Philosophical Analysis (Journal of the Korean Society for Analytic Philosophy) Vol. 31 (2014) pp. 151-181
31 (2014) pp.151-181
8]gUhhYbXUV]`]hm 7]j]` =bUhhYbh]cb UbX
h\Y 9d]ghYac`c[m cZ Df]jUWm
,FZXPSET
152 Axel Gelfert
It is something of an oddity that the concept of privacy
which hinges
on epistemological concepts such as information, knowledge, and its
communication has, for the longest time, been discussed not by
philosophers, but by legal scholars. Partly as a result of this, there seems
to have been less emphasis on providing a rigorous analysis of the
concept of privacy, and more interest in developing a cluster of somewhat
imprecise, but ‘serviceable’ legal considerations. In recent years,
philosophers especially those with an interest in social epistemology
have begun to redress this imbalance (see, e.g., [Fallis 2013] and
[Matheson 2007], and references therein), yet philosophy still has a long
way to go if it is to catch up with the legal tradition of thinking about
privacy. This paper is intended as a contribution to this philosophical
project and is organized as follows. In Section 2 (‘Models of Privacy’),
after briefly rehearsing the origins of the privacy debate in legal
theorizing, I will quickly turn to two contrasting classes of philosophical
approaches, which I shall refer to as the liberal and informational models
of privacy, respectively. Their validity will be tested against several
scenarios (Section 2.3.), and we will see that their shortcomings motivate
a third type of account (the immunity model), which posits that privacy
seeks to protect us from the judgments of others (Section 3). In Section
4, I will introduce an alternative account, which takes as its starting point
Erving Goffman’s twin notions of disattendability and civil inattention. In
a nutshell, while the principle of disattendability demands that, in public
settings, I am not to be obtrusive, civil inattention requires of us to
suspend specific attention to others and their behaviours in various
circumstances. As I shall argue, a model of privacy based on these twin
Disattendability, Civil Inattention, and the Epistemology of Privacy 153
concepts is explanatorily superior to its competitors and holds out the
promise of being able to account for both cultural diversity and the
changing face of privacy in an age of electronic surveillance. I will
conclude (Section 5) by commenting on some remaining issues, notably
the worry that the proposed disattendability/civil inattention model of
privacy might seem to unduly privilege social conservatism, and by
reflecting on the prospects of privacy as context-dependent and
domain-specific phenomenon.
As mentioned in the Introduction, the legal tradition of thinking about
privacy predates its philosophical analysis. (See, for example, [Solove
2008] and [Wacks 2010], and references therein; for an early
philosophical treatment of the topic, see [Thomson 1975], who argues
that violations of privacy are reducible to violations of other, more
fundamental rights.) More specifically, the legal tradition can be traced
back to an influential essay by Samuel Warren and Louis Brandeis,
published in the Harvard Law Review in 1890, under the title ‘The Right
of Privacy’. Warren and Brandeis’s essay attempts to give substance to
privacy as the legal ‘right to be let alone’ and was motivated by the
intrusions of the yellow press into the lives of (often famous) individuals.
More specifically, Warren and Brandeis reacted to the then novel
technology of photography, arguing that ‘the law must afford some
remedy for the unauthorized circulation of portraits of private persons’
(Warren and Brandeis 1890: 195).
In spite of its considerable influence in the legal domain, Warren and
Brandeis’s essay makes for a somewhat disappointing read, at least for
154 Axel Gelfert
the philosophically inclined reader, since its authors are not so much
concerned with the concept of privacy, but instead aim to derive a
generalized right to informational privacy from elements they take to be
implicit in the common law tradition that had always afforded persons
and property some degree of legal protection. Robert Post voices a similar
sense of dissatisfaction:
The prestige and enormous influence of the [Warren and Brandeis] piece
creates expectations of sweeping vistas and irresistible arguments. But, setting
aside the rhetorically powerful (and often quoted) passages of complaint
against the irresponsibility of the press, the article offers instead a technical
and rather dry exposition of the legal rights of unpublished authors and
artists. (Post 1991: 647)
Although ostensibly aimed at overcoming a narrow focus on property
right by establishing a broader right to privacy, Warren and Brandeis
nonetheless remain indebted to the property-based conception of rights
for example, when they argue that the ‘fiction of property’ can be
preserved by recognizing that ‘it is still true that the end accomplished
by the gossip-monger is attained by the use of that which is another’s,
the facts relating to his private life, which he has seen fit to keep private’
(Warren and Brandeis 1890: 205). It should be clear that such conceptual
contortions are largely the result of trying to shoehorn a complex and
novel question how to regulate the flow of information in an age of
sensationalist reporting into a legal tradition which, at the time, was
lacking in descriptively and normatively adequate conceptual resources.
For this reason, in the remainder of this section, I will not be looking
towards the legal tradition for guidance, but instead will draw on two sets
of considerations that emerge from political liberalism and the
Disattendability, Civil Inattention, and the Epistemology of Privacy 155
epistemology of information, respectively.
