Exchange of personal information online is usually conceptualized according to an economic model that treats personal information as data owned by the persons these data are ‘about.’ This leads to a distinct set of concerns having to do with data ownership, data mining, profits, and exploitation, which do not closely correspond to the concerns about privacy that people actually have. A post-phenomenological perspective, oriented by feminist ethics of care, urges us to figure out how privacy concerns arrive in fundamentally human contexts and to speak to that, rather than trying to convince people to care about privacy as it is juridically conceived and articulated. By considering exchanges of personal information in a human-to-human online informational economy — being friends on social networking sites — we can identify an alternate set of concerns: consent, respect, lurking, and creepiness. I argue that these concerns will provide a better guide to both users and companies about prudence and ethics in information economies than the existing discourse around ‘privacy.’Contents
1. Introduction
2. Methodology
3. Privacy and the property-based understanding of personal information
4. Personal Information reconceived in the context of relationships
5. The online-emergent virtue of interactivity
6. Being a lurker
7. Being a creeper
8. Broader applications
The exchange of information online, particularly through social networking sites (SNS), takes place simultaneously in two frames or contexts. We exchange personal information (1) as part of the social process of developing intimacy and understanding in personal relationships. Those data are also understood, particularly when approached as ‘personally identifiable information’ (PII), (2) as possessions, through the circulation of which SNS generate revenue and from the sale of which SNS can extract monetary value.
We have hitherto privileged the interpretation of exchange of data as possessions, reading back this monetary frame of reference upon personal exchanges and holding it as self-evident that persons should learn to think like corporations in order to appropriately conduct information exchanges online. Through this privileging of monetary economies, privacy becomes the end-user’s responsibility in a juridical model, and it appears as common sense that the user can be asked to agree to whatever terms of service (TOS) a business cares to ask for and that a user is fully accountable for whatever treatment follows from that contractual arrangement.
I seek to privilege instead the economy of information exchange native to the conduct of personal relationships. That economy of intimacy is the one to which people are more accustomed and which appears as commonsense in human rather than in juridical terms. It seems to me inevitable that people will think of their relationships with one another, and with brands as well, in human terms of care and support, while the property-rights-based contractual legal overlay is not at all inevitable, with many or most users having no interest, experience, or expertise in reading and thinking through TOS. Furthermore, brands actively engage in creating identities that users feel they have personal connections to and actively discourage users from dwelling on the monetary and property-based elements of information exchange online, generating prima facie responsibility for brands to meet basic norms of interpersonal decency.
If it is correct that users inevitably conceive of online information economies in terms of personal relationships, and that businesses seek to reinforce this interpretation, then it is dishonest and harmful to generate best practices by privileging instead the monetary-juridical interpretation of information exchange. In other words, if people think about information economies in terms of care, support, and intimacy, and businesses want people to think about information economies in this way, then businesses should take responsibility for thinking about their end of the relationship in terms of care, support, and intimacy as well.
By privileging a personal interpretation of information economies over a juridical one we can (a) avoid harms generated by users’ expectations of businesses that they will act in ways that fit with the personal relationships they project rather than according to the minimally decent standards set by mere legal compliance; (b) generate best practices for businesses that support businesses’ own goals of creating (monetarily) valuable brand identities and relationships; and (c) better conceive of concerns about ‘privacy’ which, under the juridical model of information exchange, doesn’t speak well to users’ experiences or concerns. Put briefly, users are frequently told they should care about protecting PII from businesses who seek to profit from them, even though users themselves are most often not terribly concerned about denying profits from companies that they regularly interact with, often like, and are happy to support and partner with. Users, instead, are more often concerned by businesses that hide data practices and that don’t act as honest, friendly partners. These ethical concerns — businesses that lurk and are creepy — are best understood in terms of economies of information exchange within personal rather than monetary relationships.
In this way, by emphasizing personal interpretations of online information exchange, I hope to provide a better basis for public debate about user and provider responsibilities in information use, and to provide a better basis for businesses to determine best practices in data use, to better benefit both users and brands.
This paper presents a portion of a large, long-term research program — a post-phenomenological analysis, oriented by feminist ethics of care, of everyday life online. In previous work in this research program, I have written on the active practices of friendship online, including the posting of “meaningless” quotidian activities, like foodstagramming one’s lunch (Wittkower, 2012a); the role of cute online media in producing an “affective supplement” to relationships of care which are experienced as affectively attenuated in the absence of face-to-face and body-to-body interactions (Wittkower, 2012b); and how social media can aid employees in avoiding bystander apathy in a critical, caring relationship to their employers and organizations (Wittkower, 2015a).
Post-phenomenological analysis is grounded in traditional phenomenology (Heidegger, 1962; Husserl, 1998; Merleau-Ponty, 2002; Sartre, 1992), but refocused on small-scale structures of experience (Ihde, 1990). Post-phenomenology seeks to objectively describe the structures which create subjective experience within particular practices of technologically-mediated living: it is a form of existentialism that leaves behind general questions of what it means in the widest sense to be mortal and embodied, asking instead about the ‘what it’s like’ of particular, concrete, and everyday ways of being. Post-phenomenology is thus the philosophical articulation of what it’s like to exist in our particular material circumstances and, in my own post-phenomenological work, I seek to describe how digital environments in particular structure our experience of others and of ourselves.
Feminist ethics of care (e.g., Gilligan, 1982; Held, 2006; Noddings, 1984) emphasizes an area of morality and the good life that is not adequately addressed by utilitarian, deontological or virtue ethics: the ethics of caring relationships between persons. Other more frequently taught ethical theories have difficulty making sense of how the favoritism shown by a parent for her child, or by friends for one another, is morally acceptable at all, since such preferential consideration violates strong moral norms of objectivity and justice. Ethics of care starts from the contrary, but strongly intuitive, perspective that it is, in at least some important sense, even immoral to start from a perspective of objectivity within those elements of a good human life that have to do with love, care, and personal relationships. Ethics of care does not seek to replace other ethical theories, but seeks instead to show that there is more to an ethical life than they easily recognize — a no-less-valid and no-less-fundamental ethics that is particular rather than universal, personal rather than procedural, and based in care rather than in justice.
