Introduction

In the last two decades of privacy debates, much of the emphasis has come to focus not merely on the individual, but also on the social dimensions of privacy. This new perspective has changed and broadened the earlier understanding of privacy that centered on the meaning and value of privacy for individuals and for individual rights. There has been consensus that the significance of privacy is almost always justified for the individual interests it protects, most importantly protections of freedom and autonomy in a liberal democratic society. Most of the authors writing on various social dimensions of privacy retain their belief in the value of privacy for individuals defended in the past to control information about oneself (Fried Reference Fried1968; Parent Reference Parent1983); to enhance personhood including an individual’s inviolate personality, integrity and human dignity (Bloustein Reference Bloustein1964; Reiman Reference Reiman and Roessler2004), as well as human flourishing and well-being (Roessler Reference Roessler2005; Moore Reference Moore2010); to allow for human intimacy (Gerety Reference Gerety1977; Gerstein Reference Gerstein1978; Cohen Reference Cohen1992; Inness Reference Inness1992); as the essential context for fundamental relations of respect, love, friendship and trust to give one the ability to mediate various social relationships (Fried Reference Fried1968; Rachels Reference Rachels1975; Schoeman Reference Schoeman1984); and to protect an individual’s ability to restrict or allow others’ access (Gavison Reference Gavison1980; Bok Reference Bok1983; Allen Reference Allen1988). Others see and defend privacy as a cluster concept allowing individuals to control information about themselves, to control access to them and their physical selves, and to preserve their ability to make personal decisions about themselves and their families, their activities and lifestyles (Allen Reference Allen1988; Schoeman Reference Schoeman1992; DeCew Reference DeCew1997).¹ Defenders of the social value of privacy, however, are now drawing attention to the additional value of social understandings of privacy. The general discussion of privacy has advanced as questions are being asked about the social dimension of privacy. Concerns over the accessibility and retention of electronic communications and the expansion of camera surveillance have led commentators to focus attention on loss of individual privacy as well as on privacy protection with respect to the state and society (Regan Reference Regan1995; Reiman Reference Reiman and Roessler2004; Nissenbaum Reference Nissenbaum2010).

According to Daniel Solove, “[b]y understanding privacy as shaped by the norms of society, we can better see why privacy should not be understood solely as an individual right … Instead, privacy protects the individual because of the benefits it confers on society” (Solove Reference Solove2008: 98). Moreover, “the value of privacy should be understood in terms of its contribution to society” (Solove Reference Solove2008: 98, 171 fn.). Solove believes privacy fosters and encourages the moral autonomy of citizens, a central requirement of governance in a democracy. These views on the instrumental value of privacy to society develop from the earlier philosophical writings on the value of privacy – that it heightens respect for individual autonomy, individual integrity and human dignity, but also enhances the value of privacy in various social roles and relationships that contribute to a functioning society. According to this contemporary scholarship, privacy norms help regulate social relationships such as intimate relations, family relationships, professional relationships including those between a physician and a patient, a teacher and a student, a lawyer and a client, and so on. Thus privacy enhances social interaction on a variety of levels, and in this way enhances intimacy, self-development and the ability to present ourselves in public as we wish. According to Solove, a society without respect for privacy for oneself and others becomes a “suffocating society” (Solove Reference Solove2007: 15).

A moving account of understanding privacy as a necessary and also indispensable condition for individual freedom as well as for a democratic society comes from a literary quotation from Milan Kundera:

The analogies between Kundera’s scenario and electronic surveillance and street cameras common in society today are clear. Kundera helps us see the interconnections between privacy as a value for individuals as well as for a flourishing and free society that is not suffocating.

What I am interested in is what this shift from individual to social dimensions of privacy means for the feminist critique of privacy. If we understand privacy in a more social way, does this have any consequences for the feminist perspective? Are there societal contexts, for instance, which show that a social understanding of privacy can more plausibly address the feminist critique? Thus, in addition to a history of the feminist discussions, I shall focus on the more systematic implications of the social dimensions of privacy for a feminist perspective. In particular, I shall argue that we can and should retain a fundamental concept of privacy, but may understand social considerations of context in a way that justifies appropriate invasions of privacy to enhance the public and collective value of privacy and social well-being.

One problem has been trying to articulate what exactly are the interests protected by privacy concerns, and what is the current and appropriate scope of privacy protection. Many have found this a serious and intransigent difficulty. It has been difficult for philosophers and others to provide clear guidelines for understanding what privacy protects and why it is important, and this is true for defenders of privacy as a value for individuals as well as those writing on the social importance of privacy.

It may be messy and difficult to find adequate words to express just what privacy governs, and it is understandable that some still believe the term “privacy” is too vague and not well enough articulated. But concepts like privacy, and equality, for example, are crucial for understanding our role as social beings and for protecting values fundamental to living lives free from unacceptable individual and governmental intrusions and surveillance.

The public/private distinction and the feminist critique

Nevertheless, the concern about just what privacy protects, and an understanding of privacy’s value for individuals as well as society, lead to additional difficulties about understanding the boundaries between the private and the public in problematic cases, and in particular the darker side of privacy raised by feminist critiques of privacy. There may be ways in which the more recent discussions of the social dimensions of privacy may help advance feminist ideas on the interconnectedness of privacy and the public sphere and may provide some guidance about how to approach and respond to the feminist critique of privacy. In the following, I first want to present important positions in the feminist debate on the public and the private, before I turn in the next section to a discussion of the social meaning of privacy and, in the conclusion, to the idea that this discussion can shed new light on the feminist position.

Early versions of the feminist critique of privacy relied heavily on Aristotle’s distinction in The Politics (Reference Aristotle and McKeon1941) between the polis, or political realm, and the oikos, or domestic realm. The political realm of governing, open to men only, was deemed by Aristotle to be a public arena, whereas the domestic realm of home and family was viewed by him to be a private arena. John Locke provides another well-known example of a historical reference to a public/private distinction. Locke invokes the distinction discussing property in his Second Treatise on Government (Reference Locke and Reardon1690). In the state of nature, he argues, one owns one’s own body and yet other property is held in common, or deemed public. When one mixes one’s labor with property – harvesting grain or catching fish, for example – that which was held in common becomes one’s private property. Although individuals are cautioned to leave “enough and as good” for others, private property acquisition is heralded by Locke as an appropriate goal. The distinction between public and private spheres of life has continued to influence and dominate much of the scholarship on privacy, perhaps to the detriment of an adequate understanding of privacy.

The feminist critique of privacy has been discussed by philosophers and legal theorists from the 1960s through the 1990s, citing privacy’s potential to shield domination, repression, degradation and physical harm to women and others. The most famous version of this critique was advanced by Catharine MacKinnon in Reference MacKinnon1989. She observes that the law of privacy fails to recognize and take into account the pre-existing oppression and inequality of women. For MacKinnon, privacy represents yet another domain where women are deprived of power, all on the suspect theory that “the government best promotes freedom when it stays out of existing social relationships” (MacKinnon Reference MacKinnon1989: 164–5).

MacKinnon is making two distinct but related claims here. The first is that women have no privacy, and hence protecting privacy provides no benefit to women. Privacy protection may even be a detriment to women, giving men the legal right to treat their wives and partners (and children) unequally or even brutally. The old rape shield laws, for example, made it impossible for women to claim their husbands raped them. The second claim is that feminism has demonstrated the importance of criticizing the split between public and private domains, and thus “has had to explode the private.” Consider each in turn.

Why is it that women have no privacy to lose or guarantee? MacKinnon’s answer appears to be that because women are violable and violated, they have no zone of autonomy within which to control their destinies. In particular, in the realm of sexuality, often viewed as a paradigm of the private, women do not have control. Men can and often do maintain their power over women in such intimate circumstances. Although sexual intimacy, and activities within the home and family, may be private in the sense of being withheld from public view and shielded from governmental intrusion, they are not private in the sense of being areas where women have control over their decision-making. That women are in fact violated in private contexts, however, implies nothing about the worth and value of protecting a zone within which they can have the power to limit intrusions and violations. In short, descriptive facts about actual limitations on privacy fail to imply anything about the normative value of seeking privacy protection for women.

MacKinnon’s second point in this passage underscores the importance of rejecting the public/private split. The public/private distinction has captured the imagination of many feminist scholars. In fact a substantial portion of feminist theory and political struggle over the past 200 years has been concerned with deconstructing the traditional notion, going back as far as Aristotle, of a public (male) political realm and a private (female) domestic realm. Some of the most influential work in feminist political theory, philosophy and legal theory takes this paradigm as its starting point in analyzing women’s oppression. Carol Pateman goes so far as to claim that the “public/private dichotomy is, ultimately, what the feminist movement is about” (Pateman Reference Pateman and Pateman1989: 188). Despite this emphasis on the public/private distinction, it is difficult to clarify what the feminist critique of it entails. Feminist scholars such as Ruth Gavison and Carole Pateman have made clear that there is no single or privileged version of the feminist critique of privacy. There is a multiplicity of interwoven ways of understanding attacks on the public/private dichotomy. MacKinnon says we must “explode” the private, so appears to believe there is no distinction between public and private because there is no private realm for women at all. But this hardly establishes that there should be no public/private distinction and that there should be no private realm for women.

In part MacKinnon is, like other prominent feminists, drawing attention to the degree to which sexual and physical violence in the family has been a degrading and life-altering experience for so many women (and children and some men). She is surely correct that abusive relationships in those traditionally private contexts are pervasive. To the extent that the private or domestic sphere is held unavailable for public scrutiny, abuse and degradation can continue unchecked. When there are legal avenues for women to combat abuse, the system often does not or cannot enforce them effectively. Moreover, on MacKinnon’s view, both the public and private spheres exhibit the social power of sexism. The subordination of women to men is evident in public, and in private it is mirrored and allowed to run its course, “inaccessible to, unaccountable to … anything beyond itself” (MacKinnon Reference MacKinnon1989: 190).

We can clearly agree with MacKinnon that the distinction can work to the detriment of women. But what is the alternative? If the line between public and private is sometimes indeterminate, does it follow that nothing is or should be private? If there is no distinction between public and private, is everything public? One interpretation of MacKinnon’s view is that we must completely reject the realm of the private and conclude that everything is public. Susan Moller Okin writes: “The protection of the privacy of a domestic sphere in which inequality exists is the protection of the right of the strong to exploit and abuse the weak” (Okin Reference Okin1989: 174). The rejection of the dichotomy is accomplished by collapsing the private into the public. Others have viewed this as a plausible reading of the feminist critique of privacy, for instance Jean Bethke Elshtain (Reference Elshtain1995: 43), and a similar understanding of the feminist critique is echoed by Ruth Gavison, who observes: “Usually when the dichotomy between public and private is challenged, the argument is that all is (or should be) public” (Gavison Reference Gavison1992: 28). Yet Gavison quickly notes that feminists often equivocate when confronted with the implications of this rejection of the public/private split:

Thus, even if women are often vulnerable and exploited in the private, domestic sphere, we may ask whether there are no contexts in which women wish to keep the state out of their lives. MacKinnon appears to say yes; nevertheless, I believe the answer must be no. Anita Allen has suggested that an analogy between privacy and liberty is helpful here. Just as the harm that results from the exercise of individual liberty does not lead to the rejection of liberty, similarly there is inadequate reason to reject privacy completely based on harm done in private (Allen Reference Allen1988: 40).

Feminists have correctly identified the ways in which the public/private distinction can be dangerous if it is used to devalue the work of women in domestic roles, to silence them politically by categorizing them as having no voice or value, and to allow the continuation of abuse and degradation under the cover of a private sphere unavailable for public censure. Thus feminists are right to urge that the distinction not be used to justify differential social and legal treatment of women.

But we need not reject the concept of privacy altogether. Some feminist authors who have wanted to rescue privacy, including Allen (Reference Allen1988), DeCew (Reference DeCew1997) and Roessler (Reference Roessler2005), have continued to focus on a concept of individual privacy. While defenders of privacy may have underestimated the emphasis on individual male power and domestic abuse, critics may underestimate the implications of state-sponsored expressions of control over women. Consider, for example, intrusions such as government sterilization programs and interventions involved in state control over welfare programs, including the withdrawal of benefits from women upon the birth of additional children. Consequently rejecting the public/private distinction by eliminating privacy obscures the difference between individual and institutional expressions of (male) power.

Thus, arguing for the rejection of privacy, according to its defenders, fails to address the need to differentiate between justified and unjustified uses of state power over individuals. Governmental regulation might refer to reasonable laws regarding family matters, such as giving women the right to charge husbands with rape. But it might, on the other hand, mean that the state will reveal and regulate all the embarrassing details. Evaluating the justifiability of state intervention may be exceedingly difficult, and exploitation and abuse should be matters of public concern. But even if we agree with feminist insights about the oppression and inequality of women, the reality and pervasiveness of abuse, and the dangers of distinguishing a private domestic realm immune from public scrutiny that preserves the status quo, there may still be value in making a distinction between public and private.

This value is what the second feminist interpretation of privacy seeks to defend.² On this alternative interpretation, rejecting the public/private divide by collapsing the private side onto the public is neither the feminist point nor an implication of the feminist position. According to this second feminist account (for instance Allen (Reference Allen1988), Roessler (Reference Roessler2005) and DeCew (Reference DeCew1997), but also Pateman (Reference Pateman and Pateman1989) and Olsen (Reference Olsen1983), see below), the boundaries between public and private need to be redrawn. They would not jettison privacy but recognize that what happens in the family is not always beyond scrutiny. An alternative understanding of the feminist critique of privacy, therefore, is that feminists merely want to reject the public/private distinction as it has been understood in the past, from Aristotle on. These feminists are emphasizing that the state must stop ignoring the unbelievable abuses that have been protected in the name of privacy; this is, they believe, a position that is not captured by the public/private distinction as it has been known and used in pre-feminist times and theories. On this account, feminists are talking about a position that bypasses the public/private distinction in a different way.

For example, Frances Olsen discusses the radical separation of two spheres of activity – the (public) market and (private) family – and their relationship to two other dichotomies, between state and civil society and between male and female. Olsen describes strategies for improving the status of women, in part because of the unequal bargaining power of women compared to men. She insists:

Olsen urges that the public/private and market/family dichotomies are a way of thinking, a human creation; they are a prism through which we have come to see our lives. It is not enough to recognize the crippling effects of such dichotomies.

Olsen worries that rejection of the public/private split will lead to an alternative system “in which the state controls every aspect of human life and nothing is personal and private” (Olsen Reference Olsen1983: 83), and she emphatically rejects this implication, saying she does not advocate replacing the present dichotomies with an all-powerful state. Instead, Olsen urges that we neither reject the humanization and connectedness of the family nor the efficient production of goods and services in the market. It is their separation and polarization that reinforces the status quo and limits possibilities of human association. On her view, “[w]e cannot choose between the two sides of the dualism because we need both” (Olsen Reference Olsen1983: 88). The preferable alternative is to transcend the dichotomies in some yet-to-be articulated way, preserving a meaningful role for an important concept of privacy.

Carole Pateman reiterates the feminist challenge to the separation and opposition between public and private spheres as central categories of political liberalism, where domestic family life is paradigmatically private. Pateman believes that “the dichotomy between the private and the public obscures the subjection of women to men within an apparently universal, egalitarian and individualist order” (Pateman Reference Pateman and Pateman1989: 20). But she emphasizes that feminists reject the claim that a public/private dichotomy is inevitable:

What is needed, on Pateman’s view, is a feminist theoretical perspective that takes account of social relationships between men and women within the context of interpretations of both the public and the private. Work by political theorists as well as practical experience from the feminist movement, has shown that women’s place in the private sphere cannot simply be augmented by extending to women a role in the public sphere. The spheres are not additive, but integrally related. In sum, Pateman views the feminist critique of privacy as stressing rejection of the dichotomy as it has been understood, but she concludes that the “separate” worlds of private and public are closely interrelated and that both are necessary dimensions of a future, democratic feminist social order. An adequate account will develop a social theory in which these categories are distinct but interrelated, rather than totally separate or opposed.

Clearly the feminist critique of privacy is multifaceted. Quite a few feminists – among them Allen, DeCew, Roessler, but also Olsen and Pateman – acknowledge the difficulties of the public/private dichotomy and the damaging effects of accepting it as it has been defended in the past. Feminists articulating this strand of the feminist critique maintain that the public/private dichotomy is misleading when it fails to recognize the interconnections between private and public life. The challenge for those unwilling to jettison privacy completely – given the lingering influence, in our culture and law, of the separate spheres analysis that women belong in the home and men in public positions – is to preserve some understanding of the two spheres while extricating them from their gendered past and connotations.

Social understandings of privacy

At this point the feminist critique connects to recent debates about the value of privacy for society and the social dimension of privacy. Drawing on Pateman’s emphasis on accounting for social relationships between men and women within the context of interpretations of both public and private, makes it clear that feminists should be interested in new social approaches for understanding privacy and hopefully addressing the feminist critique. Regan (Reference Regan1995), Solove (Reference Solove2008) and Nissenbaum (Reference Nissenbaum2010) defend social understandings of privacy and it is promising to link this approach with the feminist ideas on the interconnectedness of privacy, public and social relationships.

Priscilla Regan began to take account of the social value of privacy, noting that Westin and others saw privacy and social participation as competing desires that an individual would need to balance. Rejecting this conflict, Regan argues that threats to privacy can be individual concerns as well as social problems. “Most privacy scholars emphasize that the individual is better off if privacy exists; I argue that society is better off as well when privacy exists” (Regan Reference Regan1995: 221). Regan argues that privacy is a common value, a public value and a collective value (Regan Reference Regan1995: 213). She echoes others who have argued that privacy is valuable for a democracy by creating a place for deliberation, bolstering the voting process and helping to develop people who are capable of the kind of participation that democracy requires. Ultimately Regan’s argument is: “Privacy is becoming less an attribute of individuals and records and more an attribute of social relationships and information systems or communication systems” (Regan Reference Regan1995: 230).

Daniel Solove expands on this social understanding of privacy, worrying about the concept of privacy being in disarray, but he argues that privacy is important across the globe, and recognized explicitly by the United States, countries in the European Union, Asia-Pacific and elsewhere. Discussing examples of privacy violations leads him to note that responses that treat them as one kind of violation are not as nuanced as they should be, and he urges that being aware of context in privacy violations needs to be supplemented so that the concept of privacy is not overstretched to become an empty concept. Solove addresses different conceptions of privacy developed by legal scholars, philosophers and social scientists and argues that many of these views also fail to explain why the way in which issues deemed private are often formed within a social context.

Solove defends his own conception of privacy, understood as various kinds of good that have similar characteristics, that can be found in particular contexts and are flexible enough to be understood in different cultures and times. He writes: “Therefore, my approach to conceptualizing privacy understands it pluralistically rather than as having a unitary common denominator. In focusing on privacy problems, my approach seeks to be contextual without being overly tied to specific contexts, flexible enough to accommodate changing attitudes toward privacy, yet firm enough to remain stable and useful” (Solove Reference Solove2008: 9, Allen 2003). His emphasis on the social dimension of privacy is based in part on his argument that the value of privacy should be rooted in its utility for society, not individual rights, because its value is determined by the social importance of the activities privacy makes possible.

Solove focuses on a pragmatic approach embracing a focus on specific situations, and thus sensitive to contextual situations. He cites Nissenbaum’s framework of the “contextual integrity” of privacy in arguing for the importance of context, but he also believes that context is not enough. Anita Allen’s insight into the role of accountability is also important: people are held, or feel, accountable within relationships, not “across the board” (Solove Reference Solove2008: 48). What counts as private is not accidental on Solove’s view. To the contrary, issues are private because people want them to be.³

Solove does not find feminist critiques persuasive enough to lead to rejecting privacy, and he points out that privacy can liberate women as well (see Allen, DeCew, Gavison), thus in my view setting the stage for how his view that privacy is instrumentally valuable for specific problems and contexts deemed important in society can be helpful in replying to the feminist critique of privacy. “Privacy should be weighed against contrasting values, and it should win when it produces the best outcome for society” (Solove Reference Solove2008: 87). The value of privacy is social because it leads to benefits to society, not simply the individual. One such benefit may be in addressing domestic abuse in private contexts.

Although Solove recognizes privacy’s importance for information, I applaud the expansion of his view to be about a broader concept covering invasions including physical intrusion and decisional interference. In his conclusion Solove also addresses cultural differences in approaches to privacy. On his view different cultures, when balancing privacy problems against other values, may come to different conclusions, with some cultures putting little value on some privacy problems while protecting against others, because these privacy problems can be weighed differently depending on the historical and situational context. As we shall see, this may help address the feminist critique of privacy, but may raise other difficulties as well.

Let us next have a look at a second approach to the social dimensions of privacy. Helen Nissenbaum’s position is narrower than Solove’s work in that it focuses solely on informational privacy. She writes: “What people care most about is not simply restricting the flow of information but ensuring that it flows appropriately, and an account of appropriate flow is given here through the framework of contextual integrity” (Nissenbaum Reference Nissenbaum2010: 2, original emphasis). Thus I believe she fleshes out the idea of contextual integrity in ways that may provide more specific details about how to use a social understanding of privacy to reply to the feminist critique of privacy. Several of her key concepts can be adapted to be useful in a broader way that could include domestic and family life.

Nissenbaum notes that the public/private distinction has been a common source of our conception of privacy, bringing both a positive aspect of private space and a negative aspect of over-inclusiveness in the ways we use the concept of privacy. Citing DeCew, she agrees that the public/private division can be difficult to define and is used in different ways depending on the discipline or author, but can be defined in a useful way to use privacy as a protective barrier between individuals and government or others, as well as a line between the domestic and political realms. Yet she concludes: “The framework of contextual integrity … does not take sides in these controversial line-drawing exercises; rather, it reveals them to be symptoms of the deeper problem of invoking the private/public dichotomy to inform our understanding of privacy” (Nissenbaum Reference Nissenbaum2010: 102). Therefore, the public/private dichotomy is not necessarily problematic, but is not useful as a normative conception of privacy. On her view privacy standards have not changed, but threats to privacy have. She concludes that the private/public dichotomy leaves too many theoretical gaps to be used as a justification for privacy, yet she does not reject it outright. And she stresses that her own contextual integrity view is neutral with regard to alternative theories of privacy’s value.