Broadly speaking, liberal models of privacy seek to identify a domain
of individual behaviours, beliefs, and activities, which in virtue of their
private character should be protected from regulation (or persecution) by
the state and, more generally, from unwarranted intrusion by others. The
classic expression of the general sentiment underlying liberal models of
privacy can be found in John Stuart Mill’s On Liberty, who already
recognized in a manner that must surely seem prescient to any
21st-century reader familiar with such phenomena as cyberbullying that
society can be more oppressive than state regulation:
But reflecting persons perceived that when society is itself the tyrant society
collectively over the separate individuals who compose it
its means of
tyrannising are not restricted to the acts which it may do by the hands of its
political functionaries. Society can and does execute its own mandates: and
if it issues wrong mandates instead of right, or any mandates at all in things
with which it ought not to meddle, it practices a social tyranny more
formidable than many kinds of political oppression, since, though not usually
upheld by such extreme penalties, it leaves fewer means of escape,
penetrating much more deeply into the details of life, and enslaving the soul
itself. (Mill, Complete Works 18:219)
On the one hand, liberalism calls for real protection from substantial
interference; on the other hand, it also calls for self-restraint in our
interactions with others, even in matters that stretch the notion of what
should be considered the private realm, such as conflicting fundamental
beliefs that arise from competing conceptions of the good. As Joshua
156 Axel Gelfert
Cohen notes, ‘cultural liberalism requires that, as a general matter, we
steer clear of controversial topics about which we cannot expect to reach
agreement and that do not demand a collective decision’ (Cohen 2009:
318). This, as Cohen recognizes, gives rise to the thorny question of how
to delineate a genuinely private realm from that which is fit for public
deliberation and scrutiny, given that it is hard to see ‘how we could
possibly identify the private arena with the family, or with the economy,
or with any arena of social life, identified either spatially or
institutionally prior to normative political argument’ (Cohen 2009: 312).
Since the present paper focuses on the epistemology of privacy, I shall
not delve any deeper into its normative and political dimensions,
important though these are. It is worth noting, though, that an
epistemological analysis of the concept of privacy is bound to have
repercussions for the substantive normative debate, insofar as we first
need to develop a theoretical understanding of when the possession of
certain types of information constitutes a breach of privacy in the first
place.
Unlike the liberal model of privacy, informational models are not
primarily concerned with substantial interference with individual liberties,
but instead link breaches of privacy more directly to the possession of
private information (which may or may not have been illegitimately
obtained). A good illustration of the informational model is this definition
of privacy, given by W. A. Parent:
Privacy is the condition of not having undocumented personal knowledge
about one possessed by others. A person’s privacy is diminished exactly to
Disattendability, Civil Inattention, and the Epistemology of Privacy 157
the degree that others possess this kind of knowledge about him. (Parent
1983: 269)
Parent immediately points out that this definition pertains to ‘the
condition of privacy, not the right to privacy’ (ibid.) and, indeed, it is the
assumption of epistemological primacy of the former that is, of the
conditions that need to be in place for privacy to obtain which
distinguishes the informational model from the liberal model (whose
epistemological underpinnings are often left implicit, even while its
proponents aim to establish a right to privacy). For a breach of privacy
it is not required that personal knowledge about a person be used against
him, but rather that such information is obtained by an unauthorized party
without his consent. Klemens Kappel makes a similar observation when
he writes: ‘According to the generic epistemic theory of informational
privacy, privacy depends inversely on epistemic access.’ (Kappel 2013:
185).
However, the generic theory cannot be the full story. For one, there is
some ambiguity in how we should understand the term ‘access’. If access
is equated with mere availability that is, if it is understood in
dispositional terms
the generic theory would likely overgenerate cases
of privacy being breached. For example, when a file containing a
patient’s medical history is left unattended, so that in principle anyone
could access it, we may not wish to speak of a privacy breach unless
someone actually accesses the patient’s information. (Having said that,
there may be good practical reasons for defining ‘privacy’ in such an
inclusive manner, not least because it might encourage good data
protection habits.) Furthermore, it does not sound quite right to say that
one’s condition of privacy is compromised exactly in proportion to the
158 Axel Gelfert
number of individuals who either have access to it, i.e. inversely to
epistemic access (as the generic theory asserts), or ‘exactly to the degree
that others possess [the information in question]’ (as Parent puts it). Even
on the informational model, the term ‘privacy’ should amount to more
than just a convenient label for the degree to which other individuals
have access to, or possess, certain types of information that is, privacy
should be more than a mere ‘numbers game’.
This last point is echoed by Masahiko Mizutani who notes that a purely
dichotomous division between public and private would overlook the fact
that privacy need not be compromised if information is willingly shared
with select confidants. For example, ‘with regard to certain types of
information about myself’, I may ‘convey everything to my wife, only the
summary to my friends, and nothing to strangers’; yet, ‘in reality there
may be things I am willing to tell my friends but not my wife, or secrets
I am willing to reveal to passersby during my travels precisely because
they are strangers who do not know me’ (Mizutani 2012: 610). It seems
to me that, in the event that I share a (perhaps embarrassing) personal
story with, say, a fellow train passenger whom I will never meet again
not only is it the case that, a fortiori, my right to privacy is not
violated (given that it was my free decision to tell the story), but my
condition of privacy is not compromised either, even though the number
of individuals with access to the embarrassing information has increased.
Any account of privacy should be able to explain why access sometimes
compromises privacy, and sometimes doesn’t. If in-principle access
sufficed to undermine privacy, then it would seem that we would have
to restrict ‘true privacy’ to those private mental items to which only the
individual cognizer has access. Lest we opt for such a restrictive usage,
which would be needlessly revisionist of social, linguistic and legal
Disattendability, Civil Inattention, and the Epistemology of Privacy 159
practice, we should reflect on which kinds of access undermine privacy
and which do not.
Let us consider a few cases that will illustrate the limits of the
(access-based) informational model. Jeffery Johnson (1989) offers two
instructive scenarios, which will form the basis for the discussion in this
section.
Bugged phone. ‘You place a bug on my phone in order to overhear the most
intimate details of my conversation with my mistress. We are too smart for
you; we have already agreed to never discuss intimate things on the telephone
[...]. Your tap reveals nothing; you have gained no new personal information.