By engaging in post-phenomenological analysis of our lives online, oriented by feminist ethics of care, I hope to describe the ‘what-it’s-like’ of everyday processes in the conduct of personal relationships through a focus on the moral particularity of activities of care. My belief is that, just as a care-based perspective is able to transform some areas of normative ethics by offering alternate and far more common-sense accounts of some elements of our moral lives, so too will a post-phenomenology of technology oriented by ethics of care be able to transform our understanding of conceptual and practical issues in philosophy of technology and in Internet studies.
Here, I seek to use a post-phenomenological analysis based in ethics of care to reimagine and transform the discourse around privacy.
3. Privacy and the property-based understanding of personal information
The conventional wisdom about privacy is that ‘people today’ — millennials, or so-called ‘digital natives’ — don’t care about privacy and we need to teach these naïve (yet somehow otherwise presumptively ‘fluent’) users of digital media to worry about protecting their data and keeping their information away from others. As Grimmelmann (2010a; 2010b) has argued, the conventional wisdom is wrong on both counts: it is very clear that digital media users care very much about privacy, but “[t]he privacy they care about is social and relational” [1]. They don’t seek privacy in the sense of isolation, especially on what are, after all, social networking sites, but instead want to be able to maintain control over how they present themselves to others, and when and how others interact with them. ‘Protecting personal data’ is not relevant to privacy in a social online context — it provides solitude, not privacy, within social settings and interactions. Perhaps in the past, unlike today, having the right kinds of availability to the right kinds of people in the right kinds of way did correspond approximately with ‘locking down data’ and ‘not letting other people get your information.’ And perhaps this is why we originally adopted a property-rights model of personal information, where data is understood as a possession of ours that we want to keep and prevent others from taking.
From this property-based understanding of personal information, it seems contradictory that someone might worry about who can see this or that photo, but not worry about SNS data mining, which results in massive profit extraction based on targeted marketing in an online attention economy. Once we consider the relationships in which these informational exchanges take place, however, it is obvious that these cases give rise to different kinds of concerns. Contrary to the view from a property-rights-based perspective, it is not at all incoherent or contradictory to say — as students have told me in classroom discussions — both that “I want to make sure I look good in a photo before I let someone tag me in it,” and that “I like targeted ads because they show me stuff that I might actually want to buy.”
A property-based understanding of personal information leads us to falsely think that ‘nobody cares about privacy anymore,’ and leads us to privacy education and advocacy efforts which are unlikely to succeed because they are mismatched with users’ lived experience. It is unlikely to gain traction to emphasize how personal data is worth money and should be kept from circulation when the SNS user’s lived experience is that personal information is what you tell your best friends when you need support, or is what creates intimacy and care — in other words, that personal information has (affective) value precisely through its circulation in personal and mutually beneficial relationships.
A property-based understanding of personal information has also given rise to a legal system built around privacy that is similarly fundamentally misaligned with everyday experience and the human values central to everyday social and personal life.
When Ashley Madison, the infamous adultery ‘dating’ Web site, was hacked, I was surprised to learn that the company was trying to suppress publication of its users’ identifying information by issuing Digital Millennium Copyright Act (DMCA) takedown notices (Wittkower, 2015b) — that is, by claiming that posting the material obtained through the hack infringed on their right, protected under the U.S. DMCA and the U.S. Constitution, to earn a profit through exclusive publication rights over their original expressive or creative work. Whatever the wrongness of vigilantes outing users of an adultery site might be, it is implausible to suggest that it has anything to do with reducing the market for Ashley Madison’s own for-profit publication of their users’ names, addresses, sexual fantasies, conversations, and sexts.
We see a parallel here with the use of copyright law in protection against what is commonly called ‘revenge porn’ — the practice of posting images of nudity or sexual activity, depicting a former partner, as a hostile act intended to cause harm to her. The victim of revenge porn has little recourse if the images were initially taken by her partner with her consent, even if they are reposted without consent and along with potentially personally-identifying information. This has allowed revenge porn sites to extort money from women by charging a ‘service fee’ to remove anonymously uploaded pictures of them. Only if the image is a selfie and, thus, she (rather than her hostile ex) owns the exclusive right to publish the intimate image — for the legally-stated purpose of allowing her to profit from such publication and thus “To promote the Progress of Science and the useful Arts” (U.S. Constitution, article I, §8, clause 8) — can she issue a DMCA takedown, legally requiring the removal of her image.
In these cases, the property-based understanding of personal information allows us to protect some of the right people some of the time, but only by applying it in a context highly divergent from that for which it is explicitly intended and justified and without managing to address other, similarly wrong cases. The rightness or wrongness of revenge porn should have vanishingly little to do with which partner initially pressed the shutter on the camera.
In the case of child pornography, though, the property-rights model of personal information produces even stranger and less appropriate consequences, even though these laws, unlike copyright, are at least about the right kind of thing: preventing personal harm rather than ensuring profitability. A person below the age of majority — 18 years of age in most of the U.S. — is legally viewed as unable to consent to sexual activity because she (or he) is unable to understand the meaning and consequences of this activity, and yet can be held responsible under the law for producing child pornography if she takes a picture of herself engaged in sexual activity. A minor is thus able, under U.S. law, to be charged with producing child pornography — a law explicitly intended to protect children from sexual predators — by photographing herself, alone, engaged in a sexual act by herself. Numerous teenagers (mostly girls) have been so charged (Hasinoff, 2015).
Hasinoff (2015) has produced an excellent and detailed study of the contradictions, absurdities, and sexism of debate and legislation about teen sexting, in which, in addition to discussing the above and much more, she points out that “sexting is at once a sex act and a speech act” [2]. Hasinoff details how an overemphasis on the speech-aspect of the sext has led to a failure to appreciate the relevance of consent, leading to problems like the above, or, in other cases, how teens in long-term consensual relationships who can legally engage in sexual activity with one another can be subject to child pornography charges if they (with mutual and explicit consent) photograph their legal sexual activity [3].
