1. INTRODUCTION
Two sorts of connections between privacy and knowledge (or lack thereof) have been suggested in the philosophical literature. First, according to Alvin Goldman (Reference Goldman1999: 173), ‘epistemology focuses on the means to knowledge enhancement, whereas privacy studies focus on the means to knowledge curtailment’. Thus, he concludes that privacy falls ‘outside the scope of epistemology’. This suggests that protecting privacy typically leads to less knowledge being acquired. Second, several other philosophers (e.g. Parent Reference Parent1983: 269; Matheson Reference Matheson2007: 259; Blaauw Reference Blaauw2012; Peels Reference Peels2012) have claimed that lack of knowledge is definitive of having privacy. In other words, someone not knowing something is necessary and sufficient for someone else having privacy about that thing. Or equivalently, someone knowing something is necessary and sufficient for someone else losing privacy about that thing.
In this paper, I argue that both of these suggestions are incorrect. I argue in section 2, contra Goldman, that protecting privacy often leads to more knowledge being acquired. I argue in the remaining sections, contra the defenders of the knowledge account of privacy, that someone knowing something is not necessary for someone else losing privacy about that thing. In other words, you can lose privacy even if someone has epistemic access to facts about you that falls short of knowledge.
2. PRIVACY AND KNOWLEDGE ENHANCEMENT
In his book Knowledge in a Social World, Alvin Goldman discusses various social practices and policies from the perspective of epistemology. For instance, he looks at how legal practices can enhance the ability of juries to reach accurate verdicts. Also, he studies how scientific practices can help us to identify true hypotheses about the world. However, he does not discuss privacy practices.
According to Goldman (Reference Goldman1999: 173),
Important as [privacy] is, it does not squarely fall into the domain of epistemology as I have delineated it, because epistemology focuses on the means to knowledge enhancement, whereas privacy studies focus on the means to knowledge curtailment (at least decreasing knowledge in the hands of the wrong people). For this reason, I shall not explore this topic. I do not belittle the importance of privacy as a moral issue; it simply falls, for the most part, outside the scope of epistemology.
This particular remark has always struck me as misleading. I agree that privacy is primarily of concern to moral philosophy. But I also think that there is a place for privacy studies within epistemology (see Fallis Reference Fallis2004: 113–14).
A few years later, Goldman (Reference Goldman2002: 218–20) did suggest that epistemology might take a ‘broader perspective’ and try to identify ‘practices that would promote veritistically inferior [truth-possessional profiles] if that is what a certain task requires’.Footnote 1 If we allow that epistemology can be concerned with knowledge curtailment in this way, then privacy studies would clearly fall within the scope of epistemology. However, my problem with Goldman's original remark does not have to do with the scope of epistemology. I am fine with the idea that epistemology is exclusively concerned with knowledge enhancement (see Fallis Reference Fallis2004: 102–3). My problem is with the idea that privacy studies are only concerned with knowledge curtailment.
Much like the legal and scientific practices that Goldman discusses, privacy practices can enhance knowledge acquisition. Consider, for example, what goes on in libraries. Librarians take very seriously the privacy of their patrons. Indeed, it is a core principle in their code of ethics (see Garoogian Reference Garoogian1991: 230). So, librarians will not release patron borrowing records without a subpoena. In addition, they set up library computers so that internet caches are automatically erased after each patron. Thus, a record of the websites that a patron has visited will not exist even if it is subpoenaed.
There is certainly one respect in which such practices curtail knowledge acquisition. They involve withholding information from people who would like to know it. For instance, in their search for terrorists, the FBI would like to know which people are looking up information about bomb making, but library privacy measures make it more difficult for them to acquire this knowledge. However, such practices also make it more likely that library patrons will acquire knowledge on all sorts of topics. This is because, if people are worried that other people (such as FBI agents) can easily find out what books they have checked out, there can be a ‘chilling effect’ on their use of libraries (see Garoogian Reference Garoogian1991: 229). Thus, it is quite plausible that the knowledge enhancement outweighs the knowledge curtailment of library privacy measures.
Moreover, the library case is not unique in this regard. For instance, medical and legal confidentiality arguably enhance the sharing of information and, thus, knowledge acquisition. Thus, even if epistemology is exclusively concerned with knowledge enhancement, privacy studies fall within the scope of epistemology.