What is more important, according to Nissenbaum, is to focus on context, because privacy expectations are “systematically related to characteristics of the background social situation” (Nissenbaum Reference Nissenbaum2010: 129) and “Contexts are structured social settings characterized by canonical activities, roles, relationships, power structures, norms (or rules), and internal values (goals, ends, purposes)” (Nissenbaum Reference Nissenbaum2010: 132). Contexts will be determined by different roles, such as teacher or physician, activities conducted by people in those roles, norms specifying acceptable actions and practices guiding behavior, and values which are the goals and purposes of a given context. These attributes are a way of understanding a “snapshot” of a particular context, where the latter – values, goals and purposes – are clearly the most important. Nissenbaum urges that contexts may have both formal norms and values as well as tacit and general ones, and this can generate disagreements about privacy. Norms may work in unity or may conflict. There may be no general solution to the conflicts, but we navigate our way through such conflicts often in daily life. Contexts such as health care or the workplace will have a background of goals such as the health of patients in medical care, as well as norms, and norms may differ for different actors (one’s physician or employer) and they can change. Her point is to have norms that give appropriate guidelines of behavior, for example, for the transmission of medical information, and these may differ for one’s physician and one’s employer. But “contextual integrity is a proposed benchmark for privacy” (Nissenbaum Reference Nissenbaum2010: 140). Nissenbaum believes there are entrenched norms on information-gathering and sharing in many contexts, but those may change with new technologies.

My suggestion is that contextual integrity may be a useful framework for explaining appropriate and inappropriate informational transmission, as proposed by Nissenbaum, but may also serve as a useful framework for explaining appropriate and inappropriate behavior in the context of the domestic realm. Clearly there are formal and explicit laws and rules against domestic violence, incest, wife-battering and child abuse in the USA and multiple other countries. Moreover, in the domestic realm there are actors in roles such as mother, father, parent, child, extended family, and so on, as well as activities performed within the family in the domestic context. There are entrenched norms about appropriate parental caretaking, meeting children’s basic needs as well as enrichment and enhancement for child development and preserving respect for partners, all set within the goals and purposes of the context, for example that the family unit is, in general, meant to provide a safe space where parents and children can thrive without abuse. It seems to me that we need not reject the domestic realm as a private realm, nor reject privacy as being normatively valuable due to its protection of freedom, equality and the good life, but can recognize that spousal abuse, battering, rape, as well as child abuse violate explicit laws and entrenched norms of appropriate behavior in the family that justify external intervention from government or other public agencies including social services. The default is that privacy protection is fundamental, but considerations of contextual integrity can provide a secondary set of considerations to justify appropriate intervention.

While I use several of Nissenbaum’s concepts, there are differences between our views. My claim is that the understanding of privacy has to be consistent with the appropriate norms governing our lives (private and public). There may be a wide range of agreement about which norms are appropriate, but there will be some disagreements as well. Nissenbaum, however, seems to imply that the appropriateness of privacy contexts is inherent to the privacy norms themselves (but see Nissenbaum Reference Nissenbaum2010: 182). I believe there are more fundamental underlying norms that can be used to criticize certain privacy contexts and can explain when privacy can be overridden. While it is not completely clear what makes norms or behavior appropriate or not, I believe that we can make some determinations about appropriate norms based on the connection between privacy and its use in defending equality and freedom.

Nissenbaum’s framework allows that even entrenched norms may change to allow for progress, as she notes that for her focus on information technology there have been recent shifts in what is viewed as normal. For example, new technologies such as body screening at airports and video surveillance are now viewed as normal. “The approach I recommend here is to compare entrenched normative practices against novel alternatives or competing practices on the basis of how effective each is in supporting, achieving or promoting contextual values” (Nissenbaum Reference Nissenbaum2010: 166). Similarly, in the domestic realm of the family, a historical understanding of the “traditional” heterosexual family is now evolving to recognize and validate same sex partnerships, without altering norms about appropriate respectful and non-abusive behavior between partners. Contextual integrity looks at the purpose of the context of domestic family life, for example to see if a new practice such as same sex marriage helps or hinders that purpose. Clearly new social understandings of same sex marriage indicate that most now view it as reasonable and not abnormal, and believe it does not hinder the goal of providing a caring space and unit where all members can be supported and thrive without harm. Looking at the big picture and not just single privacy decisions, as Nissenbaum apparently recommends, can be an important way of evaluating and prescribing appropriate behavior in the domestic realm.

This approach need not lead us to totally jettison the public/private dichotomy and the fundamental value of privacy, nor to view privacy as a relativistic notion as Solove appears to do, but to understand that within private contexts there will in addition be contextual purposes and goals, norms for actors and appropriate activities and behavior within those contexts, leading to insightful ways of prescribingappropriate behavior and justifying external intervention when those norms are violated. We can recognize individual violations against women in the domestic realm but judge them as inexcusable and inappropriate in the domestic context to achieve Solove’s goal of creating the best outcome for society.

Nissenbaum believes that use of her contextual integrity framework for information access and transmission is primary, and will mean that changes over time and culture may no longer be an issue, and this will be a benefit in the domestic realm where changes in social views on consenting adult homosexuality and same sex marriage can be understood and embraced. Nevertheless, I treat privacy as fundamental with the contextual integrity framework as a secondary way to determine when invasions into a private realm can be justified. However, I do not feel that even this useful though secondary contextual integrity framework will solve all the difficult issues of privacy protection across religions and cultures. There is clearly the difficulty of determining which norms are appropriate guidelines for behavior – also a problem for Nissenbaum’s work on information technology – and there will be particularly thorny issues in the context of the domestic realm. For example, one may wonder whether parents should have the power to withhold cancer or other medical treatments from children – for religious or other reasons – when such treatments are medically sound and have been shown to increase survival rates. While some may view this as a completely private family decision, it seems extremely difficult to justify in the face of clear harm to the children. As a family decision it may be viewed as presumptively private, but it may be that explicit and entrenched norms prohibiting harm to others and norms defending some degree of paternalism in certain cases may conflict with deeply entrenched religious beliefs. Contextual integrity does not resolve the conflict on its own, but at least clarifies which norms are in conflict, and suggests that considerations of the goals and purposes of the domestic family, including how to enhance children’s abilities to thrive, are clearly important goals for society.

Consider another example. Deeply entrenched cultural beliefs, in favor of female genital mutilation, for example, seem to lead to the conclusion that even with “consent” the state should not necessarily remain out of the affairs of individuals. Otherwise it would seem that a physician must honor a woman’s rational yet culturally entrenched decision that the physician must perform genital mutilation surgery on her. Thus it seems there must be some constraints on what counts as significant and meaningful consent. Here, norms about a physician’s health care role in enhancing patient health and doing no harm may help support his refusal to accede to the patient’s request. Yet questions in such a conflict may remain.

Conclusion

So where do we stand? In the first part of this chapter I have given an outline of two interpretations of the feminist critique of privacy and the private/public distinction, one that rejects privacy and collapses it into the public realm, and one that preserves the distinction while recognizing that it has to be understood in a new way. In the second part I have presented two approaches to the social understanding of privacy and have tried to show ways they connect and help address the feminist critique. Acknowledging that there will be some extremely difficult cases, it nevertheless seems to me that recognizing the important social dimensions of privacy explained by Regan and the emphasis on context rather than individual privacy claims defended by Solove and Nissenbaum, and endorsing Solove’s broad conception of privacy extending to decisional contexts along with Nissenbaum’s contextual integrity approach, help us see ways in which a fundamentally private context such as the domestic realm can be viewed through a lens of appropriate norms of behavior and important social goals for enhancing security, personal excellence and well-being in society. Public and social goals may be intertwined in addressing violence against women and domestic abuse. As not only Allen, DeCew and Jean Cohen, but also Pateman and Olsen have argued, we need not reject important uses of the term “privacy” and its fundamental value, and we need not be chained to the ancient public/private dichotomy. Theories of privacy will need to say more about how to determine which norms are appropriate or inappropriate, and will need to move forward on the discussion of cultural differences in perspective on privacy and social norms. However, we may have found a theoretical perspective that can be endorsed by feminists that embraces the importance of privacy, yet allows privacy invasions when taking account of social relationships between men and women within the context of various areas of daily life.

Introduction

While the balance between individual privacy and government monitoring or corporate surveillance has been a frequent topic across numerous disciplines, the issue of privacy within the family has been largely ignored in recent privacy debates. Yet privacy intrusions between parents and children or between adult partners or spouses can be just as profound as those found in the more “public spheres” of life. Popular access to increasingly sophisticated forms of electronic surveillance technologies has altered the dynamics of family relationships. Monitoring, mediated and facilitated by practices of both covert and overt electronic surveillance, has changed the nature of privacy within the family. Parents are tracking children via GPS-enabled cell phone tracking software and are monitoring the Internet use of family members. Parents, siblings, and children are also posting information about their family members online, often without consent, and are creating social media profiles for others online. Prior scholarly work in philosophy and law has primarily addressed the privacy of children from third parties, usually commercial entities, and in the context of making medical decisions.¹ Less attention has been directed at exploring a more general right of privacy of one family member against parents, siblings, children, or spouses. In this chapter we do just that. In the pages that follow we consider several moral rules that determine appropriate privacy boundaries within the family. More specifically, we will consider when overt or covert surveillance of a child, spouse, or partner by a family member is morally permitted.²

Our discussion proceeds within the conceptual framework developed in prior work. In the next section we define privacy as the right to control access to, and uses of, places, bodies, and personal information and discuss prior empirical and theoretical work establishing the moral value of privacy. While admittedly contentious, we adopt an essentialist or objective account of moral value tied directly to human well-being or flourishing. For example, practices that cause disease or ill health are, on our account, morally disvaluable. Alternatively, practices that promote human health and well-being are morally valuable. The third section considers several studies that analyze the effects of different sorts of monitoring on adolescents. An established claim of child development theory is that children who are not monitored have a greater risk of engaging in problematic behavior – including drug, tobacco, and alcohol use, dropping out of school, and early pregnancy. Interestingly, children who are subject to covert spying, as well as other forms of surveillance both overt and covert, are subject to the same increased risks as adolescents who are virtually abandoned by their parents. Two-way information-sharing based on respect and caring, on the other hand, is connected to decreased risky or problematic behavior. In the final section we argue that privacy, understood as a moral and social mechanism that ensures the proper development and functioning of individuals, is essential for trust and provides the foundation for intimate family relationships.

Included in our discussion is a critique of Anita Allen’s defense of spying on one’s children, spouse, or partner. We conclude by arguing for two rules. A rule of “two-way communication” establishes a practice of trust, respect, and caring. Such a standard would help preserve private spaces within family life – spaces where children, spouses, or partners can appropriately withdraw for purposes of self-development without fear of unjustified monitoring. If in a specific instance or as a practice the two-way sharing rule fails, we retreat to a probable cause standard to determine the appropriateness of parental or other intra-familial surveillance. Probable cause is established when a parent, spouse, or partner has discovered information through logical inquiry that would lead a reasonably intelligent and prudent person to believe that their child, spouse, or partner is subject to, or the cause of, a specific and serious threat to safety or well-being. If we are correct, covert or overt surveillance would only be justified in a narrow range of cases.

Privacy: its meaning and value

While privacy has been defined in many ways over the last century, we favor what has been called a “control”-based definition of privacy (see Warren and Brandeis Reference Warren and Brandeis1890; Westin Reference Westin1967; Gross Reference Gross, Pennock and Chapman1971; Parker Reference Parker1974; Gavison Reference Parent1983; Parent Reference Gavison, Benn and Gaus1983; Allen Reference Allen2003). A right to privacy is a right to control access to, and uses of, places, bodies, and personal information (Moore Reference Moore2003, Reference Moore2008, Reference Moore2010). For example, suppose that Smith wears a glove because he is ashamed of a scar on his hand. If you were to snatch the glove away, you would not only be violating Smith’s right to property, since the glove is his to control, but you would also be violating his right to privacy – a right to restrict access to information about the scar on his hand. Similarly, if you were to focus your X-ray camera on Smith’s hand, take a picture of the scar through the glove, and then publish the photograph widely, you would violate a right to privacy. While your X-ray camera may diminish Smith’s ability to control the information in question, it does not undermine his right to control access (Moore Reference Moore2007).

Privacy also includes a right over the use of bodies, locations, and personal information. If access is granted accidentally or otherwise, it does not follow that any subsequent use, manipulation, or sale of the good in question is justified. In this way privacy is both a shield that affords control over access or inaccessibility, and a kind of use and control-based right that yields justified authority over specific items – such as a room or personal information (Moore Reference Moore2007). For example, by appearing in public and leaving biological matter behind, someone may grant access to specific sorts of personal information. We should not conclude, however, that by granting access this individual has also waived all control over any and all future uses of the biological matter or the information found within.

To get a sense of the importance of privacy and separation, it is helpful to consider similar interests shared by many non-human animals. While privacy rights may entail obligations and claims against others – obligations and claims that are beyond the capacities of most non-human animals – a case can still be offered in support of the claim that separation is valuable for animals. Alan Westin in Privacy and Freedom notes:

More important for our purposes are the ecological studies demonstrating that a lack of private space, due to overpopulation and the like, will threaten survival. In such conditions animals may kill each other or engage in suicidal reductions of the population.

Given the similarities between humans and many non-human animals, it is plausible to think that we share many of the same traits. For example, Lewis Mumford notes similarities between rat overcrowding and human overcrowding. “No small part of this ugly urban barbarization has been due to sheer physical congestion: a diagnosis now partly confirmed by scientific experiments with rats – for when they are placed in equally congested quarters, they exhibit the same symptoms of stress, alienation, hostility, sexual perversion, parental incompetence, and rabid violence that we now find in large cities” (Mumford Reference Mumford1961: 210). These results are supported by numerous more recent studies. Household overcrowding and overcrowding in prisons has been linked to violence, depression, suicide, psychological disorders, and recidivism (see Morgan Reference Morgan1972; Baum and Koman Reference Baum and Koman1976; Edwards and Booth Reference Edwards and Booth1977; Megargee Reference Megargee1977; Paulus et al.Reference Paulus, Cox and McCain1978; Cox et al.Reference McCain, Cox and Paulus1980; Farrington and Nuttal Reference Farrington and Nuttal1980; McCain et al.Reference Cox, Paulus and McCain1984; Porporino and Dudley Reference Porporino and Dudley1984; Ruback and Carr Reference Ruback and Carr1984; Clauson-Kaas et al. Reference Clauson-Kaas, Dzikus, Stephens, Hojlyng and Aaby1996; Fuller et al. 1996).

Cultural universals have been found in every society that has been systematically studied (see Murdock Reference Murdock, Hoebel, Jennings and Smith1955; Nussbaum Reference Nussbaum2000). Based on the Human Relations Area Files at Yale University, Alan Westin has argued that there are aspects of privacy found in every society – privacy is a cultural universal (see Westin Reference Westin1967; Roberts and Gregor Reference Roberts, Gregor, Pennock and Chapman1971). While privacy may be a cultural universal necessary for the proper functioning of human beings, its form – the actual rules of association and disengagement – is culturally dependent (see Spiro Reference Spiro, Pennock and Chapman1971). The kinds of privacy rules found in different cultures will be dependent on a host of variables including climate, religion, technological advancement, and political arrangements. Nevertheless, we think it is important to note that relativism about the forms of privacy – the rules of coming together and leave-taking – does not undermine our claim regarding the objective need for these rules. There is strong evidence that the ability to regulate access to our bodies, capacities, and powers and to sensitive personal information is an essential part of human flourishing or well-being.

Barry Schwartz, in an important article dealing with the social psychology of privacy, provides interesting clues as to why privacy is universal (Schwartz Reference Schwartz1968; see also Mill Reference Mill1859; Rachels Reference Rachels1975). According to Schwartz, privacy is group-preserving, maintains status divisions, allows for deviation, and sustains social establishments (Schwartz Reference Schwartz1968: 741). Privacy also preserves groups by providing rules of engagement and disassociation. Without privacy or what may be called a dissociation ritual, there could be no stable social relation. As social animals we seek the company of our fellows, but at some point interaction becomes bothersome and there is a mutual agreement to separate. Thus, having “good fences” would be necessary for having “good neighbors” (Rachels Reference Rachels1975: 331).

Schwartz also notes that privacy helps maintain status divisions within groups. A mark of status is a heightened level of access control. Enlisted men in the armed services have less privacy when compared to commissioned officers. Line level employees work without doors or secretaries who screen access to them. By protecting status divisions and determining association and disassociation rules, privacy has a stabilizing effect on groups or social orders (see McGinley Reference McGinley1959: 56). Privacy also protects and leaves room for deviation within groups. As J. S. Mill noted in On Liberty (Reference Mill1859, Chapter 2), when individuals engage in different forms of living, protected by the walls of privacy, new ideas are produced and, if good, are adopted.

Growing up can be understood as the building of a series of walls – the walls of privacy.³ Infants are without privacy. As infants grow into toddlers and begin to communicate with language, they express wishes for separation at times. This process continues as children grow into adults.⁴ Toddlers and small children begin requesting privacy as they start the process of self-initiated development. More robust patterns of disassociation continue as children enter puberty. Finally, as young adults emerge, the walls of privacy have hardened and access points are maintained vigorously.

As an example of the universal yet relational aspects of privacy, consider the following set of cases. Privacy in Native American communities is best understood within the context of a philosophy that includes two critical values: relationality and the sacred. According to Wilson: “Rather than viewing ourselves as being in right relationship with other people or things, we are the relationships that we hold or are part of” (Wilson Reference Wilson2008: 80).

Relationality means that matters of place, people, and community are inextricably intertwined. In addition, Native American knowledge systems and ways of life see the mental, physical, social, and spiritual domains of existence flowing into and informing one another. For Native people, this value system governs proper behavior in relation to people and the spaces in which they interact. As to place, “home” in this context is a sacred space, encompassing more than house. Recall that, until relatively recently, many indigenous dwellings consisted of one or two rooms that contributed to communal living practices.

There are numerous daily ceremonies and rituals that underscore the reverence that is essential to the concept of “home.” Marisa Duarte (Reference Duarte2013) expresses this relationship in the Yaqui community in the following statement:

In comparison to the Mexican American side of her family, she recognizes a critical difference: “Where [on] the Mexican American side, children are allowed to close and lock their doors, not report their daily comings and goings, and not introduce their friends to the family. It is considered rude, but not inappropriate. When those things happen in Yaqui families, the child is considered spiritually unwell and in danger” (Duarte Reference Duarte2013).

Privacy exists in Native American communities to ensure the dignity and respect of persons. Nevertheless, proper behavior, in reference to physical contact, varies by tribal community. Duarte notes that:

The impact of technology, mobile phones, and tracking devices on the behavior of Native children in reservation communities is difficult to assess, primarily because no comprehensive assessment of technology in Indian Country has been conducted. However, in contemporary reservation communities families rely on neighbors and other family members to keep track of their children’s whereabouts. People know the kids in the neighborhoods/sections of the reservations. They know where the kids hang out or go to play. They know the place. It is important to recall that families have often lived in these places for hundreds of years.

Privacy and surveillance in the family

As mentioned above, the starting point for our discussion begins with a conceptualization of the role of privacy in human development and the place for surveillance in familial relationships. Scholars in a variety of fields, including developmental and clinical science, have investigated the impact of parental monitoring in a variety of settings. Ann Crouter and Melissa Head (Reference Crouter, Head and Bornstein2002) distinguish between two concepts that have historically been confused and under-conceptualized in the relevant literature, leading to disconnects between theoretical concepts and empirical measurement: parental monitoring and parental knowledge (see also Crouter et al. Reference Crouter, Bumpus, Davis and McHale2006). Traditionally, parental monitoring has been defined as “a set of correlated parenting behaviors involving attention to and tracking of the child’s whereabouts, activities, and adaptations” (Dishion and McMahon Reference Dishion, McMahon, Ashery, Robertson and Kumpfer1998). Crouter and Head argue that empirical measures used to investigate parental monitoring have often actually measured parental knowledge, while much less research has focused on the practice of parental surveillance (the activities and technological tools employed to keep track of children) (Crouter and Head Reference Crouter, Head and Bornstein2002: 461).

Additionally, recent research has begun to identify that a high level of parental knowledge is much more closely related to “trusting parent–child relationship[s]” and a “child’s willingness to confide” in a parent than it is to the actual practice of parental surveillance. Thus parental knowledge can be described as “a relationship property” and speaks more to the importance of establishing trusting relationships than it does to the virtues (or practical benefits) of spying on family members. A “good monitor” is not a spy, but rather a parent who has made an effort to build trust with his or her child – although a child must also willingly disclose honestly (Crouter and Head Reference Crouter, Head and Bornstein2002: 461; see also Stattin and Kerr Reference Stattin and Kerr2000a).

As children grow and begin to communicate effectively with language, they express wishes for separation. More robust patterns of disassociation continue as children enter puberty. Finally, as these children become young adults, walls of privacy have become rigid and access points are secured – although the formulation of these access points may be determined partly by the level of trust they maintain with parents and/or siblings. Thus, as parental knowledge depends, to some significant extent, on trust and two-way relationships, activities that damage trust, or have the potential to do so, are inherently risky and suspect, both ethically and for practical purposes. Spying, especially covert surveillance between family members that invades established boundaries or evades walls of expected privacy, bears a substantial risk to trusting relationships (Crouter et al. Reference Crouter, MacDermid, McHale and Perry-Jenkins1990: 656).

Empirical investigation of the impact of actual parental surveillance practices may be limited, but some do exist. In a study conducted by Czeskis et al. (Reference Czeskis, Dermendjieva, Yapit, Borning, Friedman, Gill and Kohno2010), interviews were conducted with nine parent–child pairs. They were primarily concerned with understanding how technologically aided parental surveillance trends involving mobile phone monitoring impacted children and their parents. Cellular phones and other technologies allow family members to track each other’s geographic locations, record content, activate cameras and microphones on mobile devices, initiate automatic alarms when entering areas deemed unsafe or off-limits, detect and report on devices that happen to be nearby, or even measure emotions (Czeskis et al. Reference Czeskis, Dermendjieva, Yapit, Borning, Friedman, Gill and Kohno2010: 1).

