A practice–theoretical account of privacy

Abstract

This paper distinguishes between two main questions regarding the notion of privacy: “What is privacy?” and “Why do/should we value privacy?”. In developing a social-ontological recognitional model of privacy (SORM), it gives an answer to the first question. According to the SORM, Privacy is a second order quality of roles within social practices. It is a function of who is or should be recognized as a “standard authority”. Enjoying standard authority means to have the right to interpret and contest role behavior and role obligations within a specific practice (first level), as well as evaluate the normative structure, the fundamental practice norms as well as the roles and their status (second level). The SORM utilizes the concept of standard authority to explicate privacy with regard to two categories that capture the relevant phenomena of privacy: decisional and informational privacy. Within a practice, an actor is said to have decisional privacy if she as a BCR does not (or does not have to) recognize bearers of accidental roles as standard authorities. Vice versa, an actor is said to enjoy informational privacy if all other BCRs (and especially data collecting actors) recognize her as a standard authority. Additionally, the requirement of mutual recognition by the practice participants as standard authorities introduces a “weak normativity” into the theory, which can be used to identify deficient privacy arrangements within practices.

Introduction

When thinking about privacy across its different domains—decisional, informational (Schachter 2003), local (Roessler 2005), physical, or mental privacy (Floridi 2014, pp. 102–103)—at the fundamental level there are two different questions: “What is privacy?”, and “Why do/should we value privacy?”. In the course of the history of the debate about privacy,¹ most authors try to give an answer to the second question, and, by doing so, en passant answer the first.² Only a minority of authors explicitly differentiate between the two questions and answer them separately (Moor 1997; Tavani 2007; Nissenbaum 2010; Floridi 2013).

This seems to be a natural course of action, since the second question presupposes the first: By saying why we do or should value privacy, one already says something about what the notion “privacy” actually means. Vice versa, however, the same does not necessarily hold, at least not in the normative sense: a theory that answers the question what privacy is, may also give us some empirical idea why certain people value privacy. It does not, however, give any conclusive reasons as to why anyone should value it. Neither does it preclude any judgment about privacy as an instrumental or intrinsic value.³

In this paper, I will keep both questions as separate as possible in giving a social-ontological answer to the first question “What is privacy?”. By presupposing as little as possible about the second question, a theory of privacy can be open to all kinds of different accounts about the value of privacy. At the same time, however, the social-ontological recognitional model of privacy (SORM) that I advocate here retains a weak form of normativity in the sense that it observes the mutual recognition of a “standard authority” (Stahl 2013, p. 379, 2014) as the typical way, in which participants of a given practice relate to each other. By mutually conferring the deontic status of a standard authority, practices provide a “content-independent” (Raz 1986, p. 35) motivation for their participants to uphold them, thereby effectively stabilizing behavioral expectations and normatively integrating the participants. I elaborate on this weak form of normativity in section “Social pathologies”. For now, it suffices to note that by incorporating this normative dimension, the SORM can preserve a critical perspective on practices in which privacy is structurally undermined.⁴

The SORM follows a practice-theoretical approach to privacy by situating privacy within the role obligations of social practices. In difference to other practice-theoretical accounts, however, which focus on first order role obligations (Nissenbaum 2010; Roessler and Mokrosinska 2013), the SORM closely links privacy to the second order deontic status of “standard authority” (Stahl 2013, p. 379, 2014). This status allows bearers of constitutive roles of the practice to have an equal say on two levels: first, in the interpretation and contestation of the role behavior as well as the role obligations that participants of a specific practice incur (first level), and second, the possibility to evaluate the normative structure, the fundamental practice norms as well as the roles and their status as a meta-practice of critique (second level). In a nutshell, the answer the SORM gives to the first question “What is privacy?” can be stated as follows: privacy is a second order quality of roles within social practices. It is a function of who is or should be recognized as a standard authority.

I will develop this argument and some of its implications in more detail in section “The argument in full”. The next section gives an overview of the challenges that theories of privacy face. From these, I infer two general criteria that a general theory of privacy must meet as necessary conditions. In section “What are social practices?”, I briefly outline the notion of social practices and explain the social-ontological concept of “standard authority” and its implications for the integration and reproduction of social practices. Finally, section “Social pathologies” deals with the idea of social pathologies that arise from a deficient privacy arrangement at this fundamental social-ontological level. This is where the main normative force of the SORM lies.

Criteria for a theory of privacy

Any theory of privacy needs to meet certain criteria that will determine its viability. In thinking about such criteria, Floridi lists “four challenges” that apply to theories of informational privacy (Floridi 2006). It will be the first task of this section to generalize these challenges and formulate two main criteria for any general theory of privacy: phenomenon adequacy and value openness. These will serve as benchmarks for the SORM. In section “The argument in full”, I will show how the SORM fulfills both criteria, making it a legitimate contender for a viable general theory of privacy.

With respect to informational privacy, Floridi argues that a theory of privacy must (a) somehow cope with the assertion that the concept of privacy varies between cultural environments; (b) be aware of certain elaborations of individualism that may be specifically “Western”; (c) choose an adequate scope for a right to privacy, such that it is neither too narrow, nor too wide in the sense that “any informational process concerning a person becomes a breach of that person’s informational privacy” (Floridi 2006, p. 116); and (d) take seriously the problem of “privacy in public” (Nissenbaum 1998) by upholding the distinction of private personal information and public personal information.

All of these challenges belong to the realm of what I will call phenomenon adequacy. The basic challenge consists in capturing the relevant phenomena that we commonly associate with the notion of privacy, whether it be across different cultural environments, concepts of individuality, or the difference between public and private spaces. Floridi develops these challenges from the perspective of a theory of informational privacy. The criterion of phenomenon adequacy does not only hold for informational privacy, however.⁵ It is even more important for a general theory of privacy, which is not just concerned with one dimension of privacy, such as e.g. informational privacy. Such a general theory must maintain phenomenon adequacy across the different domains of privacy. Floridi himself acknowledges this fact by way of a conclusory “comment”, where he introduces a fifth challenge:

Theories of informational privacy, once mature, will make a sincere and robust effort to coordinate their findings and conclusions with those of other theories of other forms of privacy, in order to gain a comprehensive and coherent view of privacy in all its major aspects (Floridi 2006, p. 118).