It should be noted that, in Knowledge in a Social World, Goldman (Reference Goldman1999: 23) focuses exclusively on ‘weak knowledge’ rather than ‘strong knowledge’. That is, he is concerned with how practices and policies affect the number of true beliefs that people acquire. However, by using library materials, people can certainly acquire beliefs that are justified as well as true. Thus, library privacy measures arguably lead to the enhancement of strong knowledge as well as weak knowledge.
Finally, it might be suggested that Goldman did not mean to suggest that protecting privacy typically leads to less knowledge being acquired overall. Instead, he just meant that protecting privacy typically leads to some knowledge curtailment. In particular, it prevents people from knowing the information that is kept private (as when library privacy measures keep FBI agents from knowing what books people have checked out of the library). However, if that is all that he meant, it clearly would not have followed that privacy studies fall outside the scope of epistemology. In Knowledge in a Social World, Goldman himself discusses several cases where curtailing knowledge in one respect serves to enhance knowledge in some other respect. For instance, in scientific experiments, observers are kept ‘ignorant of the conditions of an experiment’ so that their observations will not be prejudiced (see Goldman Reference Goldman1999: 293–4). Also, juries are kept ignorant of certain facts about the defendant (e.g. that he has withdrawn a guilty plea) that might prejudice their deliberations (see Goldman Reference Goldman1999: 241).
3. THE KNOWLEDGE ACCOUNT OF PRIVACY
Even if protecting privacy can lead to knowledge enhancement as well as knowledge curtailment, it might be suggested that knowledge curtailment is what is definitive of privacy. More precisely, A has privacy about a personal fact p with respect to S if and only if S does not know that p.Footnote 2 Or equivalently, A has lost privacy about p with respect to S if and only if S knows that p.Footnote 3 For instance, suppose that Norm has a tattoo on his butt that says, ‘I love Vera.’Footnote 4 If Sam does not know that Norm has such a tattoo, Norm has privacy with respect to Sam about the tattoo. However, if Norm's best friend Cliff, who has seen the tattoo, tells Sam about it, Sam now knows about the tattoo. In that case, Norm has lost his privacy with respect to Sam about the tattoo.Footnote 5
This knowledge account of privacy is certainly quite plausible.Footnote 6 When we are concerned about our privacy, it seems that we are concerned about what other people know about us. But in the remainder of this paper, I argue that the knowledge account of privacy is not correct. In particular, I argue that, while knowledge may be sufficient for a loss of privacy, it is not necessary.
Some clarifications
Before we get to my worries about the knowledge account of privacy, there are a few issues that should be clarified. First, the knowledge account is an account of what it means to simply lose privacy rather than an account of what it means for privacy to be violated. In other words, it does not have anything to say about whether a person who has lost privacy has been harmed or wronged in some way. The fact that Norm has lost his privacy with respect to Sam does not necessarily mean that anyone has violated his privacy. For instance, Norm might have given Cliff permission to tell Sam about the tattoo.
Something more than epistemic access to a personal fact is required in order for a loss of privacy to count as a privacy violation. Exactly what this something more is is an important conceptual issue for moral philosophy. But for present purposes, we can set this issue aside since it is not clear that it tells us anything further about the connection between privacy and knowledge.
Second, it may not be possible for a person to lose privacy about just any fact about him. For instance, it does not seem that Norm can lose privacy about the fact that he is self-identical (see Matheson Reference Matheson2007: 265). Thus, a complete account of having (and of losing) privacy requires an account of exactly what types of facts can be private in the first place. Indeed, most defenders of the knowledge account (e.g. Parent Reference Parent1983: 269–70; Matheson Reference Matheson2007: 265; Blaauw Reference Blaauw2012; Peels Reference Peels2012) only claim that privacy can be had (or lost) about personal facts.
However, simply showing that knowledge is not necessary for a loss of privacy (as I aim do to in the remainder of this paper) does not require giving a complete account of what a personal fact is. It is enough to find a case (or cases) where privacy is lost about what clearly is a personal fact (even though that fact is not known). In particular, my counterexamples to the knowledge account all involve Norm losing privacy about having a tattoo on his butt.
