People think that there are authorities out there with direct knowledge when in fact there is no direct knowledge, there is only focal knowledge, knowledge that is located around a set of practices, discourses, institutional locations, etc. So for example a doctor doesn’t know bodies; a doctor knows a bunch of things about the body when it’s presented in the examination room at the hospital. He’s the authority at that place in that time because of the weave of social premises that support him there. I forget why I thought this was important1 . I had to get out of bed to write this because I couldn’t let this thought go.
Update: I remember now. It was because I had a mental vision along the following lines:
that a programmer as a patient might look at a doctor trying to determine a treatment for him2 and find that he recognizes the doctor’s prescription as a kind of workaround or kludge. Then I thought that the programmer would try to talk to the doctor on the level of processes, establish a common language like, but I thought that the doctor wouldn’t buy into it. The doctor wouldn’t see the workaround primarily as an absence of knowledge, but as a positive knowledge of treatment. He would say, “listen, what I know is that this practice alleviates some of the symptoms of the condition”, and that would be the basis of his authority.
What I mean is that the programmer tries to model a domain of reality, so he sees both the attempt at approximation and correspondence and the inevitable collapse of the pretense that the model re-presents the reality. A database of people is not the people–it’s not even a reference to the people, but rather to their vital statistics. Of course. How else could it be? So knowledge as a model carried in the head of someone– it’s not the thing, it’s not the authoritative image representation of the thing, it’s not the essence of the thing stored away, instead it’s what programmers call “domain specific“, it’s a specific socially-constructed protocol for interacting with a socially-constructed thing. In my vision the programmer’s disappointment doesn’t touch the doctor, because the doctor isn’t interested in the purity of a perfect model but in the efficacy of sound practices. The final step is that “people” believe in authorities at a level that is less sophisticated than either the programmer or the doctor’s (not supposing that one is more sophisticated that the other). They think that authorities just know stuff– authoritatively, without context.
I know the central situation is vague (what is the doctor ignorant about? etc.) but that’s because it’s from a daydreaming-thought-experiment of the kind I often have.