In rethinking teen sexting, Hasinoff rightly wants to refocus on consent; in rethinking ‘privacy,’ I want to refocus on relationships.
4. Personal information reconceived in the context of relationships
In the cases above, refocusing on consent and relationships means: (1) following Hasinoff (2015), claiming that the photograph itself is not rightly regarded simply as an object, but is better regarded as an act in at least some important cases; and (2) following Floridi (2013), claiming that revenge porn is not just an intrusion, but an abduction [4]. These follow from considering that the image does not just allow third parties to be privy to a prior private moment (personal information transfer), but that the circulation of the image also re-creates the sex act in new and ever more numerous contexts, with new and ever more numerous non-consensual voyeuristic participants. The sext — an image intended to be a telepresent sex act between consenting partners — becomes no longer consensual when later unilaterally shared with third parties. Sexual assault, not property rights violation, is the right paradigm in which to think about unilaterally forced non-consensual performance of sex acts, even when they are telepresent rather than face-to-face.
A property-based understanding of personal information directs us to ask about legitimate ownership and transfer of these data, leading us to troubling consequences like the hostile ex’s right to non-consensually publish identifiable images, or the illegality of a minor’s own image of herself, due to her inability to rightly transfer her own image to herself. A view to the ethics of care rather than an ethics of justice moves us to ask not ‘who has rights over these data?’ but ’what relationship is this part of?’; it leads us to focus on the particular, on interaction and consent, and on what is being established and enacted.
Note how differently, too, consent appears in the context of property transfer and in the context of personal relationships. In the former context, ‘informed consent’ guards against fraud, under a general model of contract law. In the context of personal relationships, though, consent is not a box to be checked, but instead an ongoing relationship of mutual interdependent determination on a consensus model. ‘Consent was given’ is only in the former, and not in the latter context, a sufficient justification for doing something which one party decides she no longer likes or wants. In the context of sexual consent, should a previously consenting partner change her mind (or fall unconscious or become injured), it is obviously wrong for her partner to insist upon having the right to continue sexual activity with or without ongoing consent, since consent had once been granted. Similarly, in the larger context of personal relationships, a level of intimacy granted once cannot and should not be assumed in the future. Children become older and need to be able to establish new boundaries. Friendships change and people go through different difficult times differently. Being a caring partner in a relationship of any kind requires us to respect negotiation and renegotiation of intimacy, rather than insisting upon all access previously granted.
A central part of what it is required for moving from a property rights understanding of personal information to a relationship model of information exchange is recognizing that maintaining ongoing interdependent connections with people — something that brands and corporations must do as surely as people — requires a form of consent that is negotiated and ongoing and is not simply a box to be checked.
I will not be able in this paper alone to establish a full theory of privacy newly reconceived from the perspective of an ethics of care. I hope, though, to give us a good start in that direction by looking at exchanges of personal information as the establishment or conduct of a relationship rather than as transfer of ownership of data. The former is, of course, a very commonsense way of understanding economies of information exchange in the digitally-unmediated interaction between persons; the novel aspect of this paper is the application of this way of making sense of economies of information exchange to the larger realm of privacy online — for example, to third-party data harvesting, or to human-company or human-brand interactions.
To provide a quick and simple example of the value and relevance of this approach with regard to some of our core concerns here, we might note that, while most people are distressingly unmotivated by public, justice-related concerns about privacy and ownership of data, many, many more express strong concerns when they judge that governments or companies are ‘being creepy,’ a concern whose origin is in the private realm of proper (and improper) care in the conduct of personal relationships. By looking at what counts as proper or improper economies of information exchange within personal relationships, we can (a) gain a fuller understanding of the ethics of data use; (b) gain new tools for engaging with the public about ethical issues in data use; and (c) gain new tools that can better orient government, business, and journalists in moral use of data.
In the following, though, the word ‘privacy’ will disappear entirely. In personal relationships, we use the word ‘privacy’ to address concerns that can be better and more directly articulated by talking about respect and consent. My hypothesis is that the same is true in other relationships of information exchange as well. Thus, one of the best paths forward to resolving privacy issues online is to stop talking about ‘privacy’ as such and to stop thinking about rights and property, but instead to think about relationships, and to talk about respect and consent.
We will begin by looking at virtuous interactivity in personal relationships (next section). This will include looking at vices of inappropriate interactivity: being a lurker and being creepy (sections 6–7). We will look at these virtues and vices within an online context, where they are newly an issue for us, and where we are newly required to take concrete action to maintain appropriate caring interactivity in our personal relationships. This will give us a view of ‘privacy violations’ in the concrete terms and phenomenological form in which they appear to us in our interactions with each other. We will be able to see them as part of an informational economy oriented around care, respect, and consent, rather than oriented around ownership and rights, which can then be applied to human-corporation and human-brand interactions in order to provide views on issues of ‘privacy online’ which are directed toward fundamentally human values and concerns rather than juridical constructions (section 8).
5. The online-emergent virtue of interactivity
I begin with a puzzling idea: How can a dimension of ethical behavior emerge along with new technology? This seems to imply that not only moral beliefs and practices, but morality itself changes along with technology. A contentious claim! But not a claim I am making.
Instead, the change is happening on the other side of the equation, so to speak: it is not that there is a change in the moral reasoning which ought to guide our activity within the sphere of human choice, but rather that an activity has newly entered into the sphere of human choice, calling for new moral reasoning. The values and human goods upon which this new virtue is based are not themselves new. Instead, the material circumstances affecting these goods have recently become subject to human choice in significantly new and significantly different ways, and so it is newly a requirement of practical reason to determine proper grounds of choice — for how could we have reasoned well or poorly in the past about something regarding which we had no ability to choose how to act? This new area of moral choice also gives rise to a new element of moral character, insofar as it is a new sort of virtuous habit to act well in these new sorts of choices.
