Coding the Privileged Self: Facebook and the Ethics of Psychoanalysis “Outside the Clinic”

Facebook’s Mission as Psychoanalysis “Outside the Clinic”

In 2011, Zuckerberg suggested that by providing information access and enabling people to form relationships with “more empathy” and “more understanding,” Facebook could “make the world better.” Openness and connectivity, he said, were Facebook’s “founding principles” (Brigham Young University 2011). Zuckerberg reiterated this vision in 2012, telling shareholders that Facebook’s mission “is to make the world more open and connected” (Facebook 2012). He considers his work to include a significant ethical component aimed at cultivating the virtues of authenticity, integrity, and compassion.

In Zuckerberg’s view, the Internet itself is a positive moral force, and Facebook has “the social responsibility to help spread the Internet and all the goodness that comes along with that” (Zuckerberg 2014c). To this end, in 2015 Facebook launched the Internet.org project (a multi-telecomm initiative bolstering global Internet access) to bring the “benefits of connectivity to the portion of the world that doesn’t have them” (info.internet.org/en/mission). Zuckerberg sees Internet.org as an effort to “increase social inclusion” and “close the digital divide” (Zuckerberg 2015). As long as this divide persists, humanity at large is “being robbed of the creativity and potential” of communities not yet online (Zuckerberg 2014b).

Meanwhile, Zuckerberg’s recent “Building Global Community” essay includes allusions to moral-developmental theory and explicit claims about the role of social networking in promoting psychological well-being. Facebook’s current goal is to harness the power of algorithms to strengthen the “intimate social structures we turn to for our personal, emotional and spiritual needs” (Zuckerberg 2017). This “social infrastructure” includes churches and sports teams—groups which provide “a sense of purpose and hope . . . moral validation that we are needed,” as well as “mentorship, guidance and personal development” (ibid.). Company officials like Chief Product Officer Chris Cox tout the evolution of Facebook’s mission “toward closeness—which is more than connectedness” (Nieva 2017).

Zuckerberg even proposes technological fixes for social problems widely understood to be enabled or exacerbated by the Internet. “Artificial intelligence can help provide a better approach” to issues like cyberbullying and live-streamed violence, he claims (Zuckerberg 2017). Facebook’s research also focuses on “the impact of sensationalism and polarization, and the idea of building common understanding” (Zuckerberg 2017). Such efforts have yielded positive results, he notes, though it will take years before solutions reach their full potential.

Facebook’s stated commitment to empathy and understanding may reflect Zuckerberg’s personal beliefs, but previous research suggests the company’s publicly stated social mission obscures its commercial motives (Hoffmann et al. 2016, 8–12). While Zuckerberg once claimed that Facebook was “not originally created to be a company” but to “accomplish a social mission” (Gassee 2012), he now argues that commercial organizations are the key to stabilizing “our social fabric” (Zuckerberg 2017). “Building a company is the best way to change the world,” he argues, adding that profit-seeking companies are uniquely capable as compared with nonprofits (InsideFacebook.com 2010; Startup School 2010). While acknowledging that markets fail to “fund or incentivize building the necessary infrastructure” for public-good projects, he positions such projects as charitable endeavors best subsidized by commercial organizations like his (Zuckerberg 2017). In these ways, Zuckerberg positions Silicon Valley CEOs as the new gatekeepers of psychological well-being and for individual users and the culture at large.

Analyst-engineers and the Privilege of “Integrity”

The discursive convergence of Facebook’s social mission and its commercial goals manifests in specific design and policy decisions. Facebook enables connectivity, but “commodifies that connectivity” at every step (Hoffmann et al. 2016, 9). As user policies prioritize revenue at the expense of user agency, outcomes undercut the company’s proclaimed commitment to personal development and the cultivation of meaningful communities.

