One of the things that I'm most nervous about when it comes to designing a social media alternative is having to grapple with the concept of privacy. On a purely technical level, the "privacy controls"^[1] that are actually implemented by social networks seem both too complex and too useless^[2] to bother with. On a human level, the concept of privacy has so many dimensions that I will never be able to do it justice. So I turned to one of my new favorite resources, the Stanford Encyclopedia of Philosophy, to see if they had one of their characteristically detailed and comprehensive reviews of the subject, which they did.

The first thing I learned was that privacy is the philosophical justification (and a target for critics) of a series of legal decisions about certain freedoms people have in democratic societies, such as reproductive rights and rights regarding sexual and gender expression. I don't usually think of those rights as fundamentally about privacy, and it was interesting to read about how privacy was the legal concept from which they arose. I don't intend to explore those ideas further in this context, but if you're interested, the discussion in the SEP is fascinating^[3].

The dimension of privacy that seems most relevant to social media is described in the encyclopedia as informational privacy. This conception of privacy was articulated in an 1890 article called The Right to Privacy by Brandeis and Warren^[4]. The most stunnning insight in the article is that "...the protection afforded to thoughts, sentiments, and emotions, expressed through the medium of writing or of the arts...is merely an instance of the enforcement of the more general right of the individual to be let alone. It is like the right not to be assaulted or beaten... The principle which protects personal writings and all other personal productions, not against theft and physical appropriation, but against publication in any form, is in reality not the principle of private property, but that of an inviolate personality." The authors identify the personality as the part of someone that is protected by privacy, just as the person's body is protected from assault. This implies a view of personality that includes one's feelings and sense of self, as opposed to the way we commonly use the word today, which is "the combination of characteristics or qualities that form an individual's distinctive character." They are identifying privacy as a shield that protects the personality from the harm caused by the publication of facts about a person. They go on to say that the right to privacy lasts until a person publishes or consents to the publication of those facts, specifying that "a private communication of circulation for a restricted purpose is not a publication within the meaning of the law."

The motivation for this 1890s interest in privacy was the rise of newspapers and the beginning of mass media. I have mixed feelings about that--you can definitely read the sudden interest as an expression of aristocratic fear of a lower class newly empowered to participate in important conversations. But Warren and Brandeis explicitly said that elements of the right to privacy do not extend to facts about certain kinds of public figures, like public officials, candidates for office, and a sort of blurry category of important businesspeople. The idea is that while you remain a commoner, you have a right to protection against certain things about you, or things that you express, being made public against your will, because it hurts you in some way when they are made public. When you stop being a commoner and ask for or receive public trust, the public gets a right to know things about you that enable them make an informed decision about whether that trust is well-placed. Since this view of privacy is based on individual well-being, it does not seem to apply to things like targeted advertising, where you might be shown an ad for an embarassing product, but the advertiser is not given access to your details.

An extension of this idea connects privacy to intimacy. The SEP paraphrases Charles Fried, writing in 1970 in his book An Anatomy of Values, saying "Privacy is valuable because it allows one control over information about oneself, which allows one to maintain varying degrees of intimacy. Indeed, love, friendship and trust are only possible if persons enjoy privacy and accord it to each other." This argument goes a step further than Brandeis and Warren, arguing not that privacy means protection from a type of harm, but that it enables a type of relationship that would not be possible without it. Part of the way that we make certain relationships special is by sharing things with people that we would not share publicly. The ability to not share information about ourselves is what makes choosing to share that information an expression of trust. This also extends the idea of privacy from something that is valuable to a person, because it protects them from harm, to something valuable to more than one person, because it facilitates a relationship between them.

Later philosphers continue to develop this idea of privacy as a social value. The SEP quotes Priscilla Regan, writing in 1995 "Privacy is also a public value in that it has value not just to the individual as an individual or to all individuals in common but also to the democratic political system. Privacy is rapidly becoming a collective value in that technology and market forces are making it hard for any one person to have privacy without all persons having a similar minimum level of privacy.” My favorite example of this was given in a lecture by Eben Moglen, where he describes making fun of a colleague who refused to share his social security number with some service provider, (paraphrasing) "You know that the data miners already have it, right? They know the social security numbers of everyone else born in Chicago between 1950 and 1965. Yours is the other one." This gives rise to the idea of a kind of society-wide value of privacy--the value of certain things not being known about large groups of people. The documentary The Social Dilemma (2020) makes a case that it is harmful for advertisers (especially political advertisers) to be able to target certain groups of people based on things like whether they are likely to believe conspiracy theories.

I gave this post the number 000 instead of 0 for a reason. When you want a computer to order things correctly, you need to always use the same number of digits in all your numbers, so that 11 doesn't get sorted earlier than 2. This numbering system means that I can write up to 999 posts on the subject before any of them gets sorted incorrectly. I don't really think 999 would be enough, but it's more than I'm likely to write, so it's sufficient for my lifetime. After going down this rabbit hole for a Saturday, I'm left feeling a little more optimistic than I was when I started. I don't feel like the bewildering array of "privacy controls" offered by existing social networks are a very good way to advance any of the definitions of privacy I've summarized here. I was worried that not wanting to make privacy controls like that might lead me to the conclusion that privacy itself is overrated. That hasn't happened. I think that privacy can be understood in a number of ways, each of which has value in a particular context. It is appropriate for social media tools to have privacy controls based on their context, whether that's an official public blog like this one^[5] or a chat room between friends. It doesn't seem especially difficult, when designing systems for personal use by public or private individuals, to use these concepts of the value of privacy to do an acceptably-good job of protecting what ought to be protected and valuing what ought to be valued. I'll refer back to these ideas as I build stuff, and hope to revisit and elaborate on them.