processing social relations following logics that, in their words, “hollow out the social” (p. 115). These companies convert sociability into metrics which are then analyzed, processed, and ranked in order to produce economic value.

The study by Balleys, Millerand, Thoër, Duque (2020) of processes of social recognition of teenagers on YouTube offers an interesting take on the issue. The authors combine content analyses of the videos produced by a group of teenage creators with a reception investigation of these videos by a similar age group. They argue that the processes of recognition observed in their research are twofold: “It is both a capacity to recognize oneself in others – like figures with whom one can identify with – and a need to be recognized by others, that is, to be seen as an individual with social value” (Balleys et al., 2020, p. 8). Their article demonstrates how teenagers deploy different strategies on YouTube to build a type of community of shared values that plays an important role in their identity formation. However, the research in question does not show which other types of YouTube videos this audience watch, how they find these videos, and in what forms YouTube recommendation systems and algorithms interfere in this circuit. Answering all these broad questions is not the objective of the authors, nor ought it be, as they represent a different research agenda. The point here is to argue that social media platforms like YouTube play a larger part in processes of social and individual recognition than is usually acknowledged.

In this respect, Bishop (2018) proposes that YouTube’s algorithm privileges and rewards specific classed and gendered content connected to consumerism. She also suggests, according to the case studies investigated, that YouTube creators have particular understandings of the workings and logics of the platform’s algorithm, which are then assimilated into their forms of self-presentation and other related practices. If the power associated with traditional media lies, in part, in its concentrated capacity for creating a particular symbolic meta-capital (Couldry, 2003), the power of digital platforms sits mainly on its capacity to determine, following its own economic interests, the type and frequency of content someone will be exposed to (Van Dijck, Poell, and De Waal, 2018).

3 Towards an idea of datafied recognition

Datafied platforms act directly on what Rancière (2009) calls the “distribution of the sensible” and on the configurations of the experience. What can be seen and when something can be seen on media platforms depends on decisions taken by their algorithms based on how they perceive each one of their users and their online relations. For Rancière, the individual, as a citizen, should take part in the governance of their own life in order to be properly recognized. But that demands an access to what the author denominates “a common”, which can be thought of as a distribution of space, time, and activities connected to the possibility of participation. The French philosopher argues that there is no distinction between the sphere of politics, the sphere of speech, and the social. Thus, political action does not occur only in the classic sense of the term. It happens, fundamentally, in its capacity to exercise governance over the sensible, that is, over the horizon of what can be seen, heard, said, and thought. He talks about a politics of aesthetics but not as a discussion of art theory. Instead, he is interested in a regime of identification and thought, in a mode of articulation between manners of making, forms of visibility of these manners, and modes of thought of its relations (Rancière, 2009, p. 17).

In this respect, the questions that should be asked when investigating dynamics of recognition on social media platforms should go beyond the discussion of who has the power to produce representation or who has access to the creation of media meta-capital, to also ask how the sensible is distributed, under which logics, and with what consequences. The problem now faced by such inquiries is that, unlike in the era of mass media, there is a growing interdependence of everyday life and data-driven systems. According to Couldry and Hepp (2016), “‘data’ and ‘information’ generated by systems of computers are today a precondition for everyday life”, their processing is “consequential for social life” (p. 123).

It is not only that individuals and social groups have to learn new skills in order to give visibility to their value or self-realization within a datafied society but also that the very definition of what counts as self-realization and how to achieve it changes as a consequence of the role played by the platforms. One could argue that the conditions necessary for recognition to take place depend, to a certain extent, on the construction of social life by systems based on algorithms. To advance this argument I propose here that nowhere can this be better felt than in the news feeds of digital platforms.

