This chapter offers some inconclusive reflections as to whether a fundamental concept of privacy underlies the conceptions discussed in Level 1, in a putative Level 0. The account looks for linguistic markers of privacy issues, taking the most important to be first person singular and plural possessive adjectives. The idea of a subject of a privacy claim is examined, especially as located in the self. Objections from postmodernism and Buddhism are raised and rejected, but the fact that people with disorders of the self still have privacy requirements is a more serious objection. If the self is to be the locus of privacy, then it is better seen as a marker of interpersonal bounds, rather than making a strong and controversial commitment to a metaphysical unity. The reference list is used to show the range of phenomena and structures that might be cited as subjects of privacy in ordinary language claims. The possessive linguistic markers of privacy also implicate a way of thinking about the self as the site of self-ownership, which suggests William James’ account of the extended self as particularly sympathetic. None of this is conclusive, and indeed only accounts for individual privacy, not group privacy. Finally, the heterogeneous account of privacy sketched in this chapter is contrasted with McStay’s Level 0 account based on Heidegger and Latour of privacy as an actant, with which it shares some features, despite fundamental disagreements.
The final chapter begins by summarising the thesis of the book. To illustrate how ideas about privacy change, and how the privacy discourse framework can help, the chapter looks at the COVID-19 pandemic, and how this fed into the various privacy discourses. At level 1, few new conceptions emerged. Level 2 is implicated by new uses of personal data for research and contact tracing. At level 3, our experience of others and personal space is altering, with a greater sense of threat in proximity. At level 4, attitudes to surveillance have grown more positive. At level 5, norms about working from home are changing, as are those about personal autonomy. At level 6 there are new governmental powers, for lockdowns, travel and vaccination mandates. Level 7 covers the responsibilities of the individual to the community, as well as medical research ethics. COVID has accelerated some long-term privacy trends, related to technological change. These are shifting the emphasis in modernity from choice to recommendation, which is built on a shift from privacy (for reflection on choices available) to personalisation (requiring information about individuals to be available to recommendation systems).
This chapter considers the academic discourse on privacy since 1890, and distinguishes two trends, one lasting until the mid-1970s, and a second that emerged in the 1960s and which is now dominant. It begins with an examination pf three classic papers exemplifying the first trend: ‘The Right to Privacy’ by Warren and Brandeis (1890), ‘Privacy’, by Prosser (1960) and ‘The Right to Privacy’ by Thomson (1975). These three very different and opposed papers have in common that they do not define privacy, assuming that the reader is fully conversant with what it is, and focusing instead on how it is treated in law and rights-based discourse. In the second trend, typical since ‘Privacy and Freedom’ by Westin (1967), privacy is defined, and then reasoned about. Arguments from Daniel Solove and others are considered that the result has been definitional incoherence and disarray. The chapter argues that attempts to produce more complex taxonomies of privacy do not help reduce the disarray. This change in trend in academic writing is down partly to a greater diversity of academic readership, but more importantly to the greater salience of threats to privacy from technology.
Comparing ordinary language usage of ‘privacy’ with the academic literature, this chapter creates a ‘reference list’ of non-theoretical uses of the term, covering: informational privacy; decisional privacy; private property; psychological privacy; ideological privacy; spatial privacy; attentional privacy and extrinsic privacy (or obtrusion). The reference list is meant as a heuristic device but shows that in ordinary usage the term is wide yet tractable. This contrasts with the ‘disarray’ that commentators have detected in the academic literature.
The previous two chapters exposed a divergence between messy but tractable ordinary usage of the word ‘privacy’ and an apparently incoherent academic literature. Two views on this are discerned: a narrow view, defended in this book, that the incoherence is down to different academics with diverse perspectives and imperatives; and a wider view, that the academic incoherence is symptomatic of genuine conceptual incoherence. If the latter view is correct, then there will be serious problems translating ideas about privacy into effective policy and regulation.