These “major aspects” or domains, as I will call them for the purposes of this paper, have been framed differently by various authors. However, most of them distinguish at least between informational and decisional privacy (Schachter 2003; Roessler 2005; Floridi 2014, pp. 102–103), where decisional privacy “serves to secure the scope for a subject to make decisions and take action in all his social relations” (Roessler 2005, p. 17). For the most part, it refers to some kind of freedom with regard to one’s own way of life, from choices of wardrobe to occupation, significant other, family planning etc. Informational privacy, on the other hand, focusses on the availability of personal information about a specific person, whether it be in public spaces (privacy in public), or whether it be within the confines of somehow defined private spaces. Typically, this availability is spelled out in terms of a freedom to reveal or withhold this kind of personal information (Westin 1967; Parent 1983).

In addition to these two, Roessler sets apart the domain of local privacy, which “serves to protect the possibility for spatial withdrawal upon which a subject is dependent” (Roessler 2005, p. 18). This form of spatial privacy is according to Roessler the essence of the modern life-form (2005, p. 142) and should therefore be treated as a distinct domain of privacy. The phenomena that local/spatial privacy capture, however, are either decisional in nature—e.g. non-intrusion into a private space, non-interference with life choices within that space etc., or informational—e.g. questions of surveillance within private spaces, obligations to inform with respect to authorities etc. This means that the phenomena which are commonly subsumed under the heading of local privacy are also covered either by the domains of decisional or informational privacy. This is not to say that there are no other explanatory or theory-strategic reasons to explicitly mark off these kinds of privacy phenomena as belonging to the domain of local privacy. For the task of identifying and categorizing the phenomena that a theory of privacy needs to encompass, however, the additional domain of local privacy is not necessary. Therefore, I will treat it for the purposes of this paper as a proper subset of the two combined sets of decisional and informational privacy.

The same goes for Floridi’s additional domains of “physical and mental privacy” (Floridi 2014, pp. 102–103), where physical privacy refers to the freedom of a person “from sensory interference or intrusion, achieved thanks to a restriction on others’ ability to have bodily interactions with her or invade her personal space” (Floridi 2014, p. 102). As with local privacy, the phenomena of physical privacy can be also categorized in terms of decisional privacy, e.g. bodily interactions, the possibility to physically exclude someone from one’s private spaces etc., or in terms of informational privacy, as far as personal data of one’s own body (pictures, biometric data) or private spaces (surveillance) is concerned. Mental privacy, on the other hand, focusses on psychological interference or intrusion. Phenomenon-wise, these issues can also be attributed to decisional—e.g. in the sense of psychological coercion or manipulation—and possibly also informational privacy, which more recently has become especially prevalent with respect to data aggregation about one’s mental states through big data mining and analytics in social media (Tene and Polonetsky 2013; Torra 2017).

As with local privacy, I do not claim that these categories do not have their own explanatory merits. For the purposes of mapping the phenomenal space of the notion of privacy, however, they neatly overlap with the domains of decisional and informational privacy. Therefore, with regard to phenomenon adequacy I will not treat physical or mental privacy as distinct domains. This is also true for the notion of locational privacy, i.e. “the ability to prevent undesired entities from knowing one’s past, present, and future locations” (Bridwell 2007, p. 210). The phenomena covered by this domain of privacy are also categorized by the domain of informational privacy.

Finally, Judith DeCew refers in addition to informational privacy to the domains of “accessibility privacy” and “expressive privacy” (DeCew 1997, pp. 76–77). Accessibility privacy regulates the physical access to persons and covers fundamentally the same phenomena as Floridi’s physical privacy. For this reason, in the same way that Floridi’s account of physical privacy can phenomenon-wise be spelled out in terms of decisional and informational privacy, the phenomena encompassed by accessibility privacy can be categorized by the latter. Expressive privacy, on the other hand, is defined as the “ability to decide to continue or to modify one’s behaviour when the activity in question helps define one-self as a person, shielded from interference, pressure and coercion from government or from other individuals” (DeCew 1997, p. 77). The phenomena that are captured by this dimension also pertain to the dimension of decisional privacy in the way I explicated above, as they are concerned with the possibility of a person to “make decisions and take action in all his social relations” (Roessler 2005, p. 17). If anything, DeCew’s notion is more limited, as it only incorporates decisions and actions that “define one-self as a person”.⁶

In conclusion, from this discussion about different domains of privacy it becomes clear that the phenomena which are commonly subsumed under the heading of local, physical, mental, locational, accessibility or expressive privacy are also covered either by the domain of decisional or informational privacy. For the task of identifying and categorizing the phenomena that a theory of privacy needs to encompass in order to fulfill the criterion of phenomenon adequacy, the other domains do not add phenomena not covered by either decisional of informational privacy. For purposes of clarity and parsimony, I will therefore limit the investigation to these two domains. This means that the criterion of phenomenon adequacy is fulfilled, if a theory of privacy explains the phenomena that belong to those two domains.

From the first two challenges that Floridi mentions—i.e. the varieties of the concept of privacy across different cultural environments (a), and the danger of a Westernized and Western centric notion of individuality (b)—a second criterion can be inferred: a general theory of privacy should not presuppose the value of privacy, but treat this question (the “second question”) separately, displaying a general value openness. This means that the theory can accommodate different empirical facts about values attached to privacy, as well as different normative value theories of privacy. In other words: It should in principle be possible to combine a general theory of privacy (answer to the “first” question) with any value theory of privacy (answer to the “second” question), whether they stress the individual, social, or political value of privacy (Roessler and Mokrosinska 2015), or whether they attach an intrinsic value to privacy (Fried 1984).

In conclusion, a general theory of privacy should be able meet both criteria in order to be viable. As it is the purpose of such a theory to give a unified and neutral definition of the various phenomena of privacy, the criteria of phenomenon adequacy and value openness serve as necessary conditions. While phenomenon adequacy is a necessary feature of any descriptive scientific theory, it is also essential for the descriptive elements of normative theories.⁷ Value openness, on the other hand, essentially hinges on the idea of a division of labor between general and value theories of privacy. While a value theory’s task is to explain why we do or should value privacy, a general theory is supposed to give an account of what privacy actually is in the most neutral fashion possible. The neutrality claim is grounded in the diversity of value attachments and value claims: If it is true that different cultural contexts attach different values to privacy, and if it is also true that competing yet in themselves equally coherent theories of the value of privacy exist,⁸ an answer to the question “What is privacy?” should be as neutral as possible to those diverse attachments and claims.