Despite its community-building rhetoric, Facebook’s construction of the self continues to hinge on the assumption that identity is a personal project where individuals uncover and express a static self. But, as Frosh (2010, 191) notes, “the emphasis on the individual as the object of knowledge makes specific ideological claims.” Indeed, this position reflects a privileged theoretical standpoint. For example, many Silicon Valley elites participate in self-development programs like Wisdom 2.0 and report discovering an authentic identity through mindfulness meditation. The stress-relief benefits of such programs notwithstanding, such elite programs represent “integrity bubbles” wherein “a small but satisfied group enjoys an experience of harmony unavailable to others” (Healey 2013). In fact, integrity requires us to move beyond our social persona—that is, the aspect of self prioritized by social media—and instead become aware of, and responsible for, others (Stevens-Long 2000, 169–71). This discrepancy is especially apparent in Facebook’s much-criticized “real identity” policy.

Engineering the “Real” User-patient

Facebook’s “real identity” policy exemplifies the conflation of human virtue and market values by refusing any meaningful distinction between online profiles and what Zuckerberg calls the “real self.” The policy embodies the problematic notion that informational transparency is equivalent to authenticity—a pervasive assumption within Silicon Valley’s techno-centric ideology (Healey 2017, 363–66). This notion is both inaccurate (positing the authentic self as a static entity) and ethically fraught (positing “sharing” as a normative good). Nevertheless, it forms the foundation of Zuckerberg’s self-styled role as engineer-analyst as Facebook performs its own brand of commercialized psychoanalysis outside the clinic.

For Zuckerberg, people have a “core desire to express who they are” (Rose 2011). Facebook’s architecture assumes that this core desire can, and should, be realized online. In fact, COO Sheryl Sandberg has insisted that “[t]he social web can’t exist until you are your real self online” (Rose 2011). Zuckerberg described user profiles as “where you have everything about you” and where “millions of people” go to tell “the story of their life” (F8 2011, 2011). An individual’s personal history, he says, is “all there and you can get to it all if you wanted. It’s your whole life” (F8 2011, 2011). As Facebook creates “opportunities for people to share more things,” he argues that such sharing “leads to this kind of broader social change where I think now there’s more transparency” (Computer History Museum 2010). In other words, by providing tools to realize a basic psychological need, Facebook catalyzes a broad cultural shift toward greater authenticity.

While these kinds of assertions initially seemed plausible, public backlash erupted after Zuckerberg’s defense of the “real identity” policy, when in 2009 he moralistically denounced users’ construction of multiple profiles as “an example of a lack of integrity” (quoted in Kirkpatrick 2010, 199). While he later acknowledged that “there are lots of services that are good with pseudonyms,” he insisted that Facebook’s approach was still superior “if you’re talking about mapping out a real community” (Y Combinator 2013). The issue flared up again late in 2014, when Facebook “launched a new crackdown on people using stage names, adopted nicknames, or drag names as their primary profile names” (Barmann 2014a).

Critics lambasted Zuckerberg’s naive understanding of the virtue of integrity. Reporter Jay Barmann (2014b) describes it as “ridiculously short-sighted and anti-intellectual.” Zimmer (2010) suggests that “Zuckerberg must have skipped that class where Jung and Goffman were discussed.” Indeed, those theorists understood the self as multifaceted, with no single perspective or presentation accounting for the entirety of an individual. The various facets of the self do not always cohere logically but often contradict each other. As Thompson (2015, 323) explains, the self “isn’t a thing or an entity; it’s a process.” The self consists of a “potential for growth that will naturally unfold under the right circumstances, and that can be damaged or hidden in a hostile environment” (Frosh 2010, 113).