Czeskis et al. asked questions about possible mobile phone systems that could transmit information from the child to the parent under various conditions. The researchers found that eight of the nine teens would choose to limit the disclosure of certain information to their parents, and that most parents expressed support for these decisions, especially when they involved personal information about friends or significant others. Teens also expressed interest in the ability to disclose information and to write things to friends on Facebook that they would be uncomfortable having their parents read. Six of nine teens studied had lied to their parents about where they were on at least one occasion. Claiming laziness and changing locations as primary reasons – rather than pure intent to deceive – these teens deceived their parents, who were generally unaware of these misleading disclosures (Czeskis et al. Reference Czeskis, Dermendjieva, Yapit, Borning, Friedman, Gill and Kohno2010: 7).

In an important article by Stattin and Kerr (Reference Stattin and Kerr2000a) we find compelling support for the claim that certain sorts of monitoring are harmful. Moreover, Stattin and Kerr conclude that two-way parent–child communication and sharing – not monitoring or spying – is beneficial in promoting appropriate behavior. Surprisingly, children who are monitored by parental solicitation or with the use of rule sets (you have to be home by 7 p.m.; no playing with this or that kid; etc.) have the same rate of problematic behavior as those children who are not monitored at all. “[C]ross-sectional and longitudinal studies show that poorly monitored adolescents tend to be antisocial, delinquent, or criminal … [they] also tend to use illegal substances … tobacco … do worse in school … and engage in more risky sexual activity” (Stattin and Kerr Reference Stattin and Kerr2000b: 1072). Where there is two-way communication between parents and children, when all are actively participating, including the voluntary sharing of information, there is an associated drop in the behaviors mentioned above. In a follow-up article, Kerr and Stattin conclude: “[I]t appears that the less effective strategy, and the one that has the potential of backfiring, is to try to prevent adolescents from getting into trouble by rigorously controlling their activities and associations” (Kerr and Stattin Reference Kerr and Stattin2000b: 378; see also Kafka and London Reference Kafka and London1991; Barnes, Farrell, and Banerjee Reference Barnes, Farrell and Banerjee1994; Eaton et al. Reference Eaton, Kruger, Johnson, McGue and Incono2009; Hare et al.Reference Hare, Marston and Allen2011).

Covert monitoring by parents is first perceived as non-engagement by the child. Thus, if successful and never disclosed to the child, all of the risks of parental non-involvement are present. If covert monitoring is discovered, many of these adolescents will take countermeasures (keeping two diaries, secret email accounts, etc.) and resist or defeat parental surveillance. Citing Livingstone and Bober (Reference Livingstone, Bober, Buckingham and Willett2006), Kay Mathiesen notes: “in a survey of children 9–17 years old in Great Britain, 69% said that they mind their parents restricting or monitoring their Internet use; 63% of 12–19 year olds said that they took some action to protect … privacy” (Mathiesen Reference Mathiesen2013: 264). Moreover, there is now the issue of trust that must be considered by the child – discovered covert monitoring will likely undermine the practice of two-way information sharing.

If correct, there are obvious and strong connections between flourishing or well-being and privacy for adolescents. Furthermore, problematic behavior or “poor adjustment,” including depression, violent outbursts, engaging in risky sexual behavior, and the like, increase with loss of privacy and control. Kerr and Stattin put the point succinctly:

We conclude that, as with adults, privacy – defined as the right to control access to and uses of locations and personal information – is morally valuable for children. In the following section we will consider when these privacy norms may be justifiably set aside, or trumped, by other important values.

Two-way sharing, probable cause, and Allen’s case for justified parental paternalism

Family life is not only an important setting for individual development, it is also a primary site for social development and socialization. The norms of coming together and leave-taking that occur within the family provide an important backdrop for meaningful and continuing social interactions. Thus privacy arrangements within the family have an important social function. We employ this argument to take issue with Anita Allen’s (Reference Allen2008) defense of the virtues of spying. After an analysis and critique of Allen’s position, we will defend the view that individuals within families have privacy rights that should only be set aside once specific conditions are met. While contextually dependent, the rules for setting aside privacy rights within the family will parallel the “probable cause” and “sunlight” provisions found in the legal traditions of developed societies.

Anita Allen is widely known for arguing that spying on one’s children or spouse is morally justified when specific conditions obtain. She begins with what she calls the anti-spying principle: “spying on other(s) … is prima facie unethical. Spying ought always to be approached with caution and circumspection. Regardless of its motive, spying carries an ethical cloud” (Allen Reference Allen2008: 3). Allen goes on to argue that the anti-spying principle can be overridden in a range of cases. Among others, Allen argues that obligations of caretaking and self-defense may allow for or perhaps require spying. Allen notes: “Sometimes, for limited purposes, responsible moral agents should be willing to make themselves the invisible monitors of others” (Allen Reference Allen2008: 6). When used to protect one’s children from “seriously unsafe behavior,” spying is a matter of responsible parenting. Allen would go so far as to covertly monitor one’s children, read private diaries, and search bedrooms to look for signs of trouble. Surprisingly, even protecting one’s children from poor eating habits justifies spying in Allen’s view. Allen considers McCloskey’s position that adults who target children have no privacy claims. McCloskey writes:

Allen notes that parents should spy to stop statutory rape and child abuse, but not by any means necessary. She writes: “I think a better view is that the sexual offender has legitimate privacy expectations and interests, but that not all of them would be wrongly breached by spying” (Allen, Reference Allen2008: 9–10). In a related case Allen drives this point home.

The court found that Glenn Michael, the father in the Plaxico case, had a compelling and overriding interest in protecting his child that was weightier than Plaxico’s right to privacy. Allen contends that Michael’s actions went too far – good motives, such as trying to protect one’s children from harm do not justify the magnitude of the intrusion in this case. Moreover, there were other, less invasive ways to determine the facts surrounding his ex-wife’s relationship status. A general form of Allen’s safety argument proceeds as follows:

P1.

Parents are obligated to protect their children from harm.

P2.

Information about children and their activities is needed to protect them from harm.

P3.

Overt and covert monitoring are good ways to gather this information.

P4.

Overt and covert monitoring are most plausible when coupled with a “least harmful/invasive means” rule.

C5.

So it follows that parents ought to overtly and covertly monitor their children.

This is also true of McCloskey’s child predator case. A practice of two-way information-sharing between parents and children would likely yield the requisite information to determine if further inquiry or action was needed. In instances where the sharing rule fails, we employ a probable cause rule. To examine this second rule, we consider another case provided by Anita Allen – the case of Jeanine and Albert Pirro.

Allen claims that spying in this case would be justified. Her only complaint was that Jeanine Pirro made poor decisions about who to hire as a spy and what methods to use. It seems that Jeanine hired a less than honorable private detective who used questionable surveillance methods. Had Jeanine Pirro used different methods, Allen would have defended the surveillance on grounds of self-defense. A more formal version of Allen’s argument proceeds as follows:

P1.

Out of self-interest or self-defense adults are justified in spying on their spouses.

P2.

Information about one’s spouse and his/her activities is needed to protect oneself from harm.

P3.

Overt and covert monitoring are good ways to gather this information.

P4.

Overt and covert monitoring are most plausible when coupled with a “least harmful/invasive means” rule.

C5.

So it follows that spouses ought to overtly and covertly monitor their wives/husbands/boyfriends/girlfriends.

In our view, probable cause coupled with a two-way sharing rule yields an appropriate balance in specific cases and as a general practice. Probable cause is established when a parent or other family member has discovered information through logical inquiry that would lead a reasonably intelligent and prudent person to believe that further monitoring is warranted on grounds of safety or self-protection. This basic standard is, of course, drawn from the criminal law, and generally applies to restrict state surveillance, search, or seizure. In American Fourth Amendment law, this standard outlines the boundaries of acceptable state intrusion into the private affairs of individual citizens. It prevents the state from engaging in “fishing expeditions,” by limiting state action to situations where officers have a reasonable and articulable basis to believe criminal conduct has occurred. We argue that this standard also serves a valuable purpose in governing legitimate surveillance within the family. In concert with two-way information-sharing practice, such a standard would help preserve the ability to appropriately withdraw for purposes of self-development without fear of unjustified monitoring.

Intra-familial surveillance, especially covert surveillance, is only justified when it results from failed attempts to engage in two-way sharing or when such attempts are not feasible, and is then based on evidence reasonably amounting to probable cause. Additionally, if these conditions exist, the surveillance ought to be practically implemented in a minimally intrusive fashion. In any case, making determinations about the appropriateness of possible conduct is essential if we want a theory that can be used, practically, to guide moral behavior as it occurs.

Let’s re-examine the Plaxico scenario presented above. First, in the Plaxico example, Allen contended that Mr. Michael was not justified in taking photographs of his ex-wife’s lesbian encounter in her own bedroom because the privacy intrusion was overly severe, given the assumed security threat. We agree with Allen that being raised by a same-sex couple does not, and should not, constitute a “serious risk of harm.” Nevertheless, assume there was such a worry present in this case. As a father Michael may justifiably inquire about the status of his daughter. Suppose while he is told that everything is fine and there are no safety concerns, he notices bruising on his daughter’s arms and a change in behavior. Suppose further that his daughter withdraws and refuses to share any information about what occurs at her mother’s home. After talking with teachers, neighborhood friends, and exhausting other information sources, Michael may well appeal to probable cause in justifying the use of technology to attain the requisite information. But this hardly sanctions becoming a covert peeping Tom or adopting a strategy to acquire total information awareness about his daughter, her mother, and the activities occurring when he is not present. Using probable cause along with a “least invasive means” rule promotes the values of safety, privacy, and trust.

Consider the case where a parent or spouse claims to “have a right to know everything” about a partner or child. A different way to focus on the inappropriateness of such a view and the forms of surveillance it implies is to ask if someone from an unbiased vantage point would have consented to monitoring in similar circumstances. Knowing about privacy and trust, would a rational and prudent individual unrelated to the case at hand agree that privacy should be waived on grounds of safety or self-defense? It may be true that we all want to know more about others while keeping our own secrets, but we should resist this impulse on grounds of privacy and respect.

Imagine upon returning home one day you find your partner (or child) searching through the trash, painstakingly putting your shredded notes and documents back together. In response to your stunned silence your family member proclaims: “I have a right to know everything about you – there is no reason to hide is there?” An appropriate reply to this sort of “fishing expedition” would be to highlight that privacy rights exist even within families, privacy is morally valuable and connected to human health, and that the principles of respect or trust would demand different sorts of behavior between loved ones.

The primary difference between our view and Allen’s is that we are more restrained when it comes to spying on family members. While we agree with Allen’s “least intrusive method” rule, we also employ a two-way information-sharing rule and a probable cause rule. Overt or covert spying on a family member is only justified in cases where a parent or partner has compelling evidence that safety or well-being is threatened. Allen writes: “Columbine and similar incidents around the country point to why parents are justified in inspecting teenagers’ rooms for signs of trouble … even poor eating habits can be grounds for monitoring (Allen Reference Allen2008: 7). We have argued against this view. Without good reasons, prying into the private lives of a child or partner is a violation that undermines trust and respect. As noted earlier, we can imagine cases where parents or partners have justified cause to spy, but this is only after failing to establish the sort of open relationship that would render such spying unnecessary.

Moreover, as citizens we certainly would not think that intrusions into private areas are justified because of relatively rare acts of terrorism at home and abroad. The remote threat of such travesties would not sanction governmental fishing expeditions into private lives of citizens in violation of the Fourth Amendment. Additionally, such remote threats should not be used to sanction privacy violations within the family.

Conclusion

Privacy, defined as the right to control access to and uses of locations and personal information, is morally valuable for children, adults, and within families. Moreover, privacy preserves groups, maintains status divisions, allows for deviation and personal growth, and sustains social establishments. Childhood problematic behavior or “poor adjustment,” including depression, violent outbursts, engaging in risky sexual behavior, and the like, increase with loss of privacy and control.

To outline the proper balance between various privacy and safety interests within a family, we have argued for three rules. First, a rule of “two-way communication” establishes a practice of trust, respect, and caring. Such a standard would help preserve private spaces within family life – spaces where children, spouses, or partners can appropriately withdraw for purposes of self-development without fear of unjustified monitoring. Second, if in a specific instance or as a practice the two-way sharing rule fails, we retreat to a probable cause standard to determine the appropriateness of surveillance. Probable cause is established when a parent, spouse, or partner has discovered information through logical inquiry that would lead a reasonably intelligent and prudent person to believe that their child, spouse, or partner is subject to, or the cause of, a specific and serious threat to safety or well-being. Third, in cases where monitoring is justified, a “least invasive means” rule should be employed. If correct, covert or overt surveillance within the family would only be justified in a narrow range of cases.

A call for rethinking genomic privacy

Biomedical sciences currently witness a flood of data on research subjects. Techniques such as personal genome sequencing, RNAseq expression profiling, metabolic profiling and medical imaging provide large volumes of personal biological data. These data potentially contain sensitive information, especially when they are combined with other types of health care data and lifestyle data, now often voluntarily collected by individuals with the help of smart wearable devices. The “Baseline Study” initiated by Google is indicative of the increased interest and ability to execute combined analyses on a multitude of biological and other datasets. Google is collecting genetic and molecular data from 175 participants and aims to scale this up to thousands. Smart algorithms will be applied to pinpoint how a healthy human being should look from a data point of view. This baseline will be used to identify biomarkers for disease states.

Currently, deidentification of electronic records is commonly used as a measure to protect the privacy of research participants and patients. This, however, does not provide an absolute guarantee of privacy. Privacy regulations such as the US HIPAA (Health Insurance Portability and Accountability Act) Privacy Rule therefore request measures that minimize the probability of information disclosure, rather than demanding absolute guarantees of privacy, since the latter cannot be guaranteed. According to the HIPAA Privacy Rule, a record is considered deidentified if either an expert assesses the remaining risk of deidentification to be “very small,” or if a fixed set of identifiers is removed (HIPAA Privacy Rule 2007). However, the probability that deidentified records can be reidentified is bound to sharply increase, as personal biological data are becoming ubiquitous and more easily accessible (Malin et al.Reference Malin, Loukides, Benitez and Clayton2011). Genomic data in particular prove to be prone to reidentification (El Emam Reference El Emam2011). Small sets of genomic features can already function as a unique identifier for a person. Moreover, a growing number of people have personal genotype data and even whole genome data stored in databases. In a research setting, the Thousand Genomes Project Consortium (2012) reported on the sequencing of the whole genomes of 1,092 individuals. The Personal Genomes Project aims at sequencing 100,000 individuals (Lunshof et al.Reference Lunshof, Bobe, Aach, Angrist, Thakuria, Vorhaus, Hoehe and Church2010). Public health care is also shifting its interest from genetic profiling towards full genome sequencing. The Faroe Islands started the FarGen initiative, an endeavor to sequence all 50,000 citizens (details available at www.fargen.fo/en/). The results will be stored in a database that is linked to the Genetic Biobank, which contains samples from island inhabitants and medical and genealogical records. The UK Biobank is collecting samples and medical data from half a million Britons (Gottweis et al.Reference Gottweis, Kaye, Bignami, Rial-Sebbag, Lattanzi and Macek2012), and the UK 100,000 genomes initiative aims at sequencing that many patients and their close relatives. In addition to the traditional realm of research and health care, commercial companies are building proprietary databases with consumer genetics data. Companies such as 23andMe, Navigenics and DeCode genetics experimented with genotyping services directed towards consumers. As a result of these and related activities, genetic and genomic data are accumulating in a multitude of databases.

This new data situation can lead to an increased risk of reidentification. Gymrek et al. (Reference Gymrek, McGuire, Golan, Halperin and Erlich2013) traced the identity of participants in genomic studies based on public data. Sweeney et al. (Reference Sweeney, Abu and Winn2013) used previously published methods to reidentify volunteers from the Personal Genome Project. The authors were able to identify about 40 percent of the individuals out of 579 anonymous records that contained postal code, birth date and gender. Algorithms that link deidentified family relations to named people have already been developed (Malin Reference Malin2006). Homer et al. (Reference Homer, Szelinger, Redman, Duggan, Tembe, Muehling, Pearson, Stephan, Nelson and Craig2008) demonstrated that genotyping data provides a very strong identifier, by using it to determine whether the DNA of a person is present in a DNA mixture of up to 200 individuals.

This evolution towards a ubiquitous production of personal biological data, and the related risk of reidentification demands a rethinking of genomic privacy and big biodata privacy. There is a long-standing intuition that storing and sharing genetic information deserves special caution. “Genetic privacy” is the term that is most often used to refer to a cluster of rights associated with this point of view (Roche and Annas Reference Roche and Annas2001). Central is the right to protection from non-voluntary disclosure of genetic information (Lunshof et al.Reference Lunshof, Chadwick, Vorhaus and Church2008). This right has been embedded in legislation in many countries around the world. Measures to guarantee genetic privacy range from protective (Soini Reference Soini2012) to more liberal. The frameworks were developed for a setting where genetic information is mainly used for research purposes or clinical testing. This context changes now that genetic data are frequently used beyond research. Moreover, the sheer amounts of data that are produced shift practices to whole genome analysis and other types of big biological data processing. This radically new situation calls for a careful revision of the existing practices and technologies that are used to store, mine and communicate personal genomic data (Erlich and Narayanan Reference Erlich and Narayanan2014) and big biological datasets in general (Choudhury et al.Reference Choudhury, Fishman, McGowan and Juengst2014; Sarwate et al.Reference Sarwate, Plis, Turner, Arbabshirani and Calhoun2014).

Rationales for the special status of genomic data

The privacy concerns regarding genomic data have always been prominent in public debates. This prominence can be explained by the fact that they are accorded a very special status, an idea sometimes referred to as genetic exceptionalism. Genetic exceptionalism is the thesis that genomic data are different from other types of biological and medical data, and therefore require a special status and special protection. In this view, genomic data are seen as being extraordinarily informative. Green and Botkin observe, for example, that “Right or wrong, genetic information is believed to reveal who we ‘really’ are, so information from genetic testing is often seen as more consequential than that from other sources” (Green and Botkin Reference Green and Botkin2003: 572). Article 1 of the Universal Declaration on the Human Genome and Human Rights (United Nations 1998) proclaims: “The human genome underlies the fundamental unity of all members of the human family, as well as the recognition of their inherent dignity and diversity. In a symbolic sense, it is the heritage of humanity.” A large body of literature focused on the question whether genomic data are indeed exceptional. The special biological roles of the DNA molecule were often put forward as a main argument. Such special roles are the immutable nature of the genome base sequence, the fact that it uniquely identifies an individual, the informational nature of DNA, and the fact that parts are shared among family members and ethnic communities (Green and Thomas Reference Green and Thomas1998; Sarata Reference Sarata2008; Ilkilic Reference Ilkilic2009). DNA is the central information carrier in biology and therefore potentially can reveal information about future illness or genetic defects, or unknown facts about familial relationships. Genomic sequences not only reveal information about a particular individual, but also about her or his relatives and the ethnic group the person belongs to. Taken on their own, these features do not uniquely distinguish genetic information from other types of medical data (Green and Botkin Reference Green and Botkin2003). One can argue that not one unique property in itself, but rather the combination of properties distinguishes genetic data from other types of medical data (Sarata Reference Sarata2008). It is implied in the general definition of a category that it is associated with a unique combination of properties that sets it apart from other categories. The fact that DNA-based data is characterized by a specific set of properties therefore does not provide a rationale for a special status of the data; it just implies that it is different from other categories of data. Moreover, deducing a special moral status from scientific facts about a molecule can be regarded as committing the naturalistic fallacy (Ilkilic Reference Ilkilic2009). Nevertheless, there is a persistent intuition that genetic data is special. Surveys show there is a belief that genetic information needs special consideration (Sarata Reference Sarata2008). This belief has been explained in different ways. One explanation is that the scientific discoveries in molecular biology gave rise to an aura of power when it came to DNA. Such sociological explanations leave open the question of whether there are more fundamental reasons for a special status of DNA-based data that remain overlooked, and that warrant genetic and genomic privacy.

A different set of rationales can be provided to account for the status and salience of genomic data as revealing “who we really are.” These rationales are based on (i) the information content of genetic information; (ii) the naturalistic connotation or “naturalistic load” genetic information carries, that is the fact that it is construed as a veridical and direct representation of what the world is really like; and (iii) the unique identification of individuals that genetic information makes possible, which depends in turn on the immutability and the uniqueness of the information. The first two arguments – in varying degree – not only apply to genomic data but also to other types of big biodata.

Big biological data as a substrate for social classification

Because of this widely shared conception of the special status and salience of genomic data, genomics has also inserted itself into our thinking about the social classification and categorization of persons.

At this point a distinction should be made between the use of genomic data (i) in forensic practices as evidence for (re)identifications of individuals, (ii) as a basis for classifications in clinical and clinical research practices and (iii) as a basis for categorizations in social and institutional practices. The use of DNA in forensics has been widely accepted as a highly reliable technique of establishing or confirming the identity of both criminals and victims. Genetic material is in both cases a great help in finding out “who is who.” Statistics and Bayesian probability theory have been applied to increase the reliability of our reasoning with genetic evidence in criminal justice procedures so as to prevent errors in inferences in these contexts. Genetic evidence is one of the strongest, unique identifiers of individuals in a forensic sense.