It has been argued that such an attempt at giving a single answer in the sense of a unified theory of privacy will “never achieve the goal of finding the common denominator” (Solove 2008, p. 38). Rather, we should think of privacy as a cluster concept of different yet related phenomena that cannot be addressed by a single definition. Aside from the question of the parsimony of theories (Baker 2016), cluster theories often have the problem of explaining what it exactly is that relates the different clusters within a given concept. Is it their usage in everyday language, i.e. is the cluster theory effectively giving a description of all the possible meanings of the word “privacy” in the sense of “its use in the language” (Wittgenstein 1998, § 43)? Or is it retracing the scientific debate in a sociology-of-science fashion, trying to essentially list all or the main philosophical definitions of privacy?

This is not to say that cluster theories, such as DeCew’s privacy as the bundle of interests in one’s control over information, access and life choices (DeCew 1997) or Solove’s six types of privacy as in the right to be let alone, limited access to the self, secrecy, control over personal information, personhood, and intimacy (Solove 2008, p. 13), do not have their merits. In laying out a tapestry of privacy bottom up from linguistic usage or expert intuitions, they can point to seemingly disparate usages and incorporate them without having to forego some of them to accommodate a unified theory. On the other hand, many of these seemingly disparate usages can be accounted for in a unified theory, if we distinguish between types of usages of “privacy” as (1) giving an account of privacy (answer to the question “What is privacy?”), (2) stating why we do or should value privacy, (3) pointing out what a right to privacy entails, or (4) laying out the different domains of privacy as elaborated above.

In this sense, Moor and Tavani establish their “restricted access/limited control” theory of privacy (RALC) not as different clusters, but as pertaining to different questions about privacy (Tavani 2007). “Restricted access” refers to the question what privacy is, whereas “limited control” is an answer to the question what a right to privacy entails. In the same manner, the different elements of a cluster can in many instances be assigned to different types of usages. By limiting the scope in the sense of limiting the theory to one type of usage, the cluster of seemingly disparate phenomena in many cases can be transformed into an array of answers to different questions. I cannot go here into the details of the different cluster theories to analyze whether each single aspect can be accommodated for by assigning it to different usages.⁹

Since the social-ontological recognition model of privacy (SORM) advocated here only tries to answer the first question “What is privacy?”, its scope is much more limited than what cluster theories typically set out to do. This gives me confidence that I can provide a unified answer to the question “What is privacy?”, without neglecting important aspects of “privacy” that cluster theories cover, as they belong to a different type of usage. Whether the SORM succeeds in the task of providing a phenomenon adequate answer to this question, depends on whether it can grasp and explain the phenomena in the two domains of decisional and informational privacy as outlined above. This is why I put so much emphasis on singling out the relevant domains of privacy.

What are social practices?

While sociological practice theory presents itself as a disparate and as much empirically focused as self-reflectively interested field of research, most practice theories share a set of common traits that lie at the core of their approaches.¹⁰ For example, Andreas Reckwitz defines a social practice as “an ensemble of interconnected routinely repeated activities of bodies that are being held together by implicit and shared forms of understanding and knowledge” (Reckwitz 2008, p. 192; my translation). And Rahel Jaeggi claims that practices are “habitually performed, rule-bound and socially meaningful complexes of interdependent actions that are of enabling character and with which certain aims are pursued” (Jaeggi 2014, p. 102; my translation).

From these general definitions, I will explicate six features of practices, which are central to most practice–theoretical accounts. While they are neither necessary nor sufficient in the strict sense for an approach to be considered “practice-theoretical”, they mark some kind of “family resemblance” (Wittgenstein 1998, § 67) between the different accounts. At the same time, these six features are crucial for the development of the SORM.

• (a) Practices are, as Schatzki puts it, “doings and sayings” (2002, pp. 72–88). The researcher employing a practice approach focusses on the actions as well as the interpretations of the actors within a certain practice. These are sometimes called the “materiality” (Reckwitz 2003, p. 290; my translation) or “performance aspect” (Hillebrandt 2014, p. 59; my translation) of practices: the actor’s bodies as well as physical artefacts are involved in the constitution—and therefore also in the empirical recognition—of practices.

  Although some practice theorists consider practices as the “smallest unit of the social” (Reckwitz 2003, p. 290; my translation), or the “elementary occurrences of sociality” (Hillebrandt 2014, p. 59; my translation), this claim against classical action theory is not crucial to the understanding of social practices. It is sufficient to identify social practices as the central elements of the social, not necessarily granting them an “ontological precedence” (Schmidt 2012, p. 24; my translation).

• (b) Practice theory claims that not all of the “doings and sayings” within a practice are deliberately performed, in the sense e.g. rational choice theory talks about actions. On the contrary, within a practice many actions are carried out habitually and are repeated on a routine basis. The participants in a practice oftentimes follow the rules of the practice “blindly” (Wittgenstein 1998, § 219), meaning that their knowledge of the practice norms is often “implicit” (Schmidt 2012, p. 57; my translation). In this regard practice theorists frequently mention a “sens pratique” (Bourdieu 1990, Ch. 3), a “practical consciousness” (Giddens 1984, Ch. 2), or a “knowing how to” (Schatzki 1996, pp. 91, 129–132). All these notions are trying to express the idea that within practices actors often do things without rational reflection. Rather, they follow internalized norms in a habitual manner, having acquired these norms through some form of socialization into the practice.

  From the observation of repeated behavior patterns, it is only a small step to the assertion that this behavior is in some way “rule-bound” (Jaeggi 2014, p. 102; my translation). If there is a pattern that needs to be known, interpreted, and reinstantiated, the practice establishes some kind of rules that the participants need to take into account and reproduce by either reenacting or contesting it. This implies that from the perspective of the participants there is a right or wrong way to instantiate the practice, which is embedded in the structure of the practice itself. As “implicit and shared forms of understanding and knowledge” (Reckwitz 2008, p. 192; my translation), these rules are typically acquired through some form of socialization—i.e. imitation, repetition, and learning of social norms—into the practice.

• (c) As the (implicit as well as explicit) following of rules make up the visible aspect of practices, they are ultimately constituted and integrated by social norms. In following those norms, the actors instantiate, interpret, and reproduce not only the practice, but at the same time the norms themselves. These norms can be formulated as role obligations of the roles that are available within a certain practice (Hardimon 1994).¹¹ Assuming a specific role within a given practice generates specific obligations towards the other participants, which differ across various practices. Criticism by fellow participants revolves around the adequate fulfillment of these roles in the specific instantiation—and therewith interpretation—of a certain role. Not comporting oneself like a “good” X (soccer player, doctor, father, speaker of language L etc.) refers to the implicit or explicit role obligations that pertain to the practice.¹² The critique “that’s not how we φ” then explicitly points at the norms that one has to comply with when occupying a certain role within a given practice according to the “standard interpretation” of the corresponding role obligations.