Admittedly, a tattoo is typically something that a person wants to display. Indeed, Norm presumably would like his wife Vera to see his tattoo. However, if someone gets a tattoo on a part of the body, such as his butt, that is not normally exposed, the intended audience for a tattoo is usually rather limited. He may legitimately want to have privacy with respect to most people about that tattoo (cf. Johnson Reference Johnson1989: 158).
Finally, it might be suggested that privacy is lost as soon as information about a person is made easily and publicly accessible. It does not matter whether anyone acquires knowledge as a result of accessing that information. For instance, it seems that Norm loses privacy as soon as Cliff posts on his website that ‘Norm has a tattoo on his butt’ even before anyone reads this information. It is tempting to conclude from this that knowledge is not necessary for a loss of privacy.
However, defenders of the knowledge account have a plausible reply to this sort of argument. When Cliff puts the information on his website, Norm's privacy is put at risk. But he does not actually lose his privacy until someone reads the information and comes to know that Norm has a tattoo on his butt (cf. Thomson Reference Thomson1975: 305; Matheson Reference Matheson2007: 256–8). As David Matheson (Reference Matheson2007: 257) would put it, if no one knows that he has such a tattoo, Norm still has ‘privacy relative to everyone and to those facts, even if that privacy is only precariously possessed’ (emphasis added). After all, if Norm manages to get Cliff to remove the information before anyone happens to see the website, it does not seem that his privacy was ever lost.
Admittedly, if we are concerned with our privacy, we do want it to be difficult for people to access information about us. But we only care about that as a means to an end. So, I agree with the defenders of the knowledge account that actual epistemic access is necessary for a loss of privacy. I just do not think that the type of epistemic access in question has to be knowledge.
4. KNOWLEDGE IS NOT NECESSARY FOR LOSS OF PRIVACY
The first thing to note is that true belief (or ‘weak knowledge’) is not sufficient for a loss of privacy. Suppose that Sam simply had a vivid dream about Norm having a tattoo on his butt. When he wakes up, he finds that he cannot help believing that Norm has a tattoo. But there is no connection between the fact that Norm has a tattoo and Sam's belief. In that case, it does not seem that Norm has lost his privacy about the tattoo with respect to Sam (cf. Matheson Reference Matheson2007: 264; Peels Reference Peels2012). In order for Norm to lose his privacy about the tattoo with respect to Sam, Sam's belief needs to be hooked up to Norm's tattoo in some way.
The plausible suggestion of the knowledge account is that Sam has to know about the tattoo in order for Norm to lose his privacy. However, suppose that Cliff is an unreliable testifier, but that he is telling the truth when he tells Sam that Norm has a tattoo on his butt. In that case, Sam does not know that Norm has a tattoo on his butt. But it still seems that Norm has lost his privacy about the tattoo with respect to Sam.
In response to such a story, defenders of the knowledge account of privacy might say, ‘so much the worse for reliabilist theories of knowledge’. However, suppose that Cliff is an unreliable testifier and that everyone knows that he is unreliable. In that case, Sam has an undefeated (undercutting) defeater for Cliff's testimony. But suppose that Sam loves gossip! He can't help believing it, and he can't help spreading it. In that case, Sam does not know that Norm has a tattoo on his butt according to most theories of knowledge (including purely internalist theories). But it still seems that Norm has lost his privacy about the tattoo with respect to Sam. Admittedly, if he is aware that Cliff is unreliable, Sam may be no more justified in believing that Norm has a tattoo than when he just has a vivid dream. Even so, since he hears about the tattoo from Cliff who actually saw it, Sam is hooked up to the tattoo in an important way that he is not when he just has a vivid dream.