Here, although the basic framework is a traditional Aristotelian (1984) virtue ethics, I am also advancing a theory of architecturally created latent moral ambiguities — an unusual sort of idea, calling for some explanation. The idea proceeds analogically from Larry Lessig’s (2006) idea of architecturally created latent ambiguities. Lessig argues that, in the prior “architecture” of books, some rights were given to authors and others to consumers simply because those architectures made any other option impossible [5]. If you buy a (print) book, are you allowed to lend it to a friend? If you sell a book, are you allowed to conduct market research by tracking which parts readers skip? These questions were never asked, because the architectures of the physical, printed book already definitively decided the matters: the book did not afford prevention of lending, and did not afford reader tracking. Here we have a kind of corollary to the Kantian idea that ‘ought implies can:’ inability to act with regard to a set of issues or values renders it exceedingly unlikely that we will engage in speculation regarding the morally right set of impossible and possibly inconceivable actions to take relative to those issues or values. Little ink has been spilled considering when it is inappropriate to think too loudly or at what stage in a social event it is no longer impolite to alter one’s species.
And so, prior to digital media, we had architecturally-created rights of lending, the doctrine of first sale, and rights of anonymous and private reading [6]. Now that new architectures allow publishers greater and different controls, these de facto rights have been revealed as latent ambiguities, and prior practices must either be affirmed as explicit rights, to be protected by law and by code, or we should become accustomed to no longer having them, and stop thinking of them as rights.
So too, I claim, there are architecturally created latent moral ambiguities. In off-line architectures of conduct of personal relationships, the moral value of appropriate interactivity did not need to be affirmed, as immediately apparent consequences ensured that pragmatic reasoning was sufficient to determine proper action, without recourse to moral reasoning. If you’re listening to someone, should you make eye contact, nod, and say ‘uh huh’ from time to time? When you’ve just been introduced to someone, should you take pictures of them, ask personal questions, and follow them when they leave? Making these choices never before required moral reasoning or virtuous habits, because the architectures of off-line communication ensured they would be appropriately and effectively regulated by immediate positive and negative reinforcement from the other parties involved. And so, we had a latent virtue of interactivity.
I have conducted a fuller post-phenomenological analysis of the conditions of self-presentation on SNS elsewhere (Wittkower, 2014) — here, it is most relevant to consider what I have addressed there as the promiscuous intermixing of audiences [7], the loss of unintentional architectural supports for identity management [8], and regional ambivalence [9].
In SNS, our audiences are intermixed in a way that is disorienting in relation to the expectations we develop in off-line spaces. Rather than interacting in one social context and then another, variously anchored in our constitutive communities, we are simultaneously anchored in these constitutive communities, acting in the view of, for example, our high school friends, work colleagues, church group, and ex-boyfriends. This promiscuous intermixing of audiences on SNS is bound up with and complicated by the loss of architectural supports for identity management, including simple facts like walls, time, and space. The localization of my availability for perception by others within a span of time and a certain space means that when I am in the office, I can act according to its expectations and social boundaries, and can act according to others at the club, and all who are mutually co-present at those spaces and times can be expected to have a clear and shared context for interaction that aids in determining agreed-upon norms of what kinds of information exchange is appropriate, and how quickly it should be forgotten when, for example, we have left the club and returned to the office. Even when “privacy settings” for sharing are available and effective, we newly have to, for example, “choose to wait to see our doctor only in her waiting room and not simultaneously also in our office, our college roommate’s apartment, and our mother–in–law’s living room” when sharing information online [10]. This, in turn, creates a regional ambivalence in online spaces, where it is radically unclear whether the poster intends to be in a front stage or back stage setting, in Goffman’s (1959) sense, or whether front and back regions continue to be separable in contemporary social media environments.
In these ways and others, architectures that structure, condition, contextualize, and limit information flows between persons have disappeared or become subject to choice, leaving much more interpretive and relational work to be done by persons. More context often needs to be provided by speakers online, lest they be misunderstood, and more awareness of and sensitivity to the lack of context needs to be provided by listeners online. This requires new choices about actions online of listeners as well as of speakers — even when and whether to indicate at all that they are ‘in on’ a SNS conversation that they are reading. Unless we are simply to become accustomed to ‘lurkers’ and ‘creepers,’ and stop thinking of these behaviors as inappropriate, we must affirm a virtue of appropriately responsive interactivity, and learn to act properly towards it.
To sum this up: SNSs present a newfound potential for asymmetric information exchanges in the conduct of personal relationships. In face-to-face conduct of personal relationships, information transfer between persons tends to reach equilibrium both in degree and kind, through two primary mechanisms, discussed above: (1) contexts of interaction tend to create a common set of informational expectations, and (2) constant recalibration and negotiation are possible in synchronous one-on-one or small-group interactions, guided by microtransactions of encouragement or discouragement, such as head nods or selective disattention.
On SNSs, in the absence of these contextual delimiters and cues, information asymmetries become increasingly common and in ways to which we are unaccustomed. We are called upon to develop virtuous habits of interaction allowing others to see that we are attending to them in the right way, to the right extent, and in the right circumstances.
In order to provide a fuller definition of this emergent virtue of interactivity in personal information economies, we will look at its related vices, since as with many or perhaps even most virtues, interactivity is probably best understood by contrast with its related vices: lurking, a vice consisting in an inappropriate deficit of interaction, and creeping, a vice consisting in an inappropriate excess of interaction.
A lurker remains party to an interaction without signaling her presence within it. For the lurked-upon, this lack of appropriate signals of informational reception may produce an experience of coveillance (Mann, et al., 2003), in which the lurker appears similar to an eavesdropper, an undisclosed listener-in. The lurker, on the other hand, may believe she is a normal party to the interaction if she is not aware of the technical circumstances that newly demand interactivity as a matter of moral attention and activity. This undermines consent and respect in information exchange, since it deprives the lurked-upon of the ability to know what is known about her, and the opportunity to regulate her sharing in order to shape her relationship with others.
Often, SNS communications are not targeted, but may be sent out to an entire network. The sender may not remember who exactly is within the network, and it is particularly likely that users may forget about less active members of their network, since they appear infrequently to the user. This tendency may not be apparent to those less-SNS-active users, who may instead reasonably proceed with the assumption that they are within the group of intended and known recipients. The result is that users whose SNS consumption habits significantly outstrip their SNS production habits, and who do not recalibrate for this imbalance through microtransactions of liking, commenting, or sharing, are likely to become lurkers without any awareness that they will be perceived as such by others.