The first explanation, from Deirdre Mulligan, is that privacy is an essentially contested concept (ECC), as described by Gallie. The chapter considers and rejects two important criteria for privacy being such a concept. First, privacy is rejected as an appraisive concept. Whereas with an appraisive concept such as an ECC, we agree about its value but disagree about its application in a specific context, privacy has the opposite behaviour. We tend to agree about its application, but disagree about its value. This is shown by contrasting privacy with justice, an ECC. Second, privacy is shown not to be open; its definition has not evolved over time through being contested, as with a typical ECC. New ideas of privacy have tended not to cause adjustment to older views. This is in contrast with the ECC democracy, which has evolved over time; earlier notions of democracy are not seen as democratic now. Furthermore, many apparent changes in our attitudes to privacy are the results of the evolution of privacy law in response to technical change, not the evolution of the concept itself. Privacy fails to be an ECC, but this discussion is instructive: both criteria fail because of the agreement we have about its use.
Daniel Solove’s diagnosis of the disarray of the academic privacy literature was that ‘privacy’ was a family resemblance term, in accordance with the theory of meaning of Ludwig Wittgenstein. This chapter explores this idea, looking back at the intent of family resemblance theory as a critique of Gottlob Frege’s and Bertrand Russell’s conceptual analysis. The chapter concludes with Solove that ‘privacy’ is indeed arguably a family resemblance term. However, the historical review shows that the theory cannot explain disagreement over meaning – instead it explains agreement. Hence, while Solove’s analysis was correct, it cannot do the explanatory work he required of it to explain the disarray in the literature.
A third explanation of the disarray in the academic privacy literature is needed. As a preliminary, Daniel Solove’s comprehensive critique of privacy definitions in his book ‘Understanding Privacy’ is surveyed. At least fourteen separate criteria for rejecting privacy definitions are uncovered, a large number which is symptomatic of the range of tasks privacy definitions, in law, ethics, cybersecurity and elsewhere, have been drafted to perform. Definitions have become overburdened with superfluities designed to win arguments and influence debates on the value of privacy, which have prevented consensus emerging. The chapter terms these theory-laden definitions ‘kitchen sink definitions’, and argues that much of the apparent disarray in the academic literature is down to incompatible kitchen sink definitions that go beyond the mere usage of the term ‘privacy’, misrepresenting disputes over law, value or propriety as conceptual.
There are many common differences of legal, academic and technological opinion about privacy, and the goal of this book is to address some particularly important oppositions. First, privacy as control over one’s relationships versus a state of withdrawal. Second, privacy as active and empowering for the individual versus being passive and let alone. Third, privacy as relevant only to individuals versus its also applying to (some) groups. Fourth, privacy as primarily relating to information flow versus its being more widely implicated in a range of social relationships. Fifth, privacy as a normative good versus a variably valued, context-dependent state with unpredictable costs and benefits. Sixth, privacy as constructed out of better-formed concepts (reductionism) versus a first-order phenomenon. Seventh, privacy as a human right versus an often strong preference. Eighth, privacy as a utilitarian notion whose value relates to the goods it provides versus something people want, at least sometimes, intrinsically and for its own sake. Ninth, privacy as incoherent versus clear and meaningful. Tenth, privacy as a technical construct of expert discourse versus a term in common use among competent native speakers of English requiring no expertise to understand it. In each case, the book argues for the second of the two options. The last two oppositions are its main focus.
The first level of privacy discourse is the conceptual. The chapter sidesteps philosophical debates about the nature of concepts, to develop instead an open-ended, fine-grained theory of privacy conceptions, ideas of privacy subtended by particular practices, behaviours, institutions and technologies in such a way as to bear a family resemblance to other conceptions of privacy. Following Oppenheim and Rawls, conceptions are meant to be operational, clearly connected to related terms, useful, open to empirical investigation and congruent with ordinary language use. Privacy is argued to be a state, not a process. Examples are given of new technologies subtending new states of privacy and their breach. Hypothetical and fictional cases are discussed, as well as boundary cases. Connected systems of privacy conceptions are also discussed, although it is argued that these are not typically well structured.