In both of these stories, although Sam does not know that Norm has a tattoo, his belief does come by way of someone else's (Cliff's) knowledge. But this is not necessary for a loss of privacy either. For instance, suppose that Cliff was drunk when he saw the tattoo.Footnote 7 As a result, he cannot be sure that Norm has a tattoo on his butt. Even so, if Cliff goes on to tell Sam about the tattoo, it seems that Norm loses his privacy.Footnote 8
In all of these stories, Norm has lost his privacy even though Sam's belief is not justified to the level required for knowledge. But this does not mean that such justification cannot be important when it comes to privacy. If a belief is fully justified, the loss of privacy can easily last longer and be extended to more people. For instance, if Sam's belief is fully justified, he is less likely to simply give it up when presented with contrary evidence, such as Norm denying that he has a tattoo on his butt. Also, if Sam's belief is fully justified, he is in a better position to convince others to believe it. In other words, greater justification certainly puts one at greater risk for further loss of privacy. Even so, justification that suffices for knowledge is not required for a loss of privacy now with respect to this person.
Some possible responses
As I discuss in the following section, in all of these stories there is an important epistemic connection between Sam's belief and the fact that Norm has a tattoo on his butt. Thus, it might be suggested that Sam does know, under just the right theory of knowledge, that Norm has a tattoo on his butt. However, none of the defenders of the knowledge account of privacy has offered an explicit theory of knowledge. So, the presumption has to be that they have in mind a fairly intuitive conception of knowledge. And Sam certainly does not seem to know in that sense.Footnote 9
In all of my stories, Sam finds out about Norm's tattoo by way of Cliff. But suppose that Sam sneaks into Norm's house and reads about the tattoo in Norm's diary. Norm likes to write in his diary all sorts of things about himself that sound good. Some of the claims, such as that he has a tattoo on his butt, are true. But most of the claims, such as that he has an advanced degree in philosophy, are not true. In that case, Sam is not justified in believing what he reads in the diary. But if Sam nevertheless forms the belief that Norm has a tattoo, it seems that Norm loses his privacy about the tattoo (cf. Allen Reference Allen1988: 21–2).
David Matheson (Reference Matheson2007: 263–4) has given a couple of responses to this sort of proposed counterexample to the knowledge account. First, he claims that Norm does not lose privacy because Sam comes to believe that he has a tattoo on his butt. Instead, Norm loses privacy because Sam comes to know other facts about him, such as his ‘taste in home décor’, as a result of entering his house. However, the intuition is that Norm loses his privacy about the tattoo rather than about his ‘furniture preferences’. Norm was never worried about these other facts about him being made public. As Jeffrey Johnson (Reference Johnson1989: 161) puts it, ‘though some knowledge is involved, it is incidental’. In any event, this reply to the diary case does not respond to my stories above. In all of my stories, Sam does not come to know anything about Norm. Based on Cliff's testimony, he just acquires an unjustified belief that Norm has a tattoo on his butt.
Second, Matheson points out that Sam could have acquired the very same beliefs about Norm without Norm losing his privacy. For instance, he might have formed the beliefs as a result of wishful thinking instead of as a result of reading the diary. However, this fact does not show that Norm does not lose his privacy if Sam does form the beliefs as a result of reading the diary.
Knowledge-lies
The case against the knowledge account of privacy is not limited to intuitions about hypothetical cases. We should also consider how people actually try to protect their privacy. If the knowledge account is correct, we should see people trying to protect their privacy with what Roy Sorensen (Reference Sorensen2010) calls ‘knowledge-lies’. But in the remainder of this section I argue that Sorensen has yet to identify an example of a knowledge-lie that could plausibly protect someone's privacy.
A common privacy protection strategy is to tell lies. In fact, according to John Mackie (Reference Mackie1977: 183), ‘a round, confident, lie may be the only practicable alternative to an undesirable revelation of the truth’. For instance, Dietrich Bonhoeffer (Reference Bonhoeffer1955: 330) describes a case where ‘a teacher asks a child in front of the class whether it is true, that his father often comes home drunk. It is true, but the child denies it.’ There is some disagreement about whether it is ethical to protect one's privacy in this way. But no one denies it can work.Footnote 10
On the traditional account of lying, lies always attack belief. For instance, the shepherd boy who famously shouts ‘A wolf is chasing my sheep!’ intends the villagers to believe that there is a wolf when there is no wolf. But Sorensen claims that there are knowledge-lies that attack knowledge without attacking belief. For instance, even if she does not expect to be believed outright, it might still make sense for a defendant's wife to give him a false alibi that keeps the jury from knowing that he is guilty. As Sorensen (Reference Sorensen2010: 611) points out, ‘a conscientious juror who believes the defendant committed the crime may acquit him because she does not know he committed the crime’. If the knowledge account of privacy were correct, someone could protect her privacy with such knowledge-lies just as well as she can with traditional belief-lies.