You certainly have, however, violated our privacy. Such cases cannot
plausibly be dismissed as mere attempted violations of privacy.’ (Johnson
1989: 161)
As Bugged phone makes vivid, actual acquisition of personal
knowledge is not required for a breach of privacy to have occurred since,
as the case has been described, no personal information is passed on.
Perhaps whoever intercepted the phone line was hoping for more, but
intentions on the part of third parties have no place in the informational
model. To the extent that an instance of unauthorized access has
occurred, this pertains to the mode of communication wiretapping into
someone else’s phone line
not to any sensitive information itself.
Perhaps, then, what undermines the condition of privacy in the case at
hand is the fact that, had it not been for the unusual precautions taken
by the two lovers, personal information might have easily been acquired.
This, too, however, is not easily accommodated by the informational
160 Axel Gelfert
model, at least not in its ‘generic’ formulation (see previous subsection).
But even when no new personal information can be acquired for
example, because the intruder already has all the information she could
possibly hope to obtain this does not preclude the possibility of a breach
of privacy. Consider the following case:
Voyeuristic Sarah. ‘Sarah peeks through my bathroom window in order to
admire my naked body. [
] Suppose that Sarah is both my physician and
a frequent guest in my home. As a guest she knows all about the vivid rose
paint job in my bathroom; and as my physician she knows only too well
about the sorry state of my waistline. She nevertheless clearly violates my
privacy when she peers in my window.’ (Johnson 1989: 161)
In the case as described, Sarah does not and, indeed, could not hope
to gain any new information: as a frequent house guest, she knows all
about the bathroom, and as her target’s physician she has probably seen
more of him than she could reasonably expect to glimpse through the
bathroom window. No new information is being accessed, not even
potentially, yet the manner in which Sarah accesses (known) information
clearly constitutes a breach of privacy.
One attempt to amend the informational model might be by positing
control over one’s personal information as a desideratum of privacy. On
what one might call the control model of privacy, privacy would be
determined by the extent to which we can control what others come to
know about us. As an example of a control-based account, consider this
statement by Charles Fried:
Privacy is not simply an absence of information about us in the minds of
others; rather it is the control we have over information about ourselves. To
Disattendability, Civil Inattention, and the Epistemology of Privacy 161
refer for instance to the privacy of a lonely man on a desert island would be
to engage in irony. The person who enjoys privacy is able to grant or deny
access to others. (Fried 1984: 209
10)
The ability ‘to grant or deny access to others’, however, depends on
many contingent factors, and it is far from clear that we should only
speak of ‘privacy’ when such control can be reasonably assumed.
Consider this example discussed by Martijn Blaauw:
Diary. ‘John is a very private person. All the truths about himself that he
finds important are not known to any others. Yet today, before leaving for
work, he left his diary open on his desk. His cleaning lady could open the
diary and read its contents, thus acquiring knowledge of John’s personal
propositions. And being in meetings all day, John has no means to go home
and prevent this.’ (Blaauw 2013: 175)
As Diary makes clear, John, because he has failed to place his diary
under lock and key, could not possibly prevent his cleaning lady from
reading what he has written and, in this sense, is unable to grant or deny
access to it. His (temporary) inability to control who has physical access
to his diary, however, does not in and of itself undermine the condition
of privacy at least ‘so long as the cleaning lady hasn’t perused his
diary’ (Blaauw 2013: 175). It would seem, then, that the control model
of privacy ends up collapsing into the generic informational model
according to which privacy is a matter not of access or control, but of
the actual possession of information in which case, however, it would
also inherit the latter’s problems.
Alternatively, we may wish to shift the focus from the ability to restrict
access to information to the basis on which granting or denying access,
162 Axel Gelfert
though not always successfully, may be deemed legitimate. In other
words, perhaps we should inquire into what it is that (attempted)
restrictions on the flow of private data and personal knowledge aim to
achieve. One idea, to be developed in the next section, will be that
restrictions on the spread of personal knowledge serve the purpose of
shielding an individual from the judgment of others. Before turning to
this account, however, it will be useful to reflect briefly on what we
mean by ‘personal’ information given that, in common parlance, this
label suggests that we are dealing with sensitive issues which may, for
example, be important to a person’s self-image or sense of identity. As
Kappel notes, we usually ‘think of sensitive facts as including facts about
our religious denomination, health, sexual preferences and certain parts of
our lifestyle’. In addition, ‘sometimes one’s political views are regarded
as sensitive, as are facts about one’s financial affairs’, or even ‘one’s true
views about colleagues and friends’ (Kappel 2013: 183).
Even within societies that are culturally similar (say, Western Europe
and the United States), there is often considerable variation in what is
conventionally regarded as private information
revealing one’s salary
may be taboo, whereas political affiliations may be freely volunteered by
most people, and vice versa in another country and this diversity
increases as one crosses cultural boundaries. (On this point, see
[Mizutani, Dorsey, and Moor 2004].) Thus, Parent argues that a full
explication of the notion of privacy
requires that we clarify the concept of personal information. My suggestion
is that it be understood to consist of facts about a person which most
individuals in a given society at a given time do not want widely known
about themselves. (Parent 1983: 269-270)
Disattendability, Civil Inattention, and the Epistemology of Privacy 163
While the notion of ‘personal’ (or ‘sensitive’) information retains much
of its significance, recent technological trends such as the rise of ‘Big
Data’, involving the ‘mining’ of (individually insignificant) user data for
clues as to, say, an internet user’s personal (e.g. sexual) preferences
may force us to rethink what kinds of data are relevant to the issue of
privacy. What matters, of course, is not the mere existence of such
technology, but its actual deployment as a way of gathering, collating, and
aggregating
otherwise
highly
distributed
data
points
concerning
individuals and their preferences. Much of the data that is being collected
online, by social networks and advertisers, is individually mundane and
boring, and not something that we would usually regard as constitutive
of our personality, yet it may provide sufficient clues for an algorithm to
make powerful inferences about some very personal aspects of
individuals. As Mizutani puts it, ‘many of the problems relating to
privacy in our current age are problems of a more latent nature; that is,
regardless of whether privacy is actually or violated or not, what matters
is the way in which our behavior is influenced by the mere feeling
someone might be watching’ (Mizutani 2012: 610). As I will argue in
Section 4, this idea can be made more rigorous by attending to the
relational character of one person attending to another’s actions,
preferences, and characteristics.