The lurker may come to feel that the partners are in community in an intimate ‘backstage’ setting (Goffman, 1959) while the lurked-upon is surprised, upon the breaking of the lurker’s silence, to discover that she’s ‘been in the room this whole time.’ The lurker attends to others without appropriately announcing her attendance: this diminishes the autonomy of others in managing their personal data flows relative to the lurker.
By failing to create a reciprocal co-presence, the lurker creates an informational disequilibrium, which results in very different affective orientations by constitutive parties of the relationship. Over time, the lurker may grow to feel quite close to the lurked-upon through experiences of ambient awareness which, in non-digital environments, are attended by significant mutual engagement, such as the day-to-day chats about ephemeral minutiae which are of so little importance that they take place mostly among close friends. In a non-SNS social context, work colleagues and distant relatives know about important personal and professional changes; only close friends know about your views on the merits of different sorts of pizza toppings, or your personal aesthetic disfavor for the word ‘moist,’ not because these latter attributes are sensitive or private, but instead because they are unremarkable, uninteresting, and unimportant. These trivialities, rising not even to the prominence of marginalia in the story of our lives, create intimacy even though they call for neither privacy nor secrecy. The concern with lurking (and with creeping, as we will see) is thus not a matter of control or access, but of equilibrium and exchange: the proper way of reading our concern with contextual integrity (Nissenbaum, 2004), in this case at least, is not as a public policy concern for maintaining responsive controls over privacy, but instead as a personal concern for maintaining responsive controls over intimacy.
A creeper signals presence within too many interactions, interactions of the wrong sort, or in a way disproportionate to interactants’ responses. Examples might include someone who ‘likes’ all of his niece’s spring break photos — communications within a social context where his interest in being a recipient in this nominally public informational space may appear improper — someone who regularly likes and shares posts from someone who never reciprocates, or who persists in regularly writing comments which are not responded to or, generally speaking, is overly attached.
Intensive or high-volume bilateral information transfer creates intimacy, even when the information is fully public — for example, two people simply looking into one another’s eyes steadily and at length. Such an information transfer is creepy when done unilaterally and without reciprocation, even when the information is not sensitive — again, public staring provides a good, creepy example. Creepy behavior undermines respect and consent by treating all technically available information as appropriately transferable. It fails to note that the unilateral transfer of information may create an experience of intimacy, which the creeped-upon would prefer to be mutual and consensual rather than unilateral — if ‘unilateral intimacy,’ instead of being a sufficient condition for creepiness, is not simply a contradiction in terms.
Figure 1: Instance of Overly Attached Girlfriend, original author Laina Morris [11].
This inappropriate excess of signals of informational reception may produce an experience of harassment for the creeped-upon, in many ways analogous to stalking, but, as with lurking, this may occur on SNS without the awareness of the creeper, who may believe she is a normal interactant.
In our new communicative technical ecosystem we may not yet be well-attuned to monitoring who may be sick of hearing us chime in on their personal business, even when the side-eyed dis-invitation to participate in future personal commentary would be abundantly palpable were we to conduct ourselves this way in a face-to-face context. And, as with lurking, there are reasonable prima facie reasons for the creeper to begin from and maintain the presumption that she has been invited to be a proper and welcome party to these interactions as an accepted member within the creeped-upon’s network of communicants.
By failing to appropriately scale interactions to interactant reciprocation, the creeper creates an informational disequilibrium. The creeper attends to others in inappropriate ways, becoming too present too often, and too unreservedly. This is a failure to use norms as an appropriate secondary regulatory system to code — that is, the primary architectural settings of the SNS wherein a user is included or excluded from another user’s network and allowed or disallowed technical access to a given communication through its privacy settings — failing to note that a communication’s accessibility to a communicant is not necessarily sufficient reason to conclude that the communicant is its intended or appropriate audience.
In SNS communications, we regularly speak simultaneously to multiple audiences and thus depend on recipients to self-censor their presence within interactions — a new version of the civil inattention [12] which we depend on to maintain appropriate and comfortable personal boundaries in off-line spaces where architectures are insufficient on their own. When off-line architectures are insufficient to provide appropriate interpersonal boundaries, we supplement those architectures with moral norms by, for example, not taking part in a conversation between two people standing in front of us on the subway, despite being all but required to be a party to that conversation by the amoral facts of proximity and the physics of sound. So too, we must recognize that not all conversations to which we are party on SNS, through the not-entirely customizable architectures of sharing and posting, are conversations in which we are in fact welcome. Care and respect for others demand that we not treat contingent non-denial of access as equivalent to consent, but that we are instead mindful of the limits architectures may place on choice, treading lightly or seeking out explicit consent when it is unclear whether we have been invited, whether we have failed by oversight to be dis-invited, or whether our dis-invitation was simply impractical or impracticable.
With these discussions of lurking and creeping in hand, virtuous interactivity in personal economies of information exchange can be given a brief definition. To be virtuously interactive is to attend to others in ways that appropriately signals your appropriate attention — that is, with appropriate frequency and with appropriate response to others’ recognition of your attention, respecting them as an equal partner in constructing the relationship, and limiting information appropriately by consent. In practical terms, much of the cultivation of this virtue has to do with these qualifications of actions as ‘appropriate,’ and while this may appear to be a weakness in this definition, it is in Aristotle’s view (although not precisely in his phrasing) a feature rather than a bug. What is appropriate for each of us and in each situation and in each relationship will and should change. Aristotle points out that an appropriate amount of food for Milo (a wrestler) would be vastly excessive for most other people [13]. So too some of us may thrive by living out in the open, freely sharing personal thoughts and details, while others may be happier and healthier keeping more to themselves and to a few intimate others — and in either case, the appropriate level of interactivity for a user’s best friend is probably vastly excessive for a user’s new work contact.
We have a great deal of experience with propriety in the conduct of personal relationships — although it may well be the case that some of us are far better at determining propriety than others — and so the main challenge for human interactants is becoming accustomed to having to proactively engage in interactive behaviors within new technical environments which do not architecturally prompt such behaviors.