• (d) From the possibility of criticizing each other with respect to the norms of the practice, it follows that the norms of the practice must in principle explicable, even though it may be true that “physically” following a rule within an actual practice “cannot be equated with the rule as it is written down” (Hillebrandt 2014, p. 62; my translation). Normative obligations within a social practice might be tacit knowledge all the way down, but will surface and be “made explicit” (Brandom 1998) in the event that disagreement between the participants about the expected behavior arises.

  Since it is not obvious, however, that one of the—potentially many—conflicting interpretations of the implicit practice norms is more adequate than any other, according to many critical accounts of practice theory it is the task of the social theorist to analyze the practice in order to explicate the norms that are implicit within the doings and sayings of the participants (Bourdieu 1977; Habermas 1995; Honneth 2014b; Jaeggi 2014).¹³ The explicit reflections of certain participants about the practice norms may not only stand in contrast to their implicit rule following, but also may not be the most adequate self-description of what they are actually doing within the practice. This is not to say, however, that the sociologist perspective necessarily trumps the self-interpretations of the participants, but rather that it can add a second layer to them that provides some normative potential for critique (Celikates 2009; Stahl 2013).

• (e) As we have seen, practices are constituted and integrated by bundles of norms. These norms are spelled out in terms of role obligations of roles within the practice. In addition to the first order deontic status of role obligations, the very reality of conflicting interpretations with respect to the right way to instantiate the practice, as well as the possibility that participants criticize each other, gives rise to a second order deontic status of a “standard authority” (Stahl 2013, p. 84; my translation, emphasis added) with regard to the norms of the practice.

  This means that the participants recognize each other on a social-ontologically more basic level as equally accountable with respect to the distribution of roles and corresponding role obligations. Although different roles will distribute normative status positions very unevenly (think about the judge in comparison to the defendant or even the prosecutor and the attorney), what these roles entail is open to a “standard” criticism by everyone occupying a constitutive role within the practice. A constitutive role is a role without which the practice cannot be instantiated, in contrast to purely accidental roles such as onlookers and the like. For example, the practice of shopping differs considerably in its instantiations, but to count as shopping, there has to be at least the role of the buyer/customer and the seller/shopkeeper.

  Because all bearers of constitutive roles reproduce (in the sense of interpret, affirm, and contest) the practice norms via their performances within their roles, they quasi automatically play a part in the “game of giving and asking for reasons” (Brandom 1998, p. 89), which will ensue as soon as there is disagreement about the right interpretation of norms. Therefore, the concept of a standard authority claims that in our everyday practices bearers of constitutive roles typically recognize every other bearer of such a role as someone to whom she is accountable with regard to her role privileges and obligations. This point is crucial for the social-ontological account of privacy that I am advocating here, as will become clear later on. For now, it suffices to understand that the idea of a standard authority works on two levels:

  On the first level, every bearer of a constitutive role, a status I call being a “participant of the practice”, is accountable to every other participant in the discharging of her role obligations within the practice. This means that every participant has the standard authority to criticize any other participant of the practice. In turn, this mutual recognition incurs the meta-obligation to answer to this criticism and justify one’s own behavior. On this first level, the standard authority entails a second order deontic status, insofar as it incorporates the fundamental possibility of all participants to equally interpret the practice and criticize each other.

  On a second level, the role privileges and obligations of each constitutive role are themselves subject to interpretation, criticism and justification by all participants of a certain practice. They have—by way of their reenactment, interpretation and contestation of the practice norms—a say in the constitution and reproduction of the practice. The idea of standard authority claims that at the social-ontological level this say typically is an equal say. Although the doctor has considerably more role privileges than the patient, she is together with all other participants equally accountable as to what these privileges entail. On this second level, the second order deontic status of standard authority constitutes a meta-practice of critique that complements the practice itself. This meta-practice is part of the ontological structure of every social practice.

• (f) The notion of “constitutive roles” directly refers to the question of the individuation of practices. How does the practice theorist know that a certain ensemble of “doings and sayings” makes up a specific practice? How is it that the participants’ interpretations and contestations are directed at the same practice and not for example at two similar but different practices with slightly differing role obligations (Applbaum 1999, Ch. 3)? For the purposes of the social-ontological approach to privacy, I will individuate practices by their goals.¹⁴ When giving interpretations, participants of practices typically revert to “standards of excellence” (MacIntyre 1985, p. 187) that belong to roles—i.e. being a “good” X. These standards in turn only hold in light of the goals or values that the participants are pursuing collectively with or within the practice. Without these goals, there would be no standard within the practice to evaluate deviating interpretations of practice norms.

  Moreover, the notion of a meta-practice of critique implies a yardstick for the critique of the existing role obligations, which cannot itself be part of those role obligations. Rather, such a benchmark must point to some goals or values, which are to be realized with or within the practice. In other words: If the practice participants were to contest the way role privileges and obligations are distributed within e.g. the practice of judicial hearings, they will sooner or later have to refer to the function of judicial hearings and the fundamental values incorporated into its practice norms. Therefore, analyzing the function of a given practice—for the greater scheme of practices that it is part of, but also for the participants involved—gives the sociologist or social theorist a starting point, from which to conduct her observation of the doings and sayings.

The argument in full

A standard social role model of privacy (SSRM) explains the intersubjective aspect of privacy in practices by way of norms that are institutionalized as role obligations within social practices (Benn 1984; Roessler and Mokrosinska 2013). In this respect, it draws on social practice theory as I explicated it above, to show that privacy is fundamentally a relational category which is normatively structured by the relevant practice it is embedded in: “What others, such as students or bankers, know about me, and I about them, is also principally guided by the type of relationship I am in and the roles enacted within that relationship” (Roessler and Mokrosinska 2013, p. 777). What privacy entails is fundamentally determined by the context (Nissenbaum 2004, 2010; Barth et al. 2006), i.e. the practice or bundle of interlocking practices.

