User:Teratornis/How Wikipedia defeats Brooks' law

Wikipedia appears to defeat Brooks' law, namely: as more people join the Wikipedia project, Brooks' law predicts their average efficiency should decrease. Yet this does not appear to happen. As Wikipedia grows ever larger and more complex it may actually be growing more efficient.

To-do: explain why I think this is so.

Some working references
This section will go away as I develop this essay. These are just some links and notes to cement the ideas I will develop.


 * Brooks' law - well, of course.
 * User:John Broughton/Editor's Index to Wikipedia - the antidote. Wikipedia is a tool for writing documents, and documents are the antidote to Brooks' law. Wikipedia attracts lots of documenters, who naturally proceed to document their way out from under Brooks' law. In contrast, very few programmers like to document, so it is no wonder that they usually fail to self-organize similarly on software projects (hence, Brooks' law).
 * WP:HD (permanent link) - a comment on why I consider template substitution harmful in a Brooks' law sense (which is not to dismiss other arguments in favor of substitution, but merely to point out the potential cost).
 * Avoid instruction creep - an essay which contains some valid points, but misses the Brooks' law point entirely (also see my comments on the talk page).
 * To-do: complement (i.e., render more complete) the above essay with my "do not fear complexity" ideas. That is, if a system B is "better" than a system A, B is almost always more complex than A. Examples: a Boeing 747 is "better" than a Wright Flyer, and far more complex; an iPhone is "better" than an old touch-tone phone, and far more complex; Linux of today is "better" than CP/M of the 1980's, and far more complex; the human brain is "better" than the nervous system of a cockroach, and far more complex; get the picture?
 * Sometimes a simpler system can be better than a complex system, for example if the complex system is poorly designed or riddled with obsolete relics. However, to improve a system that is already well-designed, usually you have to add complexity. More is more.
 * Discuss how people learn (through repetition, as understood by people as different as Adolph Hitler, Daniel Dennett, drill sergeants, and any advertising agency), and the common misconceptions about learning ("people don't read instructions").
 * Find a source for Dennett's claim: "Every time you read or hear something, your brain makes another copy of it."
 * Wikipedia (like virtually all Open Source Software projects) relies on remote collaboration. Since the participants are not concentrated into one geographic location, they cannot rely on meetings and spoken language as traditional bricks and mortar organizations do. This forces participants to rely almost exclusively on written language, which has the important side effect of building up a knowledge base that newcomers can then easily read, without requiring the more-experienced participants to recite knowledge to them in the form of an oral tradition. That is, Wikipedia relies on literate communication rather than the pre-literate tribal communication that predominates in many if not most non-scholarly human ventures. Of course tribes that lacked written language were very limited in what they could do, as the oral tradition could not cope with much complexity.