Things are far more uncertain regarding propriety in the conduct of relationships between persons and non-persons, and it is here that this analysis might bear the most fruit. In the majority of brand and corporate interactions, public relations and social media professionals are able to draw on intuitions drawn from personal relationships in order to be appropriately responsive to human interactants. However, this is complicated by the singularity of the human and multiplicity of the non-human party, leading to a significant information gap regarding the history of interaction within the relationship — so, for example, a corporate social media representative has access only to the immediate present interaction, while the brand she represents may have been actively or passively present in a long history of interactions on the part of the human involved. In the case of a suicidal posting on the Mountain Dew Facebook page (Figure 2), although Abraham later described his post as “a joke” (Greene, 2011), it may be unexpected but not implausible that someone would have a history of interaction with the brand of a nature which would in fact cause him to turn to it while in the thrall of the helplessness and uncertainty of suicidal ideation. It is worthy of further note that, among all the persons present in this interaction, it is the Mountain Dew representative — the only person interacting as a non-person — who acts in the most appropriate and humane manner.
Figure 2: Suicidal note posted to Mountain Dew’s Facebook page [14].
In other kinds of cases, however, it may be less clear that a non-person is able to act in ways that fit more-or-less within the ethics of care in the conduct of personal relationships, because of some combination of factors, including (1) that the interaction might occur within a primarily economic or technical impersonal context, to which care seems ill-suited; (2) that the interaction might include new kinds of action only recently made possible, as in the discussion of latent ambiguities and virtues above, which are of a newly personal nature and which we are therefore unaccustomed to thinking of in the context of moral choice; and (3) that the history of interaction in this context, while newly personal, may have been historically ‘just business’ or historically non-interactive, leading to a newly incorrect preconception of that sphere of activity as subject only to universal and public concerns of justice rather than interactive and personal concerns of care.
A systematic exploration of such applications is well beyond the scope of this preliminary articulation of the virtue of interactivity, and I hope that a small sampling of some such cases will suffice to indicate this direction of future research and the utility of the preceding model to such investigation.
8.1. The right to be forgotten
Google seeks to deliver the results users find most valuable, not the results that would result in an insensitive and decontextualized keyword search. By taking into account a user’s location, search history, and other attributes when available, Google has sought to customize search results to deliver not the most objectively relevant results (whatever that might mean), but the most subjectively relevant results. In so doing, Google has very clearly invested in a model of responsive interaction and it is unsurprising that users may expect attributes of care within a service which takes on attributes of a research assistant along with those of a card catalog. Since, however, it does so by using data in ways of which the user is unaware and cannot modify, like a lurker, it denies the user the ability to be aware of and regulate their relationship, as a respectful, consensual relationship requires.
Furthermore, in contrast to the architectural ‘forgetting’ that takes place in ephemeral non-digital media, the presentism of Google search produces a tone-deaf insensitivity to the natural process of personal change. Like an old friend who seems to only know the person we regret to admit that we used to be, and who tags us in drunken photos from decades ago in full view of our family and office mates, Google can’t take a hint.
Considered together, and considered along with the efficacy of Google’s Web crawlers and the general inability of persons appearing as content on others’ Web sites to delimit access through a robots.txt, Google is creepy, and, insofar as we use it to google others, threatens to make lurkers of us all. Google stares at us in a disrespectful, non-consensual way, unresponsive to messages we may send about what kinds of looking are welcome or unwelcome.
Now, of course, in this case and in others, there may be competing interests from different ongoing relationships. For example, we may want Google to recognize and be responsive to who we are, but our concern with ‘the right to be forgotten’ should be limited by others’ legitimate interests in information access. Google’s relationship of trust and partnership in information access with other users means that when they search for our criminal record, Google will tell them about it, even though maybe it shouldn’t necessarily be the first thing Google says about us as soon as our name comes up. Relationships are tough, and you have to make difficult choices sometimes.
8.2. Responsivity in media ethics
Previous journalistic architectures required reporting about people by producing non-responsive artifacts. This strongly contrasts with the conversational and interpersonal process of the journalistic interview and, even when reporting does not follow from an interview, is often uncomfortably ill matched with the personal nature of the information being reported. The non-responsivity of journalistic artifacts is increasingly troubling as more news takes place or finds a place online, becoming subject to the presentism already addressed in the Google case above.
We should explore new ways of thinking about journalism as reporting with rather than reporting about, giving subjects of reportage ways of responding to, participating in, and intervening in that reporting, and allowing that reporting to be responsive to changes over time — even if this responsivity is limited to something like providing subjects the option to add a link of their choice at the top of an old story as an ‘update.’
Consent to being subject to reportage — being interviewed, for example — meant one thing during an era of print, when the interviewee’s relationship with the publication (and hence the publisher) was sharply limited in time, with the newspaper or television report appearing ‘current’ only for a matter of days, then disappearing from public dialogue and usually public knowledge, existing thereafter only in shadowy, archival form. New media reports tend to appear ‘current’ far longer, popping up as soon as an interviewee’s name is entered into a search engine, especially if she is not a public figure who is often reported on and who has her own Web presence. Like the circulation of sexts in revenge porn, the journalistic report keeps bringing a past, contextual self into the present again and again, without that context, and without consent for each re-contextualized re-enactment of that prior performance. The kind of consent required by a respectful and bilateral long-term relationship requires different considerations and negotiations than what was required for the mostly short-term relationships interviewees and subjects had with articles, journalists, and publications in an age of print-only media.
8.3. Appropriate transparency in end-user licensing agreements (EULAs)
Even with traditional commodities, we enter into ongoing relationships in which an ethics of care can be expected to emerge in the consumer experience. A particular vacuum cleaner or bag of oranges will have its particular attributes and stand or fall on its own merits, but our relationship with a brand continues beyond each product, and brands seek exactly the sort of consumer trust which is inclusive of both (a) the public justice-based concerns of truth in advertising, fair pricing, and product safety, and (b) the care-based concerns of brand-identity, trust, and positive affective associations.
