What Social Media Could Be: Normative Frameworks for Evaluating Digital Public Spaces

Abstract

Increasing attention has been placed to the societal downsides of social media, and appropriately so. Less attention has been paid to the qualities to which social media should aspire. We contend that this is critically important. Not only must social media, and social media scholars, identify and reduce negative outcomes, but we must also critically engage with what is desirable. The purpose of this theoretical essay is to propose a normative framework for digital public spaces. We lay out four categories, and 14 sub-categories, of normative ideals to which social media could aspire. It is our hope that chronicling these qualities will allow scholars to more critically reflect on their normative assumptions when they research social media and will encourage practitioners to think about how social media could be built with these ideals in mind.

A plethora of research looks at what is wrong with digital spaces or online forums where people interact with one another, such as social media sites. It is clear that digital spaces often are rife with rudeness and incivility (Chen, 2017; Coe et al., 2014; Su et al., 2018), facilitate the rampant spread of false information (Pennycook & Rand, 2020; Shin et al., 2017), and provide a haven for the most virulent forms of speech, like misogyny and racism (e.g., Sobieraj, 2021). Concerns about online addiction abound (e.g., Kim et al., 2019; Mieczkowski et al., 2020), as do worries about what data platforms collect and what they do with these data (e.g., Sarikakis & Winter, 2017). Digital spaces are often cast as malevolent places where predators loom, chicanery reigns, and the imagined virtual public sphere is but a halcyon dream that can never be realized.

We do not debate that a sinister side of digital public spaces exists. We acknowledge the internet has hardly lived up to the early imaginings that saw it as an equalitarian utopia that would mute offline hierarchies and allow people to connect across differences by redefining “dominant relationship patterns that are culturally instigated” (Ebo, 1998, p. 3). We also agree that the challenges that plague digital spaces are not easily solved and have troubling and far-reaching implications for individuals, society, and, indeed, the world. Yet our goal with this essay is to put forth a normative framework for what digital spaces could be, rather than bemoan their baser proclivities.

In doing this work, we draw inspiration from Althaus’s (2012) critique that the normative criteria employed when evaluating the media are often asserted without critical reflection. Here, we aim to not only chronicle various normative criteria but also discuss how these criteria can sometimes conflict with each other. Our work is in the vein of other projects that try to determine quality criteria for democratic functioning, such as the quality criteria for evaluating democratic decision-making posed by Price and Neijens (1997). The point of departure here is that we focus on digital public spaces and the criteria to which they could aspire. We also build on the literature that has considered how changes to the design and architecture of social media could make them a space for “genuine dialog” that is characterized by positive attributes, such as mutuality and empathy (Kent & Taylor, 2021, p. 2; see also Kent, 2013). However, our focus is proposing a framework for evaluating the social media experience, including the content, design, and perceptions of users.

We begin by asserting an overarching argument for why social media have unrealized potential for normative value in a democracy. Then we propose 14 indicators, which we call signals, that describe the type of digital space that could be beneficial for individuals and democratic society. These signals fall into four main categories: welcome, which involves how digital spaces can be hospitable; connect, which includes bringing people together productively; understand, which focuses on making sure that people have the information they need; and act, which deals with activism and resilience. We link these 14 signals into a normative framework of what the digital space could look like and critically evaluate how these signals interact.

Public-Friendly Digital Spaces

When originally conceived, social media were seen as spaces for entertainment and interpersonal connection, some targeting narrow segments of the population. In 2004, Facebook was limited to Harvard University students and then to other college students before it was opened to the general public (boyd & Ellison, 2007). Twitter, in its infancy, prompted users to tell their friends, “What are you doing right now?” Responses seemed “oddly personal and tame compared with today’s dizzying Twitterverse—a cacophony of news and information, political talk, and sincere connections between people, interspersed with vitriol, hate speech, abuse, and disinformation” (Sobieraj et al., 2020, p. 1,647).