Second, the use of genomic data in clinical diagnosis, therapy and research is ubiquitous. Our knowledge of diseases and health problems has benefited tremendously from our study and understanding of the human genome. But it also became obvious that the picture is far more complex than initially anticipated. The complexity of the information embedded in DNA proved to go far beyond the approximately 20,000 protein coding genes. Studies over big populations show that phenotypic traits are often associated with large sets of genomic features. The non-protein coding part of the genome plays a significant role, as highlighted by the ENCODE project (ENCODE Project Consortium 2012). In most cases a complex network of multiple genes, gene control elements and so on contribute to the phenotype. The secrets of epigenetics and proteomics are being unraveled and add to the complexities. For many diseases with a genetic component it is anything but clear that genetic material allows for easy and quick clinical diagnosis with high reliability. Reasoning from genomic data implies feature selection: out of the 3.2 billion base pairs of the human genome, characterizing features and discrete categories need to be extracted. For instance, in the case of personal genomic information, the process starts with a biological sample taken from an individual. The DNA in the biological sample is extracted, prepared and run through a sequencing machine, which will generate a series of signals that correspond to the sequence of bases in the DNA fragments. The signals are recorded, stored in a computer system and assembled via bioinformatics tools into a genomic sequence. The end result is a series of character strings on a computer hard drive. The features in this personal genomic sequence can then be analyzed in order to derive personal genomic information: statements about disease susceptibilities or character traits, for example. This process of interpretation often relies on the output of statistical inferences like Genome Wide Association Studies (GWAS). At the highest level of abstraction, interpretation of a personal genome sequence results in symbolic statements. For example: “Person A belongs to the group of people that has a higher risk of developing disease D.” This statement divides the population in two groups: the group non-D that has a low risk of developing the disease, and the group D that has a high risk. Given the complexity of biological systems and the underlying probabilistic nature of the inferred relations, this statement is a simplification of a complex molecular and biochemical reality. Interpretation of genomic data often renders probabilities, rather than deterministic certainties or memberships of clearly distinguishable classes. A classical example is the presence of the ApoE4 allele that relates to an increased risk of developing Alzheimer’s disease. The information needs to be interpreted as probabilistic: one can be carrier of the allele and nevertheless grow old without developing the disease. The same mechanism holds true for other types of personal big biodata. For metabolite data or brain images, for example, features also need to be extracted in order to put these data to use. Reduction of big biological datasets to symbolic statements allows for talking in a common language – closer to ordinary language – about what these complex datasets can mean for us. It is a necessary process if we want to use these data in everyday assessments, comparisons or categorizations. We need to keep in mind, though, that such statements are always simplifications of an underlying complex biological reality.

Third, as far as the evidentiary role of genetic information in social classifications (e.g. entrepreneurial, leadership, impulse control, alcoholism) is concerned, the story is even less straightforward and the following considerations need to be taken into account. As indicated, interpretation of big biological datasets implies a data reduction, in which complex characteristics are used to classify individuals in terms of membership of discrete social or clinical categories. Some of these categories represent naturalistic properties in a natural population; others, however, are social constructs that may be socially controversial and contested, categories such as “Attention Deficit Disorder,” “homosexuality,” “weak impulse control” and “alcoholism.” In social, institutional and non-scientific discourse the certainty and special evidentiary role that genomic data can play in forensic identification and in a clinical context, as discussed above, cannot be assumed to carry over to social classification practices. Simply because genomic data can help to identify a person uniquely with high degrees of certainty, or may help to underpin statistical inferences in clinical research, we cannot infer that symbolic interpretations and the corresponding social labels attach with the same degree of certainty to a given individual. This process implies jumping beyond the evidence and endorsing claims that often cannot be justified. Nevertheless, social categorization and classifications based on a person’s genomic data may give rise to entrenched perceptions or “frames” of a person’s social identity and may eventually constrain an individual’s ability to choose how he or she presents and defines himself or herself in social contexts. The genetic basis of novelty-seeking behavior can be taken as an example. Some allelic variants of the dopamine receptor genes were associated with differences in dopamine binding and with phenotypes related to novelty-seeking behavior (Padmanabhan and Luna Reference Padmanabhan and Luna2013). A genetic test can thus in principle be designed to label individuals as either carrying this allele or not carrying this allele. However, genetic evidence will not warrant tagging individuals with the labels “novelty-seeking” or “not novelty-seeking.” The presence of the allele does not say much about the novelty-seeking tendencies of a particular individual. At best, it may indicate a statistically higher likelihood of displaying such behavior. Without caution, the information on whether an individual carries the allele could get translated into a claim about how a person is really, namely being eager or not eager to engage in novel experiences.

We refer to this cognitive attitude as epistemic immodesty. Epistemic immodesty (Van den Hoven Reference Van den Hoven, Van den Hoven and Weckert2008) is defined as the making of knowledge claims about persons that are not fully supported by the evidence available, with potentially significant cost or negative moral consequences to the target person while ignoring how the person involved would like to be identified. In epistemic immodest judgments one draws conclusions about who a person is, or one claims to be acquainted with one or more of his or her properties on the basis of limited or irrelevant evidence. Because of the reasons mentioned before, big biodata inherently carry the risk of giving rise to immodest claims on what one knows or can know about a given person. Classification of people based on biological characteristics transforms statistical claims into symbolic claims about who a person is. These claims are strengthened by the naturalistic load of the classifications and by the unique identifying power of genomic data. Moreover, genomic studies often concern morally salient characteristics, increasing the proneness to epistemic immodesty. As an example, a study on patients with bipolar disease revealed a genomic region that was indicative of an increased proneness to suicide attempts (Willour et al.Reference Willour, Seifuddin, Mahon, Jancic, Pirooznia, Steele, Schweizer, Goes, Mondimore, Mackinnon, Perlis, Lee, Huang, Kelsoe, Shilling, Rietschel, Nöthen, Cichon, Gurling, Purcell, Smoller, Craddock, DePaulo, Schulze, McMahon, Zandi and Potash2012). Epistemic immodest claims can arise when such probabilistic findings are put to use beyond the field of science. When applied by a future employer or to customers in a consumer genetics setting, for instance, the focus is not on identifying generalizations out of the data of many research subjects. In this different pragmatic context, the focus shifts towards the labeling of particular individuals by putting the generalized data to use. It can be foreseen that the claim to know that person will give rise to moral judgments that may negatively affect the person. The immodesty is constituted by the fact that the judgment fails to take into account the subjective experience of individuals and the way they want to be identified.

Genomic privacy and the right to self-determination

The question then is how to deal with personal big biodata and how to respect and protect them so as to prevent making data subjects vulnerable in the light of the availability of the data. Personal big biodata need to be protected to prevent harms, inequities and discriminatory practices of which patients and data subjects may become victims (for a taxonomy of moral reasons for data protection see Van den Hoven (Reference Van den Hoven, Van den Hoven and Weckert2008)). It is obvious that individuals can be harmed, wronged, disadvantaged, targeted and exploited in numerous ways on the basis of the access that others have to their genetic information. With respect to the right to control access to one’s own big biodata, the requirement of informed consent is still the central point of all existing data protection legislation. There are four types of moral reason for this: 1. Prevention of (information-based) harm, 2. Equality of opportunity and fair treatment in markets for commodified personal data, 3. Informational justice and discrimination and 4. Moral autonomy (Van den Hoven Reference Van den Hoven, Van den Hoven and Weckert2008).

First of all, it is clear from the discussions above that personal big biodata can potentially be the cause of information-based harm. Information-based harm is here defined as harm that could not have been (easily) inflicted if particular information had not been available. Second, another type of moral wronging may occur when big biodata are commercially exploited without proper benefits to the data subject, or without him or her even knowing about this. Third, a form of moral wrongdoing occurs when the data are produced in a research or clinical context, but become available in a very different social sphere, such as the world of insurance or potential employers (market) or police or criminal courts (criminal justice). The use of information about a person at a time and place where it is deemed inappropriate or is irrelevant is a form of discrimination. The prevention of discrimination calls for controlling the boundaries of social spheres in which the big biodatasets are produced and used. Also the different practices (forensic, clinical research, social or institutional), their different associated standards of care in reasoning and methodology, the governance and their norms for the fair allocation of access ought to be observed and separated.

Autonomy is the fourth moral reason to be concerned with the protection of personal big biodata. Some types of personal big biodata such as genomic data and brain structures are largely immutable during a person’s lifetime. Such immutable data that is intimately linked to a person vastly increases the impact when releasing this data, since it can impair a person’s ability to determine how he or she presents himself or herself to others. The focus of privacy concerns in this case is to prevent one specific form of wronging moral persons, namely the fact that they are made subject to social sorting, categorization and classification, on the basis of personal genomic data or personal big biodata in general. When people are classified in this way and given the perceived reality of genetic exceptionalism, there is a lack of plausible deniability of the evidence for these classifications on the part of an individual (assuming he or she is not an expert in genetics). This interferes with the moral autonomy of persons, the moral right of persons to define themselves and shape their own identity, present themselves and have their chosen public persona recognized and respected by others. Controlling the way one presents oneself provides the necessary space for self-determination. To be recognized as a being who is able to conceal information is fundamental to being perceived as a “self-representing being,” as argued by Velleman (Reference Velleman2001). Norms of privacy dictate that certain things should be allowed to be concealed. If the ability for self-presentation is impaired, one is naked in the sense that one is exposed in a way that fundamentally impacts one’s standing as a social agent.

In the case of personal big biodata that are immutable, change in a person’s profile can only occur at the level of the interpretation of these data. It is thus important to instill mechanisms to control the process of the inscription of meaning to the data. However, this process is largely outside the control of the person. A first reason for this lack of control lies in the fact that new interpretations emerge as technologies and insights evolve. This fact is acknowledged, for instance, in the informed consent form of the Personal Genome Project: “because the science in this area is evolving, and data will be collected on an on-going basis by the PGP, the risks involved due to your participation in this study, as well as the likelihood and severity of such risks, will change over time” (Personal Genome Project 2014). Second, information systems that are used to store personal data can constrain the freedom of a person to manage his or her own identity (Manders-Huits and Van den Hoven Reference Manders-Huits, van den Hoven, Fischer-Hübner, Duquenoy, Zuccato and Martucci2008; Manders-Huits Reference Manders-Huits2010) irrespective of the type of personal data that are stored and processed. In the case of personal big biodata, these systems represent a person via a digital record, which can contain the person’s genomic sequence, medical record and so on. Digital representations reduce a person to his or her representation in the system, and allow for the creation of “types” of people by clustering such representations. Together with the fact that data in these systems tend to persist, digital identities result in a reduced ability to reshape one’s identity. Central to this moral autonomy, reason for data protection (Van den Hoven Reference Van den Hoven, Van den Hoven and Weckert2008) is Bernard William’s proposition that respect for persons implies attempts at moral identification, namely the identification of a person as someone who is engaging in self-identification and who ought to be identified as such. This is an important aspect of what it means to respect someone: we owe the other an attempt at identification (not clinical or forensic) as the person who he wants to be identified with. The object of knowledge claims therefore becomes a more complex one, namely an objective physical human being who is characterized in terms of a panoply of biomarkers and a genome, but who has a subjective conception of him- or herself. Self-identification encompasses the ability and freedom to determine and reorient oneself. In databases this ability can be constrained when a person’s identity is reduced to a digital record. Defining one’s course of life can be hampered when the person has no ability to control this digital information.

Epistemic modesty is thus a moral source of restraint in the process of the inscription of meaning when this process of inscription of meaning is out of the control of the subject. In addition to the avoidance of information-based harm and the prevention of inequity and discrimination, genomic privacy provides a person with the freedom to shape his or her own social and moral identity, and to relativize or completely undo previous and external determinations made by others and actively engage in self-presentation in a range of social contexts. The epistemic or immodesty that is so easily produced by practices associated with collection and utilization of personal big biodata should be counteracted by instilling and institutionalizing forms of epistemic modesty about claims that we know “who someone really is” if we know this person’s genomic blueprint.

Ensuring privacy as a property right in one’s own biological data, and creating markets for personal biological data, may not work to ensure a fair processing of these data. Markets where people sell and buy personal biological data are highly problematic, as is the case, for example, in markets for organs. The items changing hands are unique and in a relevant sense belong to the individual, or can be seen to be co-constitutive of the individual. And the transfer is irreversible, since once the data is out, it is practically impossible to get control over it again. The idea of a market also implies freedom, informed consent and full information, which are all precisely what is at stake. Genetic data protection by means of data markets and transferable property rights therefore seems a bad idea. An alternative to a market is the “information altruism” as highlighted by Lunshof et al. (Reference Lunshof, Chadwick, Vorhaus and Church2008). In this setting, the individual shares his or her genomic data in order to support scientific progress, aiming at generating benefits for many. Information altruism is a peculiar form of altruism since the audience to whom one provides the information is unknown and global, and so are the potential beneficiaries. By subscribing to the open consent form, the research subject allows the storage of his or her personal genomic data in publicly accessible databases without any guarantee of anonymity, privacy or confidentiality. Withdrawal from a study is possible at any time, but the research subjects need to acknowledge that their data might not be completely removable. They also need to acknowledge that the information release might not be to their benefit and can even harm them. An information-altruistic policy can therefore lead to the vulnerability of research subjects.

The tension between personal and societal interests is bound to be a central topic in big biodata privacy debates (Knoppens Reference Knoppens2009). The individual benefits from a strong protection of his or her big biodata. The community benefits from improved health care that is the result of a better understanding of biological data. Inaccessibility of personal genomic data for the research community can hamper scientific progress and the related societal benefits. There is a need for solutions that ensure protection of the individual, while allowing researchers to mine the data. Whether technical solutions can dissolve this moral dilemma is being investigated (Ayday Reference Ayday, De Cristofaro, Hubaux and Tsudik2013). For instance, Baldi et al. (Reference Baldi, Baronio, De Cristofaro, Gasti and Tsudik2011) applied encryption technologies to full personal genome sequences. Such technologies allow for the in silico execution of genome tests, without disclosing the information outside of the intended audience. In this setting, a person can have personal genomic data available on an electronic carrier in an encrypted way. The person then can consent to let a service provider run an algorithm on his or her genomic data, without having to disclose the data to this service provider. Such encryption strategies can provide the basis for innovations in health care and consumer genetics that ensure genetic privacy, while providing researchers proper access to analyze the data.

In any setting, it will be crucial to create room for self-determination, and for mitigating the effects of epistemic immodesty, by giving persons more means to control not only the access to their big biodata, but also by giving society the means to ensure proper interpretation and usage of these data. Such means imply pluriformity in available models, frameworks and rival theories, and access to countervailing interpretations, checks and balances in actions based on interpretations of genomic data. Good governance of personal genomic data is a way to instill epistemic modesty in users. Practices should also extend to training and support for genetic counselors, in systematic reviews of the interpretations offered by genetic services, the required indication of probabilities in claims, creation of modes of contesting available categorizations and the dissemination of scientific information in society.

Privacy, personal data and theneutrality of the market

“Personal data” can mean any information concerning the “personal or material circumstances” of an identified or identifiable individual, or a “data subject” in the discourse of law (see Wacks Reference Wacks1989, Reference Wacks2010: 110ff; Wacks).² Let me quote the definition given recently by the OECD Privacy guidelines:

Thus, what we call personal data can originate from very different sources, and these may be data that are more or less “private” or sensitive from the point of view of the individual person herself. This sensitivity can vary depending on the contexts in which it is used as well as on the distance it has to what a person would call her personality (Wacks Reference Wacks2010: 111ff, 118). As we know, however, the question of whether there is something like a “biographical core” of data and the question of how one objectively determines sensitive data is contested (Millar Reference Millar, Steeves, Kerr and Lucock2009: 104ff; see also Nissenbaum Reference Nissenbaum2010: 120ff). So, no matter whether the sensitivity of personal data should be dependent on contexts or conceptualized around a non-contextualized “core,” the idea of sensitivity illustrates the normative dimension of personal data and their protection comprising both a legal and a moral dimension. Yet from the perspective of the market, these differences are important only insofar as they put possible legal limits on data collection, analysis and trade. Apart from that, the market is largely interested in what it can learn about the person as a consumer and as a possibly interesting object of advertising.

When I speak of personal data as a tradable good, I mean exactly this sort of digitalized information, which is produced through the Internet and is collected, processed and so on by interested parties. During the last decade or so, personal data have been harnessed in order to develop a gigantic – “big” – data market: the amount of data stored in databases and the possibility of processing them through increasingly sophisticated means is almost incomprehensible, and the market in databases is one of the most profitable (Craig and Ludloff Reference Craig and Ludloff2011; OECD 2013; for many examples see Marwick Reference Marwick2014; Tanner Reference Tanner2014). As Meyer-Schoenberger and Cukier explain, the value of such data is constantly rising as the cost of digital storage falls, as the means of processing become more precise and as targeting becomes increasingly specific and, therefore, increasingly profitable (Mayer-Schoenberger and Cukier Reference Mayer-Schoenberger and Cukier2013: 98ff). Accordingly, the economic value of personal data cannot be overestimated, and the quantification and monetization of personal information constitutes an economic value that will only increase in the future.

Now the question is, of course: what is wrong with a market in personal data? A market arranges the exchange of goods or services on the basis of self-interest and supply and demand, which determine the value/money for the goods in question. Markets are convenient and efficient, they coordinate behavior and a society without markets is hardly imaginable. However, doubts concerning the perfect neutrality of markets have been raised since their inception (Herzog Reference Herzog2013). Over the past few years, such doubts have been accentuated following the global financial crisis of 2008, which led even conservative critics to express some skepticism about the workings of financial markets. Of course, such concerns and broader concerns addressing the very idea of the neutrality of markets have prominently been taken up in the debates of moral philosophy for a long time. So what can we say about the harm that such markets can cause?

Three problems with markets in everything

One of the classic positions arguing for constraining the market sphere is that of Michael Walzer. In his book Spheres of Justice from Reference Walzer1983, he developed the idea of complex equality based on different societal spheres as distributional realms that should be maintained and limited if justice is to be done. The meaning of the different societal goods – education, hard work, money, leisure time and so on– placed in the different spheres, is essential for the respective criteria of their distribution, and since the meanings vary widely, so too do the criteria for distribution. Walzer’s theory forms the background for many contemporary critics of the market because of the idea of spheres – compartments – that should not be succumbed to the logic of markets, and because of the idea that goods might lose their meaning and significance for us, if they were commodified.³ However, as has been argued frequently, the mere reference to different spheres and alleged common understanding of meanings cannot explain precisely on which reasons the separation between the spheres should be grounded (Dworkin 1985; Radin Reference Radin1996: 46ff; Satz Reference Satz2010: 80f).

So let us look more closely at the limits of marketization with the help of the three questions I sketched earlier: the first one on the very idea of what commodification of personal data means for persons and their agency; the second question concerning the moral problem as to the possible injustice of markets; and the third one concerning the ethical problem of the possible consequences of markets in personal data for the individual and social good life.

Quantified relations and quantified selves

In the previous section I reviewed three different perspectives on the limitation of markets in general and specifically of markets in personal data: the question of the commodification of everything, the question of social injustices as a consequence of markets and the question of possible transformations of social relations through the logic of markets. In the following, I shall come back to these arguments and reasons against the neutrality of markets in order to further deepen the understanding of the effect of markets on our social life.

It is usually the predictable consumer as object that serves as the point of reference for market transactions. In this sense, the industry of databased marketing endangers and threatens individual agency – subjects are not understood as agents but as perfectly predictable data objects. Of course, consumers’ data can be mined with very different goals, not solely for improving behavioral targeting. But when we focus on the goal of using a profile to entice a person to purchase specific goods, then it is probable that the more precise the profiles, based on millions of personal data, the more predictable and susceptible to manipulation the subject becomes. Let me come back in more detail to the question of friendships and social relations on the social media: of course we should allow for the fact that people have very different relationships called “friends” on social media. But at least a section of these relationships are conceived of by the subject as friendships or intimate relationships, such as families, belonging, as Habermas puts it, to the lifeworld. I have already argued in the previous section that these relations can change when the subjects are pushed to see themselves primarily as data objects and no longer as subjects in relations: since personal data that are collected through social media are useful and valuable for any data broker and are thus commodified, the relationship between the subject and her friends on social media might be transformed. As Steeves argues, we have to “better understand how commercial mining of the social world restructures social relationships and restricts the kinds of identity performances available to young people online … The detailed individual profiles that result enable marketers to integrate mediatized messages into children’s social environment, through behavioural targeting and “natural” or immersive advertising” (Steeves Reference Steeves, Roessler and Mokrosinska2015). This analysis obviously applies not only to children and young adults. From the perspective of the interests of the data market, I am driven to view the information I share with my friends no longer as intimate communication “oriented towards understanding” but as yet another item to be used by companies to send yet another advert to, to strategically suggest to me to go for yet another “hit.” This is a change in perspective from the first person participant to the third person perspective of the observer and precisely the shift from communicative to strategic action that we saw being analyzed and criticized in Habermas’ theory. Of course, this does not always have to happen: persons will not always understand one another through this third-party lens; they might even engage in subversive practices, trying to undermine market logics. However, the possibility of the transformation of these social practices has to be diagnosed as such if we want to take seriously what I referred to earlier on as “discontent.”

Thus one aspect here is the change in the meaning of the relationship if it is predominantly seen as a relation between data subjects; as Marwick has also pointed out, the sheer struggle for “status” on social media can entail the commodification of personal information, thus changing and harming the meaning of the relationship (Marwick Reference Marwick2013; see also Kennett and Matthews Reference Kennett and Matthews2008; Fuchs et al.Reference Fuchs2012). Again, I want to point out that we do not have to assume that these pieces of personal information are completely commodified: even incomplete commodification can form a threat to the communicative functions of these social relations. That this process and development should have consequences for a person’s own idea of differences in self-presentations in different relations, and therefore for a person’s identity and personality, is not surprising. I therefore want to go one step further and have a look at what the market in personal data could mean for the identity and self-relation of the subject.