Moreover, all other things being equal, there are reasons to prefer telling knowledge-lies. For one thing, it is generally easier to simply keep someone from knowing something than to actually get her to believe the opposite. In other words, a belief-lie is typically overkill if you just want to keep someone from knowing something. In addition, if Sorensen (Reference Sorensen2010: 612) is correct that ‘most of what is wrong about lying concerns corruption of belief’, then it is less wrong to tell a knowledge-lie than to tell a belief-lie. So, if the knowledge account of privacy is correct, we should expect to see a lot of people protecting their privacy with knowledge-lies. Thus, if there are no privacy-protecting lies that attack knowledge without attacking belief, it counts against the knowledge account.
Sorensen has not yet given us an example of a lie that plausibly protects privacy just by attacking knowledge. In most of his examples, including the witness lying to the jury, the liar wants to manipulate behavior. In a similar vein, the child might want to keep the teacher from interfering in his life (e.g. by calling in Child Protective Services). But if a lie is intended to manipulate behavior in this way, then it must target belief. For instance, while the child might not expect the teacher to fully believe that his father is not a drunk, he must intend to keep the teacher's degree of belief high enough so that the expected utility of not interfering is greater than the expected utility of interfering (cf. Fallis Reference Fallis2011: 354–60).
Sorensen (Reference Sorensen2010: 609) does give one example where the liar is not trying to manipulate behavior and the lie does not attack belief at all: ‘Denying you are hurt can prevent a tormenter from knowing he hurt you. If you believe your enemy is vain, you anticipate he will ignore this evidence. So you tell the lie without hope of preventing his true belief that he hurt you.’ In a similar vein, the child might not intend to manipulate the teacher's behavior in any way. Although he expects the teacher to ignore what he says, the child might just want the teacher not to know that his father is a drunk.
However, it is not clear that such cases provide support for the knowledge account of privacy. If the torturer is so vain that he will ignore evidence to the contrary, it is not clear that he can know that he is hurting you. Similarly, if the teacher will simply ignore the child's denial, it is not clear that he can know that the child's father is a drunk. As Elizabeth Fricker (Reference Fricker, Lackey and Sosa2006: 242) points out, ‘sensitivity to defeating evidence … is usually taken to be the hallmark of belief which amounts to knowledge’. So, this sort of lie does not attack knowledge after all (see Fallis Reference Fallis2011: 360–1).Footnote 11
5. COGNIZANCE AND LOSS OF PRIVACY
Since true belief is not sufficient for a loss of privacy, and since knowledge is not necessary, there must be some other connection between the fact and the belief that is sufficient for a loss of privacy. Indeed, in all of my stories from the previous section, even though Sam does not know that Norm has a tattoo on his butt, there is such a connection. In particular, there is a causal connection between Sam's belief and the fact that Norm has a tattoo on his butt. The fact that Norm has a tattoo on his butt leads to Cliff's testimony, which in turn leads to Sam's belief.Footnote 12
While Goldman's causal theory of knowledge (CTK) was ultimately unsuccessful as a theory of knowledge, it may be just what is needed in a theory of privacy. In other words, my claim is that A loses privacy about a personal fact p with respect to S if ‘the fact p is causally connected in an ‘appropriate’ way with S's believing p' (Goldman Reference Goldman1967: 369). Let us refer to this sort of connection between the fact and the belief as cognizance.Footnote 13
In the remainder of this section, I identify the principal ways in which the concept that I have in mind here coincides with (and a few ways in which it differs from) what Goldman had in mind. While this does not constitute a complete account of cognizance, it should be sufficient in order to establish that this sort of causal connection between a fact and a belief can result in a loss of privacy.
Cognizance and perception
If the sort of causal connection appealed to in the CTK results in a loss of privacy, it makes sense of why the knowledge account of privacy is so attractive. Being cognizant is very close to knowing. After all, cognizance was (at least for a while) a serious contender in the philosophical debate over what knowledge is. Thus, it is not surprising that we would mistakenly think that knowing is always required for a loss of privacy. In addition, it makes sense that this sort of causal connection would result in a loss of privacy. The prototypical example of a loss of privacy is when someone looks in your window and observes you. And perception is certainly the principal type of causal connection that Goldman (Reference Goldman1967: 358–9) had in mind when he developed the CTK.
