Notes on Formal and Informal Knowledge

Readings:

From Computers and Social Change

Key Concepts for Midterm

  1. formal
  2. informal
  3. algorithm
  4. heuristic
  5. gatekeeper
  6. bureaucracy (details | the_game | game_solution )
  7. rationalization (details)
  8. Taylorism
  9. work to the rule

Formal Procedures

Formal, written procedures can be usually be implemented in computer and communication systems using algorithms. Frequently formal procedures are slow tedious compared to people showing or telling one another something. People who attempt to reduce complex human activity to linear sequences of simple steps can make major errors by omitting some crucial action. The introduction to Jorge Luis Borges' poem La Luna is an interesting comment on the reductionist impulse.

Informal Communications

Informal actions are negotiated in an ad hoc fashion. While they frequently fail to be as efficient as rationalized procedures, they are usually better at dealing with unanticipated problems. They are also subject to abuses like bribery, favoritism, and discrimination. For example, in informal communications, people have to convince a human gatekeeper to let them in. When gatekeeping is implemented in technology, gatekeepers have fewer options for action. Gatekeeping procedures can be frustrating and even coercive.

Formal and Informal Knowledge in Bureaucratic Organizations

The use of computers to rationalize factories and other organizations was anticipated by Babbage in the mid 1800's. Taylorism was a workpace design philosophy introduced in Frederick Winslow Taylor's book The Principles of Scientific Management.

Many of our efforts to computerize work in bureaucratic organizations has produced computer models of inefficient bureaucratic procedures. Often these are less efficient because they remove the possibility of informal communication and problem solving. Work to the rule is a way to bring most organizations to a halt. Yet this is precisely what the worst of our computerized businesses have been designed to do. Even when software algorithms do implement the desired actions, the result tends to be so buggy as to be embarrassing according to computer science theorist Edsger Dijkstra.

perrolle@ccs.neu.edu