My Response to "Galbraith Games Anthropology"

I am not familiar with the work of Barry Oshry, so this response is based on Ben's description. I strongly endorse the idea of scientifically studying our organization instead of simply trying random things based on untested theories. However, I doubt that the Log-based approach Ben promotes can do that effectively.

The Log is Not Effective Science

Ben says the "Anthros" wrote down everything that was ever said or done and used this information to set up more effective communities. Through the Log, Ben hopes to gain some of the benefits of this system. I don't think the Log provides any benefits for several reasons:

• The logged events are too general to use as experimental data
• The log only records the least useful data about the events
• It does not record details (actions, words) which may be important
• It does not include many important events (spontaneous meetings etc.)
• Filling those gaps requires more participation from every member

We can afford to have one or more members spend ALL their time studying our organization and propose changes - if the results of accepted changes are more valuable than the effort put into the study. We cannot afford to have one or more of our members spend SOME of their time studying the organization if it will not produce results.

If all of us got involved in logging our activities, we would have a very complete record of major events (mostly meetings.) This would require much more work and we would still not have the detailed, useful data which could be used to scientifically improve our organization.

On the other hand if Ben worked full time as an "Anthro" studying Galbraith Games, he would have to commute quite a long distance and we would have to do all of our work in a central location where he could record what happens.

Recommendations

I believe the Log is simply ineffective. One positive case of the Log data being used to improve our organization would disprove that theory, but until the theory is disproved, the Log should not be promoted as a way to scientifically study our organization.

I believe that more effective methods of studying our organization are possible.

As long as we have severely limited contact and manpower, I think the most effective methods will involve collecting only the most essential information. This document discusses this problem - the signal-to-noise ratio - in more detail.

A Sketch of each specific job that we do would provide some useful information that the Log does not: The source and purpose of an activity, as well as detailed requirements and ongoing progress.

What methods can you think of?