Admittedly, some developmental approaches understand identity coherence as “a normative psychological state” and failure of such achievement as a “psychological disturbance” (Frosh 2010, 100). This approach is common in earlier iterations of psychoanalytic theory and persists in some recent approaches to sexual development (ibid., 109). Yet critical approaches insist that identity is not static but “a fragile and contingent process” (ibid., 102) unfolding within morally fraught contexts. We locate ourselves within specific, and usually multiple, identity positions which “may be quite fluid and even mutually contradictory” but in which we “become heavily invested” (ibid., 99; see also Phoenix and Rattansi 2005). Our multiple identities are “lived as real, but are in fact constituted in the flux of contradictory impulses” (Phoenix and Rattansi 2005, 102). Just as authenticity is not equivalent to total transparency, neither is psychological integrity equivalent to perfect consistency in one’s self-expression (online or otherwise). Furthermore, the construction of identity is not a solitary project but a matter of group belonging—especially where group identities are ascribed and marginalized (Phoenix and Rattansi 2005, 100). Some identities may appear, and may be lived, as more “real” than others insofar as they fit within preconceived normative bounds. Integrity is therefore not a simplistic matter of personal moral rectitude but a relational commitment requiring a willingness to take others’ concerns into consideration and insist on their fair treatment (Beebe 2000, 12; Calhoun 1995, 260).

The analytical error embedded in Facebook’s real identity policy can pose real risks. Enforcement of the policy burdens the user-patient with “a demand for a certain mode of narrative integrity that is simply not possible without doing violence to the actuality of the fragmented psyche” (Frosh 2010, 146, summarizing Butler 2005, 42). The lesbian, gay, bisexual, transgender, and queer (LGBTQ) community clearly perceived this danger and the real identity policy sparked particularly intense outrage among drag queen users (Barmann 2014a, 2014b), many of whom maintain a separate profile for their drag persona. Transgender users also protested, as in Hoffmann’s (2016) argument that transgender folks “use fake names to manage their own conflicting audiences by making themselves visible to those who may offer love and support, but invisible to those who may wish them harm.” In this sense, Facebook’s real identity policy bolsters the company’s bottom line “at the expense of the wellbeing of its most vulnerable members” (Hoffmann 2016). Zuckerberg’s dismissal of such concerns exemplifies how normalizing discourses “remove the other’s otherness” through processes where “as soon as the other claims a right to maintain its difference, it is regarded as hostile” (Frosh 2010, 131).

Meanwhile, users who accept the premise of the real identity policy—that their profile should portray one coherent identity—face a more subtle risk. A seemingly coherent public-facing self always exists within the more complex structure of a developing self. Here Jung (1966, 192) offers the concept of the persona, which is “a kind of mask” designed to outwardly present some aspects of the self but not others. Meanwhile, Goffman ([1959] 1973) distinguishes between “front stage” performances intended for a general public and “back stage” performances which assume a particular audience in a private setting. The social-self is thus merely one aspect of the whole self, and healthy self-development involves not total transparency but defensive maneuvers of concealment (Balick 2014, 114). Such maneuvers provide a “continuity of being” that protects against “impingement, virtual or otherwise” (Balick 2014, 17). Problems arise “when the individual identifies with their persona at the expense of the other attributes of their personality: when they believe the persona to be ‘the whole thing’” (Balick 2014, 18).

Unhealthy identification with the persona may exacerbate preexisting narcissistic tendencies. Narcissism is not simply a self-conscious regard for one’s own virtues. Only in extreme cases does it involve a sense of entitlement, a tendency toward exploitation of others, a lack of empathy, or a pattern of grandiosity (Balick 2014, 79). It is in fact “not self-love at all, but a compensatory and vulnerable preoccupation with a reflection of the false self” (ibid., 79). For most users, what Facebook affords is the construction of a persona that is reflected back by others. These reflections “can be experienced as a ‘fix’ for the narcissistic needs in all of us” (ibid., 81). As the outward-facing persona tends to receive the recognition that the “true” self seeks, platforms like Facebook amplify users’ narcissistic tendencies (Carpenter 2012), even if they do not “cause” narcissism (see Twenge 2013). Beyond individuals, such needs exist as part of a broader cultural ethos. Johnson (1987, 155) argues that “less extreme variations” of narcissism are “endemic to the culture,” existing on a continuum including relatively healthy self-confidence (quoted in Balick 2014, 80).

Ultimately, the desire for recognition cannot be satisfied through likes, shares, retweets, or other such affordances, because it is not the core self that is recognized (Balick 2014, 67). Insofar as its platform architecture incorporates the fundamental error of the “real self” policy, Facebook’s affordances may do more to amplify these issues than to catalyze the achievement of authentic selfhood.