The news feed is a space curated by digital platforms which aim to give access to a personalized construction of social life. Facebook, for example, defines it as “a personalized, ever-changing collection of photos, videos, links and updates from the friends, family, businesses and news sources that you’ve connected to on Facebook”.^[1] How a news feed is curated, personalized, and how it defines what “matters most to its users” is of huge importance and, at the same time, a relatively opaque issue. Bucher (2018) rightly suggests that the news feed “is political insofar as it is exercising a position of governance” (p. 67). It relates to the “different ways of being in the world”. In a nutshell, news feeds construct regimes of visibility as much as regimes of invisibility, which have huge influence on the building of social life.

This was evident in the now infamous Facebook/Cornell “emotional contagion” project (Kramer, Guillory, and Hancock, 2014), a massive-scale experiment in which almost 700,000 users of the social media platform were exposed to emotional expressions on their news feeds. The objective of the experiment was to discover whether people would change their posting behavior when in contact with different emotions. Although much of the outcry regarding the experiment came from its appalling lack of ethical principles (Puschmann and Bozdag, 2014), it is important here not to lose sight of its results.

To be precise, Kramer et al. conducted two parallel investigations with distinctive groups of people. In the first one, they reduced the exposure to negative emotional content on the users’ news feed, whilst in the second group, they did the opposite, that is, reduced the positive emotional material. The authors concluded that both investigations indicated a clear emotional contagion. A decrease in the expression of positive emotions in a person’s news feed, for example, made that person publish less content with positive emotional disposition, and vice versa. In this respect, the news feed can have a buffer effect in the creation of particular types of emotion on Facebook users. The experiment also demonstrated that less exposition to either form of emotion in the news feed resulted in less expressiveness and engagement, of any nature, of that person on the platform in the following days (Kramer et al., 2014, p. 8790).

The results of this experiment are testament to the power of the digital social platforms. It is a power connected to the capacity for categorizing the world but also the power to build, for example, different emotional dispositions in the social world. Facebook can directly interfere in how reality is presented to its users, not only in the form but also in the content, in order to provoke specific responses from them. Ultimately, one can imagine this capacity also interfering in the production of visibility and recognition of individuals using this type of technology.

It is also crucial here to draw attention to the fact that the experiment took place specifically in the news feed of the participants. In other words, all the content that was deliberately omitted in their news feeds never became unavailable in the profiles of the people who produced it. This is evidence of the role played by the news feed as the main space for intervention of the platform. Bucher (2018) argues that we must consider the algorithm systems behind the news fees as political devices as they reflect particular views on “how the world is to be ordered” (p. 67). For Facebook, as with other social media platforms, it is crucial that the news feedss work to validate a form of “participatory subjectivity that hinges on continued and ongoing engagement with the platform” (p. 155).

If, on the one hand, the democratization symbolized by the free and easy access to Facebook’s news feed makes possible the expression of social esteem and recognition of the diverse voices that make up society, it nevertheless restricts and conditions intersubjective relationships within dynamics that privilege practices closely linked to capitalist logics. The news feed operationalizes what Honneth and Margalit (2001) call “expressive gestures of recognition”, that is, the demonstration of “emotional readiness to morally engage with the addressee” (p. 122). Although Honneth and Margalit refer to smiles, respectful greetings, or nods when exemplifying these expressive gestures in the face-to-face world, it is possible to imagine that the acts of following, liking, or sharing content on social media platforms can also be seen as a type of expressive gesture of recognition.

Honneth and Margalit also suggest that a person has to be visible in order to be recognized. In fact, and this is also key here, visibility in the literal sense is also a prerequisite for invisibility in the figurative sense (Honneth and Margalit, 2001, p. 114). This means that for a person or a social group to experience a lack of recognition, which is similar to being figuratively invisible, they have, prior to anything, to be physically visible. Honneth and Margalit remind us, for instance, that nobility was traditionally allowed to be naked in front of their maidservants because the latter were figuratively invisible, despite their physical presence. The servants were there but not recognized as worthy individuals capable of provoking discomfort or shyness within the noble. Their lower status as persons hindered their capacity to affect individuals from a higher rank.

