Avery seems to have started a little dialogue with me on his weblog, with a series of essays. I'm going to continue writing on this topic because it is (a) interesting and (b) something I like thinking about.
Avery's thoughts have meandered down the path of power, which is great because that's where I wanted to walk down as well. As a Free Software developer, I'm quite aware that the lack of concentration of power is a big problem with Free Software projects. So I ask myself, what is power and how is it concentrated?
Looking at the general business literature, you can classify organisational power into two branches: position power and personal power. The literature distinguishes between individual power and group power, but I'm unconvinced that this is a useful divide, since you can treat a group as being a person suffering from multiple-personality disorder. I'm even going to posit that any group that doesn't act like a reasonable person is ineffectual and has barely any power at all.
Position power is very 陽: it is the power that comes from external sources; power that comes from an office. Examples of this type of power are: authority due to rank and position, the ability to award rewards, the capacity to administer punishments, and the possession of privileged information.
Personal power is quite 陰: it is power that comes from within; an inherent power. Examples are: the ability to persuade people, the possession of superior domain knowledge or judgement, respect and reputation, as well as sheer charisma.
I think most organisations suffer because the mean balance of position and personal power is incorrect. Typically, there will be a bias towards one type of power to the detriment of the other. For instance, in the United States military, position power is very important and permeates the entire structure. Orders are to be followed because a higher ranking officer has issued them. In the Debian project, people must be painfully persuaded that a particular action is correct, and days are wasted arguing points and premises.
These two extremes are quite detrimental because they are merely projections of true power, and not the real thing. To be more concrete: position power without personal power is transitory. Eventually, people are sick of unjustified requests and seemingly random demands. Personal power without position power is ineffectual. Decisions are eventually reached, but may no longer be relevant.
What to do? What to do? Luckily, we are thinking of these two powers in 陰陽 duality. Instead of being conflicting opposites, you can see that no one power can truly exist without the other, because they are shadows of true power. What is a good way of having both?
You use trust! Trust allows you to mix personal power and position power, by using the former to justify the latter. People willingly co-operate with the authority because it is merited and legitimate. Since trust is fleeting and fragile, efforts must be made to assure people that their trust is warranted. You can do this by exposing privileged information and persuading people that decisions were correct, by allowing them to verify the reasoning themselves. Rewards should be given in a way that people agree upon, and punishments effected by the entire group. By encouraging collective confirmation of concentrated power, organisations can harness true power to become more effective.
In Avery's terminology, using true power amortises (but sadly doesn't eliminate) the effect of Quantization on Exclusivity. That is to say that people become more willing to suppress their own personal good for a collective good that is also in their interests. Understandably, this is far more difficult in North American society where individualism is highly prized. In cultures that are strongly collectivist, studies have shown that people are far more willing to come together under a leader towards a common goal.
However, trust is hard work to earn and hard work to maintain. Like a house of cards, it takes tremendous effort to build but a misplaced word can knock it down. That is why people who have been entrusted with power need to continually prove that this trust is deserved. Mistakes made should be quickly acknowledged and corrected, in order to prevent a grave loss of morale.
An important point to bring up here is to draw in here from Avery's discussion is membership control and its effects on groups. The standard way to control membership in a group is to select who can join and who must leave. I'm all for discriminating for traits that are beneficial to the common goal. Allowing anyone and everyone to join leads to an overwhelming mediocrity. But we must beware of naïvely following this rule: "if someone agrees with what you're doing, you hire them. If they disagree, you fire them." You need to examine how this person behaves in the context of the common goal; not just personal interaction. If someone agrees with your goal, but disagrees with you personally, you should strongly consider hiring that person. This individual could give you valuable insight that you might never come up with. If someone disagrees with your goal, but constantly agrees with you, this person is a sociopath and you should fire them. Then get a restraining order.
This doesn't just apply to people in positions of trust trying to manage a group. It's also good advice reciprocally. If you've placed someone in a position of trust, but the circumstances have changed such that this trust is no longer well placed, it is only wise to stop letting this person make decisions. This is especially important in fields that are fast-moving, because for the group to succeed, you can only make so many choices based on false premises before you're doomed to failure. Likewise, if the group recognises someone is especially trustworthy, then give them enough power so that their good decisions can have a positive effect. It is in this way that you keep the leadership of your group effective, because the leaders have motivation to stay relevant and confirm their legitimacy.
By definition, a group that follows this will have effective leaders and a strong centralised power structure. And this allows the singly responsible leader to exert control, because it is demonstrably warranted. Not only is it in the rational best-interest for everyone to follow these decisions, but also the emotional best-interest. This is a very important consideration, because humans are highly irrational beings.
I will have to write further meditations on trust: its effects, advantages and disadvantages.