“Controlling” User-patient Agency

As Facebook continued to enforce its real identity policy, Zuckerberg’s public statements suggested increasing flexibility. After noting that “identity and anonymity have always been heated topics of debate,” one interview summary concluded that he “seems eager to relax his old orthodoxies” and quoted him as saying, “we’re at the point where we don’t need to keep on only doing real identity things” (Bloomberg Businessweek 2014). He went on to suggest that the acquisition of anonymous services like Instagram had resulted in the company being “a little bit more balanced” (Bloomberg Businessweek 2014). Such rhetoric spurred follow-up commentary under headlines such as “As Facebook turns 10, Zuckerberg changes his mind about anonymity” (Smith 2014).

This discursive shift makes sense in light of the relationship between Facebook’s real identity policy and its business model. As van Dijck (2013, 200) explains, “platform owners have a vested interest in pushing the need for a uniform online identity to attain maximum transparency, not only because they want to know who their users are, but also because advertisers want users’ ‘truthful’ data.” Like any entity dependent on ad revenue, Facebook’s core product is its users, and product value increases with Facebook’s ability to serve targeted ads. Linking users to a single “real identity” maximizes both the amount and value of data. This dynamic is especially important in aggregate terms, as the ability to identify correlations between categories of users enables the creation of accurately predictive user profiles even in the context of incomplete user-submitted data.

An understanding of the relationship between identity, data, and revenue allows for a finer parsing of Zuckerberg’s statements on privacy and anonymity. Facebook’s toleration of anonymity on acquired apps like Instagram and WhatsApp does not disrupt the collection of data tied to a single identity. User profiles are linked to Facebook accounts, making a user’s “real identity” known to Facebook even when hidden from other users. As long as user-generated data can still be monetized, Facebook is willing to tolerate anonymity.

Typically, Facebook edges toward such tolerance only after its preference for user transparency creates a backlash. Over time, Zuckerberg began to respond to privacy concerns with rhetorical gestures toward greater “control” for users. This strategy emerged in response to public backlash against Beacon, an ad program that connected Facebook profiles to activity on third-party websites (see Havenstein 2007). In 2009, Facebook settled a class action lawsuit which alleged that Facebook failed to provide “adequate notice and choice about Beacon” and its exploitation of user data (Brodkin 2009). Beacon was shut down, and Facebook agreed to provide US$9.5 million in funding to a nonprofit privacy organization (Brodkin 2009). In 2010, Zuckerberg explained that control is “a core tenet of Facebook” (Computer History Museum 2010). A year later, he reiterated that “People want control . . . and as long as we roll out products, we continue on the controls” (Rose 2011). So pervasive was the pointed use of “control” that Facebook engineers spoke publicly “as if a memo had been circulated that morning instructing them never to use the word ‘privacy’” (Zimmer 2014).

Importantly, Zuckerberg’s goal in providing “the right tools” for such control is that over time, “people become comfortable sharing more things” (Building 43, 2010; see also Hoffmann 2014). Zuckerberg’s rhetoric creates “the appearance of more control” by touting new “tools” allowing users greater choice in their online sharing (Freishtat and Sandlin 2010, 516). Crucially, such statements reduce privacy to matters of social interaction among friends while ignoring third-party data collection. While “control” is among the most frequently appearing terms in the Zuckerberg Files, it refers primarily social privacy (users’ ability to decide how content is shared with other users) and only secondarily to informational privacy (how meta-data may be used by third parties). In fact, user control over data decreased over the company’s first ten years as new advertising platforms began “spreading users’ information beyond their direct control” (Zimmer 2014). While our analysis confirms these previous findings, we add that Facebook’s discourse of control obscures the company’s opaque and nonconsensual exploitation of data in ways that impinge on psychosocial processes of identity formation.