Towards the end of the preceding section I argued that we should
inquire into what it is that attempts to restrict the flow (or accessibility)
of private data and personal knowledge aim to achieve. As we saw in the
various cases discussed above, privacy can be breached even when no
164 Axel Gelfert
personal knowledge is made accessible (Bugged phone), or when no new
personal knowledge is acquired (Voyeuristic Sarah), yet no such breach
needs to happen even when extensive personal knowledge is in plain
view (Diary). It may be preferable, therefore, to focus not on the flow of
information itself but on the uses to which it is put.
Johnson has suggested that one reason why we are especially
concerned about personal information is that it may lead to unfavourable
opinions that others form about us, due to their disapproval of certain
types of behaviour (or certain facts about ourselves) that we normally try
to hide from public view. As Johnson argues,
all examples of privacy have a single common feature. They are aspects of
a person’s life which are culturally recognized as being immune from the
judgment of others. (Johnson 1989: 157)
The basic idea is that, wherever there are cultural norms of privacy in
place, societies have implicitly agreed to treat the requisite information as
not fit for the purpose of assessing others on its basis. (Such collective
normative agreement, of course, does not imply that no one will judge
another person on the basis of personal information about him or her;
indeed, it is precisely because many people are judgmental that personal
information stands in need of protection!) Johnson’s immunity model of
privacy according to which privacy is a matter of being immune from
the judgment of others echoes Mill’s point about the oppression that
comes with the experience of social disapproval. Where Mill argued that
society ‘practices a social tyranny more formidable than many kinds of
political oppression’, Johnson pinpoints the social mechanism by which
society achieves this pervasive influence: through the judgments of others.
Johnson claims that his account fares better in explaining cases not
Disattendability, Civil Inattention, and the Epistemology of Privacy 165
covered by the liberal and informational models of privacy. Consider the
following case:
Golf on Sunday. ‘I routinely play golf on Sunday mornings. I do not take
precautions to hide this information. My neighbor in seeing me leave for the
course every weekend does not violate my privacy. If, however, she takes it
into her head to lecture me on my heathen ways and how my time on
Sundays could be better spent in more spiritual pursuits, she has now intruded
into a sphere of my life that is private. Her possession of personal
information is not the issue; her negative judgment very much is.’ (Johnson
1989: 161)
On the informational model, there is little that would mark out this
case as a breach of privacy: no attempt is made by me to hide the
information that I routinely play golf on Sundays, and my neighbour’s
gaining that information by looking out the window does not by itself
breach privacy. Yet, it seems to me that, in the scenario as described, it
is the neighbour’s lecturing me, not her negative judgment per se, which
infringes upon my privacy. In this sense, the Golf on Sunday case
appears to be covered by the liberal model, and the specific contribution
of judgments on the neighbour’s part (except as the psychological cause
of her unwarranted interference) remains unclear.
The
immunity
model
(or,
as
Johnson
prefers
to
call
it,
judgment-of-others model) also becomes implausible when applied to
certain clear cases of privacy violations. Consider the case of Bugged
phone, where the intercepted interlocutors had devised an elaborate
system of rule to make it impossible to extract information. As Johnson
recognizes,
166 Axel Gelfert
[t]here is a sense in which this counter-example to the information model
could be marshalled against the judgment-of-others model. If the tap reveals
no new information about our relationship, it is not immediately apparent
how the eavesdroppers have formed any new judgment about us. They have
certainly not, however, remained emotionally neutral. The very acts of
monitoring my phone conversations implies a great deal of (emotional)
interest in this aspect of my life. (Johnson 1989: 166)
This defence of the role of judgments on the part of the eavesdropper
strikes me as deeply implausible. For the wiretapping of the phone to
constitute a breach of privacy, no emotional investment on the part of the
interceptor needs to be posited: Imagine the phone conversations were
being recorded by a jaded surveillance specialist who ‘has seen and heard
it all’ and who is deeply bored by (or, lest boredom be counted as an
‘emotional investment’, totally indifferent to) the affair, no matter how
‘juicy’ (or upsetting to others) its details may be. Surely, this would still
count as a severe breach of privacy. Indeed, it seems to me that violations
of the condition of privacy do not require the involvement of any human
judgment. For example, the gathering of information could be outsourced
to physical instruments (such as recording devices), and whatever analysis
is performed may be carried out entirely by algorithms which, as the
whistleblower Edward Snowden has revealed, is the standard mode of
operation of the United States’ security agencies and their allies. Its
automated nature, however, does not render such interception any less of
a breach of privacy.