Today, an estimated 3.8 billion people worldwide use some type of social media platform ¹ and the multiple ways in which social media are used extend far beyond how these platforms were originally conceived. Social media have become an important source for news, politics, and democratic discussions (Newman et al., 2020). Politicians use social media to tailor campaign messages (e.g., Gerodimos & Justinussen, 2015), foster intimacy with the public (McGregor et al., 2017), and serve as a virtual megaphone (e.g., Wells et al., 2016). As these examples illustrate, social media now frequently serve as public spaces. Unlike offline public spaces, which are purposefully designed to accommodate different people and facilitate productive interactions, social media did not have this original purpose. Imagine if social media, however, were designed like a well-planned public park or piazza (Pariser, 2019). Sidewalks would be organized to make it easy for people to gather. Water fountains and seating would be plentiful but not impede the flow of foot traffic. Trees would be just where people need them to stay cool in the midday sun. The way the space looked and felt would tell people how to use it. Although this is an imperfect metaphor—public spaces most certainly are not always sites of societal functioning—it is a pragmatic metaphor for what we aim to do in this article, namely, outline criteria for well-functioning digital public spaces. To do so, we draw from criteria that have been developed based on ideals for offline public spaces. Our hope is that by providing a framework of normative criteria, we can inspire more critical engagement with norms, as Althaus (2012) suggests, and more research into normative outcomes that have been understudied or that could conflict with one another when thinking about digital spaces.

Notably, we are not the first to consider how social media could become better. Kent and Taylor (2021), for example, argue for changing the architectural and design features of social media to make them more hospitable to dialogic engagement, which is characterized by relationship-based interactions where people take turns, are autonomous, and treat others as valued, particularly in regard to activism. They propose a series of architectural features to facilitate dialogic engagement, including requiring that users have public profiles and agree to site rules and that platforms allow a variety of interaction buttons where users can “agree” with or “respect” a post, rather than just “like” it (see also Stroud et al., 2017). Place and Ciszek (2021) caution that dialogue cannot occur unless platform architecture takes into account that different groups have varying levels of societal power, so the goal should be “making spaces for the multivocality that represents the lived experiences of those who have been, and continue to be, erased” (p. 8). Our work builds on this literature by proposing a framework for evaluating the social media experience. Although there are design elements that would yield more normatively desirable experiences based on our criteria, we do not pre-specify them as there could be variation across time and the type of platform in terms of what elements work best to create a desired end. We do, however, provide examples of the signals in practice.

Specifically, we consider how social media algorithms could be changed to promote societal good, as opposed to focusing on boosting engagement or de-emphasizing what is problematic. Algorithms, the “black box” of social media (Diakopoulos & Koliska, 2017), could be trained to surface content in line with normative criteria such as those outlined here. Beyond algorithms, features of platforms also could be designed to emphasize societal functioning (e.g., Stroud et al., 2017) as opposed to features that maximize business-related, attention-focused criteria like time on site.

A model focused on attracting attention creates a challenge on social media because attention is finite. Emotions, needs, and social identity, among other factors, influence what attracts attention (e.g., Stroud, 2017). Content that is eye-catching (Wu, 2016), novel, threatening (Marcus et al., 2000), or deviates from the norm (Shoemaker, 1996) tends to grab attention. As a result, people often gravitate to negative content (e.g., Bene, 2017; Smith & Searles, 2014), so it seems reasonable that algorithms would promote negative content to capture users’ attention (Feezell et al., 2021). Yet, other factors draw people to content, such as timeliness, proximity (Shoemaker, 1996; Wise et al., 2009), relevancy (Chakraborty et al., 2017), and personal interests (e.g., Bolsen & Leeper, 2013). Thus, it is feasible the public could be drawn to more normatively valuable content with the help of algorithms that optimize for good and platform architecture that emphasize normatively desirable outcomes. Social media could provide people with content that is more useful or valuable, rather than most provocative. That raises the question, of course, of what is useful or valuable or good. Our answer is to catalog 14 signals that make up a normative framework, each detailed below. ²

Building a Normative Framework

In our framework, we adopt a normative theory approach. Rooted in the writings of early Greek philosophers like Plato and Aristotle, normative theory employs specific arguments to make a point, rather than drawing on empirical research in a more positivistic tradition (e.g., Bauböck, 2008). Normative theory concerns what is both feasible and desirable in society, suggesting that inquiry need not be value-free. Assessments of the media often include normative assumptions that are not always explicit and often are not critically evaluated (Althaus, 2012). By presenting a catalog of normative considerations for digital public space, we aim to generate more critical attention.