Depending on how the SSRM is spelled out, it can cope with both the criteria that I outlined earlier for a general theory of privacy. On the one hand, it can capture all relevant phenomena of privacy, as it does not prima facie exclude any practices as irrelevant to the phenomenon of privacy. On the other hand, it retains value openness, either by situating the value of privacy within the practice, i.e. subscribing to a “practice positivism” (Applbaum 1999, Ch. 3), or by giving a separate account as to why we value privacy. The latter is the approach that e.g. Roessler and Mokrosinska take when they assert that privacy is not only “morally valuable for the individuals involved, but which also directly serve to promote social integration” (Roessler and Mokrosinska 2013, p. 776).

As a description about social phenomena, I wholeheartedly agree with the SSRM and the social theory behind it. A practice-theoretical explanation of social interactions is best-suited to make sense of the part that social norms play for social integration, as well as the idea of habitual rule-following tied to social roles. Furthermore, such an explanation cannot only show how different contexts generate different privacy requirements by explicating these in terms of role obligations within specific practices. At the same time, the practice-theoretical account makes clear that privacy norms are “also necessary for the constitution and regulation of social roles, relationships and, more generally, social practices” (Roessler and Mokrosinska 2013, p. 774).

The SSRM, however, cannot specify in further detail the social-ontological connection that it postulates, by claiming that privacy norms constitute practices. Rather, in order to establish this connection, it has to rely on answers to the second question, i.e. why one should value privacy. In other words: Whether privacy norms are in effect constitutive for all kinds of social practices, as maintained by the standard model, is a normative claim about why we as a society should value privacy norms. For its justification, it either depends on a value theory of privacy, which the SSRM itself cannot and does not want to give; or it hinges on the empirical evidence that the functioning of modern societies essentially depends on the existence of privacy norms within its practices.

More importantly, still, the SSRM cannot even answer the first question: Neither from the fact that privacy norms are embedded in social roles within practices, nor from the assertion that they are constitutive for the practices, the SSRM is able to individuate which norms are privacy norms. As a result, the standard model is in essence not an account of what privacy actually is. Roessler and Mokrosinska implicitly acknowledge this shortcoming by giving a freestanding definition of informational privacy that is largely independent of the social-ontological claims the SSRM makes: “Information is private when a person can control access to it herself, or when she can be at least relatively secure in her expectations of being able to monitor access to and disclosure of information” (Roessler and Mokrosinska 2013, p. 772).

Ultimately, their version of the SSRM does not give an answer to the first question (What is privacy?) at all, but rather to the second: We should value privacy, because it is constitutive for our social practices. For this reason, their answer to the first question does not draw on the social ontology of practices at all, but rather on notions of control/access that pertain to individuals and not to practices.

In contrast, the social-ontological recognitional model of privacy (SORM) that I advocate in this paper, provides a social-ontological answer to the first question: Privacy in this account is a second order quality of roles within social practices, which assigns practice roles the second order deontic status of a “standard authority” (Stahl 2013, p. 379, 2014) with regard to the norms of the practice. This status means to have an equal say in the interpretation of the practice participants’ behavior with respect to their role obligations (first level standard authority). Put differently: By enacting the practice, the participants perform a series of “doings and sayings” that are expected of them with regard to the role they are embodying. In these performances the participants interpret their role and thereby affirm or contest their role obligations. Deviations in interpretation can and often will be made explicit, i.e. contested and criticized, requiring justification on the part of the critiquing as well as criticized participant. In this “game of giving and asking for reasons” (Brandom 1998, p. 89), we implicitly recognize each other’s performances as skilled interpretations that are backed up by validity claims (Habermas 1984, vol. 1, Part III).

Additionally, the notion of standard authority implies a second level authority in the sense that the role obligations themselves are equally interpretable and contestable by all practice participants. This flows naturally from the first level mentioned in the last paragraph: If participant A of a practice p criticizes a certain performance by participant B, and an argument about the right interpretation of the corresponding role privileges or obligations ensues, the argument may revolve around B’s performance. In this case, A and B agree on the norms, but disagree whether the performance adheres to the norms. They both could, however, also disagree about the norms. In this case, we move from the first level authority to the second. While this happens naturally in the act of making the practice norms explicit, it is worth separating the two analytically, as the second level authority will do much of the argumentative work for a notion of privacy a little later on.

To sum up: In addition to the first order deontic status of role bearers with role obligations, practice participants incur the second order deontic status of a standard authority, which gives rise to a first- and a second-level interpretational status. This status is characterized by the ability to make authoritative claims to the validity of one’s performative interpretations of one’s own or other’s role obligations. By virtue of this authority, A’s performances, even if criticized or contested, are recognized as attempts at valid interpretations of the right way to instantiate the practice. This means that every criticism of her performance has to resort to some form of justification with regard to the norms of the practice. In other words: If B were to criticize A with regard to her interpretation of her role as a doctor, B would have to make a claim that “this is not how doctors should behave”, to which A can reply with her own interpretation.

Although the first order role obligations may distribute authority within the practice highly unevenly (judges have more role privileges than defendants), the second order status of standard authority is generally distributed equally among the participants. This is the case by virtue of the social-ontological workings of social practices: Each performance perforce reenacts a certain role that the performer is instantiating with a given practice, thereby interpreting the practice norms. Typically, the other participants take these performances as attempts of valid claims at interpreting the practice. In practices, we naturally assume that the other participant is a skilled performer, i.e. knows her role and the corresponding obligations, and tries to discharge them in her performance.

This is where the weak normativity of the SORM comes into play. From this observation of the typical mutual recognition of the practice participants as skilled performers, the notion of standard authority claims that—in order to count as an adequate instantiation of a practice—all participants must have an equal say in the second order interpretation of the roles, role obligations and whether they are discharged correctly within the practice in question. In other words: In a practice that is structured and instantiated adequately, the participants must recognize each other as standard authorities. This means that they have to mutually accept each other’s interpretations and contestations as equally relevant interpretational contributions to the practice, even though they may disagree with regard to the content.

This kind of normativity can be labeled as “weak”, as it is not the result of a normative theory about social practices, but rather the result of how social integration works. As Hart argues in his distinction between “being obliged” and “having an obligation” (Hart 1994, Ch. V), a social order cannot be integrated, stabilized, and reproduced by coercion alone. Hegel makes a similar point, when he asserts that “representational thought [Vorstellung] often has the impression that force holds the state together, but in fact its only bond is the fundamental sense of order which everyone possesses” (Houlgate 2008, p. 268 Addition).