Two further worries about the immunity model merit attention, both of
which concern the extent to which we can reasonably demand not to be
judged by others. When discussing the Golf on Sunday scenario, Johnson
argued that, whereas my neighbour’s possession of personal information
Disattendability, Civil Inattention, and the Epistemology of Privacy 167
about my golfing habits ‘is not the issue[,] her negative judgment very
much is’ (Johnson 1989: 161). To the extent that negative judgments may
often form the basis of unwarranted interference by others as in the case
of the fundamentalist neighbour liberal models of privacy may be able
to accommodate such cases. Perhaps in order to distance his own position
from liberal models, Johnson elsewhere argues that immunity is not
limited to negative judgments only:
One should not infer, however, that the concept of privacy only addresses the
negative judgments of others. The professional model, who has no reason to
fear a negative evaluation of her body, can still have her privacy violated by
the voyeur. [...] When we claim immunity from the judgment of others, this
applies to all judgments, positive, negative or neutral. (Johnson 1989: 163)
Yet, this notion of immunity seems too expansive by far: Surely it can
neither apply to all areas of life, given that robust judgments concerning
others are required in many contexts
not least for practical projects that
depend on social cooperation for their success nor can the mere
subjective expectation of immunity from judgment be sufficient to ground
privacy. In connection with the hypothetical case of a married couple’s
public embrace being plastered all over the front page of a newspaper,
Johnson writes:
It is also, at least in this case, an area of your life in which you have a right
to expect immunity from the judgments of others, even if the data had been
correct. (Johnson 1989: 162)
Most of us, no doubt, would concur that the newspaper’s publication
would constitute not only a violation of the right to privacy, but would
168 Axel Gelfert
also undermine the legitimate expectation that a state of privacy about a
couple’s personal matters be respected. But the case only serves to raise
another question: Which are the areas of our lives where we can
legitimately appeal to immunity from the judgments of others? One
response would be to say that any answer to this question must be based
on normative agreement about which areas of individuals’ lives society
deems worthy of protection in which case we find ourselves back at
Cohen’s point that, ‘prior to normative political argument’, there is no
telling which areas of life should enjoy privacy and which should be
available for scrutiny. Beyond political and legal ways of managing
privacy, however, the underlying question ‘What is it about certain
ways of judging others that makes them (as opposed to other types of
judgment)
intrusions
into
other
people’s
privacy?’ still
awaits
elucidation.
A second, related worry concerns Johnson’s conclusion that all
judgments that others form about us ‘without our consent’, as it were, are
illegitimate. For this contradicts the core enlightenment idea of testing
one’s judgments against those of others and, by extension, having one’s
judgments contested by others. (Immanuel Kant provides us with a clear
expression of this idea when he remarks in his discussion of testimony
as a source of knowledge that ‘Man always wishes to test his judgment
on others’ and that it would be a form of ‘logical egoism’ to dismiss the
need ‘to compare one’s opinions with those of other people’; see the
discussion in Gelfert 2006.) Importantly, such a willingness to at least
consider criticism should also extend to (some) unsolicited judgments: we
would deprive ourselves of much useful criticism if we were to dismiss
all unsolicited judgments of others as an illegitimate assault on privacy.
Disattendability, Civil Inattention, and the Epistemology of Privacy 169
Delineating the areas of life in which one can legitimately appeal to
privacy in order to ward off unwarranted intrusion, unwanted attention,
and unsolicited judgments of one’s behaviour by others should be a
central part of any philosophical account of privacy and cannot be
relegated to an afterthought. Importantly, such an account should not only
list the areas of life that various societies have deemed worthy of
protection, but should also aim for some degree of unification, for
example by identifying a common feature or mechanisms that tells us
what would be objectionable about privacy intrusions in the various cases.
(The immunity model proposes one such feature the formation of
unsolicited judgments on the part of others but fails for the reasons
discussed in the previous section.)
In order to develop an alternative model of privacy, I shall help myself
to Erving Goffman’s notion of disattendability. Disattendability, on
Goffman’s understanding, correlates with a form of civil inattention that
we accord others in public places, across a range of acceptable
behaviours. Unpacking this suggestion, as we shall see, will lead to
important questions regarding the notion of acceptability as well as the
nature of publicity. For the moment, however, let us follow Raymond
Geuss in characterizing a public place as
a place in which I can expect to be observed by ‘anyone who happens to be
there’, i.e., by people whom I do not know personally and who have not
necessarily given their explicit consent to entering into close interaction with
me. (Geuss 2001: 58)
Precisely because, in public settings, we cannot rule out the possibility
170 Axel Gelfert
of our being exposed to others, we need to regulate the ways in which
attention is being deployed in our dealings with each other. Civil
inattention towards others is a way of recognizing that they have an equal
claim to making use of public space and that our being co-present in the
same space neither requires justification nor incurs special responsibilities
towards each other. According to Goffman, it is through routine markers
of civil inattention e.g., temporary (brief) eye contact between strangers
that ‘individuals exert respectful care in regard to the setting and treat
others present with civil inattention’ (Goffman 1971: 331), so that public
order can be maintained.
At the same time as we accord civil inattention to others, public space
is governed by a norm of disattendability: Since, in many public settings,
we find ourselves surrounded by others who did not specifically choose
to associate with us, their co-presence imposes on us the demand ‘not to
make undue claims on people’s fears or trigger their embarrassment and
disgust’ (Miller 1997: 1999). Elevating this demand to the level of a
behaviour-guiding maxim, we may then speak of the principle of
disattendability, which is aptly described by Geuss as follows:
The principle of disattendability states that in such contexts and places I am
to be unobtrusive. That is, I am to allow the other whom I may encounter
to disattend to me, to get on with whatever business they have without
needing to take account of me. I am not to force myself on their attention.
(Geuss 2001: 58)
The principle of disattendability can be made even more vivid by
reflecting on some of its more blatant violations. Thus, Geuss recounts
the story of Diogenes of Sinope, who
Disattendability, Civil Inattention, and the Epistemology of Privacy 171
was in the habit of masturbating in the middle of the Athenian market place.