But it is important to emphasize that direct perception is not the only type of causal connection that can result in a loss of privacy. If Sam looks in Norm's window and sees the tattoo while Norm is dressing, Norm certainly loses his privacy about the tattoo. However, suppose that Sam is wearing a device over his head that uses information from sensors to simulate exactly the visual experiences that he would have had if the device were not blocking his vision. In that case, Sam is not seeing his surroundings (see Goldman Reference Goldman1967: 359; Dancy Reference Dancy1985: 173–4). But if Sam points his head at Norm's window and ‘sees’ the tattoo, he still perceives the tattoo in an attenuated sense. In other words, Sam is still cognizant of the tattoo. Thus, Norm still loses his privacy.
In addition to perception (mediated and unmediated), Goldman (Reference Goldman1967: 362–3) claims than an appropriate causal connection can involve testimony. Basically, if someone who is cognizant of some fact tells you about it, you are also cognizant of that fact even though you did not directly perceive it. Thus, as in my stories from the previous section, if Cliff sees the tattoo and tells Sam about it, Norm loses his privacy with respect to Sam.Footnote 14
Some fine-tuning
But just like the CTK itself, the concept of cognizance requires some fine-tuning. Not just any causal connection between the fact and the belief will result in a loss of privacy. For instance, suppose that, when Norm's wife Vera first sees the tattoo, she is so horrified that she runs out into the street screaming incoherently. Sam, who is on his way to Cheers, is so distracted by this strange behavior that he falls and hits his head. When he wakes up, he finds that he has acquired (on the basis of no evidence at all) a number of strange beliefs, such as that the Chicago Cubs have won the World Series and that Norm has a tattoo on his butt. In this case, while Sam's belief that Norm has a tattoo is caused by the fact that Norm has a tattoo, it is not an appropriate causal connection (cf. Feldman Reference Feldman2003: 83). Norm no more loses his privacy with respect to Sam than he does when Sam simply has a vivid dream about the tattoo.
Moreover, even if the causal connection between the fact and the belief is of the right sort, the believer has to have some evidence for the belief even if it still falls short of knowledge. For instance, suppose that Sam sees a photo of Norm's tattoo that Fred's tattoo parlor has used in its advertising. Even though Sam does not have any evidence that the tattoo belongs to Norm (e.g. the photo does not say that the tattoo belongs to Norm and he does not know that Vera is Norm's wife), he decides that the tattoo does belong to Norm. In that case, Sam perceives Norm's tattoo (at least in an attenuated sense) and this leads to his belief that Norm has a tattoo. But again, Norm no more loses his privacy with respect to Sam than he does when Sam simply has a vivid dream about the tattoo.
But it should be noted that the sort of causal connection that leads to a loss of privacy is probably not exactly what Goldman had in mind when he was developing the CTK. According to Goldman (Reference Goldman1967: 363), ‘a correct reconstruction is a necessary condition of knowledge based on inference’. That is, a knower must have an accurate understanding, at least in rough outline, of the causal chain leading from the fact to her belief (cf. Feldman Reference Feldman2003: 84). In all of my stories from the previous section, Sam has such an understanding. However, this does not seem to be a necessary condition for a loss of privacy. For instance, suppose that the device that simulates the visual experiences that he would have had if the device were not blocking his vision is placed on Sam's head without his knowledge (e.g. while he is sleeping). In that case, when Sam points his head at Norm's window and ‘sees’ the tattoo, he may not know that Norm has a tattoo on his butt because he is mistaken about how the fact led to his belief. But Sam is still cognizant of the tattoo and it seems that Norm has still lost his privacy.