Put in terms of social psychology, this means that in order to normatively bind the participants long term, a practice must provide a motivation for them to uphold it. The deontic status of a standard authority provides one aspect of this motivation on a fundamental level.¹⁵ Participants take part in and reproduce social practices, as long as they can regard themselves at least implicitly as standard authorities with respect to the practice norms. Therefore, the demand that the participants recognize each other as standard authorities may be normative in nature, but it is “normatively weak” in the sense that this mutual recognition is required for practices to function effectively: They stabilize behavioral expectations by way of social integration through norms in the form of role obligations.

In essence, this is the social-ontological aspect of the SORM: mutual recognition as standard authority is necessary for practices to do their social-ontological work as social structures, i.e. to integrate individual actors as participants and stabilize behavior and behavioral expectations through normative demands on the participants. Without this recognition, practices will cease to be reproduced, i.e. deteriorate, become pathological (cf. section “Social pathologies”) and disintegrate in the long run. From this, I can rephrase the weak normative claim of the SORM thusly: to truly participate in a practice means to be recognized by the other participants as standard authority, and to recognize them as such in return.

But if the recognition as a standard authority means to be able to participate in the practice, does this not make the concept of standard authority a social-ontological explanation for inclusion and participation rather than for privacy? Does this explanation not incorporate phenomena that do not belong to the notion of privacy and is therefore too wide in terms of phenomenon adequacy? This is where the two main dimensions of privacy come into play. The concept of privacy intersects at two specific points with the social-ontological idea of standard authority. In spelling out this idea with respect to the two main dimensions of privacy (decisional and informational), I can give a social-ontological answer to the question “What is privacy?”:

If it is true that within a certain practice p all participants, i.e. all bearers of constitutive roles (BCR), have to recognize each other as standard authorities, decisional privacy lies in the fact that they do not have to recognize bearers of accidental roles (BAR) as such. For example, when I enter a church, the attendants as BCRs have the right to criticize my choice of wardrobe as inadequate for this kind of sacral space. In recognizing them as a standard authority, I have to give justifications for my behavior. In front of the church, however, people just passing me by do not have this right (assuming that I am about to enter the church and they therefore only have accidental roles with regard to the practice of visiting a church). This is not to say that they may not criticize me, but since they enjoy no standard authority, I incur no obligation to react to this criticism.

In other words: decisional privacy is a function of the freedom not to respond to the criticisms of BARs within a practice. On the level of the meta-practice of critique, which—as I have shown in the last section—complements each practice, decisional privacy depends on assigning constitutive and accidental roles. When we exercise our standard authority on this second level and evaluate the structure of a given practice itself, agreeing on a specific role as being purely accidental defines the relational space between the bearers of this role and all BCRs as one of decisional privacy. On a social-ontological level, this is the practice-theoretical answer to the question: “What is decisional privacy?”

With regard to informational privacy, norms that govern access to information and information processing are—like all other practice norms—subject to the standard authority of the practice participants. This means that all participants of the practice may criticize other participants for not adequately discharging their informational privacy obligations, which are part of their respective role descriptions. For example, as a patient I have the second order status (i.e. as bearer of a BCR a standard authority) to criticize my doctor, if she freely distributes my medical records. In addition, all participants have an equal say as to what kind of informational privacy norms should be attached to which roles within the practice. Following the example, this means that the notion of standard authority provides me with an equal voice as to how informational privacy with respect to medical records is to be spelled out.¹⁶

Furthermore, this second order critique in form of a meta-practice also serves to evaluate, contest, and justify the distribution of constitutive and accidental roles, much like in the case of decisional privacy. With regard to the practice of medial examinations, the participants evaluate e.g. the role of the assistant by reacting to his performances as valid claims to the right interpretation of his role. This does not mean, however, that all practice-internal interpretations as to what counts as a constitutive role are equally valid. There is a social-ontological limit on the self-interpretations of the participants: From a social-ontological perspective, a constitutive role r is a role without which the practice in question could not exist in its current form. Although in reality there will be many highly contentious cases, if it can be shown that r is constitutive for practice p, then self-interpretations that assign only an accidental status to r with respect to p, are flawed.

This limit is essential for the understanding of informational privacy, especially in practices of collecting and processing personal data. Since these practices cannot exist without actors that are datacized, i.e. whose personal data is collected and processed, these actors have to be regarded as BCRs. Therefore, by way of the weak normativity that comes with the notion of standard authority, they are owed an equal say in the norms of the practice. This entails the right to obtain justification for their interpretations and contestations of informational privacy norms on the two levels explicated above. On the first level, justification is owed by the data collecting and processing actors with regard to the enactment of their role obligations. If e.g. data privacy regulations and terms of service are not honored, the SORM claims that the social ontology of practices assigns the datacized participants as BCRs a standard authority to criticize these performances.

In order to evaluate and possibly criticize these performances as interpretations of the practice norms, the other participants—especially the datacized individuals—need to be able to appreciate them fully. Therefore, the equal say that the notion of standard authority confers cannot just be realized as formal consent to the terms of service. Rather, the artificial cognitive barriers in the sense of incomprehensible fine prints, unnecessary access privileges, and other privacy-by-design violations must be removed or at least reduced to a level where a substantial exercise of one’s standard authority is possible. I will address this issue in more detail in the next section.

On the second level, the standard authority bestows all BCRs the right for justification with respect to the structure and content of the obligations themselves. As standard authorities in data collecting and processing practices, all BCRs should have an equal say when it comes to how the scheme of informational privacy norms is structured, fleshed out, and enacted. In other words, they should have an equal say in e.g. what informed consent amounts to (Ulbricht and Weber 2017) and where and how privacy-by-design principles should come into play (Schaar 2010).¹⁷

To sum up the claim of the SORM: Privacy is a second order quality of roles within social practices, which assigns all bearers of constitutive roles of the practice (BCRs) the second order deontic status of a standard authority in the two domains of decisional and informational privacy. By virtue of this authority they have an equal say with regard to the practice norms. On the first level, they may criticize all other BCRs in their enactment of their respective role obligations, while on the second level they can evaluate the structure of the practice itself, i.e. criticize the existence and structure of roles, role obligations etc.