Our framework proposes 14 signals that are grouped into four categories: welcome, connect, understand, and act (Table 1). These categories and the signals that flow out of them are all injunctive or prescriptive (Burgoon, 2015; Cialdini et al., 1991; Kallgren et al., 2000), meaning they theorize what social media ought to be. Each offers an aspiration of what would be normatively valuable in a democratic society, so that social media enrich human communication. However, each category and signal are also linked to what is descriptive or predictive (Burgoon, 2015; Cialdini et al., 1991; Kallgren et al., 2000) of social media today. In other words, each signal seeks to solve a problem that mires the digital space. For example, we urge that social media welcome people because often the vitriol, misinformation, and harassment online render them unwelcoming. We envision social platforms as places that connect people because all too often people feel divisiveness in these spaces. Understanding should be facilitated on social media because the reality is these platforms are frequently spaces where people are confused by the plethora of conflicting information. Social media should facilitate action, we argue, because sometimes people feel frozen in place by the jarring nature of online discourse.

Table 1.

Normative Criteria for Evaluating Digital Space.

Welcome	Connect	Understand	Act
■ Invites participation ■ Ensures safety ■ Encourages humanization ■ Keeps information secure	■ Cultivates belonging ■ Builds bridges ■ Strengthens local ties ■ Makes power accessible	■ Elevates shared concerns ■ Shows reliable information ■ Fosters civic competence ■ Promotes thoughtful conversation	■ Boosts community resilience ■ Supports civic action

Our model also has a logic to it with the first category, welcome, leading to the second, connect, and then to the third, understand, and the fourth, act. But it is not strictly linear. For example, welcoming could in some cases lead directly to understanding, or action could improve understanding. In some cases, action may draw people to social media, and then they will feel welcomed because of that activism. Thus, while we envision a logic to our model, our main initiative is to theorize the specific signals that make up this model, not to predict that social media always follow a rigid format. We draw liberally from the political communication literature in developing our signals because this subdiscipline has a developed literature on deliberative democracy that examines how heterogeneous groups can, and do, communicate (e.g., Mutz, 2006). We also draw from other aspects of democratic theory scholarship, namely, theories of participatory and liberal democracies (Helberger, 2019). However, we propose these signals in the spirit of pragmatist philosopher Dewey (1888), who proposes that all of a person’s everyday activities can be more democratic. Even a digital garden club can advance these ideas. Thus, our signals would be applicable to both political and non-political digital spaces.

Welcome

In theorizing ideal digital public spaces, the most basic—and essential—responsibility is to foster a place where, at the very least, people feel welcome. By welcome, we mean people feel safe entering this space as well as comfortable remaining in it. We propose four signals regarding whether a space is accessible and hospitable. It must be a place that invites participation, ensures safety, encourages humanization, and keeps information secure.

Invites Participation

By invites participation, we mean social media could create opportunities for all people to take part (Wilner, 2021c). Certainly, social media in theory allow anyone to take part, but, in reality, certain groups have been minoritized in society, and they can feel unwelcome on social media (e.g., Sobieraj, 2021). Other groups are left out because they lack the technology or knowledge to use social media (Van Dijk, 2020). Groups that may fall into either or both of these categories include the poor, older adults, ethnic and racial minorities, women, people with disabilities, rural residents, and the socially isolated. Inclusion is important if we want a society where diverse viewpoints surface and that does not discriminate. Exposure to minority viewpoints can improve decision-making, increase information-seeking (Nemeth & Rogers, 1996), and bolster learning (Karlsen et al., 2017). It can also enable people to appreciate that a variety of opinions exist on a topic, lead to attitude changes (Stromer-Galley, 2003), and prevent further inequalities (Helsper, 2008). Foss’ (2009) concept of invitational rhetoric aligns with what we mean by invites participation, as it rejects the elitism and privilege of social media and instead sees all people as having “immanent value” (p. 570) as sentient beings who should feel welcome online. An exemplar of this signal would be the non-profit E-Democracy, which hosts more than 50 local forums for 17 communities in three countries with a goal of making diverse groups feel welcome. ³