Honneth (1995) describes how minority groups and excluded segments of western society struggled in the last couple of centuries to be legally and symbolically recognized as worthy citizens. In fact, one could argue that this is a permanent quest driving the actions of common individuals who want to see their personal and professional achievements valued by society. This is not different in the age of social media platform ubiquity. What is peculiar in the present context, however, is that individuals are only certain that they are figuratively visible, that is, that they are valued by others, when they see some kind of positive response from friends or followers, even when these responses are simple expressive gestures of recognition. People want to be visible online through their publications, reactions, and updates, but they wish to do so in a way that their followers respond accordingly via, for example, comments, shares, or likes. In short, social media platform users want to be visible in both senses, literally and figuratively. But, in order to do so, they have to accept and internalize some broader economic principles informing social media platforms. The latter are structured in such a way that they can collect and process as many data as possible from their users in order to create economic value for the companies. Not only must people agree to have their own personal data gathered and shared by the platforms, but they have to fit their sociability into a language valued by the algorithm. It is thus impossible to separate the working dynamics connected to the creation of visibility and recognition in social digital platforms from the production of data and financial profit.

4 The news feed and the social space of datafied recognition

While considering the importance of social media for contemporary processes of recognition, this article proposes a more detailed discussion of some typical features shared by these platforms. It is clear that Tik Tok, Facebook, Instagram, Twitter, and YouTube have specificities that differentiate one platform from the other. Nevertheless, they all share some common governing principles that underline our discussion on datafied recognition. These technologies are heavily dependent on a curatorship carried out by their algorithms to decide what and when to show to their users. The news feeds are particularly important in this task. It is in the social space of the news feeds where the processes of capturing, translating, valuing, and rewarding the publications become concrete to the users of these platforms. This is the basis of such technologies and systems that strive to make people stay connected for longer periods of time while, at the same time, predicting their actions and consumer choices.

Ajana (2018) argues that the wide use of quantitative assessments of the individual is no longer restricted to government, institutional, and business spheres, as in the past, but is now part of the everyday life of the common subject. She proposes that the development of data-driven technologies has made possible the widespread adoption of self-tracking devices, such as Fitbits and Apple Watches, as well as social media platforms that use metric principles to assess the most diverse social practices. What is most striking in the phenomenon, however, is not simply the enormous volume of data produced today, but, as Ajana (2018) points out, the rationalities and discourses created from them. After all, datafication has to do with reorganization (Mayer-Schönberger and Cukier, 2013)

The deepening of the quantification of the individual’s actions and social practices, within a scenario where identities are increasingly valued from numerical perspectives, seems to be particularly telling. The introduction of the “like” button by Facebook in 2010, for example, paved the way for the conversion of sociability into metrics, which not only optimizes the monetization of individual scores and data produced across platforms but also facilitates strategies for creating engagement. Gerlitz and Helmond (2013) state that both the “like” button and other social plug-ins are potential triggers for numerous processes aimed at producing more online participation. According to the authors, people engage more with content on platforms that demonstrate a greater number of likes and shares, a process that clearly favors a feedback logic.

Bucher (2018, pp. 77–78) also argues that the more reactions a Facebook post generates, the more “edges” are assigned to it by the platform. An edge is a type of attribute that was originally composed by three dimensions, each with its own score: the type of interaction (comments, shares, likes, etc. have different weights); the temporality of the interaction (more recent interactions usually have greater weight), and who was responsible for carrying it out (greater affinity between a user and a follower or friend on the network results in more weight). To estimate the relevance of an object within Facebook – which can be a published image, a post, a video, or an external link – the algorithm multiplies the values of each edge it generates to create an overall score, which is then used to decide the hierarchy of this object in the news feed of the users on the platform. The weights of these valuations, as well as decisions about the ranking of objects, are dynamic elements defined by artificial intelligence systems that seek permanent optimization capable of generating more engagement and time spent on the platforms by users.