Those brands which request or require personal information, or which seek control over our (virtual) actions, enter into a private realm in which we should expect users to adopt increasingly care-based kinds of concerns. Certainly, for example, a company which stores or transmits our photographs (e.g., Facebook) is found in our day-to-day conduct of identity and community in significant, homey, and intimate ways, and building a relationship of trust between the brand and its human hosts seems central to the brand’s survival and success.
Even a mere moment of reflection on what a care-based approach to the relationship between brands and persons might look like should make clear that posting a privacy policy and making users click through a license is not the kind of presence or transparency required for a respectful, consensual, bilateral, and mutually-determining relationship. Virtuous interactivity in an ongoing personal customer-provider relationship of this kind must involve regular check-ins, which treat the user — and explicitly and believably promise to treat the user — as a partner in a mutually beneficial consensual relationship.
8.4. Personalized services
Targeted advertising seeks to strike a difficult balance between being ineffective and creepy. Viewing the brand-person relationship through a perspective of care can help to approach users and consumers in an effective and less troubling way by offering users/consumers more responsive and transparent opportunities to invite and dis-invite brands into their lives.
We can find a couple of limit cases to see what’s at stake here. First, there is the famous, although possibly apocryphal [15], case of Target’s surprisingly effective targeted advertising toward expectant mothers. As reported in the New York Times:
About a year after Pole created his pregnancy-prediction model, a man walked into a Target outside Minneapolis and demanded to see the manager. He was clutching coupons that had been sent to his daughter, and he was angry, according to an employee who participated in the conversation.
“My daughter got this in the mail!” he said. “She’s still in high school, and you’re sending her coupons for baby clothes and cribs? Are you trying to encourage her to get pregnant?”
The manager didn’t have any idea what the man was talking about. He looked at the mailer. Sure enough, it was addressed to the man’s daughter and contained advertisements for maternity clothing, nursery furniture and pictures of smiling infants. The manager apologized and then called a few days later to apologize again.
On the phone, though, the father was somewhat abashed. “I had a talk with my daughter,” he said. “It turns out there’s been some activities in my house I haven’t been completely aware of. She’s due in August. I owe you an apology.” [16]
This is clearly an excess of interaction, and is indeed probably a paradigm case of corporations being creepy. The store was certainly not invited into the intimate informational space where this young woman did not (yet) invite her own father, and this is certainly not the sort of news that he wanted (or could have reasonably expected) to hear first from a multinational corporation.
There is also clearly an aspect of the lurker in all this. Brands tend not to make clear how much they are listening in on our interactions — building up, mining, exploiting, and selling customer profiles through loyalty cards, purchasing habits, trackers gathering Internet browsing habits after we leave their Web pages — and so when a tellingly specific banner ad or, in this case, mass-produced mailer, signals that the corporation has been there all along, listening in, it is unsurprising that we feel surveilled and eavesdropped-upon.
But the issue here is not that we simply want businesses to not get up in our business at all. It’s that we want them to do so in a responsive, respectful way. With targeted advertising in particular, this takes the form of personal recognition.
Consider a different limit case. What if Amazon simply suggested that you should listen to Beyoncé and the Beatles, and read Tom Clancy and Stephen King? Or what if Netflix kept insisting that you should watch the newest Transformers movie, Tom Cruise vehicle, or ‘holiday’ (read: Christmas) special? We want these services to respond to us not simply by indicating what a lot of people like, but instead by responding to the particularity of the behavior we have exhibited. Telling us what’s popular might be interesting or useful, and by all means do listen to Beyoncé and the Beatles, but recognition of us personally is required to tell us what people like us also enjoy, and that provides a responsive aesthetic partnership that offers distinctive benefits to both users and companies.
When these services don’t use our data in a way that indicates responsive recognition of who we are, they simply fail to serve their function — and when they mostly do well, but occasionally misrecognize how we identify ourselves through our taste-cultures, we in fact can find this insulting. I’ve heard friends distraught by the selections showing up on their Pandora radio stations, showing that, despite these occasional off-base selections, the service has succeeded in gaining their trust in its judgments based on their perception that it recognizes them. When sufficient trust in the insights of the service have been gained, especially when users have invested significant effort and self-disclosure in training these algorithms through the detailed and ongoing work of providing ratings and up- and down-votes of content, users may in fact find the service’s judgment to have real bearing upon their character. “What have I done,” they ask themselves with some amount of self-recrimination, “to make Pandora think that I would like Nickelback?”
It is this sort of personal recognition which personalized services, including targeted advertising, should strive for — the kind in which the customer feels appropriately recognized by the service provider. This means being recognized in her specificity to whatever degree she is comfortable with, in a way that is clear to her, and which is responsive to her control. If the user is able to access and correct profiles about her, and if it is transparent how this information is being used, and how (and that) it benefits the consumer for these profiles to be both detailed and accurate, it seems like customers will often gladly provide additional information and, in fact, as in cases like Pandora and Netflix, invest considerable time in providing personal preferences in order to get more responsive recommendations from the predictive system.
This final case in particular may be a useful example of why an ethics of care has already been adopted by consumers in response to non-person interactants and why it should also be adopted by those non-persons in relating to consumers. Business models which depend upon users providing personal information and which seek to involve themselves in intimate situations and environments cannot sustainably continue to treat their human partners as mere economic actors subject to contract law in purely impersonal and commercial interactions. Non-person actors in these informational economies are asking for persons to engage in informational exchanges characteristic of personal relationships of care, and must demonstrate care in return if these business models are to be effective. ‘Don’t be evil’ sets too low a bar — as does avoiding force and fraud, or acting according to rights and contract.
In an age of corporate interactions, corporations must be interactive, and must be virtuously interactive: they should attend to user-customers in a way that appropriately signals their appropriate attention with appropriate frequency and with appropriate responsivity to user-consumer recognition of their attention, building consensual relationships that show respect for user-consumers. They must, in short, demonstrate moral recognition and care.
About the author
D.E. Wittkower is Associate Professor in the Department of Philosophy and Religious Studies at Old Dominion University.