Ensures Safety

We conceptualize this signal as a means to protect people from online harm or danger (Fiesler, 2021) because harassment (Chen, 2017; Coe et al., 2014) and bullying are common online (Festl & Quandt, 2013). These virtual attacks can lead people to withdraw from digital spaces, making them less diverse (Sambasivan et al., 2019), and harming people’s perceptions of minoritized groups (Weber et al., 2020). Physical harm that stems from online attacks has also been well documented: Abusers use technology for surveillance and intimidation (Freed et al., 2018), while hacking and revenge porn have led to suicides (e.g., Valdés, 2019). Efforts to boost safety include detection and deletion of content violating publicly disclosed rules. Reddit, for example, allows individual subreddits to create their own content rules.

Encourages Humanization

Encourages humanization means recognizing the humanity of other individuals by affirming their human characteristics (Wilner, 2021b), such as intelligence, rationality, and individuality (Haslam, 2006), and seeing them as deserving of moral treatment (Opotow, 1990). This signal addresses a descriptive norm on social media where people dehumanize outgroups (e.g., Harel et al., 2020), which can lead to genocide, slavery, torture (Opotow, 1990), and discrimination (Leyens et al., 2007). Humanization, however, enables cooperation, forgiveness, acts of heroism, and peacebuilding after conflicts (Bandura, 2002). An example of this signal would be when Spaceship Media brought together hundreds of American women with diverse political views in Facebook groups and helped them see the humanity in each other (Schmidt, 2018).

Keeps Information Secure

This signal relates to preserving the confidentiality of information, accurately setting people’s privacy expectations, and protecting information (Renaud, 2021). It addresses the descriptive norm of violations of information security, including identity theft, doxing personal information, and installing malware or viruses. Consequences of these violations can be dire and long-lasting, including anxiety and physical distress (Sharp et al., 2004). Addressing this signal requires platforms to make their security options more transparent and understandable (Renaud & Shepherd, 2018). For example, companies like Twitter, PayPal, and Google offer two-factor authentication to help their customers secure their information.

Connect

Once people arrive at digital spaces, and feel welcome, these spaces need to foster bonds among people. Clearly, the fact that digital spaces could connect people has always been part of their architecture (boyd & Ellison, 2007), but we submit that connecting is more than just bringing people together. It involves four signals: cultivates belonging, builds bridges, strengthens local ties, and makes power accessible.

Cultivates Belonging

The earliest humans had to affiliate to take on large tasks such as hunting or fighting predators to survive, and this need to belong remains strong and evolutionarily adaptive (Baumeister & Leary, 1995). Belongingness is associated with lower stress levels, decreases in depressive symptoms, lower mortality and morbidity rates for many mental and physical illnesses, and higher self-esteem (Cockshaw et al., 2014; Holt-Lunstad et al., 2010). Digital spaces are particularly suited to encouraging camaraderie between people (Chen, 2011) and fostering affective publics (Papacharissi, 2015) that can cultivate this sense of belonging. In line with this signal, digital public spaces could be designed to encourage belonging and to ward off the threat of disconnection (Curry, 2021). An illustration of this signal is the website Meetup, where people join interest groups for real-life gatherings.