Within a practice, an actor is said to have decisional privacy if she as a BCR does not (or does not have to) recognize bearers of accidental roles as standard authorities. This does not preclude that she has comparatively little decisional privacy within a specific practice, e.g. at work, especially vis-à-vis her boss as one of the main BCRs. It is important, however, that her performances are regarded as attempts at valid claims to the right interpretation of her role, even by her boss. In doing so, the employees may be able to gradually change their role obligations through contesting interpretations of their roles.¹⁸

Vice versa, an actor is said to enjoy informational privacy if all other BCRs (and especially data collecting actors) recognize her as a standard authority. This means that she has an equal say with respect to the norms that govern information about her. This is especially (but not exclusively) prevalent in data collecting and processing practices: Although as a datacized individual she naturally assumes a constitutive role in data collecting and processing practices, she is often not recognized as a standard authority and therefore has no equal say. As with decisional privacy, the SORM says very little about the actual privacy arrangements within a given practice. It may be the case that within a certain practice, informational privacy is very limited, especially with respect to certain other BCRs (think of the doctor-patient example). However, the SORM claims that this is legitimate, as long as all BCRs enjoy a standard authority and have an equal say in the evaluation of the particular privacy scheme.

This is in essence where the difference between the SORM and the SSRM lies. While the SSRM spells out privacy as a set of norms in a given context (Nissenbaum 2010) or within a certain practice (Roessler and Mokrosinska 2013), the SORM does not disagree. It refrains, however, from explicating these norms, leaving their construction, interpretation and reproduction to the practice participants. Instead, it evaluates whether all participants (i.e. BCRs) have the real opportunity to do so. It claims that on a more fundamental social-ontological level, privacy means the equal say with respect to norms about autonomous decisions and information distribution within a practice. Therefore, even in a practice where the participants enjoy little informational privacy—e.g. the problem of “privacy in public” (Nissenbaum 1998)—they are said to have informational privacy as long as they enjoy a standard authority with respect to the norms that govern informational privacy within this particular practice.

By specifically addressing the two dimensions of privacy, which I have singled out as the phenomenon-wise encompassing categories, and giving an account of how the mechanism of a standard authority differs in each dimension, the SORM is phenomenon adequate in the above-mentioned sense. Descriptively, it claims that privacy in both domains amounts to having an equal say vis-à-vis other BCRs—or not having to respond to BARs respectively—, which moves it somewhat in the direction of control theories (Westin 1967; Rachels 1984; Fried 1984; Roessler 2005). However, an equal say in the norms that govern life choices (decisional privacy) or personal information (informational privacy) is not the same thing as having actual control over one’s life choices or personal information. The latter is a function of the actual privacy norms within a practice. The SORM does not make any claims on these: Individuals within a practice can enjoy privacy even if they do not withhold any control over life choices or personal information, as long as they have an equal say in the arrangement of the norms governing both decisional and informational privacy.

With respect to the criterion of value openness, it does not presuppose why we do or should value privacy individually. Rather, the SORM is compatible with answers such as that privacy is crucial for our autonomy, authenticity, or intimacy. An objection to this assertion may be raised with regard to the weak normativity the SORM implies. While it explains privacy only by recurring to social-ontological mechanisms, one might demur that the demand for mutual recognition as standard authority depends on the value of stabilizing social practices by motivating their participants to uphold them. Against this objection, I claim that while the function of stabilizing practices can be viewed as having value, this value is only instrumental with respect to the more fundamental value that is realized with a specific practice. In other words: If practice p realizes certain values v_1 − n and the participants therefore seek to uphold it, they need to make sure that all participants recognize each other as a standard authority. Otherwise, p cannot be stable and is therefore not able to realize v_1 − n. As the SORM does not presuppose anything about v_1 − n, it remains value-open in the sense explicated in section “Criteria for a theory of privacy”.

Social pathologies

One of the main advantages that come with the weak normativity the SORM entails, is the possibility to criticize privacy arrangements within practices as deficient or pathological. A social pathology in this sense is a structural misrepresentation of a practice or bundle of practices, such that it does not award all participants the status of standard authority. In this situation, any “inconsistency between explicit and implicit norms, for systemic reasons cannot be dealt with as an explicit conflict about norms” (Stahl 2013, p. 367; my translation).¹⁹ From this structural inability to incorporate the interpretations and contestations of all practice participants, “disruptions and disorganizations in the reproduction” (Honneth 2014a, p. 18; my translation) of the practice in question will result. If the performances of some participants are systematically disregarded and not viewed as valid claims to an adequate interpretation of their role obligations, this will eventually lead to “defective social relations” (Roessler and Mokrosinska 2013, p. 780), which in turn will ultimately disintegrate the social practice.

Both the standard social role model of privacy (SSRM) and the SORM claim that structurally defective privacy arrangements within a practice will have this effect. However, while the SSRM argues that the normative relations between the role bearers within a practice p itself are deficiently structured—therefore the problem lies with the privacy norms of p—the SORM maintains that this is a result of an underlying deficiency in the distribution and mutual recognition of the standard authority in p. Because not all BCRs are recognized as standard authorities (or some BARs are erroneously recognized as such), deficient privacy norms will arise. Following this argument, the SORM—unlike the SSRM—cannot only explain why there exist deficient privacy relations within p, but also why they are structural as opposed to just accidental.

Analogous to the two levels of standard authority, the SORM can distinguish between first and second order social pathologies. On the first level, a social pathology occurs when some constitutive roles are structurally excluded from the interpretation of the practice norms, or vice versa some accidental roles are included. This means that not all BCRs of a specific type of constitutive role are recognized as standard authority, or that some BARs holding a specific type of accidental role are wrongly recognized as a standard authority. On the second level, the meta-practice of critique is somehow distorted. A second order social pathology ensues if some participants (i.e. BCRs) are structurally excluded from the possibility to define and contest the normative structure of the practice itself, i.e. the distribution of constitutive and accidental roles.

While a first order pathology structurally excludes BCRs (or unwarrantedly includes BARs) from criticizing either behavior on the basis of role obligations, or the role obligations themselves, a second order pathology deficiently assigns the second level standard authority to have a say in the distribution of constitutive and accidental roles. When breaking down this social-ontological claim to the two main dimensions of privacy, a social practice may become pathological with respect to decisional privacy in two ways: A first order pathology typically results in the systematic confusion of BCRs and BARs, effectively recognizing some BARs as standard authority. An example for this would be if the practice of visiting a church would systematically recognize random bystanders in front of the church as BCRs with respect to the practice, and would therefore accept them as a standard authority on the clothing norms of the practice.²⁰

In contrast, a second order pathology amounts to the structural inability of all participants to engage in the interpretation and assignment of BCRs and BARs respectively. In this case, the participants do not only wrongfully believe that bystanders should have the authority to criticize them for their wardrobe choices, even if they do not. Rather, the practice is structured in a way that prevents them from exercising their authority to assign constitutive and accidental roles in the first place. They have no effective voice within the complementary meta-practice of critique.

