Skip to content
Coordinates of ThoughtHome of Philosophy
Join
Discussion

Contextual integrity explains why my aunt’s clinic group chat felt wrong

My aunt is a nurse at a small clinic outside Kuala Lumpur. Last month she showed me the staff group chat, where a colleague had posted a photo of a patient's rash to ask the doctor whether it looked like shingles. No name, no face. The doctor answered within minutes and the patient was treated well.

Nothing in the photo was secret in the usual sense, and the purpose was good. But it bothered me, and I could not say why using the public/private distinction I work with every day. The rash was "private", but it was shared with people already treating the patient, so that cannot be the whole story.

Helen Nissenbaum's idea of contextual integrity (worked out in Privacy in Context, 2010) gave me words for it. Information flows are governed by norms that belong to particular contexts: who is sending, who is receiving, whose information it is, what kind of information, and on what terms it moves. Medicine has strong norms about all of these. A consult between clinicians fits them. A photo that now sits on a dozen personal phones, backed up to who knows where, does not, even if no one ever misuses it.

I also find the idea echoed in my own tradition. The Qur'an forbids spying on one another (49:12), and the ethics I grew up with treats what you learn about someone in a trusted role as an amanah, something held in trust, not a fact you now own.

What I am unsure about is whether contextual integrity is too conservative. If every norm is tied to existing practice, how does it criticise a bad practice that everyone already follows?

LikeReplyFollow5 replies

Join the conversation: members reply here, and joining is free. Reading is always free.

Join free to reply Sign in

5 replies

  1. Bilal Rashid

    Fellow

    I meet this often. The tradition's word I reach for is satr, covering. There is a well-known hadith that whoever covers a fellow Muslim's faults, God will cover theirs. Patients tell visitors things they would not tell their families, and the trust depends on the information not travelling. Nissenbaum's framework describes what the satr ethic asks of us without needing the theology, which I find useful when I explain it to the other volunteers who visit with me.

    Helpful · 2
  2. Tsion Bekele

    Fellow

    From the hospital side: the consult was legitimate, so the problem is entirely in what you call the terms of the flow. In our hospital the rule is that clinical images go into the record, not onto phones. Patients consent to being seen by the team. They did not consent to becoming a file in a messaging app's cloud. Your instinct was right, and it is exactly the kind of case the framework handles better than "public or private".

    Helpful · 3
  3. Wanjiru Njenga

    Fellow

    On the conservatism worry: I don't think Nissenbaum leaves it there. As I understand her, a new flow that breaks entrenched norms is not automatically wrong. You also ask whether it serves the ends and values of the context itself. Medicine aims at healing and depends on trust, so a new flow can be judged by whether it serves those. That gives the theory a critical edge.

    My own work is the hard case. Phone records used for credit scoring: the data was generated for communication and now feeds lending. Same bytes, new context.

    1. Nur Aisyah Rahman

      Fellow

      Wanjiru, that answers my worry, thank you. And Bilal, satr is the word I was missing. I told my aunt about the image rule at Tsion's hospital. Her clinic is now arguing about whether to adopt it, which I count as progress.

      Helpful · 2
  4. Sana Rizvi

    Fellow

    Journalists hear "but it was already public" all the time, usually from people defending something indefensible. Contextual integrity is the best answer I know to that line. An address known to your neighbours is not the same as an address published beside a story calling you a traitor. The information is identical; the flow is not.

    Helpful · 1