Builds Bridges

By building bridges we mean fostering social connections between clusters of individuals who might not ordinarily connect (Van Duyn, 2021). Such groups could be based on location, religion, race/ethnicity, political ideology, or other commonalities. Bridges between these groups constitute “weak ties” (Granovetter, 1973) that leverage bridging social capital, an information exchange between people who do not hold a pre-existing connection (Putnam, 2000). Social media can be particularly adept at these types of exchanges; however, the harassment, bullying, and dehumanization that occur on social media (Coe et al., 2014; Festl & Quandt, 2013; Harel et al., 2020; Sobieraj, 2021) can undermine these benefits. Building bridges between groups can reduce political polarization (J. K. Lee et al., 2014) and health disparities (Poortinga, 2012), as well as increase political participation (Giugni et al., 2014). An exemplar of this signal is “Germany Talks,” ⁴ a newspaper-created platform that matches people with their ideological opposites and provides conversation guides for in-person meetings.

Strengthens Local Ties

Strengthens local ties is similar to building bridges, but it relates to bonding social capital, which involves bolstering connections between people who are already linked (Putnam, 2000), and its focus is on geographically proximate others. This signal addresses forces that divorce people from their local communities, such as the closure of local papers. This signal involves surfacing content that will help people to improve their local ties to physical places and the communities associated with them (Riedl, 2021). These ties are emotional, structural, and, to some extent, participative. Such ties improve people’s emotional well-being, including their sense of relaxation (Scannell & Gifford, 2017) and social support (Krämer et al., 2021). These ties can increase civic engagement (Stefaniak et al., 2017), lead to better health (Burgess et al., 2019), improve public safety (Scannell & Gifford, 2010), satisfy critical information needs (Friedland et al., 2012), and prepare people for disasters (Klinenberg, 2015). For example, NextDoor specifically connects people within their real neighborhoods.

Makes Power Accessible

Makes power accessible means that social media facilitate a dynamic where the public is heard by those in power, including government, business, and other institutions, and has access to the decision-making process (Jennings, 2021). The democratic ideal is that public voices are reflected in policy, and people’s influence over policy does not depend on their social positions. Of course, this does not often occur because the hegemonic societal power structures that render some with less power operate online much as they do offline (Ebo, 1998). Facebook’s Town Hall function, for example, displays users’ local, state, and federal representatives, with buttons for contacting them. LinkedIn also allows people to connect with those in power.

Understand

Ultimately, if social media are welcoming and allow people to connect, they should increase users’ understanding of the people, places, and information around them. People do not have to all agree with each other, but society as a whole would benefit if our digital spaces helped people agree on fundamental truths and appreciate the differences in context that shape different worldviews. This category includes the following four signals: elevates shared concerns, shows reliable information, fosters civic competence, and promotes thoughtful conversation.

Elevates Shared Concerns

Elevates shared concerns means social media should increase the visibility of issues that are important to significant subsets of the population yet are relatively unknown to the broader public (Wilner, 2021a). This signal addresses the descriptive norm on social media whereby popular, or trending, content is surfaced for others to see. Instead, social media could privilege content that people should know about, raising the profile of critical information from the news media, lawmakers, and other stakeholders. This could result in people being better able to discuss issues with each other (Chan & Lee, 2014), more people perceiving these issues as important, and lawmakers and regulators prioritizing these topics (e.g., Boydstun, 2013). An example of this signal is how Twitter worked with the World Health Organization (WHO) during the COVID-19 pandemic to highlight information about the virus. ⁵

Shows Reliable Information

Critically, social media must show reliable information, meaning displaying or emphasizing facts that can be verified using the best available evidence reasonably available to the creator or publisher (Wilner, 2021d). Because the spread of false information is a descriptive norm on social media (Vosoughi et al., 2018), and false beliefs are difficult to change (Nyhan & Reifler, 2019), social media must supply people promptly and frequently with reliable information. The spread of false information has led to disastrous outcomes, such as mob violence (Banaji et al., 2019). Reliable information, however, can help people participate in democracy (Eltantawy & Wiest, 2011), manage their health (K. Lee et al., 2014), and perform well in their jobs (Benesch, 2012). To enact this signal, platforms could alter their algorithms so that they surface more trustworthy sources to more users and remove problematic accounts; organizations like NewsGuard and the Trust Project are attempting to do this work.