Even with all of this fine-tuning of the concept, however, it is not clear that cognizance itself is necessary for a loss of privacy. It seems possible for someone to lose privacy even when there is not a causal chain leading from the fact to the belief. In most cases, a fact causes some evidence, which in turn causes a belief. For instance, in all of my stories, the fact that Norm has a tattoo on his butt leads to Cliff's testimony, which in turn leads to Sam's belief. However, sometimes there is just a common cause for the fact and the evidence. For instance, suppose that Sam sees Norm going into Fred's tattoo parlor and that Sam knows that Fred never lets anyone leave without getting a tattoo. As a result, Sam concludes that Norm has a tattoo on some unexposed part of his body. In that case, Norm's intention to get a tattoo leads to both Norm having a tattoo and to Sam's belief. However, the fact that Norm has a tattoo is not part of the causal history of Sam's belief. Even so, it seems that Norm has lost his privacy about having a tattoo.
In developing the CTK, Goldman (Reference Goldman1967: 364–5) claimed that there can be a causal connection between the fact and the belief even if there is not a causal chain leading from one to the other (see also Feldman Reference Feldman2003: 84). For instance, Goldman would say that Sam knows that Norm has a tattoo as a result of seeing Norm go into the tattoo parlor. There is a causal chain from Norm's intention to get a tattoo to the tattoo itself, and there is a causal chain from Norm's intention to Sam's belief, and Sam has an accurate understanding of these causal chains. However, the concept of cognizance should probably not incorporate such an expansive notion of causal connection. Although it seems that Norm has lost his privacy about the tattoo when Sam sees him go into the tattoo parlor, it is not because Sam perceives the tattoo itself even in an attenuated sense.
In any event, even if we adopt Goldman's expansive notion of causal connection, it is still not clear that cognizance is necessary for a loss of privacy. Several critics have pointed out that the CTK has trouble explaining how we can have knowledge of universal facts, facts about the future and negative facts.Footnote 15 For instance, as Jonathan Dancy (Reference Dancy1985: 34) points out, ‘my belief that all men are mortal is caused, but not by the fact that all men are mortal; if any facts cause it, they are the facts that this man, that man, etc., have died’ (see also Feldman Reference Feldman2003: 84–5).Footnote 16 Yet despite the lack of a purely causal connection between such facts and our beliefs in them, privacy about at least some of these facts can be lost. For instance, Benjamin Braddock in The Graduate wants to keep the fact that he is a virgin private. Indeed, he lies to Mrs Robinson to protect his privacy about this negative fact.
6. IS TRUE BELIEF NECESSARY FOR LOSS OF PRIVACY?
So far, I have been following the knowledge account of privacy and taking for granted that true belief is necessary for loss of privacy. However, it is possible to question the necessity of the truth condition and the belief condition as well as the necessity of the justification condition.Footnote 17
The truth condition
Some philosophers (e.g. Allen Reference Allen1988: 21–2; Johnson Reference Johnson1989: 162) claim that one can lose privacy about falsehoods. So, for instance, they would say that, when Sam reads Norm's diary, Norm loses privacy about the false claims as well as about the true claims. In addition, some legal scholars have argued that you have lost privacy when you are put in ‘a false light in the public eye’ (see Johnson Reference Johnson1989: 162). Also, privacy scholars claim that access to inaccurate information (and not just accurate information) is one of the potential problems with privacy violations (see Fallis Reference Fallis2004: 114). The unregulated flow of personal information can make it very difficult to correct errors and such errors (e.g. in medical reports and credit reports) can cause serious harm.
Even so, I think that referring to such cases as privacy violations may just be a loose way of speaking. Many people – and not just philosophers who are committed to the knowledge account of privacy – talk as if one can only lose privacy about facts. For instance, Dorothy Parker famously said, ‘I don't care what's written about me, so long as it's not true.’ In a similar vein, Diane on Cheers tells Sam that she does not care what he tells people as long as he does not ‘discuss my private life with everyone that comes in’. She is a bit perturbed when Sam tells Coach that ‘she's a hooker’. Having certain things said about you can be harmful regardless of whether they are true or false. However, as William Parent (Reference Parent1983: 269) notes, while having false claims made about you may constitute libel or slander, it does not necessarily constitute a loss of privacy.
Admittedly, there are some compelling cases in favor of the view that one can lose privacy about falsehoods. For instance, suppose that a Danish filmmaker is embarrassed about his Nazi heritage.Footnote 18 When a journalist reads his private diary, it seems that he loses privacy about being descended from Nazis. However, as his mother later reveals to him on her deathbed, the man that the filmmaker believed to be his father was not his biological father. So, he was not actually descended from Nazis.