He was not pathologically unaware of his surroundings, psychotic, or
simple-minded. Nor was he living in a society that stood fairly low on what
we take to be the scale of our cultural evolution [...]. Rather, we know that
the Athenians objected to his mode of life in general and to this form of
behavior in particular. We know this because the doxographic tradition
specifically records Diogenes’ response to a criticism of his masturbating in
public. He is said to have replied that he wished only that it were as easy
to satisfy hunger by just rubbing one’s belly. (Geuss 2001: 57)
Evidently, Diogenes’ masturbating in public being difficult to ignore
when confronted with it violated the principle of disattendability.
Clearly, Diogenes should have limited this activity to a more private
space. Perhaps, then, what motivates a distinction between the public and
private realm and, in turn, creates a need for shielding the latter from
intrusion is the fact that certain morally permissible actions and
activities (and information pertaining to them) would violate the principle
of disattendability if they were to be dragged into the public realm.
Can the principle of disattendability form the sole basis for an account
of privacy? Recall that, as Goffman puts it, the principle serves to
transform ‘normative expectations into righteously presented demands’
(Goffman 1963: 2) on others. This should raise some alarm bells for
anyone who cares about protecting the private realm from undue
moralization by others. After all, the principle of disattendability presents
itself, in the first instance, as an imposition of a duty of restraint on the
individual, who ought to behave in such a way as not to attract undue
attention to himself. Yet, what is regarded as ordinary, routine, and
normal in a given setting is heavily dependent on contingent cultural
factors and conventions. It might seem, then, that privacy needs to be
172 Axel Gelfert
‘earned’ by behaving in a conformist fashion. And indeed, critics have
pointed out that disattendability is a socially problematic notion. As
Margaret Urban Walker puts it:
Civil disattendability shades off rather too readily into norms of
‘respectability’ that load hierarchical social arrangements, sheer prejudice, and
socially sanctioned contempt for and exclusion of certain groups or their
‘ways’ from specific social locales. (Walker 2006: 124)
This is a legitimate worry, which I will address in the final section of
this paper. It would be hasty, however, to dismiss the relevance of
disattendability (and its correlate, civil inattention) to the project of
developing a robust account of the notion of privacy. For, as we have
previously seen, other proposed models of privacy also have their share
of problems, and it will be instructive to see how the disattendability/civil
inattention model fares by comparison. For one, it is important to keep
in mind that disattendability is only one element of the model, civil
inattention being its correlate: Where the disattendability principle
demands that we behave ourselves so as to be civilly disattendable, civil
inattention requires that we suspend specific attention to others and their
behaviours or characteristics. It is this second demand that we should
accord others, who are simply going about their business and living their
lives, the requisite degree of civil inattention, so as to not disrupt their
routines, or the routines of public order in general which may provide
at least a tentative starting point for an alternative model of privacy.
In order to better assess the prospects of an account of privacy based
on disattendability and civil inattention, let us turn to some of the test
cases discussed in Section 2.3. Consider Voyeuristic Sarah, the case of
the voyeuristic family friend and doctor. As we saw earlier, Sarah gains
Disattendability, Civil Inattention, and the Epistemology of Privacy 173
no information by watching her target getting undressed in his bathroom,
since she is already well-acquainted with all the relevant facts. And, as
Johnson notes, even if new information were to be gained ‘my outrage
at this violation of my personal privacy is not focused on the knowledge
Sarah gained’ (Johnson 1989: 161). Johnson concludes
rightly, in my
view that the informational model fails to adequately account for this
breach of privacy. How does Johnson’s immunity model fare? In line
with its guiding idea, according to which privacy is a matter of being
immune from the judgments of others, one would expect that, what
constitutes the breach of privacy in the present case, would be the
formation of unsolicited (and potentially negative) judgments by Sarah.
But this seems implausible: After all, which new judgments could Sarah
form, given that no new information has become available? If, in the
Voyeuristic Sarah case, the informational model fails, as Johnson argues,
then so does his immunity model, and for the same reason: where all
information has been properly taken into account, there is simply no room
for new judgments. By contrast, on the disattendability/civil inattention
model I am proposing, Sarah’s voyeurism is a breach of privacy, not
because of any new information she gains or any judgments she forms,
but in virtue of her paying undue attention to a routinized aspect of her
target’s everyday life: although her target has done nothing to attract her
attention, his reasonable expectation not to be attended to as he goes
about his business is being violated.
Similar considerations apply to the Bugged phone case: Engaging in
everyday phone conversations is a routine part of modern communication;
when I speak to my secret lover on the phone, both of us are operating
in a way that neither intrudes upon others nor, all else being equal,
should attract the attention of a third party. Indeed, enabling two and
174 Axel Gelfert
only two interlocutors to engage in conversations without intruding on,
or being intruded upon by others, is arguably part of the proper function
of telephony as a technology. (Things would be quite different if both of
us were to engage in intimate conversation by shouting from the
rooftops!) For a third party to intercept such a conversation would be to
pay undue attention to what is, in essence, a properly disattendable
activity on our part.
Finally, in the Golf on Sunday scenario, what constitutes the invasion
of privacy is not the neighbour’s acquisition of information about me per
se, nor the fact that she forms a negative judgment concerning my lack
of spirituality (which she would form in any case, even if she had chosen
not to confront me). Rather, according to the disattendability/civil
inattention model, it is the obtrusiveness of her intervention the fact that
she lectures me on matters of morality, when I have done nothing out of
the ordinary that renders it a violation of my privacy. (Note that, by
contrast, it would hardly constitute a violation of privacy if she had
volunteered, say, an unsolicited but unobtrusive comment about the
weather ‘looks like a fine day for golfing’ small talk being an example
of a culturally specific way of acknowledging one’s neighbour without
violating the norm of civil inattention.)