Fosters Civic Competence

Fosters civic competence refers to increasing people’s ability to perform democratic duties, such as voting, communicating with their representatives, and sitting on a jury (Harness & Wilner, 2021). This signal is necessary because vitriol and misinformation on social media may undermine civic competence. Civic competence is vital for a thriving democracy because it gives citizens the ability to make good decisions, as well as an understanding of how to use civic tools. In general, when people are more politically and civically knowledgeable, they are better able to perform their role as voting citizens (Shah et al., 2009), evaluate arguments (Zaller, 1992), and resist illogical persuasive appeals (Luskin, 1990). This signal maps onto ideas from liberal democratic theory, where being a good citizen focuses on acts like voting (Helberger, 2019), and it also incorporates ideas from participatory democracy (Mutz, 2006) with its focus on other forms of civic action besides voting. An example of this signal is Google’s Voting Information Project, which provides local ballot information for users. ⁶

Promotes Thoughtful Conversation

This signal suggests that ideal digital public spaces should encourage exchanges that exhibit elements of deliberation, such as when people are receptive to others’ perspectives or share the reasoning behind their opinions (Murray, 2021). This signal is needed because deliberative discussions are not the norm on social media (Denisova & Herasimenka, 2019). This signal relates to the ideals of deliberative democracy because fostering more deliberative exchanges can create a more informed public that is exposed to opposing perspectives in a way that may help them see those views as legitimate (Mutz, 2006). The result is that people may be more likely to cooperate with opposing groups (Chambers, 2003) and engage civically (Valenzuela et al., 2011). We argue that online discussions can be heated or contain emotional arguments because to prohibit this may undermine the “democratic merit of robust and heated discussion” (Papacharissi, 2004, p. 260). Yet, verbal aggression that attacks people as opposed to ideas (Infante & Wigley, 1986) would undermine this signal. An illustration of this signal was Ceasefire, a website that required that users provide reasons for their opinions, be open to change, and reply to others in substantive ways. ⁷

Act

The last piece of our model is the action that comes out of the understanding, the connecting, and the welcoming of digital spaces. By action, we mean being both proactive and reactive. This breaks down into two specific signals that suggest that social media boosts community resilience and supports civic action.

Boosts Community Resilience

This signal suggests that social media could help geographic communities, such as a city or a neighborhood, better rebound from significant adverse events, such as natural disasters, acts of terror or violence, and public health emergencies (Whipple & Lee, 2021). For example, Facebook and WeChat played pivotal roles of emotional and informational support during Hurricane Harvey, which ravaged Houston, Texas, in 2017 (Li et al., 2019). Resources such as health services, jobs, trusted information sources, and political partnerships (Norris et al., 2008) are critical aspects of resilience, and social media can help to surface these resources and connect people to them. This results in better mental health and well-being for people (Ungar & Theron, 2020). Another example of this signal is Google’s SOS alerts, which provide information and emergency contacts during a disaster.

Supports Civic Action

By supports civic action, we mean the actions people take to identify and address issues of public concern (Delli Carpini, 2000) to make a difference in civic life (Ehrlich, 2000). Such actions can include volunteerism, donating to charities, voting, attending rallies, going to community meetings, and getting involved in political or social justice organizations (Delli Carpini, 2000). Thus, this signal relates to participatory democracy (Helberger, 2019; Mutz, 2006), which focuses on these specific acts. Civic action is considered a positive value for society because it is linked to individual benefits, such as increased interpersonal trust (Theiss-Morse & Hibbing, 2005) and improved health (Grimm et al., 2007), and societal advantages, such as reduced crime rates (Rosenfeld et al., 2001). Social media features, like hashtags on Twitter (Freelon et al., 2018), and events on Facebook (Papa, 2017), facilitate civic action by helping people join together around political or social issues, but bad actors can co-opt these features to undermine these efforts (Cole, 2015). This signal suggests that social media platforms should encourage features that facilitate civic action and take steps to ensure these features are not easily sabotaged (Tenenboim & Jennings, 2021). For example, California Counts hosted a Voter Cram Session where participants learned about upcoming ballot propositions. ⁸