But we can plausibly bite the bullet here and insist that the filmmaker's privacy is not lost about his being descended from Nazis. If the journalist publishes this false claim, the filmmaker may very well be harmed and libelled. But since this claim is not a fact, it is not something that the filmmaker can lose his privacy about. When the journalist reads his diary, the filmmaker's privacy is only lost about the fact that he believes that he is descended from Nazis.
Moreover, there is a serious difficulty with claiming that one can lose privacy about falsehoods. Just as you do not lose privacy merely as a result of someone having a true belief about you, presumably you do not lose privacy merely as the result of someone having a false belief about you. In order to for you to lose privacy about a personal fact, the belief must be hooked up to that fact in an appropriate way. But if you were to lose privacy about a falsehood, what would the belief be hooked up to and how would the belief be hooked up to it?Footnote 19
The belief condition
It might also be suggested that one can lose privacy about a personal fact even if no one believes that fact. First, there are cases where privacy seems to be diminished even though no new belief is acquired. For instance, suppose that Sam takes a photo of the tattoo through the window while Norm is dressing and then proceeds to study the photo every day. Sam believes (and knows) that Norm has a tattoo on his butt as soon as he takes the photo. No new belief is acquired each time that he studies the photo. But it seems that Norm's privacy is diminished each time that Sam studies the photo (cf. Johnson Reference Johnson1989: 161).
Admittedly, the first several times that Sam looks at the photo, he may acquire new beliefs. For instance, he may learn things about the tattoo that he had not noted originally, such as the exact dimensions of the ‘V’ in ‘Vera’. But these are not the sort of facts that Norm is worried about being made public. In any event, after looking at the photo over a long enough period of time, Sam is not going to acquire any new beliefs about Norm, but Norm's privacy will still be diminished.
Second, there are cases where privacy seems to be diminished even though there is no belief about the fact at all. For instance, it could be that, when Cliff tells him that Norm has a tattoo on his butt, Sam cannot believe his ears.Footnote 20 (Norm just does not seem like the sort of person that would get a tattoo.) Even though Sam does not believe that Norm has a tattoo, it seems that Norm has lost his privacy about the tattoo with respect to Sam.
In addition, suppose that Norm does some online research about how to remove his tattoo. When an automated system for targeted advertising subsequently analyses his internet searches, Norm starts receiving ads for Turlington's Lower Back Tattoo Remover (as advertised on Saturday Night Live). Even though no actual person believes that Norm has a tattoo, it seems that Norm's privacy has been diminished.
In all of these cases, even though there is no belief (or at least no new belief) about personal facts, a cognitive agent is still appropriately causally connected to a personal fact. In particular, a cognitive agent arguably perceives Norm's tattoo at least in an attenuated sense.Footnote 21 Thus, unlike truth, belief may not be necessary for a loss of privacy.
Admittedly, belief is often important when it comes to a loss of privacy. Someone believing a personal fact about us is usually what is required for us to be harmed by privacy violations. For instance, if Sam does not believe that Norm has a tattoo on his butt, he is much less likely to pass this information along to other people. So, as discussed in section 3 above, we try to preserve our privacy about a fact by keeping people from believing it. For instance, even though a lie cannot destroy the causal connection between a fact and a belief, it can destroy the belief. But while someone believing a fact is a big part of why privacy violations are bad, it is arguably not necessary for a loss of privacy.
7. CONCLUSION
Philosophers have suggested two sorts of connections between privacy and lack of knowledge. First, Goldman suggests that protecting privacy typically leads to a decrease in knowledge. However, I have argued here that protecting privacy can often lead to an overall increase in knowledge. Second, several other philosophers claim that a lack of knowledge is definitive of privacy. However, I have argued here that you can lose privacy even if someone has epistemic access to facts about you that falls short of knowledge. In particular, you can lose privacy as long as there is an appropriate causal connection (along the lines suggested by the CTK) between a fact about you and someone's belief about you. In fact, you can probably lose privacy even if that someone does not actually form a belief about you.Footnote 22