“Self-knowledge through numbers” is the advertising motto of the Quantifiedself-movement. On their website (http://quantifiedself.com) you find everything about self-tracking: measuring, noting, registering and putting down all of your activities 24/7. The Apps⁴ you are supposed to use are mostly available free of charge and the data you collect can not only be shared with the group of friends in your quantified self group, but also with companies and data brokers. As Laura Phillips writes:

On the one hand, this form of self-observation stands in the long and venerable tradition of diaries and self-observation (Passig Reference Passig2012). Not every tracking of one’s own personal data constitutes the distancing of the subjects from themselves in a harmful or inimical, damaging way. Self-observation, we know, can have very positive effects, individually and in social contexts, and not every self-tracker shares her data with the companies, or even with friends. However, the reasons why we could understand the self-tracking and objectification of the self as possibly having alienating effects and as yet another form of commodification is, first, that the perspective on the person herself can be changing because she is turning into an observer of her own behavior, and is not any longer the acting subject as the first person (Lupton Reference Lupton2013, Reference Lupton2014). The idea of self-knowledge, then, is prone to degenerate from the imperative “know thyself” into knowledge of numbers about oneself as a moving object: “self-knowledge through numbers.”

Second, the collection of data about oneself can go hand in hand with the data being collected, mined and sold (prominently in the case of health data), as we have seen in the Fitbit example and can observe in many more examples (Lupton Reference Lupton2013, Reference Lupton2014). Honneth, in his theory of reification, characterizes reified social relations as violating ethical principles in not respecting the other person in her individuality. Instead, the other persons are seen “as ‘things’ or ‘commodities’” (Honneth Reference Honneth and Jay2005: 19). He subsequently applies this idea to the relation of the subject to herself, the self-reification (Selbstverdinglichung) which the subject cannot escape if and because the social relations in which she finds herself are commodified in the way described (Honneth Reference Honneth and Jay2005: 78ff). Marwick, who also criticizes “self-commodification,” suggests a similar argument in her chapter on Lifestreaming (Marwick Reference Marwick2013: 117, 205ff): the digital self assumes that everything she does is translatable into data that consequently demonstrate to her who she “really” is. This supports my argument that a market in these data can have detrimental effects in changing the perspective on ourselves, the meaning we give to self-knowledge and, therefore, to agency and autonomy. Even if we should be cautious not to exaggerate the threats that lie in the as yet incomplete commodification of our personal data, it certainly is right to point out wrong directions.

Let me conclude: my intention in this contribution was to approach the problem of the protection of privacy or personal data in the digital age through the question of what the market in personal data could mean to persons and their identities, to the flourishing of social relations as well as to social justice in a society. And I hope to have demonstrated that reference to debates on the neutrality of markets and their moral limits provide us with substantial arguments that help to protect the personal data of individuals in their different social contexts. I have already pointed out in the introduction that the critique of the commodification of personal data does not necessarily mean that we could not consider ideas of “incomplete commodification” (Radin Reference Radin1996; O’Callaghan Reference O’Callaghan2013): personal data could play a meaningful role in friendships and other non-market practices and relations, and yet, within limits, be treated as a commodity in a data market. Other examples of incomplete commodification demonstrate the possibility of having both, an object or relationship that is commodified while at the same time being meaningful. Consider work: work is a commodity, but can also be meaningful and autonomous for the person herself, and as a practice (Radin Reference Radin1996: 105). Consider books: they clearly are a commodity, but they are perfectly meaningful objects at the same time. Radin suggests viewing commodification on a continuum between two endpoints: the complete commodification of an object and its complete non-commodification. Incomplete commodification of personal data would then take up a middle position on this continuum.

Note, however, that we have different ways to conceptualize this incompleteness: in Radin, the commodified and non-commodified meaning of an object can peacefully coexist, making its meaning incompletely commodified (think of work, again). With personal data, we should conceive of incomplete commodification in a different way: we should conceive of some data being easily commodified, with other sets of data resisting commodification (as I in fact have argued in the above sections). We could also make a difference between contexts: in some contexts, for instance on social media, commodification or commercialization is clearly more tenuous than in other contexts.

So what can we learn from the debate on the moral limits of markets? The perspective from the market in personal data demonstrates the need for moral norms that are not primarily based on individual consent and control of privacy violations, or, for that matter, on the idea of contextual integrity of information flows, but that take into account how the market in personal data can transform our social lives in ways that are harmful, detrimental or injurious. This moral and ethical dimension of the consequences of the marketization of personal data can only be captured when we take up the perspective of the moral limits of markets. Personal or private data are de facto a tradable good; and even if these norms do not tell us where precisely to stop with commodification, the task for ethics and political theory is to criticize tradability if it becomes harmful and dangerous for people and their social relations. The norms that guide the limitation of a market in personal data are necessary if we want to hold on to the idea of a flourishing and well-lived social life.

Defining and describing privacy

Before proceeding, however, it may help briefly to clarify some points of terminology and methodology. A great deal of philosophical and legal debate about privacy concerns the best way to define it (Allen Reference Allen1988: chapter 1; Moore Reference Moore2010: chapter 2). However, the main reason why it is hard to define privacy – the absence of a set of necessary and sufficient conditions that distinguish privacy from allied concepts – suggests that the fuzziness of our concepts of liberty, equality and rights, rather than some particular obscurity in the concept of privacy itself, likely explains why the boundaries of privacy are hard to fix. No definition of privacy will remove that problem. However, for the purposes of this chapter, we can think of privacy as referring to some combination of seclusion and solitude, anonymity and confidentiality, intimacy and domesticity. Whatever else the word ‘privacy’ is used to describe, it is used to describe these three groups of words; and whatever else talk of privacy as a moral or political right is meant to illuminate, it is normally meant to illuminate our rights and duties in these.

Democracy

Just as privacy has many meanings, whose merits are controversial, so with most of the other concepts with which we must work, including that of democracy. I will therefore follow standard contemporary usage in referring to democracies as countries whose governments are elected by universal suffrage and where people have an equally weighted vote. I will also assume that democracies require ‘one rule for rich and poor’ and for governors and governed – that they are constitutional governments. I also assume that democracies enable people to form a variety of associations through which to advance their interests, express their ideas and beliefs, and fulfil their duties as they see them. They are therefore characterized by protection not just for political parties, unions, interest groups and churches, but also by the protections they secure for soccer clubs, scientific societies, families, charities and like-minded people.

Freedom or liberty, equality and rights

Clarifying the way I will be using the word democracy helps to explain the ways I will be using words such as‘freedom’ and ‘equality’. Completely different things have been taken to epitomize freedom and equality. I therefore suggest that we take whatever forms of liberty are uncontroversially necessary to democratic government as examples of freedom; and we take whatever forms of equality are uncontroversially necessary to democratic government as examples of equality. So, taking some familiar features of democratic government can help us to clarify our ideas about freedom and equality, and can give us a shared reference point for resolving disputes about the relationship of privacy, liberty and equality.

As with freedom and equality, so with rights: we can use standard democratic rights to illustrate people’s legal and moral rights, bearing in mind that the precise relationship of the legal and moral is a matter of controversy in most democracies. We can therefore think of the right to vote as both a moral and a legal right – a right that, in democratic countries, is legally protected partly because people are morally entitled to participate in forming their government. Problems clarifying the idea of a right, therefore, can be resolved in the first instance by thinking about familiar democratic rights – whether legal or moral.

Democracy and methodology

Finally, it will be helpful to clarify what I mean by talking about ‘a democratic conception of privacy’, and the nature of the argument that I will be presenting.

By ‘a democratic conception of privacy’, I mean an interpretation of the nature and value of privacy, and of its implications for public policy, which is based on democratic principles, ideas and institutions. I do not assume that there is only one form of democracy.⁴ On the contrary, I imagine that there can be more liberal, republican, socialist, utilitarian and communitarian ways of interpreting central democratic ideas, rights and values, and of embodying these in customs and institutions. However, I am concerned with what must be common to any form of democratic government and society, rather than what might distinguish them. That is, I assume that democratic moral and political principles provide the appropriate perspective for determining what rights and duties to attribute to individuals and, therefore, what forms of privacy, if any, are to be treated as part of the structure of democratic politics. In turn, I assume that the claims to privacy – and to political participation, for that matter – that are necessary for democratic government provide the appropriate starting point for thinking about the claims of those who do not live under democratic governments, or who are stateless or not yet members of any political society at all. This is necessary to ensure that our reflections on people’s moral rights and duties adequately reflect their legitimate interests in democratic government, whatever their current circumstances, interests and desires.

The foundational distinction for my approach to privacy, then, concerns the differences between democratic and undemocratic governments, as we best understand them, rather than the differences between consequentialist and deontological moral theories, or between liberal and republican political principles (Lever Reference Lever2014b: 188–90). Given how little we know about democracy, and the imperfect character even of what we think we know, the analysis of the myriad forms of undemocratic government strikes me as a necessary guide to what democracy requires. As there is nothing inherently democratic about republicanism, liberalism, socialism or utilitarianism, nor of consequentialist and deontological moral theories, none of these seem a particularly helpful starting point if we want to understand the nature and value of privacy on democratic principles. So, while it may be unusual to take the differences between democratic and undemocratic governments as our starting point for moral and political reflection, I believe that it provides our best chance of finding common terrain on which to resolve the competing philosophical, empirical, political and legal debates about privacy and freedom of expression in our different societies.

Oliver Sipple and the ethics of ‘outing’

Oliver Sipple was a former US Marine, injured while serving in Vietnam. Sipple lived in San Francisco, and on 22 September 1975 he joined the crowd gathered outside the St Francis Hotel to see President Ford. He was standing beside Sara Jane Moore when she pulled out a gun to shoot the President. Sipple managed to deflect her aim, and to prevent further shots. The police and the secret service immediately commended Sipple for his action at the scene. President Ford thanked him with a letter, and the news media portrayed Sipple as a hero.

Harvey Milk, San Francisco’s openly gay City Councillor and a friend of Sipple’s, saw this as his chance to strike a blow for gay rights. So, without consulting Sipple, he leaked the fact that Sipple was gay to Herb Caen of the San Francisco Chronicle. Caen duly published the news, which was picked up and broadcast around the world.

Although he was known to be gay among members of the gay community in San Francisco, and had even participated in Gay Pride events, Sipple’s sexual orientation was a secret from his family, for whom it came as a shock. Outraged, Sipple sued the Chronicle for invasion of privacy, but the Superior Court in San Francisco dismissed the suit. Sipple continued his legal battle until May 1984, when a State Court of Appeals rejected his case on the grounds that Sipple had, indeed, become news, and that his sexuality was part of the story. Sipple died in February 1989, aged 47.

Several things seem to be wrong with outing Sipple. The first is that Milk’s failure to ask Sipple for permission to talk to the press seems exploitative and contemptuous. Even if one’s sexuality were altogether unremarkable, one might object to having it broadcast to all the world; and if it were likely to make one notorious, the subject of hateful abuse and, even, violence, one might well hesitate to have it widely known, even if one felt no shame about it. Second, Sipple’s case highlights how easily we can be deceived (or can deceive ourselves) into thinking that we know more about other people’s lives and interests than we do. Most cases of outing do not involve one friend outing another, but are motivated by anger at what is, or seems to be, the hypocrisy, injustice or selfishness of someone else. So Sipple’s experience suggests that those doing the outing are very likely to underestimate the harm that they inflict on others – both on their immediate victims and on those who care for, or depend upon, them.⁵ Hence, outing will often be unjustified on instrumental or consequentialist grounds – because its benefits are uncertain, unpredictable and, such as they are, may be achievable in other ways. By contrast, the harms are usually considerable, unavoidable and the full extent of the damage from outing can be easy to underestimate.

Outing means using someone simply as a means to one’s own ends. Strikingly, the Sipple case suggests that this can be morally troubling even when those ends are ones that the victim shares, and has actively endorsed. And this interesting feature of the Sipple case points, I think, to the political dimension of ethical objections to outing, and to the ways that these differ from a consequentialist weighing of likely benefits and costs, or a Kantian concern with the ways that people can be misused by others. Those are objections to outing that we might have regardless of the society we live in, or our assumptions about the legitimacy of democratic government. By contrast, a political perspective on outing centres on the power that outing involves, and the difficulties of justifying this type of power from a democratic perspective.

Outing involves one person or a group claiming the right to make potentially life-changing decisions for a competent adult, although they have not been authorized to do so, are typically in no position to make amends for any harms their actions cause, and cannot be considered either impartial or expert judges of the claims that they propose to overrule. Such unilateral, unrepresentative and unaccountable power over others is difficult to reconcile with democratic political principles, which limit the extent, form and justification of the power we can exercise over others. Moral and political objections to absolute government, therefore, help to explain what is ethically troubling about outing, even when it achieves legitimate objectives, including ones that have the support of its victim.

The Estlund challenge

Of course, to say that ‘outing’, as usually practised, is at odds with democratic commitments to accountable, representative and participative government is not to say that it is not also at odds with the freedom and equality of individuals. Rather, it is a way of specifying in what ways and why outing violates people’s freedom and equality, given that at a purely formal level outing pits my claims to freedom and equality against yours, and therefore seems to provide no reasons to condemn – or to favour – outing.

We cannot settle for purely formal conceptions of freedom and equality if we care about democratic government, because attention to how power is distributed and used is essential to creating the conditions in which people can share in the authorization of collectively binding decisions. The importance of this point for the ethics of outing – and for claims to privacy and freedom of expression more generally – becomes apparent once one considers what I will call ‘the Estlund challenge’ to my analysis of outing. If concerns with a lack of accountability, representation and participation are at the heart of democratic objections to outing, Estlund asked, would not outing be acceptable if settled on as a policy through suitably democratic political procedures, such as majority votes by a government elected by a majority of the electorate?⁶

The Estlund challenge gains its appeal from the fact that legitimate government would be impossible if all error or injustice were forbidden. Democratic legitimacy, therefore, must be consistent with some injustice as well as some error. However, the fact that some unjust decisions are consistent with democratic procedures and legitimacy does not mean that all are, and it is genuinely hard to see how a policy of outing could be reconciled with the idea that governments must protect the legitimate interests of all citizens.

No amount of voting, for instance, will make ‘employment at will’ a democratic form of employment contract, given the forms of power involved in the ability to hire and fire workers at will.⁷ Likewise – or so I would suggest – no amount of voting will make ‘outing’ at will consistent with democratic forms of freedom and equality. The reasons for this are at once simple and complex. They are simple, in so far as such unaccountable power over others more closely resembles the power of an absolute monarch over his/her subjects than the powers appropriate to people who see each other as equal and, therefore, owed an explanation for behaviour that harms them. Hence, as I have argued elsewhere (Lever Reference Lever2011: chapter 3), inadequate protections for the privacy of American workers, consequent on the use of work contracts that entitle employers to fire workers for ‘good reasons, bad reasons and no reason at all’, illustrate the importance of distinguishing democratic from undemocratic forms of privacy.

But the reasons why outing and employment at will are undemocratic are complex, as well as simple. They are complex, because democratic government is, itself, a complex political ideal and practice and one that can be instantiated in many ways. It is therefore rare for there to be only one democratically acceptable way to organize or distribute power amongst citizens, whether we consider their relations as voters, as producers and consumers, or as family members. Democratic objections to employment at will, I suspect, do not depend on the thought that it is a particular individual – an ‘employer’ – who is able to fire you, but on the thought that no one should be able to have such untrammelled power over something so critical to well-being and social status as the ability to earn a living. On the other hand, the fact that most employment practices put the ability to hire and fire in the same hands and, usually, in the hands of those who supervise and regulate our work lives, clearly exacerbates the undemocratic aspects of employment at will. Hence, centralizing the power to ‘out at will’, or whenever it might lessen prejudice, by granting that power to a corporate body of some kind, rather than to unaffiliated individuals, is unlikely to alleviate democratic concerns about the powers involved in outing and may, indeed, exacerbate them.

Of course, what powers, in practice, are distributed by legal rights to hire and fire at will, or to disclose sensitive information at will, depend on the context in which employment and outing occur. But even under the most favourable circumstances – a developed welfare state, a lack of stigma attached to unemployment – employment at will involves a dramatic ability to disrupt the lives of workers, their relationships to others, their way of life and their sense of themselves. Under less favourable circumstances, the results can be devastating.⁸ Likewise, what harms one can actually inflict by outing someone depends very much on the extent to which information disclosed in one context can be reused or broadcast in others, and on the sorts of penalties that exist and are enforced for the misuse of information (Anderson Reference Anderson and Markesinis1999: 139–67). Nonetheless, as I have argued, the nature of the power implicit in outing is usually inconsistent with democratic norms of government, even under favourable circumstances in which the information generated by outing cannot be endlessly used, reused and broadcast forever.

Contra Estlund, then, a democratic vote is insufficient to render outing legitimate, given the forms of power over others that it usually involves. Nor is a democratic vote necessary to the justification of outing in those cases where democratic concerns for freedom and equality might justify it. We do not need a democratic vote to be justified in publicizing evidence of bribery, corruption or serious illness in relation to powerful figures. Serious ill health in a powerful politician, for example, is a matter of legitimate interest for citizens, in so far as it can affect the outcome of important deliberations, the politician’s ability to think calmly in crises and their ability to cope with the stresses and exhaustion that politics at the highest levels often involves. Likewise, it seems perfectly fair for journalists and citizens to ask Tony Blair, then UK Prime Minister, whether he had given his children the combined Mumps Measles and Rubella (or ‘German Measles’) vaccine – the MMR – given anxious debate around its safety at the time, and Blair’s public statements of confidence in it. Were it not for the privacy of his children, I would have thought it legitimate for journalists or citizens actively to seek out such information.

Blair’s statement of confidence in the MMR vaccine was not really necessary to allay public anxieties at the time, given the overwhelming weight of medical evidence in its favour, and the possibility of disaggregating the triple vaccine into its component parts so that children did not have to have the injections in one go. There was, therefore, no justification for infringing the privacy of Blair’s children in order to allay public anxiety, or to substantiate Blair’s readiness to fit actions to words. But it is easy to imagine cases where politicians’ claims to privacy would seem less powerful – if, say, Blair had been trying to reassure us about the safety of British beef and its freedom from ‘mad cow disease’,⁹ rather than the safety of a vaccine that could be taken in ways that were not controversial at all. A willingness to eat British beef and to feed it to one’s children would be an appropriate way of substantiating one’s claims, as prime minister, to confidence in its safety – perhaps the only way, in the circumstances, to show that one means what one says. It would therefore be appropriate for the press to ask what the prime minister’s family were eating, and to try to find out if no answer were forthcoming, whatever the claims to privacy of the PM’s family.

Democratic principles, then, mean that individuals and the press must be free to publish true personal information about politicians, in so far as this bears on their willingness to live by rules that they urge on others, or aim to impose on them. These interests in expressive freedom and in democratic government are constraints on the ways that governments can regulate privacy, because they mark the boundaries between democratic and undemocratic forms of politics, rather than forming one democratic choice amongst others. It is therefore wrong to suppose that a democratic conception of privacy means that voting is necessary or sufficient to resolve the ethics of outing. Democracy is not reducible to voting, and our interests in privacy include interests in being seen and treated as the equal of others.

Generalizing from the Sipple case: privacy and the ethics of publication

The Sipple case suggests that democratic concerns for freedom, equality and responsibility mean that people ought to have broad, though not absolute, rights over true information about themselves, whether the point of publicizing that information is to enlighten others, to entertain them or to advance a legitimate moral or political cause. Publicizing sensitive personal information, even if it is true, undermines people’s privacy, and threatens their social standing and equality with others. It turns some people into instruments for public amusement or edification regardless of the damage that this may do to their self-respect, their ability to command the respect, trust, affection and loyalty of others, and regardless of its impact on third parties.¹⁰

Such publication, we are often told, is justified by the moral failings of the victim, whether those failings involve acts of hypocrisy, ingratitude, sexual infidelity, attention-seeking or, indeed, illegality (Dacre Reference Dacre2008).¹¹ But while our interests in controlling sensitive information may be self-serving, there is more to our interests in privacy than that. Control of personal information enables us to protect the feelings of other people, as Sipple’s case shows, and to respond to their needs and concerns, even when we do not share them. Such control enables us to act with tact, discretion, respect, and out of a sense of duty, whether or not confidentiality protects our own interests. It enables us to distinguish what is owed to those who have cared for us from what is owed to those towards whom we have no special duties. In short our interests in confidentiality are not reducible to interests in avoiding embarrassment, pain, shame or indignity, but include interests in meeting the needs and claims of others for whom, with all their limitations of imagination and sympathy, we may feel love, as well as obligation.

Protection for privacy, therefore, can promote personal as well as political freedom and our ability to form a variety of personal and political ties to others. Whether our expressive interests are artistic, scientific, sexual or religious – and whether our medium of communication is gestures and behaviour or words and pictures – protection for privacy protects our ability to explore the world and our place within it, and to communicate what we have found to others without exaggerating its importance or having to vouch for its truth, beauty or utility. Democratic claims to privacy for expressive and creative activities, as Warren and Brandeis recognized (Reference Warren and Brandeis1890: 193) do not depend on their economic or artistic value, nor on Millian concerns with the conditions necessary for genius to flourish (Mill Reference Mill1869).¹² Rather, they reflect the claims of even ordinary, unremarkable, individuals freely to develop and exercise their expressive capacities and to do so as the political equals – not the superiors or subordinates – of others.

Rights to privacy are not absolute, however; nor do they invariably trump rights to freedom of expression if, and when, the two conflict. If you are well known and obviously ill, for example, you can expect to be the object of gossip and speculation. But it hardly follows that you should therefore have to anticipate what, now, is almost inevitable: the public broadcasting of such gossip, and its treatment as a means to fame and fortune by strangers. Likewise, if you are well known and seen to be staggering around drunk, or hanging out with people who are notorious, you can expect to be regarded unfavourably by those in the know. As John Stuart Mill emphasized, such knowledge and personal condemnation is the inevitable consequence of social life in a free society (Mill Reference Mill1869: chapter 4). What is not inevitable, however, is the industrialization of gossip and of its marketing to a mass audience as a form of entertainment, titillation and education. Such industrialized gossip is hard to justify morally, even if we are inclined to think that it should be legal. As Stanley Benn argued, it is wrong to treat an entertainer’s life simply as a source of entertainment (Benn Reference Benn and Schoeman1984: 233). Doing so wrongly treats the entertainer as a person with no feelings that can be hurt, and no aspirations or plans that can be harmed by our intrusive attention. So whatever legitimate purposes the publication of gossip serves can usually be met without humiliating and degrading people, however foolish or complicit they may have been in their humiliation.