This model, described in detail a few years ago by Sanghvi and Steinberg – engineers responsible for developing Facebook’s news feed – , has already undergone numerous modifications and has since become more complex, as demonstrated by De Vito (2016) in his analysis of the patents approved by the platform over the years. The very concept of EdgeRank, used by Facebook in the early 2010s, fell into disuse after a short time. However, even though the ranking of different types of publication in the news feed has incorporated other dimensions, such as prior engagement, platform priority, content quality, capacity of creating more interactions, among many others, one essential characteristic seems to remain stable over the years: a constant creation and redefinition of affinity scores and scales between users and objects on Facebook. In early 2021^[2], the platform published an updated technical explanation of its news feed algorithms, which also mentions the use of information regarding past engagement of its users as well as diversity rules aimed at offering a mix of types of content. Facebook literarily uses thousands of signals, a term more in vogue nowadays, to evaluate what a person might find more relevant in their news feed.

These signals and affinity scores are the backbone of social media platform algorithmic systems. Bucher (2018, p. 11) argues that Facebook’s valuation of online friendships, based on different affinity criteria, “serves an essential role in sustaining the social networking system itself”.

The measurement of friendships on Facebook and other social platforms, however, goes beyond the computational systematization of connections. It is a vital part of the visual experience of its users, which, as Grosser (2014) points out, influences their evaluation of what is read, seen, and watched. Thus, the visualization of the number of likes or shares received by an object directly influences a user’s decision to react to that same object. In other words, the more eloquent the numbers of engagement or expressive gestures of recognition a publication receives, the more it will be perceived as relevant by users. It is hard to deny that this type of logic puts pressure on users to get more reactions to their publications. There are a number of studies suggesting negative effects on mental health caused by “compare and despair” attitudes on social media (Hampton, Rainie, Lu, Shin, and Purcell, 2015).

It is also worth noticing that while digital platforms facilitate the establishment of intersubjective connections in the most diverse ways and transform them into metrics, they simultaneously monitor and watch over such connections, even if this occurs in imperceptible ways by users. Exchanging likes, making comments, posting and watching photos and videos online are actions that result in a huge amount of data that are processed in order to transform users’ intersubjective relationships into products (Couldry and Mejias, 2019), a consequence of the expansion of this culture of metrics.

5 Concluding remarks on the datification of recognition

There are, of course, several works problematizing how social media platforms relate to processes of self-realization and identity formation which do not draw on recognition theory. Karppi (2018) and Paasonen (2021), for instance, explore the ambivalences of such technologies in relation to their capacity to foster social equality, whilst, at the same time, relying on “affective manipulation within data capitalism” (Paasonen, 2021, p. 14). The economic and technological frameworks described by them and other researchers (like Zuboff, 2019) are enacted by algorithms, which have to constantly decide “what^[3] is relevant, in a generic sense, and who the user is, so that this generic rule about relevance can be applied individually” (Araújo and Magalhães, 2018, p. 4). In this context, recognition theory, especially under Honneth’s approach, offers an interesting entry point to the debate because of its normative effort to define what is of value, in the personal and social spheres, and its preoccupation with the constitution of identity.

Works analyzed in this paper showed how social media platforms like Facebook, YouTube, and Instagram are important spaces for deliberation and social representation. These technologies can help connect plural voices and assist the formation of new communities, where subordinate groups can challenge existing conventional narratives about them. But the huge capacity of datafied platforms for collecting and processing data gives them an unprecedented capacity for framing the social. Corporations responsible for creating digital platforms are changing the very nature of the patterns of interactions (figurations) in social life (Couldry and Hepp, 2017).

It is, therefore, proposed here that future works investigating recognition on social media platforms should pay closer attention to the dimensions of capturing, translating, processing, and rewarding media practices involved in the processes of data production, which are embodied in the news feeds. The latter gives concreteness to the governance performed by the platform algorithms, which, in Lim’s terms, signals to “a transition from identity politics to ‘personal identity economics’” (Lim, 2021, p. 1). In this context, practices of sociability aimed at increasing recognition in social media platforms should be understood within a broader movement of datafication of society, which deepens the presence of capitalist logics in social life.