Milwaukee.
OK, I'm in the next session at AoIR 2009, and Michael Blanchard makes a start by introducing Foucault's idea of governmentality. He believes that Deleuze's statement that we now live in societies of control is problematic - the societies of discipline that Foucault has introduced have been replaced with societies of control, but there was never an idea that there was a clear succession from sovereignty to discipline to control; these three were always a triangle.
Digital media amplify disciplinary power; the use of digital media carves out the individual as a more identifiable reality, as is evident when we consider the use of databases. Governmentality, by contrast, pertains to a mode of power that produces populations, the body which it works over is more virtual. There is still a political anatomy of detail (which is what discipline is described as), but governing produces from this a very different body with a more virtual presence.
Digital media have a role in governing populations, then, but it is interesting to examine what forms of governing it has opened up. Michael points to the recording of fingerprints and photos in the US immigration line, which all go to a database, and this is of use in governing procedures which protect the US populace from outsiders. Earlier digital machines such as the Hollerith (?) calculation device were invented to process US census data, and fulfilled similar functions.
These tools make visible a specific level of detail that is useful for governing things - a level of detail that operates at the level of probabilities. This now operates at a space that is explained best as a space of variations rather than as a milieu. The more detail we have about something, the more we know about it, and the ability to govern the object we have thus carved out is described by this.
Foucault describes the concept of the milieu as something delineated by four characteristics: it is what we need to account for action at a distance of one body on another - the conduct of conduct. The milieu is not the seat of power (you can't have power, you can only exercise it) - the milieu is where power is exercised, and where results from this exercise emerge. This works on a level of givens, it is not like discipline, but works on an understanding that there are natural givens which those responsible for governing need to take stock of.
A second characteristic is that in working with problems of governing you work on probabilities; the third is that the future is open - you don't fit a model, but aim to discern what is good or bad, dangerous or safe for a population. Many things in a milieu work on many other things, and that is the final characteristic: one thing is the effect of another thing at the same time that it is a cause of another.
These characteristics are necessary to understand the use of digital media and its impact; digital media may magnify these characteristics. The problem of governing is the problem of managing the unknown in a space in which those governing are not responsible for everything that is happening and where certain things cannot be changed. Digital media elicit a scene before the event.
There now is a machinery in place which demonstrates this: high-definition cameras are now installed in Boston's Fenway Park baseball stadium, which detect movements of players and the ball and on this basis statistically record a detailed picture of the game. In the analogue universe, this was possible for the defensive team only by counting the number of times a player fields the ball, and the number of errors (these were the only tools for governing the team); through developing software which uses the data recorded by the cameras, by contrast, a new form of governing the team can now emerge.
This is an analogy for wider population surveillance - so far as actions can be coded through software, this creates, a priori, a profile of the activity which is ready for tracking actual activities, for creating a target that is understood as the population. This describes a space of possible variations, and thus opens up a whole new field of problems - we are able to see before the event, in terms of possibilities and variables, but this is a problem of 'too many normals', of too many things that appear normal, and thus also of what emerges as abnormal. How do we possibly deal with all this detail, then? We now start planning to govern and secure ourselves against things we cannot possibly imagine to happen.