Privacy, freedom of expression and the press

It would therefore be wrong to confuse freedom of the press with freedom of expression, or to suppose that privacy and freedom of expression are antagonists, locked in a zero sum game, in which gains to the one can only come at the expense of the other. Our interests in being able to express ourselves freely, and to communicate with others, are varied and not reducible to interests in untrammelled access to other people’s ideas and experiences. Protection for people’s privacy, therefore, means that it should be legally possible to demand and win damages for wrongful invasions of privacy, and that the press should be regulated in a way that respects people’s claims to privacy. Hence we cannot resolve conflicts over the respective claims of privacy and press freedom by assuming that the one is intrinsically more valuable than the other.¹³ Instead, we will have to identify and evaluate the expressive and privacy interests at stake when conflict arises, bearing in mind that if the right to publish in a democracy does not depend on literary, moral or political merit, respect for privacy is not just for the virtuous, sensible or the uninteresting. In some cases this means that autobiographical accounts of people’s lives will have claims to invade the privacy of other people, which will be lacking in journalistic and biographical accounts of seemingly similar subject matter.

There is, for example, little to recommend the average ‘kiss and tell’ story, recounting the one-night stand, or lengthy affair of someone who is not famous with someone who is. The format does not lend itself to much variation or reflection, but provides an excellent vehicle for personal grudges, self-justification and self-congratulation. However, citizens must be free publicly to describe their lives and affairs, and to use their lives as art, as science and as an example to others. Because our lives are bound up in the lives of others, it follows that if we are legally entitled to describe and publicize the details of our lives, there is much about the lives of others that we must be legally entitled to publish also, and that we must be able to publish without their permission. Otherwise, most people would find it nearly impossible freely to describe, discuss and publicly to explore the significant events, relationships, constraints and opportunities in their lives. It must therefore be legal to publish stories, autobiographies and reports that are of questionable quality and taste, and that exhibit moral failings such as selfishness, complacency, insincerity and dishonesty, so long as they are not libellous, defamatory or extortionate.

‘Kiss and tell’ stories, I would suggest, are an example where the privacy interests of those who wish to avoid publication are unlikely to justify legal constraints on a person’s ability to publish ‘their story’, and to profit financially from the legal freedom to do so. Even if their moral or aesthetic quality, on the whole, is poor, the subject matter of these stories – what it is like to enter a privileged social circle and to be the lover of someone famous when one is oneself unknown – is a legitimate object of personal reflection and public communication in a democracy. It is therefore difficult to see how kiss and tell stories could be made illegal on democratic principles, simply because their first-person narratives are unlikely to meet with the approval of one of the parties to the ‘kiss’.

Protection for privacy still has a role in determining other aspects of the publication of ‘kiss and tell’ stories, even when it is insufficient to prevent publication. For example, it may be desirable to limit how intensely, and how frequently, journalists are allowed to pursue and try to question third parties to such stories, such as children and spouses, even if this makes it more difficult to question the story’s author and its main subject. In the UK, for example, the families of those caught up in a media frenzy suffer from press behaviour – packs of journalists and photographers following them around; endless ringing of their doorbells and of their phones; the inability to leave the house without being surrounded by a scrum of journalists – that looks very much like harassment and that is likely to be frightening for children, and even for the adults involved. No one’s right to self-expression justifies such behaviour; nor is there any ‘right to publish’, or to know what people are feeling or thinking, that does so either.

It may be also desirable for newspapers to report the sums they offered and subsequently paid for their ‘kiss and tell’ stories, to inform their readers whether they were the ones who solicited the story, or merely agreed to publish it, and so on. Were these standard practices, readers would be better placed to judge how far newspapers are being used to carry out a grudge or feud, and how actively they are instigating stories that, under the guise of autobiography, publicly describe and evaluate the private life of well-known figures. Reporting the fees that such stories command may increase their supply for a while, and the invasions of privacy that accompany them. But it is public knowledge that selling one’s story of sex with the famous is a way to make money and, even, to launch a career, so there is no reason why newspapers should not disclose the sums involved and the way that they are negotiated. Doing so would promote public understanding of the economics of a lucrative branch of journalism, and would make it easier to understand the market price, if not the value, of privacy.

However, autobiographical justifications for publishing privacy-invasive material do not automatically apply to third-person publications, or publications by strangers, whether biographical or journalistic. Where celebrities do not wish to relinquish their privacy, and have taken steps to secure it, it is hard to see why journalists should be entitled actively to pursue them, and to publish stories about their sex lives. Such stories may be entertaining, even informative, but curiosity about the sex lives of consenting adults cannot explain why people who are otherwise entitled to privacy should be deprived of it. Hence, the reasons why it should be legal to publish kiss and tell stories, invasive of people’s privacy though they are, do not apply to those cases where none of the people involved wish to relinquish their privacy.

This is not because autobiography is more important, more expressive or more interesting than biography, nor that it is morally superior to write about oneself than about other people. Often the reverse is true. Any democratic conception of freedom of expression will provide significant protections for journalistic and biographical accounts of people’s ideas, actions and experiences, and the importance of protecting such expression will very likely justify limits on the privacy of politicians and of other people who hold positions of political power and influence.¹⁴

In writing your authorized biography, for example, it may be appropriate for me, with your consent, to discuss a formative love affair even if it has hitherto been secret, and the other partner wishes it to remain so. Whatever the ethical considerations of a biographer in such circumstances, it should surely be legal for me, with your consent, publicly to report experiences, sentiments and beliefs that you are entitled to publish yourself. Likewise for my unauthorized biography of you: I should be legally entitled to publish, with their consent, intimate information and opinions that others are entitled to publish as part of the story of their lives.¹⁵ However, when it comes to stories about the sex lives of celebrities or would-be celebrities, as I have argued, people have a stronger claim to publish stories about their own lives, even if this means publishing details about the lives of others, than they do to publish stories about people who, however fascinating, have neither the desire nor the obligation to relinquish their privacy.¹⁶

Conclusion

In this chapter I have argued that people have important personal and political interests in confidentiality, which are intimately related to democratic ideas about the way power should be distributed, used and justified in a society. On that view, ordinary people, with their familiar moral failings and limited, though real, capacities for sensitivity, altruism and wisdom, are entitled to govern themselves and, in so doing, to take responsibility for the lives of others. This suggests that they are not in need of constant hectoring or supervision in order to act well, although they are rightly accountable to appropriate public authorities for their exercise of public powers, their use of public resources and their respect for others’ rights. Hence, as we have seen, people’s claims to confidentiality do not depend on the usefulness of that confidentiality to others, or on the moral, aesthetic, or even the political and economic worth, of the things that they wish to do with it.

Privacy in politics – I

Most defenses of privacy in liberal-democratic politics are of an instrumental character. With regard to the privacy of office holders, the argument is that keeping personal matters of office holders private may contribute to reasoned political discourse. Dennis Thompson and Thomas Nagel argue that the preoccupation with the personal affairs of office holders may detract from policy issues and burden discussions on matters that are the proper object of political regulation with unnecessary disputes. Targeting the extensive media coverage of Bill Clinton’s sexual affairs, Thompson emphasized the negative consequences of publicizing the personal lives of office holders:

Nagel resorted to similar observations to argue that “[t]he public–private boundary keeps the public domain free of disruptive material” (Nagel Reference Nagel1998: 20). Next to sexual affairs of office holders, disruptive material that the privacy norms are meant to keep out of the political domain, Nagel argued, includes “private racism, sexism, homophobia, religious and ethnic bigotry, sexual puritanism, and other such private pleasures” (Nagel Reference Nagel1998: 30). As he put it:

This instrumental approach to the role of privacy in politics is subject to two problems. First, it is not clear whether privacy plays this instrumental role in all liberal-democratic societies. Nagel’s and Thompson’s arguments are anchored in the historical and cultural contingencies of specific societies, such as the excessive interest of the American public in the irregularities in the sexual lives of politicians. To the extent that such features are characteristic of some societies but not others, these defenses of privacy are of limited scope. Second, on the instrumental approach, the political status of privacy is vulnerable. Privacy as a political right does not stand on its own but only as a way of securing the quality of political discourse. We could do without privacy if other and better ways to secure the quality of political discourse were found.

In the liberal-democratic tradition there are approaches to privacy that move beyond the instrumental approach, such as those of Corey Brettschneider (Reference Brettschneider2007) and Annabelle Lever (Reference Lever2012; Reference Lever2015).¹ Privacy, in their view, is of intrinsic rather than merely instrumental value in that it is implicated in the foundational value commitments of liberal-democratic governance, namely freedom and equality. These approaches, however, do not address the privacy of government officials. In what follows, I investigate the link between privacy on the one hand and, on the other hand, freedom and equality with the aim of addressing this issue. I place my argument in the framework of political liberalism. Drawing on previous work (Mokrosinska Reference Mokrosinska2014), I present privacy as constitutive of the liberal-democratic model of political legitimacy as based on the idea of public justification. The concept of public justification emphasizes the significance of disagreement between citizens as regards conceptions of the good life in determining which moral principles are appropriate for governing the political domain. The concept of public justification refers to principles that are neutral between the different moral views about the good life that divide citizens. My argument links the concept of privacy to the concept of public justification as understood in this way. I employ this argument to fix the boundaries of the domain of privacy that is internal to liberal politics, and to distinguish between those personal affairs of government officials that are a matter of public concern and those that are not.

In investigating the place of privacy in political liberalism, my aim is not to defend political liberalism. It should prove interesting to see, however, that this prominent political theory is committed to acknowledging privacy as an internal element of political life and to investigate the implications of the liberal account of privacy for the privacy of office holders.

Public justification

Much of political sociology views the political domain as a domain of power relations (see Lukes Reference Lukes, Bottomore and Nisbet1978). However, while descriptively correct, this account is incomplete. To the extent that political relations are relations of power, power is exercised in the name of a collective, which implies that there are reasons that justify the exercise of power in the eyes of its subjects. To admit such reasons into an account of political relations is to say that political relations are not merely relations of power, but relations in which the exercise of power enjoys public justification.

Public justification is at the core of the liberal concept of political relations. The idea that “the social order must be one that can be justified to the people who live under it” is among the theoretical foundations of liberal political thought (Waldron Reference Waldron and Waldron1993: 57–8; Rawls Reference Rawls1996: 137). The lack of public justification deprives an association of political legitimacy and, indeed, of a political character because, for liberals, the lack of public justification characterizes the non-political condition of a state of nature: “The moral flaw of the state of nature … is that we act without [public] justification” (Gaus Reference Gaus1996: 182).

There are a number of competing accounts of public justification in liberal thought (Gaus Reference Gaus2003). For my purposes, it will suffice to identify its most frequently recurring features. First, public justification does not refer to just any set of beliefs supporting government action that may prevail among individuals in a given historical period. Public justification is a response to the disappearance of such widely shared beliefs and a way to bypass disagreement between individuals who hold competing views of the good life. As Charles Larmore puts it, public justification refers to “principles of association which individuals have reason to affirm together despite deep substantial disagreements setting them apart” (Larmore Reference Larmore and Freeman2003: 380). Exactly which issues of mutual concern should be decided by an appeal to principles that all can reasonably accept is a matter of controversy between liberals. On the classic Rawlsian approach, the requirement of public justification is limited to “constitutional essentials and questions of basic justice” (Rawls Reference Rawls1996: 227–30). Others, like Larmore, claim that the rationale for public justification requires applying public justification to issues of daily politics including, for example, issues such as education, housing, city planning and land use, the organization of health care or employment, tax regulations, investment incentives, transportation policy or regulation of the entertainment industry. Given that the point of public justification is to legitimize the use of coercion to those subject to it, almost all state action is in need of public justification, since almost all state action is backed by coercive power (Larmore Reference Larmore1999: 607–8, cited in Quong Reference Quong and Zalta2013: 7). Although I cannot discuss this position here, I will endorse it for the remainder of the chapter.

Second, various accounts of public justification converge on the rationale for seeking public justification. Why bypass disagreement? For liberals, disagreement involves two dangers. First, the diversity of private judgments poses a practical challenge to the stability of cooperation between individuals. From this perspective, public justification creates the conditions for peaceful coexistence and stable cooperation between individuals who profoundly disagree. Second, disagreement is a moral problem because it undermines the equal moral status and moral sovereignty of individuals. When there are several competing conceptions of the good in the community, unilateral enforcement by an individual of her favored view is inevitable. However, such a unilateral imposition infringes on the equal freedom of those who hold competing views (Gaus Reference Gaus1996: 182–4). From this perspective, public justification creates relations of equal freedom, moral symmetry and mutual respect between individuals.

I have isolated the concept of public justification as a core concept in organizing the liberal model of political association. By defining the political realm in terms of public justification, liberals construe politics as “the final recourse for people who cannot agree” (Macedo Reference Macedo1991: 53). Does this model commit liberals to grant any special status to privacy in politics?

Privacy in politics – II

For liberals, the political realm is the common ground upon which people, despite deep differences setting them apart, can stand together in a way that enables stable cooperation and equal freedom between them. Not every personal view of the good life, not every lifestyle, commitment or action, and not every piece of personal information has a place in the political realm defined as common ground. Some lifestyles and commitments, for example religious beliefs or sexual morality, concern issues that are objects of irresolvable controversy. Public exposure of contentious issues would drag us into disagreement and destroy rather than establish common ground. Information searches meant to bring such personal commitments under collective attention would have a similar effect. To establish common ground, the material brought under collective attention should allow the people concerned to find principles with which to conduct their lives together and bypass their disagreements. Therefore, the liberal commitment to public justification constrains the material that individuals and groups acting in their political capacity bring to the attention of others. First, this commitment limits the considerations that individuals might wish to employ in deciding matters of mutual concern (e.g. when casting votes in elections or pressing group demands on common resources) to considerations that reasonable others could accept. Second, it requires that individuals engaged in the processes of decision-making that concern the organization of their life together refrain from pressing claims in terms that others could not accept.

Based on the understanding of public justification outlined above, the commitment to public justification in politics rules out certain personal commitments and actions as objects of mutual interference among individuals acting in their political capacity. Similarly, it rules out certain personal information as an object of mutual scrutiny. On pain of endangering politics as a common ground between people who profoundly disagree about matters of worldviews, lifestyles, beliefs and commitments, such material, and the corresponding information, should be withheld or, if known, left unacknowledged. In sorting out which material is appropriate and inappropriate for individuals to introduce into the political forum, public justification sets out rules of concealment and disclosure between individuals acting in their political capacity. Insofar as this requires that individuals withhold certain personal material, I submit, political relations based on public justification involve privacy arrangements. Here I refer to these arrangements as liberal privacy. Insofar as these rules prescribe withholding personal commitments, views of the good life and lifestyle, they are decisional privacy arrangements. Insofar as they prescribe withholding certain personal information, they are informational privacy arrangements.

The depoliticization and privatization of substantive views about which citizens disagree could be taken to suggest that all disagreement is relegated off liberal politics. This is not the case. What is relegated off politics is an appeal to reasons that others cannot accept, and not all cases of disagreement are of that sort. Now it is possible that people disagree but appeal to mutually acceptable reasons. The disagreement arises here from what Rawls called the burdens of judgment: due to the complexity of evidence or the variety of life experiences that bear on judgment, individuals may interpret, apply and rank reasons differently (Rawls Reference Rawls1996: 54–8). Insofar as individuals formulate their conflicting positions in terms of reasons that all sides can accept, reasonable disagreement properly belongs to the political domain. A classic example is the discussion on abortion, in which both pro-life and pro-choice advocates formulate their claims in terms of reasons that their adversaries can accept, namely the value of life and the value of freedom, respectively.

What is private in liberal politics?

I have argued that defining the political realm in terms of public justification as outlined above commits liberals to say that privacy arrangements are internal to liberal politics. What, then, is private in liberal politics?

Certain spheres of activity, such as the sphere of domestic life, have been traditionally marked as private.² The realm of privacy I have isolated in liberal politics is not defined in terms of any substantial concerns. The border between the political and the private is constructed out of reasons that people can and cannot reasonably accept as governing their life together; that is, reasons that meet and fail to meet the test of public justification. Public justification sorts out the material that falls in and out of the political realm. Failures to provide reasons that others can reasonably accept therefore identify the material that counts as private from the perspective of liberal politics.

What people can and cannot reasonably accept is a matter of well-known controversy among liberals. However, there is consensus that one fails public justification if one rejects the aim of pressing one’s claims on others in terms of reasons that others could accept. Claims that fail public justification in this way are, in the liberal idiom, unreasonable (Rawls Reference Rawls1996: 49). The unreasonable, then, outlines a domain that, on pain of violating the integrity of liberal politics, should be held back from the political forum and, if known, left unacknowledged. Generically, all views that deny the equal status and freedom of some (groups of) individuals fall into this class. Race and gender discrimination, say, are examples of unreasonable commitments and actions. Incapable of organizing the common life, such views should be confined to private quarters. Appealing to such views in political arguments injects material into the political realm that violates equal freedom and endangers liberal politics as a common ground. The same holds for probing into people’s personal affairs with the aim of placing their unreasonable views in the spotlight of collective attention and judgment: when unreasonable views are kept private, the integrity of liberal politics demands that they be left unattended. In this second sense, the privatization of the unreasonable is a version of the liberal tolerance paradox: just as the commitment to equal freedom commits liberals to tolerate the intolerant as long as their intolerant views are kept off the political realm (Rawls Reference Rawls1971: 220), so it requires them to tolerate the unreasonable as long as their unreasonable views are kept off the political realm and do not influence it.

A qualification is in order. I have said that the integrity of liberal politics presupposes a domain of privacy that removes from the collective attention material that is incompatible with the commitment to equal freedom, such as racism or sexism. By removing such commitments from the collective attention, decisional privacy serves liberal politics as a common ground, as does informational privacy, by removing from the public forum corresponding information that is dysfunctional to it. Is this argument not a plea to protect racists, for example, making them free to engage in acts of violence against another racial group? Or to make sexists free to engage in acts of discrimination against women? Does it not commit us to grant privacy protection to environments in which they can engage, undisturbed, in hate speech? In response to this concern, recall that liberals ascribe priority to political rights and duties over other commitments that individuals may have (Larmore Reference Larmore1990: 349–51; Rawls Reference Rawls1996: 30–1). This implies that, whatever private commitments individuals might hold, violation of others’ political rights by exercising private commitments is never private. Thus, insofar as adherence to private commitments leads to discrimination in a political sphere, liberals do not divest themselves of the right to interfere (Hartley and Watson Reference Hartley and Watson2010: 20). The same holds for cases in which the pursuit of matters classified as private causes harm to others: in the spirit of John Stuart Mill’s harm principle, liberals are prepared to allow the state to intrude in order to prevent harm that individuals cause to others. If no harm or rights violation is involved and the unreasonable is kept out of the political realm, however, surveillance or interference is unjustified.

I have argued above that privacy is implicated in the concept of public justification and that privacy is, therefore, internal to liberal politics. The object of privacy is material that is incapable of public justification. One class of material that should be kept private for the sake of the integrity of liberal politics is the unreasonable. However, the scope of failures of public justification is broader than the one sketched here, and so is the scope of privacy in liberal politics.

Beyond the fixed core of liberal privacy

Failures of public justification are not confined to renouncing the aim of justification. One person may strive to justify her claims to others but fail nonetheless. This is the case when the justification one offers appeals to beliefs that others, who adhere to different worldviews, cannot be expected to endorse. For liberals, substantive claims failing public justification in this way should not be invoked in the political decision-making process. Their proper place is in the privacy of personal or associational life, but not in the realm of liberal citizenship. In effect, many substantive beliefs and commitments, important though they are to people’s self-understanding, will be depoliticized and set aside as private issues. Exactly what material is depoliticized and set off as private depends on the model of public justification one endorses.

There are two general approaches to public justification (Chambers 2010). On the first approach, public justification is a constraint on the content of reasons to which individuals can appeal in the political domain; it admits only those substantive views upon which all reasonable worldviews could converge (Ackerman Reference Ackerman1989). The substantive views that divide individuals fail the public justification test and should be set aside as private:

On this content-oriented model of public justification, in a society characterized by pluralism and disagreement religious beliefs or sexual morality are granted the status of privacy: given that they are the object of disagreement, they fail the public justification test. Privacy insulates such matters from public exposure, scrutiny and interference: they should be held back from the political forum and, if known, left unacknowledged. The same holds for other material about which people, who adhere to different worldviews, can disagree.

On the second approach, public justification, rather than being a restriction on the content of reasons, is a constraint on the process of reasoning whereby citizens arrive at substantive decisions (Chambers 2010). What material is excluded from the political forum depends on the manner in which people engage with each other’s arguments and respond to them. As one exponent of this approach argues, what material is let in and out of the political forum depends on (1) whether objections based on public reasons are advanced against appealing to it in political decision-making processes and (2) whether there is any convincing way to answer these objections (Lafont Reference Lafont2009: 132). For example, “citizens can publicly advocate for a ban on same-sex marriage on the basis of religious reasons against homosexuality,” provided that they address “any objections against such policy based on the political value of equal treatment. Unless next time around they are willing to accept unequal treatment themselves, they must come up with a convincing explanation of how is it that ‘separate but equal’ is an acceptable policy as regards this group of citizens but not others” (Lafont Reference Lafont2008). In this procedure-oriented model of public justification, then, citizens cannot determine in advance what material is capable and incapable of justification; that depends on what reasons have survived the scrutiny of public deliberation. Thus, appeals to religious beliefs in advocating a ban on same-sex marriage are depoliticized and set off as private only if citizens advancing religious reasons against homosexuality fail to answer objections that such a policy violates the political value of equal treatment. If no objections are raised, or if the objections are answered convincingly, the political rationale for privatizing these issues is absent and no privacy considerations insulate them from the public gaze (Lafont Reference Lafont2009: 132).

I have argued that the liberal model of political relations relies on a boundary between the private and the political. The boundary tracks the distinction between reasons that can or cannot reasonably be accepted by people who are motivated to justify to one another their claims to power. With respect to the scope of liberal privacy three more comments are in place.