In other respects, too, the disattendability/civil inattention model is
explanatorily superior to its rivals the explanandum, of course, being the
various culturally specific ways in which human societies allow their
members to negotiate their social interactions with one another. First, it
accounts for a class of cases which, as Johnson himself realizes, ‘seem
to present great difficulty for the judgment-of-others [=immunity] model’.
For example,
Disattendability, Civil Inattention, and the Epistemology of Privacy 175
in spite of the well-known variation in sexual practices around the world,
anthropological data indicates that sexual matters, in the main, belong to the
private sphere. Societies have rules for concealment of the genitals, and
restrictions on the time and manner of genital exposure; only a handful of
societies practice complete nudity. (Schneider 1977: 57)
This poses a problem for the immunity model: ‘The problem here is
that our concern with nudity, sexuality, and other cultural taboos does not
seem to be a desire for immunity from the judgment of others, as much
as, the simple desire not to be observed.’ (Johnson 1989: 165; italics
added.) In defence of his immunity model, Johnson argues that such cases
still involve ‘a concern with judgment, rather than mere observation’, and
he points to the use of sexual taboos in the language we use and the ‘talk
of shame and embarrassment’ (ibid.) in this regard. The connection
between linguistic taboos and our desire for privacy in sexual matters,
however, is shaky at best, and it is far from clear that the latter would
disappear in less judgmental societies. By contrast, the disattendability/
civil inattention model takes the ‘simple desire not to be observed’ as
basic: Rather than treating our desire, in intimate situations, not to be
attended to by third parties who have no legitimate claim on our attention
(and vice versa) as standing in need of explanation, I propose that it be
seen as a core expression of the twin principles of disattendability and
civil inattention.
The disattendability/civil inattention model also holds out the promise
of accounting for the historical contingency and cultural diversity of
privacy as a social practice, while at the same time identifying a common
underlying mechanism. In an article on Japanese conceptions of privacy,
Masahiko Mizutani, James Dorsey and James Moor point out that
traditional living arrangements make the spread of personal information
176 Axel Gelfert
virtually inevitable. Japanese homes are
separated into rooms by sliding lattice-work doors (called ‘shooji’), often
covered with nothing more than paper. Because sound travels through these
doors quite easily and as Japanese homes are rather small, there is almost
surely someone in an adjoining room most of the time. In spite of this,
convention has long held that conversations overheard through closed shooji
are not to be repeated or acknowledged in any way. One can amae (presume
indulgence) from anyone who might overhear the conversation; those who do
hear will exercise enryo, restraint, and not act on or repeat what was
overheard. (Mizutani, Dorsey, and Moor 2004: 124)
In such living conditions, it is virtually impossible to create conditions
‘of not having undocumented personal knowledge about one possessed by
others’ (Parent 1983: 269), as the informational model would demand,
nor does it seem we can suppress judgments of others, which they might
privately form about us. What we can demand, though, is that we be
treated with a certain level of civil inattention. (While the concept of civil
inattention was initially proposed for the public realm, there is no
difficulty to conceive of it as coming in degrees and as extending also
to, say, individuals living under the same roof, where it would serve to
give each other some ‘private space’.) This appears to be exactly the
function of the Japanese concept of enryo, restraint, which, as Mizutani
et al. note, ‘does much to maintain the privacy of individuals within
groups’ and is accompanied by an ‘“as if” convention for protecting
privacy’ (Mizutani, Dorsey, and Moor 2004: 126). It should count in
favour of the disattendability/civil inattention model of privacy that it can
explain the functional point of conventions such as enryo, which emerge
even in situations that are empirically not amenable to either the
Disattendability, Civil Inattention, and the Epistemology of Privacy 177
containment of information per se (as the informational model would
require) or the prevention of other people’s judgments (as the immunity
model would demand).
In this paper, I have argued for an alternative to traditional models of
privacy, which either have tended to emphasize access to (or control
over) information per se, or have aimed at preventing the formation of
(potentially
negative)
judgments
by
others.
By
contrast,
the
disattendability/civil inattention model I am proposing emphasizes the
relational aspect of privacy, which is reflected in our reasonable
expectation to be accorded civil inattention by others as we go about our
daily lives, without drawing undue attention to ourselves. A state of
privacy, on this account, is a state of affairs where disattendability on the
part of an individual is matched by the civil inattention the social
environment accords him or her. This need not be limited to how an
individual can expect to be treated in a public setting, but also extends
to other aspects in life. For example, in the workplace we can reasonably
expect that colleagues refrain from attending to, say, our love life,
political orientation, or family issues. Which degree of attention is
legitimate is clearly context-dependent and domain-specific, and the
proposed model can accommodate such variation across contexts and
domains.
A nagging worry remains, however, and stems from the earlier
criticism of disattendability as an oppressive means of enforcing (often
dubious) standards of ‘respectability’, ‘civility’, or the like. (See Section
4.) By linking privacy to civil inattention which, in turn, has to be
178 Axel Gelfert
‘earned’ through exhibiting disattendability, the proposed account might
seem to encourage the ‘hiding’ of aspects of an individual’s behaviour,
or way of life, which are merely deemed unacceptable by the majority,
when in fact privacy should serve to protect individuals from unwarranted
intrusion by others. If this worry sounds rather abstract, consider the
following example. In many socially conservative societies, a gay couple
holding hands or exchanging public displays of affection would no doubt
attract attention and might be frowned upon; by definition, such
behaviour would thereby not conform to the norm of disattendability. Yet,
most proponents of a right to privacy (this author included) would deem
such public disapprobation unjustified. Does the disattendability/civil
inattention model then merely serve to enshrine existing (and potentially
oppressive) social norms, by declaring any behaviour that might attract a
significant degree of attention as ‘fair game’ and not worthy of privacy
protection? Not quite. This is because the proposed model, while aiming
to clarify the concept of privacy, is largely explanatory in character; as
such, it does not pass judgment on which actions or behaviours are, or
aren’t, deserving of privacy. Clearly, we must leave open the possibility
that there are a great number of things that ought not to attract attention,
even if, in fact, they do. The existence of double standards for public
displays of affection between heterosexual and same-sex couples is a
good case in point. In short, not every violation of disattendability is a
moral violation, and not every suspension of civil inattention is justified.