One problem with Facebook’s discourse of control is that it exploits users’ latent desire for omnipotence. This desire is rooted in an early-childhood phase where the infant believes it controls its environment, and others within it, without distinguishing others as others (Balick 2014, 107). Adults continue to waver in and out of believing they are, or wishing themselves to be, in control. Social media platforms tap into such desires by “allow[ing] the user to control what happens in their interface and achieve psychological distance if necessary” (Balick 2014, 107). Zuckerberg’s concern for making users more “comfortable” confirms this point. However, this affordance also encourages users to “imagine others as you wish them to be, constructing them for your purposes” (Turkle 2011, 187, quoted in Balick 2014, 108). Insofar as Facebook’s definition of “control” conceals issues of algorithmic manipulation (e.g., the use of geo-fencing to restrict access to the LGBTQ “Pride” button for users in certain locations), this dynamic may undermine the achievement of mutual recognition between users.

Another problem lies in Facebook’s misleading claim that third parties cannot directly access user data. Zuckerberg has insisted that “We do not and never will sell any of your information to anyone” (Zuckerberg 2010) and that “no information ever goes to advertisers” (Brigham Young University 2011). Instead, Facebook simply serves ads to audiences whom advertisers request. It does this, however, by collecting user data for the construction of online marketing profiles of which users are largely unaware. Although nonanonymized user data may be withheld from advertisers (or from third-party platforms accessed through Anonymous Log In), marketing profiles nevertheless play a determinative role in how users are positioned in relation to those advertisers, to other users, and to Facebook itself. As Facebook’s discourse of “control” results in decreased attention to informational privacy, users facilitate the construction of passive online identities which may impinge on processes of self-formation.

Engineering the User-patient’s Passive Identity

While profiles are shaped by users’ choices, they are also structured by algorithms written by Facebook and third-party engineers. Users are at a disadvantage in this iterative process, because the platform’s algorithms are opaque and many features are compulsory. Users unwittingly co-create a “passive online identity” that may impinge on both individual and collective processes of identity formation (Balick 2014, 111). The paradoxical result is that while control over our online profiles may increase, our ability to control others’ perceptions of us is “far less under our control—than ever before” (Palfrey and Gasser 2008, 34; quoted in Balick 2014, 139).

For example, Facebook’s News Feed algorithm is designed to reflect what the company deems “important” or “interesting” to users, from “global news” to “updates about your friends” (Zuckerberg 2014c). Such algorithmic curation has important implications for “how we interact as complex subjects” (Balick 2014, 74), especially given the convergence of user-generated “organic” content and advertising (i.e., “paid content”). As Facebook’s director of product management explained, “Just like we connect people with people . . . we connect advertisers to people” (Zuckerberg 2014a). On an earnings call, Zuckerberg explained that “our goal here is to make ads as interesting and useful as your friends’ content” (Facebook 2014). For their part, users (including commercial organizations) can “boost” their visibility by paying Facebook to feature their content in others’ News Feeds. Commercial interests therefore impinge on the dynamics of identity formation in ways that are invisible to most users.

Socioeconomically marginalized groups are more vulnerable to virtual impingement by passive, highly commercialized online identities. News Feed algorithms may neglect news about or of particular interest to minority users, as when protests in Ferguson, Missouri, failed to trend as quickly as on Twitter. Facebook’s handling of hate speech is likewise concerning, as in 2013 when the company faced boycotts for its failure to monitor misogynous content. More recently, internal documents revealed that censorship algorithms appear to define white men as a protected group but not African American children (Angwin and Grassegger 2017).

As platforms encourage users to share personal information, marginalized groups also risk “exposing too much unprotected unconscious material to an external world that might not respond with care” (Balick 2014, 19). Passive identities are constructed also by insurance companies and banks whose exploitation of user data sometimes amounts to economic coercion (Healey 2014, 199). Facebook patented a technology allowing lenders to review credit ratings for users’ authorized connections (i.e., “friends”) and reject loan applications if the average credit rating is below a minimum score. Critics decry this program as discriminatory because it makes vulnerable users “more prone to using alternative predatory systems such as payday loans” (Zombeck 2015; see also Sullivan 2015). While Facebook has yet to act beyond the acquisition of the patent, neither has it responded to such concerns.