First, the circumstances in which the norms of liberal privacy obtain correspond to the range of application of public justification. Here, following Larmore, I said that the rationale for public justification commits us to say that the requirement of public justification binds in matters of daily politics and not only with respect to the political fundamentals as suggested by Rawls. Adopted for the purposes of my argument, this position implies that the norms of privacy bind in an equally broad political spectrum.

Second, as with public justification, the norms of liberal privacy do not bind individuals in the non-political sphere of what Rawls called the “background culture” of civil society, that is, the culture of “daily life, of its many associations: churches and universities, learned and scientific societies, and clubs and teams” and professional groups (Rawls Reference Rawls1996: 220). As long as individuals act in civil society, unreasonable and comprehensive personal beliefs failing the public justification test need not be kept private (they may be public with respect to the members of a given group or association that share them). Their status in the non-political domain of civil society is not determined by the political account of privacy I outline in this chapter. In that regard, the traditional theories of privacy (for example, those defending privacy in terms of individual autonomy).

Finally, within their domain of application, norms of privacy bind equally on all individuals acting in their political capacities, both citizens and representatives. This is because, following Rawls (Rawls Reference Rawls1996: 220), the requirement of public justification binds on all individuals equally; the norms of privacy implicated in public justification reflect that.³

How much privacy for public officials?

Having argued that privacy is internal to liberal politics, I return to the debate surrounding the scope of privacy for public officials. If, as I claim, the rationale for privacy is linked to the demands of political practice, so is the scope of privacy that individuals enjoy when acting in their political capacities. In other words, the scope of privacy due to public officials corresponds to the scope of privacy implicated in the liberal model of politics.

What does that imply regarding the example of a cabinet member making a racist joke when at the dinner table with his family, with which I opened this chapter? Norms of privacy implicated in liberal politics require individuals to bracket their unreasonable views when acting in the political realm. With respect to the material that is so bracketed, norms of privacy have others refrain from attending to it. Thus, insofar as the cabinet member’s racist beliefs do not show up on the job and do not engage him in practices that discriminate against or harm others, the privacy norms internal to liberal politics would rule against bringing this incident to collective attention.

Similar conclusions hold for revelations concerning those aspects of the personal lives of government officials the controversies over which reflect disagreements about substantive worldviews between individuals. Examples are sexual morality, religious beliefs or matters of lifestyle such as recreational drug use. To the extent that such views are depoliticized and set aside as private in liberal politics, then François Hollande’s love affair, Bill Clinton’s adultery or cheating at golf, or Barack Obama’s past cocaine use are beyond the legitimate business of the public. Irrespective of what the public wants to know about public officials, the integrity of liberal politics imposes constraints on what the public is entitled to know. To insist, like Schauer, that such information may be relevant to citizens’ voting decisions is to allow citizens to appeal in their political judgments to material that is dysfunctional to liberal politics. To vote for or against politicians on the grounds of their sexual choices or religious beliefs is to recognize that a particular sexual morality or religion can generate principles capable of governing the common life. This approach to politics violates the liberal commitment to equal liberty.

Compartmentalized lives

On my argument, broader comprehensive views of office holders including their unreasonable beliefs should be set aside as private. On pain of compromising the integrity of liberal politics, such material should be removed from the public gaze and, if known, left unacknowledged. Many, however, are concerned that office holders cannot sufficiently bracket their private views from the performance of their offices, the result being that material failing public justification enters the political realm. This concern is particularly acute with regard to unreasonable beliefs. Some suggest that the unreasonable views of office holders, rather than being relegated to the domain of privacy, should be subject to public censure. As Susan Mendus put it:

If this objection is right, the racist joke incident featured in my opening example disqualifies the individual as an office holder because it indicates racist attitudes that inevitably affect his performance in office. I will devote the remainder of this chapter to discussing this objection.

First, the claim that people cannot separate their public responsibilities from their broader ethical and metaphysical beliefs places one in opposition to one of the core assumptions of political liberalism. Liberal citizenship does presuppose that people can distinguish between, and thereby compartmentalize, different modes of reasoning. In particular, when deciding fundamental political matters, individuals would accept the restrictions of public reason, but when deliberating about other matters they need not adhere to such restrictions. Rawls’ claim that individuals put aside their comprehensive views when acting in their capacity as citizens is the first and most influential articulation of the compartmentability thesis (Rawls Reference Rawls1996: 10–13). Insofar as Mendus’ objection is not meant to challenge the liberal framework on this point, it should not be understood as denying the compartmentability thesis, but rather as holding that the thesis does not apply to people holding unreasonable views. It is not clear, however, whether one can press this point on either empirical or conceptual grounds.

To argue that people holding unreasonable views cannot separate them from their performance in office, one would have to demonstrate that there is an empirical correlation between racist, homophobic, sexist and other personal prejudices and the performance of public duties. It is not clear, however, that such an empirical correlation can be found (Dobel Reference Dobel1998: 120). While there are undoubtedly cases of politicians acting on their unreasonable beliefs, there is no empirical evidence to warrant a claim that they instantiate a rule. Thompson concedes that “[m]any people, especially politicians, are quite able of compartmentalizing their lives in the way this [objection] denies” (Thompson Reference Thompson2005: 236). As Erin Kelly and Lionel McPherson argue, it is not uncommon for people to “cabin off” unreasonable views:

The absence of empirical correlation aside, it is not clear that political liberals are prevented, on conceptual grounds, from saying that people holding unreasonable views could compartmentalize them in a way that liberal citizenship requires. In his introduction to the paperback edition of Political Liberalism, Rawls distanced himself from his earlier position according to which individuals supporting liberal political principles would also have to be able to endorse them from their broader philosophical views; he recognized that there will inevitably be people who are cooperative in supporting a liberal political regime but who are also illiberal in their comprehensive views (Rawls Reference Rawls1996: xxxix). This claim, as commentators submit, implies that a commitment to liberal political principles is a matter of practical rather than philosophical or epistemic commitment (Kelly and McPherson Reference Kelly and McPherson2001; Garthoff Reference Garthoff2012: 191–2). Loosening the link between the practical commitment to liberal political principles and broader philosophical or epistemic commitments opens up a space for political liberals to argue that holding illiberal commitments and supporting a liberal regime is possible, namely people may hold illiberal views merely as epistemic commitments and refrain from presenting them as authoritative for political purposes (Garthoff Reference Garthoff2012: 191). To concede that illiberal comprehensive views need not spill over into the political standpoint is to concede that people whose broader value commitments contain racist, homophobic, sexist and other illiberal ideas can set them aside when acting in their political capacities. Kelly and McPherson endorse just this position to plea for tolerating sexists:

The concern that people holding unreasonable views are not able to bracket them when acting in an official capacity turns out to have neither strong empirical nor conceptual support. The endorsement of the compartmentability thesis commits liberals to saying that holding unreasonable views does not automatically disqualify individuals from occupying public office: what matters is not whether public officials hold certain views, but whether those views inform their office performance. To the extent that unreasonable views do not inform the job, they are not a matter of public concern. Neither are they the proper object of media reporting.

Preventive screening?

Of course, if unreasonable views do show up on the job, this fact is a matter of public concern and the proper object of media reporting. The same holds for cases in which holding unreasonable beliefs engages individuals in practices that harm or discriminate against others. What is the proper focus of public inquiry in those cases?

The focus of public inquiry follows the focus of public concern. If the proper object of public concern is not whether public officials hold unreasonable views, but whether unreasonable views show up on the job, then media reporting should be directed not towards the views officials hold, but towards the policies they implement and decisions they take. What, however, if office holders driven by their unreasonable views engage, in their off-duty capacity, in practices that involve harm or discrimination, such as when a cabinet member belongs to a private club that engages in discriminatory practices by supporting private schools denying admission to children of certain minority groups? In such cases, attention to office performance will not do. Is the citizenry entitled to subject their off-duty activities to public screening?

Screening personal lives of public officials with respect to practices that involve discrimination and harm to others’ interests does not pose a dilemma for my argument. As I indicated earlier, the pursuit of matters classified as private can become a matter of public concern when it harms others. Thus, personal pursuits involving harm to others are not protected by privacy. Second, the pursuit of matters classified as private can become a matter of public concern when it involves violation of others’ political rights or leads to discrimination in the political sphere. One of the most fundamental political rights in liberal-democratic societies is the right to equal moral status. Supporting private schools denying admission to children of certain minority groups violates that right.⁴ In violating others’ rights to equal consideration, such personal associations are then not private. As they are not private, no privacy considerations ban their public screening.

I have said that office holders’ engagement in practices that involve harm, discrimination and violation of political rights voids the private status of their off-duty pursuits and eliminates an objection against their surveillance. In political practice, screening of the personal lives of public officials or candidates for public office is often undertaken without prior knowledge of misconduct on their part. Indeed, it is undertaken only to establish whether any misconduct has taken place. Whereas some authors defend this policy (Dobel Reference Dobel1998: 129–33), my position does not go all the way toward supporting preventive screening. When screening of the personal lives of public officials or candidates for office is undertaken without prior knowledge of misconduct on their part, we are dealing with cases of surveillance of practices whose private status has not yet been voided. In that case, screening of these domains remains problematic.

One way to deal with this problem is to say that well-grounded suspicions of misconduct are sufficient to void the private status of the otherwise private pursuits of office holders. On this view, well-grounded suspicions of, for example, discrimination against a minority group would eliminate the privacy-based objection against screening. The success of this strategy depends on how the suspicions are supported. First, exactly what counts as well-grounded suspicions should be determined in a way consistent with the nature of liberal politics, namely it should be subject to public justification. Second, it matters how the facts that serve as grounds for suspicions are obtained. Suspicious facts cannot be obtained in the course of surveillance of the personal pursuits of office holders as long as the private status of these pursuits has not been voided. In order to screen the personal whereabouts and doings of public officials, suspicious facts voiding their private status must be available beforehand. This leaves us with two classes of facts that can serve as grounds for suspicion. On the one hand, suspicions could be supported by facts that relate to individuals’ public performances, for example previous work experience. Second, suspicions could be supported by facts that relate to individuals’ performances in the private domain but are available without intruding into their private pursuits. Examples in point could be an outcome of a legitimate investigation targeting third parties or a report of a criminal act. For example, if, in the course of a legitimate investigation of the school financial administration, the police discover that a private discriminatory club conferred a substantial gift on it, then this fact justifies screening that club’s members. In this case, however, the facts that reveal the official’s engagement in the discriminatory practices of the club and eliminating an objection against screening his off-duty associations are obtained without first intruding into his associations.⁵

Public scandals and blackmail

I have said that personal affairs of office holders should remain private insofar as they do not inform their performance in office. One could argue, however, that even if private lives do not have a direct effect on their performance in office, they may have an indirect effect. Indirect effects could result from other people’s reactions toward the official’s private life. Public reaction could undermine confidence in the office, impair its ability to win support for its policies and weaken an official’s political position. An argument referring to the possibility of using private information for blackmailing public officials is a version of this argument. The expectations of a negative reaction by the public to revelations concerning their private lives might motivate office holders to submit to the blackmailer. The problem with this argument is that the expected reaction of the public – public scandal – is not always justified in terms of liberal politics. As Thompson puts it: “It is not enough … to point to the public scandal that an official’s conduct may cause. The question must always be asked: should the public be scandalized?” (Thompson Reference Thompson1987: 140). In terms of my argument, if material fails public justification and is, thus, relegated to the domain of privacy, public scandal following its revelation is unjustified in the sense that it is impermissible for citizens to base their decisions and political judgments about the office holder on such information. As such information cannot be used as a valid consideration for judging the public performances of office holders (e.g. it cannot serve as a ground for legal complaint), it cannot properly be used as material for blackmail. To the extent that public reactions are unjustified, they should not count as a reason for exposure and should be disregarded (Thompson Reference Thompson2005: 238).

The normative status of privacy

I have argued that concerns with the integrity of liberal politics set limits to the publicization of personal lives of public officials. I close this discussion by examining what is at stake in respecting the divide between private and public realms in liberal politics.

I have argued that, in advancing the interests of public justification, privacy is a condition of political legitimacy. Political liberals tie the concept of legitimacy to an obligation that falls on individuals as members of political societies. Rawls speaks of a “duty of civility,” which is a moral duty to explain how the principles and policies an individual advocates can be supported by reasons that everyone can reasonably accept (Rawls Reference Rawls1996: 217). If, as I have argued, privacy is implicated in public justification, then respect for privacy is an aspect of that duty.

Does this imply anything about enforcement of privacy norms? Public justification is not merely one consideration that individuals, groups and institutions should take into account when acting in the political realm, but it is the foundation of the political realm (Wall and Klosko Reference Wall and Klosko2003: 10). The constitutive role of public justification in the political order can be seen as an argument in favor of the legal enforcement of those constraints on actions performed in the political domain that public justification entails. If privacy is implicated in public justification, this would also provide an argument in favor of its legal enforcement.

The early Web and its limitations

As envisioned by its creator, Sir Tim Berners-Lee, the Web was designed such that “once someone somewhere made available a document, database, graphic, sound, video, or screen at some stage in an interactive dialogue, it should be accessible (subject to authorization, of course) by anyone, with any type of computer, in any country” (Berners-Lee et al.Reference Berners-Lee, Fischetti and Michael2000: 37). As standards developed and the technology became mainstream, websites became general information resources through which people requested and received information (Abelson et al. Reference Abelson, Ledeen and Lewis2008: 58). At this early stage, the referentiality of links between different webpages functioned as a primary means of web-based communication; “users” of the Web demonstrated activity merely by choosing what to watch and read (Taddicken Reference Taddicken, Fuchs, Boersma, Albrechtslund and Sandoval2012).

Despite being a predominantly consumption-driven media, early websites did have the capability of fostering intense and reflective discourse between individuals. Though early access to online-promoted speech was limited by access to, and familiarity with, some sophisticated technology (computers, modems, and Internet browsers for consumers, as well as knowledge of HTML, FTP, and other server-related technologies for producers), individuals could publish content online, and carry on conversations with other similarly experienced computer users (Kies Reference Kies2010: 42–4). Content, once distributed on websites, could subsequently be debated in other channels, such as on listservs or bulletin board systems.

However, Web 1.0 lacked a number of key deliberative features. Despite the potential for “off-channel” conversations, the technical publishing system could not incorporate all those involved in the discussion and, as a result, lacked discursive equality (Kies Reference Kies2010: 40–2). Consequently, key features that constitute deliberation itself – reciprocity, reflexivity, plurality – were only nascent, and were not distributed across the spectrum of web users more generally. With the development of Web 2.0 and more contemporary web environments, these technical hindrances broke down to the point where even those speaking different languages could engage in discourse with one another on a global scale.

Web 2.0 and networked publics

Advancing from Web 1.0, the so-called “Web 2.0” created platforms where individuals and groups could write, comment on, and edit content using more technically accessible tools. Examples include wiki pages, blog posts, and early social networking tools. Such platforms function as “a set of tools that enable group forming networks to emerge quickly. It includes numerous media, utilities, and applications that empower individual efforts, link individuals together into larger aggregates, interconnect groups, provide metadata about network dynamics, flows, and traffic, allowing social networks to form, clump, become visible, and be measured, tracked, and interconnected” (Saveri et al. Reference Saveri, Rheingold and Vian2005: 22). Essentially, Web 2.0 services established distinct platforms enabling communication and collaboration between parties who were geographically distant.

According to boyd (Reference boyd and Marwick2011: 46), Web 2.0 possesses the following characteristics: persistence (online expressions are automatically recorded and archived); replicability (content made out of bits can be duplicated); scalability (the content of networked publics is potentially visible); and searchability (content in networked publics can be accessed through search). Empirical research suggests that there are some key differences in the ways that different social media are used for political purposes (Hindman Reference Hindman2009). As noted by Kaye (Reference Kaye and Papacharissi2011: 224):

In effect, while web platforms can facilitate communication and collaboration, there is no necessary relationship between participating on such a platform and becoming more involved in political deliberation. However, by reducing the technical challenges and costs of communication, and increasing the availability of Internet-capable devices, the potential for a broader population to debate any issue has arisen. Whereas initial iterations of the Web required some basic HTML coding skills and perhaps familiarity with how to upload the code to a server, Web 2.0 established an online environment where anyone with an email address and web browser could quickly write and disseminate their thoughts online.

Web “3.0” and political deliberation

If Web 2.0 can be characterized as constituting sets of discrete communications platforms, Web 3.0 can be better understood as establishing a social infrastructure, deeply integrated with wireless mobile computing systems. Fuchs (Reference Fuchs2010) argues that 3.0 systems are characterized by a web of cooperation, wherein service users “read existing texts or create new ones (cognition), they discuss how texts could be changed, appended, and enhanced (communication), and they together produce new content (cooperation)” (Fuchs Reference Fuchs2010: 131). Data sources are disparate and extend beyond classic means of interacting with computer systems to include speech recognition, optical character analysis, crowdsourced semantic search engines, and other techniques that extract data from “unstructured” content sources. Mobile devices are key to this iteration of the Web, bringing remote sensing tools (such as picture and video cameras, global positioning, and audio recordings) to the Web through direct sensor-to-Web interfaces.

These features of Web 3.0 facilitate, as never before, a more dispersed, interactive, and potentially deliberative Internet insofar as a range of communications technologies permit a wider number of people to discuss issues across a wide breadth of platforms, including YouTube videos, written blogs, Facebook or Twitter, or forums such as Reddit. And as a result of contemporary “embedding” features, it is possible to weave together different modes of communication into a common space: YouTube videos can function as comments on blogs, Tweets sharing an insight or confrontational position can be embedded in Facebook pages, and so forth. As a result, it is possible that a wide range of individuals from a spectrum of society can, though may not, participate in political debate. Moreover, requirements for written literacy are diminished as individuals increasingly take part in debates using rich multimedia formats as well as more traditional means of online communication. And these varied content formats can all be used within the same “thread” of a conversation. Web 3.0 breaks down the geographical and literary barriers insofar as individuals within and outside a geographical community who believe they will be affected by a given issue can participate in the discussion.

At the same time, the use of these devices generates mass amounts of “unstructured” content that facilitate cognition, communication, and cooperation, but also provides enormous potential for surveillance. The monitoring of deliberation between individuals using web-based communications systems becomes much easier as communications take advantage of commercial communications platforms or integrate external software analytics programs. Even without reading or hearing the content of a person’s communications, it is possible to map subscribers’ circle(s) of friends and associates, communications patterns, the geolocation from where (and when) a person accesses the service, the willingness of individuals to browse “controversial” content and associate with suspected “deviants,” and more (Danezis and Clayton Reference Danezis, Clayton, Acquisti, Gritzalis, Lambrinoudakis and Vimercati2007; Diffie and Landau Reference Diffie and Landau2007; Strandburg Reference Strandburg, Acquisti, Gritzalis, Lambrinoudakis and Vimercati2007). Moreover, such surveillance can begin at a broad level, by capturing data that identifies classes of persons, and then subsequently drill down to focus on specific individuals who are of interest (Solove Reference Solove2008: 4). Such surveillance has the capacity to shape individuals’ willingness to communicate, particularly over politically sensitive subjects (MacKinnon Reference MacKinnon2012), a central argument underpinning the litigation against the surveillance of bulk telephone records by the US National Security Agency (Electronic Frontier Foundation 2013).

Surveillance online does not just affect individuals, then, but theoretically all users (and affiliates of users) of different web communications systems. The surveillance of Web 3.0 therefore stands at odds with its democratic potential. Web 3.0 may increase the range of parties involved in political deliberations, but it also increases exponentially the potential for surveillance and the risks to privacy. How might the privacy literature come to terms with the potentials and harms of the “social Web”?

Privacy boundaries and thesocial Web

Dominant conceptions of privacy predominantly based upon liberal-democratic foundations attempt to understand privacy through the frameworks of information control, spatial violation, and expressions of individual autonomy and liberty. This privacy paradigm has deep roots in Western liberal political theory, resting on a conception of society comprising relatively autonomous individuals who enjoy a private sphere that is distinct from a public sphere of government and the state. Autonomous individuals are understood as more-or-less rational actors who can separate “public” from “private” behavior, and who require “a modicum of privacy in order to be able to fulfill the various roles of the citizen in a liberal democratic state” (Bennett and Raab Reference Bennett and Raab2006: 4). Although expressions of how such separations should be theoretically and practically achieved vary, scholars and practitioners differentiate between places, practices, and the revelation of particular kinds of information when drawing privacy boundaries for ethical, instrumental, and political reasons (Bennett Reference Bennett1992).

A spatial boundary is often implicit when we attach verbs such as “invasion” or “intrusion” to uncomfortable privacy experiences. The classic definition of privacy offered by Samuel Warren and Louis Brandeis (“the right to be let alone”) differentiated between what individuals should, and should not, expect to be outside of reasonable public inquiry (Warren and Brandeis Reference Warren and Brandeis1890). Since their initial proposal, the language of privacy has often adopted an explicit or implicit spatial dimension. Contemporary legal notions of a “reasonable expectation of privacy” are often concerned about the place or space within which the subjective privacy claim is being asserted, and whether society would see that such a claim is justifiable (Austin Reference Austin2012).

For others, the privacy boundary is drawn in terms of the specific behaviors or actions that should be shielded from intrusion, such as sexual behaviors, medical matters, or similarly “intimate” actions. In John Stuart Mill’s (Reference Mill1859) terms, there should be certain “self-regarding” activities of private concern, contrasted with “other-regarding” activities susceptible to community interest and regulation. A related way to think about the boundary is in terms of individual decisions and choices. Per this understanding, privacy is essential for enabling liberty and autonomy, both of which permit individuals to behave as responsible adults in liberal democracies (DeCew Reference DeCew1997: 41). This concept of decisional privacy has been relied upon, especially in American constitutional law, to protect decision-making surrounding abortion, contraception, and the right to rear children in accordance with one’s own religious convictions (Allen Reference Allen1988).