In much the same way, social pathologies work within the dimension of informational privacy: Analogous to decisional privacy, a first order pathology signifies the confusion of constitutive and accidental roles. However, in the case of informational privacy, it is usually the other way around. Instead of accidentally including some BARs as standard authority, in the domain of informational privacy typically actors whose personal data is collected and processed, are structurally excluded from the practice within which the data is collected and processed. This means that they are systematically not recognized as BCRs and therefore have no standard authority as to which data is collected and how it is used. Floridi aptly describes this kind of pathology thus:

We expect websites to monitor and record our activities and do not even seem to mind for what purpose. It is not that we do not care about privacy, but that we accept that being online may be one of the less private things in our life (Floridi 2014, p. 110).

If Floridi is right, in many practices of digital data collecting and processing the participants implicitly accept that they are not fully recognized as BCRs and therefore can only exercise their standard authority in a deficient way. Clickwrap and incomprehensible end user agreements, excessive access privileges, and other privacy-by-design violations build up cognitive barriers that systematically prevent the participants from exercising their standard authority. However, as datacized persons, they are BCRs and therefore should enjoy a standard authority (weak normative claim of the SORM). Systematically excluding them from this authority constitutes a pathological practice.

With the ever-increasing diffusion of digital technologies into other social practices—for example through the internet of things—these pathologies will disseminate into more and more practices. As the disuse of digital devices is increasingly onerous and becomes less and less of an option, the question whether the participants even want to be able to trade their data for free content becomes more prevalent (Roessler 2015). Systematically obscuring this question by exploiting psychological mechanisms (Shampanier et al. 2007) and offering free apps in return for personal data, amounts to a pathological practice if it effectively prevents the participants from exercising their standard authority. This is not only a pathology with regard to informational privacy, however, but equally with respect to decisional privacy, as the decision to forego informational privacy for the benefit of accessing free content and using free apps (as informed consent), is structurally undermined.

While a first order pathology within the dimension of informational privacy usually means a structural disregard of the datacized actors as BCRs, to the effect that they “become less clear on who their actions are accessible to and in what circumstances their actions may be reviewed” (Patton 2000, p. 184), a second order pathology affects their possibility to evaluate the normative structure of the practice itself. As with the dimension of decisional privacy, there is a systematic non-recognition of standard authority that prevents the BCRs to have an equal say in which role should count as constitutive and which should count as accidental.

This can be elucidated by the example of mass surveillance practices (Patton 2000; Parsons 2015; Stahl 2016). In these practices, typically the monitored persons do not only have no say in what data is collected and how it is used, but also have no or very limited say in who should have a say. Even if the majority of participants in these practices were not as complacent and unalarmed by the mass surveillance activities of intelligence agencies and secret services, it is unclear if these institutions can even be monitored effectively. If this is not the case, the other participants of these practices—i.e. the majority of society, as far as we know from the Snowden disclosures—do not even have the possibility of an equal say.

The nature of social pathologies makes it often very difficult for the participants to discern that they are subject to such a pathology. From their point of view, the structure of a pathological practice seems to be justified and the practice itself the “right way to φ”. This is exactly what Floridi points to when he is concerned that we have come to accept the online part of our lives as less private than the offline part. In order to recognize the pathological structure within a practice or bundles of practices, it usually needs a sociological point of view that adds a second layer to the self-interpretations of the practice participants. A sociologist or social theorist will be able to show why a certain practice is pathological, i.e. why the structure of the practice is constructed in a way as to systematically prevent its participants to realize and articulate its deficiencies that I have spelled out here in terms of inadequate relations of mutual recognition as standard authority.²¹

Conclusion

With the social-ontological recognitional model of privacy or SORM, I have established an answer to the first question “What is privacy?” by linking it closely to the social-ontological concept of “standard authority”. While privacy norms (or their lacking) are part of all practices, on a socially more fundamental level, privacy is a function of who is or should be recognized as a standard authority with regard to their autonomous decisions (decisional privacy) and the information distribution (informational privacy) within a practice. This means to have the right to interpret and contest role behavior and role obligations (first level), as well as to evaluate the normative structure, the fundamental practice norms, and the roles and their status (second level). Especially the distribution of constitutive vs. accidental roles on the second level plays an essential part in explaining privacy in terms of social practices and role obligations.

In general, the recognition as a standard authority means to be able to participate in the practice, which prima facie makes “standard authority” rather a concept of social inclusion than of privacy. However, privacy prominently comes in at two points of this concept, which are specific to the two dimensions which encompass the relevant phenomena of privacy: Decisional privacy means to be able to exclude certain roles from having a standard authority, i.e. recognizing them as accidental rather than constitutive. As BARs do not enjoy a standard authority within p, their interpretations and criticism do not have to be addressed by the BCRs of p. Informational privacy, on the contrary, must include all actors whose personal data is being collected and processed as BCRs within p. Failure to do so results in a practice which is deficient with regard to informational privacy, no matter what the first level practice norms dictate.

In explicating the concept of privacy in social-ontological terms, the SORM has a variety of advantages: First of all, it meets the two criteria, phenomenon adequacy and value openness, which are necessary conditions for any general theory of privacy. Secondly, it explains privacy on a fundamental level as a social-ontological recognition relation between practice participants. This relation is constitutive for the functioning—i.e. stabilization and reproduction—of practices. Third, this constitutive relation generates a weak normativity that enables an evaluation of practices as adequate or deficient in reference to the social-ontological recognition of standard authority. Deficiencies can be spelled out in terms of first and second order social pathologies, reflecting the constitutive role that privacy plays for social practices. In addition, from this relation the SORM is able to explicate why there may exist deficient privacy relations within a practice, and also why these deficiences are structural as opposed to just accidental.

In contrast to other relational theories of privacy, the SORM is able to shed light on this relation and explicate on a social-ontological level, why privacy is so important for the stabilization and reproduction of social practices. By doing so, it not only adds a social-ontological insight on the question “What is privacy?”, but also gives a fundamental social-ontological reason why we should value it: If we value our social practices, we need to uphold them, which we will only be motivated if we as participants are recognized as standard authorities. This recognition with respect to autonomous decision and life choices, as well as the distribution of information is what we call “privacy”.