Discussion

Our goal in this research was to theorize a normative framework for evaluating what digital spaces, such as social media, could be. We catalog four categories of signals for digital public spaces: welcome, connect, understand, and act. We conceptualize each of these categories and the 14 signals that flow out of them as injunctive or prescriptive norms—what the internet ought to be—and contrast each with descriptive or predictive norms—what social media actually are much of the time. Our framework proposes that it is vital that social media platforms optimize their spaces through algorithms and platform architecture to foster normatively valuable content and experiences for the public in support of a democratic society. We believe this is essential because social media were created in a piece-meal way without an overarching notion of what they should be. Our normative framework provides this structure for scholars and the public to evaluate these spaces. We summarize here what our framework proposes and suggest implications for this work.

Our framework proposes normative ideals for a digital public space that is welcoming to everyone, and by welcoming we mean: invites participation, ensures safety of participants, encourages humanization, and keeps people’s information secure. These spaces also must make it easy for people to connect by cultivating a sense of belonging, fostering the construction of bridges between groups, strengthening local ties, and making the powerful more accessible. This connection, we believe, will lead to greater understanding through elevating shared concerns, showing reliable information, and creating online experiences that foster civic competence and promote thoughtful conversations. Ultimately, social media can make it more possible that people act by boosting community resilience and supporting civic action.

The normative criteria outlined here are neither exhaustive nor mutually exclusive. Our hope is that this outline can help scholars and practitioners identify normative goals that dominate current scholarship or platform design, such as showing reliable information, and expose others’ goals where there is more work to be done, such as encouraging humanization. It also begs for revision and updating as scholars and practitioners consider novel normative implications of digital public spaces.

The framework serves to expose challenges in normative criteria that could motivate social media platforms. For instance, allowing everyone to participate could make social spaces feel less hospitable to some. Strengthening local ties certainly can be useful for building trust, support, and intimacy (e.g., Krämer et al., 2021), but in other situations, such as providing new information, bolstering weak ties between proximally distant people may be more valuable (Ellison et al., 2014; Granovetter, 1973), which is why we include building bridges between distant others as a signal. Indeed, a benefit of social media is how they allow people to connect with non-proximate others around a shared goal or cause and form online communities (e.g., Freelon et al., 2018). Cultivating belonging could mean allowing homogeneous groups to form, which could work at cross-purposes with building bridges across networks. Promoting thoughtful and deliberative conversation could reduce civic action (Mutz, 2006). We urge scholars and social media platforms to consider how to harness the power of these tensions. It is outside the scope of this essay to articulate specific affordances from individual platforms that could help or undercut these signals. But we urge platforms to take these signals into account as they update their algorithms, architecture, and designs, and develop new digital spaces.

Our signals do not suggest specific architectural or design changes; rather they offer a normative framework for assessing what outcomes such changes should work toward. In doing this work, we build upon scholarship that examines how design and architectural changes can improve public space (e.g., Kent, 2013; Kent & Taylor, 2021). Our contribution is to articulate a vision for social media that takes into account the descriptive norms of the space and links each to an injunctive norm of what the space could be.

Practically, this normative framework could prove useful to platforms. It is worth noting that the framework is agnostic as to whether these ideals are taken up by existing or new platforms. We are skeptical that any existing platform lives up to all of these signals, as none were designed with them in mind and the profit motive that drives most social media platforms can work at cross-purposes. However, as we highlighted earlier, some platforms have features that speak to some of the signals. The framework is equally agnostic on how social media could design for these normative ideals. It could be done algorithmically by surfacing content that speaks to these goals. Alternatively, it could be included in new products, such as the COVID information provided by many of the major social media platforms.

In summary, social media are neither utopian dreams, nor dystopian nightmares. Rather, like any type of reflexive modernization, they offer dialectical challenges where a seemingly positive change results in a “mass of unintended side effects” (Beck et al., 2003, p. 2). Social media make people feel better but also worse; they simplify people’s lives but also complicate them (Masullo et al., 2020). Our framework suggests a way for social media scholars and platforms to more thoughtfully consider the normative implications of their work.