And finally, the boundary can be drawn in terms of information, and particularly between sensitive or non-sensitive forms of information. The former are assumed to be inherently private and worthy of higher levels of protection. Examples in most legal formulations include information on race, sexual orientation, health, and political affiliation. Other scholars and jurists have developed complicated (and often debatably subjective) conceptions of a person’s “biographical core” and what facets of this core are more or less deserving of privacy protection (Kerr et al.Reference Kerr, Binnie and Aoki2008; Millar Reference Millar, Kerr, Lucock and Steeves2009). However, what is in the biographical core is not necessarily sensitive; and what is sensitive is not necessarily core. These disputes go to the heart of contemporary debates about what is and is not personal data and therefore what can be subject to regulation under data protection law.

A notion of privacy that flows from certain assumptions about the boundary between the self and the other plays an important role within liberal democratic theory because it: prevents the total politicizing of life; promotes freedom of association; shields scholarship and science from unnecessary governmental interference; permits the use of a secret ballot; restrains improper police conduct such as compulsory self-incrimination and unreasonable searches and seizures; and serves to shield institutions, such as the press, that operate to keep government accountable (Westin Reference Westin1967; Schwartz Reference Schwartz1999). This privacy paradigm underpins the passage and implementation of “data protection” or “informational privacy” laws that allow individuals rights to control their information and impose obligations on organizations to treat that information appropriately (Bennett and Raab Reference Bennett and Raab2006).

A notion of privacy based on a boundary between the self and the other certainly resonates, but also has limitations when applied to web-based communications platforms and when used to challenge the specific strategies of surveillance that accompany social networking environments. As will be made clear, although boundary-based conceptions of privacy can identify and offer remedy for some of the harms on the social Web, they are less helpful in repulsing the most routine forms of surveillance within this environment.

What lies outside the bounds of boundary concepts?

Most privacy models conceive risks as stemming from personal information about individuals (data subjects) being collected and processed by organizations (data controllers) that cross certain boundaries. The Web 1.0 model tended to reflect this basic relationship between controllers and subjects. However, Web 2.0 confused the boundary, because individuals themselves create data on web platforms and environments, and thus also become data controllers. These platforms and environments are themselves built by private corporations. Consequently, organizations are not necessarily creating the content so much as creating the tools of content creation; actual content tends to be subscriber-generated.

Privacy or data protection laws impose obligations on data controllers, and grant rights to data subjects. In the context of Web 2.0 services the advisory body of European Data Protection Agencies, charged with overseeing data protection law, defined SNS providers as “data controllers under the Data Protection Directive” on the basis that “these services provide the means for the processing of user data and provide all the ‘basic’ services related to user management (e.g. registration and deletion of accounts)” (EU Article 29 Working Party 2009). However, even in 2009 these distinctions were becoming problematic because of the growing tendency to use Web 2.0 platforms for public, rather than private or semi-private, communications between friends, family, and colleagues. It is thus difficult sometimes to assert that a privacy boundary was crossed in a web platform or infrastructure. Where an individual is “responsible” for the disclosure, she will often be limited in her ability to assert that a boundary was violated because that individual in question made that information public. These problems are only exacerbated in an era of global data processing, where the Web 2.0 environment spans many organizations and many legal jurisdictions (Bennett et al. Reference Bennett, Parsons and Molnar2014a).

Boundary problems also arise when certain actions are facilitated by technological design, such as the persistent monitoring of nearby wireless access points or the geolocation of a subscriber relative to other subscribers. What then does it mean to protect certain kinds of sensitive information in one place rather than another? Is it a violation for the mobile phone affiliated with a given web platform to report on the close presence of two subscribers, especially if they do not know that their devices are reporting on their presence to the infrastructure owner? Or does the violation only refer to the recording and reporting on the specific intimate action? Must we differentiate between locations where such intimate actions take place (such as bedrooms or bathrooms, or halls used for union meetings) and the performance of such an action, in such a place? Or do we instead rely on the fact that individuals may have consented to such reporting and, as such, must assume responsibility for the surveillance?

Of course, the protection of space and behavior is arguably both under the “umbrella” of information, and the control of information across said boundaries. Yet defining what constitutes “personal” information is a contradictory maze between what privacy regulators ascribe as personally identifiable, what individuals understand as identifiable, and what the companies operating the services themselves perceive as legally or operationally deserving of protections (Bennett et al. Reference Bennett, Parsons and Molnar2014a). Moreover, given that many theories of privacy have been instrumentally established to protect personal information, they are challenged by the increasing capture and processing of metadata – the data about the data, typically including identifiers such as users’ IP addresses, their operating systems, and any information gained from cookies. Such information can subsequently be used to identify individuals and their personal browsing habits as well as to track their physical location. In a study of twenty-four SNS, not one identified any element of metadata as personally identifiable information, nor did the SNSes give users any expectation of privacy regarding their metadata. The broader privacy implications associated with the capture and processing of metadata are rarely addressed (Bennett et al. Reference Bennett, Parsons, Molnar, Gutwirth, Leenes and De Hert2014b) unless in response to public campaigns (Opsahl Reference Opsahl2013).

The strain between traditional privacy frameworks and how Web 2.0 and 3.0 platforms and infrastructure operate becomes clear when examining the services provided. Users of SNS, through the course of their online communication, often encounter other individuals who have not joined the service or who have not even chosen to disclose information on the platform. In such scenarios the individual subscriber and the platform or infrastructure owner would be considered controllers. Moreover, there is a question concerning what constitutes genuine spaces, actions, or information that deserve privacy on political grounds, particularly when data provided or collected by web platforms and infrastructures can be recombined to reveal political affiliation, movement patterns, sexual orientation, or other characteristics. When all these patterns may affect how a person might be engaged politically, then what needs protection? Perhaps the attempt to ascertain what types of data need protection is a hopelessly lost effort on the social Web (Mayer-Schönberger and Cukier Reference Mayer-Schönberger and Cukier2013).

So, instrumentally, a problem has arisen: the companies that operate social networking platforms and environments must grapple with the massive collection and dissemination of their users’ personal information. Moreover, companies must ascertain the legal basis for collecting information about individuals who are not subscribers to the service or platform in question, but whose information is disclosed by actual subscribers. Individuals must, at the same time, struggle to find out if their “biographical core” is being inappropriately revealed in the course of using, or being mentioned on, these platforms. Surveillance scholars often rightly criticize privacy for missing the broader harms and concerns linked to social surveillance for commercial, personal, and state purposes. The belief that privacy creates some kind of space or bubble around an individual that cannot be intruded upon without justification (Stalder Reference Stalder2002) is seen as insufficient to capture how privacy is, today, “compromised by measured efforts to position individuals in contexts where they are apt to exchange various bits of personal data for a host of perks, efficiencies, and other benefits” (Haggerty and Ericson Reference Haggerty and Ericson2006: 12). In response to these deficiencies a rather different framework has developed that focuses on how privacy enables communities and personal development by attending to intersubjective and community bonds.

Intersubjectivity and privacy theory

Responding to the deficiencies surrounding boundary concepts of privacy law and policy, some scholars and jurists have focused on how privacy is needed to strengthen community and facilitate intersubjective bonds needed for democratic action. For these scholars it is not enough to protect distinctly political places, behaviors, and information, but to shield the plurality of actors who are involved in the act of behaving politically. Privacy, then, is as much or more a common right than an individualized one. In what follows, we outline the dimensions of intersubjective concepts of privacy and then explain why these theories also face challenges in the face of Web 2.0 and 3.0 environments.

Approaches to privacy that try to stipulate a boundary tend not to capture the processes through which people engage in social action and become actors capable of social action more generally. Such relationships, in particular their political instantiation, are taken up in Regan’s approach to privacy as a social value that is necessary to secure a democratic society’s economic and political interests. She recognizes that the value of the term is truly realized upon considering its aggregated societal benefits (Regan Reference Regan1995: 220–8). More specifically, she posits three interrelated social values to privacy: as a “common value” – something we all have an interest in; as a “public value” – as essential to a democratic system of government; and as a “collective value” – or an indivisible or non-excludable good and one that cannot be allocated through market mechanisms. There are, therefore, social values to privacy as, according to Regan, it is “becoming less an attribute of individuals and records and more an attribute of social relationships and information systems or communication systems” (Regan Reference Regan1995: 230).

Regan is not alone in bringing the social importance of privacy to the fore. Recent efforts have demonstrated that even liberal privacy scholars such as Westin have been mischaracterized in the literature. Steeves, for example, maintains that Westin “argues that some collective benefit is not a sufficient reason to invade privacy … From the start, Westin’s full legislative program accordingly questioned whether or not surveillance should be tolerated by the public, based on its effect on social relationships” (Steeves Reference Steeves, Kerr, Lucock and Steeves2009: 195). Indeed, Steeves holds (along with Regan) that privacy cannot merely be “traded off” for other benefits, but needs to be understood as a social construction through which “privacy states” are negotiated. Surveillance, then, is problematic on the grounds that it “objectifies the self, collapses the boundaries between social roles and negates the conditions necessary for inter-subjectivity” (Steeves Reference Steeves, Kerr, Lucock and Steeves2009: 208). Steeves’ analysis coheres, broadly, with Schoeman’s view that while autonomy is required to be self-expressive and demands certain privacy considerations, the point of autonomy is not to disengage from one’s relationships but to enhance and form new and deeper relationships (Schoeman Reference Schoeman1992). For these scholars privacy is not exclusively focused on individuals who live with boundaries about them, but upon how those same individuals necessarily parse and develop the connections between and amongst each another (Etzioni Reference Etzioni1999).

For Nissenbaum, a right to privacy is not about the ability to control information, but instead reflects “a right to live in a world in which our expectations about the flow of personal information are, for the most part, met; expectations that are shaped not only by force of habit and convention but a general confidence in the mutual support these flows accord to key organizing principles of social life, including moral and political ones” (Nissenbaum Reference Nissenbaum2009: 231). In focusing on social life, Nissenbaum argues that our expectations of privacy must be coherent with the situations in which events, actions, and information are expressed. Behaviors, actions, words, or information might deserve a reasonable expectation of privacy in one context but not in another. Social norms and conventions largely dictate, or help describe, the privacy norms applicable to an individual and their community, and vary on an ongoing basis.

When social norms are in flux, however, as a result of emergent and viral technologies the privacy norms may be disturbed and contested. In her anthropological research, Nippert-Eng found that “[c]hanging technologies, expectations, and the habits that incorporate them mean that the need to attend to the problem of social accessibility is highly likely to persist in the future. The need to personally attend to this aspect of achieving privacy will not go away” (Nippert-Eng Reference Nippert-Eng2010: 209). Turkle’s (Reference Turkle2012) book, Alone Together, documents many of the challenges (and failures) for people trying to negotiate these norms among themselves, especially when using social media platforms and infrastructures, suggesting that establishing whether a person has exceeded a privacy norm or expectation remains fraught with difficulty. Furthermore, infringements on individuals’ and communities’ expectations of privacy are often realized significantly after the fact. The result is that the harms can become internalized realities and, in the context of rapidly changing technologies, possibly engrained as normal operating practices for the contemporary web-based systems. As a result, the practices that generate such harms can be extremely difficult to reverse.

An intersubjective approach to privacy, then, focuses on ensuring that the bonds that form communities and support the development of individuals vis-à-vis their relation within communities are secure from inappropriate monitoring. This approach emphasizes the relations that constitute populations and their collective capacity to engage in politics. When surveillance naturalizes “Othered” identities, and subsequently identifies relationships on this basis, there is a chilling effect upon the willingness of citizens and residents to develop deeper understandings about one another, which, as a result, further weakens social bonds. While such effects may be most felt by individuals who retreat from associating with others on the basis of an individualized fear of being part of a group, what is centrally being chilled is the capacity to engage in relationships and the building of trust within and across communities. Such relationships are the targets of this mode of surveillance when data on individuals are captured as a result of community associations (Strandburg Reference Strandburg, Acquisti, Gritzalis, Lambrinoudakis and Vimercati2007).

Intersubjective approaches to privacy that are principally attentive to the bonds between individuals in communities align in many ways with the character and purpose of social networking. Web 2.0 and 3.0 also confound traditional boundary distinctions. The harms that are experienced are often challenging for any particular individual to recognize and articulate, because they are linked to how the surveillance affects communities as a whole. By focusing on the community-based conditions that enable the flourishing of human dignity, an understanding of privacy that acknowledges the link between private development and community involvement is made richer. However, this approach to privacy still suffers challenges when we begin to consider the emerging and complex risks within the social Web.

Intersubjective privacy on the social Web

Whereas specific risks and harms to individuals can often be identified and acted upon, it can be more difficult to recognize actionable harms to communities. Yet for intersubjective conceptions of privacy to carry weight, they must surely identify some recognized and actionable harms on a community or a social level. But these are the very harms that can be most challenging to both detect and report in the context of Web 2.0 and 3.0, where data collection, dissemination, and analysis can take place without the individual’s or the community’s knowledge.

Moreover, what defines a community is often unclear. For example, when Google smartphones are used by those traveling to a political protest to report road conditions on Google Maps and shared through Google’s Web 3.0 infrastructure, what is the “community” to be understood and protected? Is it all Google smartphone users? Or just localized Google smartphone users? Or users of Google Maps attending the political event in question? Without a clear definition of what constitutes a community, it is challenging to ascertain the specific harms associated with a group. In the context of establishing intersubjective bonds or social privacy norms, it is essential to know which bonds between which people are being monitored (and influenced by such monitoring) in order to evaluate whether a discriminatory form of surveillance is taking place, and whether the privacy practices being applied to the data are suitably calibrated to the groups’ norms.

Further, if people do know that a privacy violation of some sort has taken place on the social Web, how then should they calculate or report that violation? Targeted communities face a choice of either continuing to experience the violation, and communicate and account for that harm on the very platforms and infrastructures believed to be the source of the harms, or else maneuver to alternative (and perceived “safer”) communications systems. Network effects operate as powerful disincentives to leaving a network that has coalesced over time.

Even if a population has decided to (and can) remain and account for the normative or empirical harms experienced, there is the challenge of translating that negotiated understanding into actionable policy or law. In effect, while parties can deliberate to come to a consensus about the harms experienced, their ability to do anything about those harms can be limited when they are associated with a community writ large rather than specific individuals. Historically in the United States and other jurisdictions, experiences of general chilling effects born of suspicions of surveillance have been insufficient for courts to overturn or establish injunctions on either government or corporate practices (Kravets Reference Kravets2010; ACLU 2013). And while some corporations have expanded the uses of encryption to secure the personal communications of their individual subscribers (Gallagher Reference Gallagher2013), there has been little appetite by corporations thus far to shield their users, as communities, on the grounds of their associations.

More recently, however, we have seen a broader willingness for courts to question whether corporate and government monitoring of communications coheres with established constitutional laws, indicating that even if individuals cannot specifically articulate harm, the harm may still have affected the communities with whom the individuals are affiliated (Feiler Reference Feiler2010). Many civil liberties and rights groups have waged a coordinated effort to articulate principles that would secure both individual and communal rights and expectations of privacy (see www.necessaryandproportionate.org). In the wake of the revelations by Edward Snowden, both communal and individual harms are clearly being recognized. While the national security-related revelations of 2013 and 2014 are extremely significant, it remains to be seen whether they will lead to reforms in what legally constitutes inappropriate state and corporate surveillance.

In summary, then, there are deficits in applying the social conceptions of privacy to account for the contemporary uses of web platforms and environments. At one level, as we have suggested, communities and contexts have merged. Therefore, an analysis of privacy as contextual integrity, as proposed by Nissenbaum, provides little guidance when the environment itself fuses contexts and communities, and disrupts norms and expectations. According to Regan’s framework, what does it mean to say that privacy has social value on the social Web? How can it be a common value of a community, when violations may be occurring to emergent, and not yet normatively formed or recognized, collectives? Moreover, we are not necessarily dealing with privacy as a collective or indivisible good, when increasingly individuals are being encouraged to “choose” their individual privacy settings and protect more or fewer personal data than their peers; under these assumptions, privacy on the social Web is clearly “divisible.”

Perhaps, then, privacy should be seen as a public value, or a good that advances a particular form of community, promoting a public or a common realm – a “community of one’s peers” in Hannah Arendt’s (Reference Arendt1958) philosophy? In conclusion, we suggest that it might be possible to see more clearly the role that privacy might play on the social Web by thinking of privacy in these terms: as a public value, rather than as a common or a collective value. In Regan’s terms, “privacy may be essential to a democratic political system because some commonality among individuals is necessary to unite a political community and the development of commonality requires privacy” (Regan Reference Regan1995: 227). More specifically, we suggest how an integration of deliberative democratic theory might provide an important theoretical foundation, rarely considered in the privacy literature, upon which to promote privacy and respond to the harms made possible by surveillance on the contemporary Web.

Conclusion: deliberative democratic theory and the social value of privacy

While both boundary-based and social theories of privacy operate as helpful means to understanding the significance of surveillance on Web 2.0 and 3.0, deliberative democratic theory provides another useful theoretical foundation from which to critique web surveillance and defend the essential normative freedoms of association and speech. Specifically, by reorienting the very conditions of political analysis away from the harms experienced by either individuals or communities to supporting the deliberative requirements for a flourishing democracy, a more holistic accounting of the implications of surveillance-based harms can be developed.

Deliberative democratic models, such as that of Habermas (Reference Habermas, Cronin and DeGreiff1998a, Reference Habermas, Cronin and DeGreiff1998b), take the pluralist character of human subjectivity seriously, insofar as individuals are ethically obligated to consider attitudes and practices that are associated with their own as well as their community’s best interests (Young Reference Young2001: 672). In the context of political deliberation on the Web, these ethical obligations are practically supported by web-based communications systems that are designed to let persons of different political, cultural, and economic positions, and who are affected by a given issue, to communicate with one another to reach an understanding of a problem and a solution. That the Web tends to retain such discourse for extensive periods of time, if not permanently, means that settled issues can be reopened as new participants enter into the deliberative process. Thus any consensus is temporary and potentially more inclusive than more transitory discursive processes. Moreover, some communications systems will enable high impact contributions to surface to the top of the discussion, whereas those less useful, or outright malicious, contributions can be suppressed from popular view (though not removed entirely). Given that the social Web functions as a space of individuals working with, and within, communities, the Habermasian model lets us focus on individual and public autonomy and rights simultaneously, instead of prioritizing the individual over community or vice versa.

Habermas emphasizes that individual and public autonomy co-originate, and thus are equally necessary to establish the basic laws of a nation-state that secure individual freedoms. To establish such basic laws in the first place, individuals must be able to exercise their public autonomy as collaborative participants in a political system. In effect, public autonomy, made possible by engaging in public deliberation with other members of the public, presupposes that individuals regard themselves as privately autonomous, and that they can shape their freedoms through exercising their public autonomy (Habermas Reference Habermas, Cronin and DeGreiff1998b). Any theory of privacy that is ontologically premised on either individual or community autonomy and rights as the basis for legal privacy protections misunderstands the ontological genesis of basic law. A theory of privacy attentive to Habermasian theory would, therefore, recognize first the importance of securing both private and public autonomy, and would also lead to laws that ensure both public deliberation and private rights.

This deliberative democratic model allows us to critique surveillance that excludes or otherwise inhibits participation in politics. Stymying expression within a political community undermines the legitimacy of law itself, since equitable assent cannot be acquired from all citizens (Habermas Reference Habermas1996). Where surveillance, discrimination, manipulation, or other coercion manifests on the social Web, a harm is registered regardless of whether it is focused on a specific individual (or small set of isolated individuals), or upon broader established communities, or on emergent communities. In effect, this theoretical foundation advances a social understanding of privacy by promoting the development of both communities and individuals. Deliberative democratic theory, then, is consistent with a “holistic” accounting of privacy since it recognizes the equal moral and political weight of individuals along with the integrity of communities.

We are not suggesting, however, that adopting a Habermasian approach to conceptualizing privacy, which is mindful of the individual and the public alike, would force the rewriting of privacy scholarship generally. Rather, under this model, harms that affect large communities in a political association (such as one entire ethnic group) are as significant as harms that affect individuals’ private rights. Thus it is possible to rely on deliberative theories of privacy to register the broad social harms associated with community-based surveillance while, at the same time, prospectively affording individual-based tools to redress those harms. This linkage is possible because community-based harms have an associated derivative (and negative) effect on the rights of the individuals involved, just as negative impacts on individuals have a correlative effect on the communities within which they operate. Consequently, where individuals refuse to engage in collective discourse because of illegitimate or coercive surveillance, their communities experience harms, thus affecting the full potential of the group’s deliberations.

As noted earlier, one of the greatest boons of the contemporary Web is that it has significantly broken down what counts as a mode of deliberation. Participation is less predicated on access to sophisticated technologies given the widespread adoption of mobile communications devices. Moreover, even reading and writing skills are of diminished importance, with individuals able to share photos, videos, and other multimedia to express their particular positions and engage in debate with others. Whereas text and written literacy was initially a prerequisite for participating in deliberation online, this is less the case given the structure and composition of the contemporary and emerging social Web. Where and if, however, either commercial or state surveillance hinders or diminishes individuals’ willingness to communicate, or communities’ willingness to knowingly come into being, then the deliberative potentials of the Web are undermined along with the privacy interests of those interested in contributing to deliberations. Thus privacy and the deliberative potential of the Web are inescapably interrelated.

The focus on deliberative theory facilitates a critique of surveillance that has debilitative effects on the political rights of individuals and communities. Because this theoretical approach recognizes community and individual rights as co-original, it reveals a more comprehensive understanding of the sources and effects of privacy harms. Moreover, given that the Web is characterized as a mixed-media communications environment wherein political participation is facilitated between individuals and communities that might not otherwise meet or communicate, it behooves established and emerging democracies alike to protect this environment in a way that identifies and responds to contemporary privacy claims. Traditional individualistic or intersubjective approaches to privacy are part of these response mechanisms, but must be encapsulated within a more holistic model of political participation in order to respond to the routine and accelerating modes of surveillance that all users of the Web now experience. Deliberative democratic theories that focus on the co-originality of individual and public autonomy and, accordingly, the co-emergence of individual and community rights offer a foundation from which critical theories of privacy and surveillance can be based. They also more effectively respond to the realities of communication and deliberation on the social Web.