(It is important to note that this works both ways: as we know only too
well, some moral outrages such as child abuse are often allowed to
persist because those in the know wrongly perceive them to be a ‘private’
matter.)
A number of authors have pointed out that privacy is an essentially
Disattendability, Civil Inattention, and the Epistemology of Privacy 179
cultural notion. As Johnson aptly puts it:
What is considered private is socially or culturally defined. It varies from
context to context, it is dynamic, and it is quite possible that no single
example can be found of something which is considered private in every
culture. (Johnson 1989: 157)
The proposed disattendability/civil inattention model acknowledges this
contingency of our privacy conventions, while at the same time
identifying a common mechanism. It is flexible enough to accommodate
both cultural variation and technological change, insofar as gratuitous
(unwarranted) attending may take many forms (including, say, the
algorithimic monitoring of electronic communications). But the model
should not be misunderstood as providing a moral justification for
conformism, let alone for the suppression of minority viewpoints and
uncommon ways of life, in the name of maintaining ‘disattendability’. On
the contrary, it brings into sharper focus the need for a vigorous defence
of the ‘visibility’ of minority practices and behaviour, given that a
defence on the grounds of privacy alone is unlikely to be sufficiently
robust and might even run the risk of downplaying the significance of
public exclusion or of trivializing the much larger moral struggles to be
won.
Acknowledgments. I am grateful to the participants of the Episteme
conference ‘The Epistemology of Privacy and Secrecy’, held at TU Delft
in May 2012, and of the NUS-Kyoto Applied Philosophy Workshop in
February 2014, especially Masahiko Mizutani, for useful discussions. I
would also like to thank Luke O’Sullivan (Singapore) and two
anonymous reviewers for their comments on an earlier draft of this paper.
180 Axel Gelfert
Martijn Blaauw. (2013) “The Epistemic Account of Privacy”, Episteme
19 (2), pp. 167-177.
Joshua Cohen. (2009) Philosophy, Politics, Democracy: Selected Essays.
Cambridge, Mass.: Harvard University Press.
Don Fallis. (2013) “Privacy and Lack of Knowledge”, Episteme 19 (2),
pp. 153-166.
Charles Fried. (1984) “Privacy [a moral analysis]”, in Philosophical
Dimensions of Privacy: An Anthology in Ferdinand D. Schoeman
(ed.), Cambridge: Cambridge University Press, pp. 203-222.
Axel Gelfert. (2006) “Kant on Testimony”, British Journal for the
History of Philosophy 14 (4), pp. 627-652.
Raymond Geuss. (2001) Public Good, Private Goods. Princeton:
Princeton University Press.
Erving Goffman. (1963) Stigma: Notes on the Management of Spoiled
Identity. Englewood Cliffs: Prentice-Hall.
Erving Goffman. (1971) Relations in Public: Microstudies of the Public
Order. New York: Basic Books.
Jeffery L. Johnson. (1989) “Privacy and the Judgment of Others”, The
Journal of Value Inquiry 23, pp. 157-168.
Klemens Kappel. (2013) “Epistemological Dimensions of Informational
Privacy”, Episteme 10 (2), pp. 179-192.
David Matheson. (2007) “Unknowability and Information Privacy”,
Journal of Philosophical Research 32, pp. 251-167.
John Stuart Mill. (1963-1991) The Collected Works of John Stuart Mill
in John M. Robson (ed.), Toronto: University of Toronto Press.
William Ian Miller. (1997) The Anatomy of Disgust. Cambridge, Mass.:
Disattendability, Civil Inattention, and the Epistemology of Privacy 181
Harvard University Press.
Masahiko Mizutani. (2012) “Privacy, Ethics of” in Encyclopedia of
Applied Ethics (Second Edition) in Ruth Chadwick (ed.), San
Diego: Academic Press, pp. 609-615.
Masahiko Mizutani, James Dorsey, and James H. Moor. (2004), “The
Internet and Japanese Conception of Privacy”, Ethics and
Information Technology 6 (2), pp. 121-128.
W. A. Parent. (1983) “Privacy, Morality, and the Law”, Philosophy &
Public Affairs 12 (4), pp. 269-288.
Robert C. Post. (1991) “Rereading Warren and Brandeis: Privacy,
Property, and Appropriation”, Case Western Reserve Law Review
41, pp. 647-680.
Carl Schneider. (1977) Shame, Exposure, and Privacy. Boston: Beacon
Press.
Daniel J. Solove. (2008) The Future of Reputation. Gossip, Rumor, and
Privacy on the Internet. New Haven: Yale University Press.
Judith Jarvis Thomson. (1975) “The Right to Privacy”, Philosophy and
Public Affairs 4 (4), pp, 295-314.
Raymond Wacks. (2010) Privacy: A Very Short Introduction. Oxford:
Oxford University Press.
Margaret Urban Walker. (2006) Moral Repair: Reconstructing Moral
Relations after Wrongdoing, Cambridge: Cambridge University Press.
Samuel D. Warren and Louis D. Brandeis. (1890) “The Right to Privacy”,
Harvard Law Review 193 (4), pp. 193-220.
Department of Philosophy, National University of Singapore
phigah@nus.edu.sg
Axel Gelfert