E-mail: dwittkow [at] odu [dot] edu
Notes
1. Grimmelmann, 2010a, p. 800.
2. Hasinoff, 2015, p. 4.
3. Hasinoff, 2015, p. 141.
4. Floridi, 2013, pp. 49–51.
5. Lessig, 2006, pp. 189–191.
6. Lessig, 2006, pp. 190–196.
7. Wittkower, 2014, §4.6.
8. Wittkower, 2014, §4.7.
9. Wittkower, 2014, §4.8.
10. Wittkower, 2014, n.p.
11. Creator of this instance unknown; available at http://www.quickmeme.com/img/74/7486d782a9e655c51c1508500bac226907bf735b106835d10f97f7823fdb9c7e.jpg.
12. Goffman, 2010, pp. 4–5fn.
13. Aristotle, 1984, p. 1,747 [1106a35–1106b6].
14. Available at http://blogs.villagevoice.com/runninscared/2011/10/man_who_wrote_a.php.
15. Siegel, 2013, pp. 53–57.
16. Duhigg, 2012, n.p.
References
Aristotle, 1984. “Nicomachean ethics,” In: Complete works of Aristotle: The revised Oxford translation. Edited by J. Barnes. Princeton, N.J.: Princeton University Press, pp. 1,729–1,867.
C. Duhigg, 2012. “How companies learn your secrets,” New York Times Magazine (16 February), at http://www.nytimes.com/2012/02/19/magazine/shopping-habits.html, accessed 3 September 2016.
L. Floridi, 2013. The ethics of information. New York: Oxford University Press.
C. Gilligan, 1982. In a different voice: Psychological theory and women’s development. Cambridge, Mass.: Harvard University Press.
E. Goffman, 2010. Relations in public: Microstudies of the public order. New Brunswick, N.J.: Transaction Publishers.
E. Goffman, 1959. The presentation of self in everyday life. Garden City, N.Y.: Doubleday.
N. Greene, 2011. “Man who wrote about suicide on Mountain Dew’s Facebook page says ‘It was a joke’,” Runnin’ scared: A Voice news blog, Village Voice Blogs (29 October), at http://blogs.villagevoice.com/runninscared/2011/10/man_who_wrote_a.php, accessed 3 September 2016.
J. Grimmelmann, 2010a. “Privacy as product safety,” Widener Law Journal, volume 19, pp. 793–827.
J. Grimmelmann, 2010b. “The privacy virus,” In: D.E. Wittkower (editor). Facebook and philosophy: What’s on your mind?. Chicago: Open Court, pp. 3–12.
A. Hasinoff, 2015. Sexting panic: Rethinking criminalization, privacy, and consent. Urbana: University of Illinois Press.
M. Heidegger, 1962. Being and time. Translated by J. Macquarrie and E. Robinson. Oxford: Blackwell.
V. Held, 2006. Ethics of care: Personal, political, and global. Oxford: Oxford University Press.
E. Husserl, 1998. Ideas pertaining to a pure phenomenology and to a phenomenological philosophy: First book : General introduction to a pure phenomenology. Translated by F. Kersten. Dordrecht: Kluwer.
D. Ihde, 1990. Technology and the lifeworld: From garden to Earth. Bloomington: Indiana University Press.
L. Lessig, 2006. Code: Version 2.0. New York: Basic Books.
S. Mann, J. Nolan, and B. Wellman, 2003. “Sousveillance: Inventing and using wearable computing devices for data collection in surveillance environments,” Surveillance & Society, volume 1, number 3, pp. 331–355, and at http://ojs.library.queensu.ca/index.php/surveillance-and-society/article/view/3344, accessed 6 September 2016.
M. Merleau-Ponty, 2002. Phenomenology of perception. Translated by C. Smith. New York: Routledge.
H. Nissenbaum, 2004. “Privacy as contextual integrity,” Washington Law Review, volume 79, number 1, pp. 119–157.
N. Noddings, 1984. Caring, a feminine approach to ethics and moral education. Berkeley: University of California Press.
J. Sartre, 1992. Being and nothingness: A phenomenological essay on ontology. Translated and with an introduction by H. Barnes. New York: Washington Square Press.
E. Siegel, 2013. Predictive analytics: The power to predict who will click, buy, lie, or die. Hoboken, N.J.: Wiley.
D.E. Wittkower, 2015a. “Social media and the organization man,” American Philosophical Association (APA) Newsletter on Philosophy and Computers, volume 14, number 2, pp. 16–20.
D.E. Wittkower, 2015b. “Opinion: Ashley Madison hack reveals need For new approach to guard intimate data,” Passcode, cybersecurity blog of Christian Science Monitor (21 July), at http://www.csmonitor.com/World/Passcode/Passcode-Voices/2015/0721/Opinion-Ashley-Madison-hack-reveals-need-for-new-approach-to-guard-intimate-data, accessed 3 September 2016.
D.E. Wittkower, 2014. “Facebook and dramauthentic identity: A post-Goffmanian model of identity performance on SNS,” First Monday, volume 19, number 4, at http://firstmonday.org/article/view/4858/3875, accessed 3 September 2016.
doi: http://dx.doi.org/10.5210/fm.v19i4.4858, accessed 3 September 2016.D.E. Wittkower, 2012a. “‘Friend’ is a verb,” American Philosophical Association (APA) Newsletter on Philosophy and Computers, volume 12, number 1, pp. 22–26.
D.E. Wittkower, 2012b. ”On the origins of the cute as a dominant aesthetic category in digital culture,” In: T. Luke and J. Hunsinger (editors). Putting knowledge to work and letting information play. Transdisciplinary studies, volume 4. Second edition. Boston: Sense Publishers, pp. 212–221.
Editorial history
Received 5 September 2016; accepted 6 September 2016.
This paper is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.Lurkers, creepers, and virtuous interactivity: From property rights to consent and care as a conceptual basis for privacy concerns and information ethics
by D.E. Wittkower.
First Monday, Volume 21, Number 10 - 3 October 2016
https://firstmonday.org/ojs/index.php/fm/article/download/6948/5628
doi: http://dx.doi.org/10.5210/fm.v21i10.6948
