The Politics of Legitimate Governance



An Open Letter to the Membership Committee of the Open Source Initiative .

Re: Membership Policy

Let me take a step back and try to articulate why I think this issue of membership is coming up now, in the context of larger challenges facing the OSI. Caveat: I am not by training or position a political scientist, though I do play one on the Internet. :-)

Roughly speaking, there are three ways communities can make legitimate decisions:

a) Informal consensus
b) Formal voting
c) Designated autocrat

Of course, there's numerous variations and combinations of these, but this should suffice for our purposes. At the risk of oversimplifying even further, I would add that:

 a) Consensus works when the community can socialize a shared understanding of the problem

 b) Voting works when there exists mechanisms to ensure "fair" representation across winners and losers

 c) Autocracy works when the person(s) in power are trusted by the community for the task in question

Of course I'm sure you could all find loopholes in these admittedly vague assertions, but hopefully it gives a flavor of the central issues involved, at least from my perspective.

Why does this matter? I believe that the OSI worked so well in the past because:

 a) The OSD did in fact reflect a broadly-shared community understanding

and

 c) The community -- based on past experience with the OSD -- trusted the OSI to manage that wisely

That is, as long as the OSI stayed within community norms, on a topic for which they had credible expertise, (almost) everybody was happy.

However, what has changed -- and the reason I got 'bounced' here from license-discuss -- is that the OSI decided it needed to be proactive in an area where there was *not* a pre-existing community consensus.

Were they right? I don't know. Yes, there *is* consensus that license-proliferation is "a" problem. However, I don't believe there is a *broad* consensus that:
i) it is an *urgent* problem 
ii) the OSI has characterized it properly
iii) it is related enough to OSD issues that OSI should/must take it on
iv) the proposed solution will be effective
v) the benefits of the solution outweigh its costs

I don't want to debate those issues here (that's what I promised to do on license-proliferation-discuss :-), but merely point out that the OSI does NOT have (currently) an effective mechanism for determining whether their position does in fact reflect community understanding. Clearly, the reaction on license-discuss seems ample proof that a consensus does not (yet) exist. In that case, what are the OSI's options? I see only four:

   1. Admit that it cannot deal with the problem, since there is no consensus, and give up

    2. Invest significant time and effort in developing a consensus -before- enforcing a strong stand

    3. Establish a fair voting mechanism, to determine the community's will (even, perhaps especially, if divided)

    4. Decide the issue is too important to wait for consensus, and set itself up as an autocrat

Anything else? I don't think so, but I'm willing to hear realistic alternatives. But for now, allow me to assume this is true and move on.

Within this framework, I suspect that the OSI Board felt that their actions were in line with #2: i.e., their perspective was "close enough" to the community consensus, that the community would naturally recognize that these decisions were in the community interest, and a natural extension of their existing trust, and not take offense.

Alas, if so, I feel they gravely misjudged the situation. Not maliciously, I hasten to add, but severely. In fact, the trust about the OSI's ability to correctly perceive the 'community sense' behind the OSD only *partially* transferred to their actions on license proliferation. Worse, the limitations (which, to be fair, may be technical rather than political) in how it was communicated further exacerbated the problem. With the end result that a vocal minority of the community (at least of those paying attention) are worried that the OSI Board is acting like #4 (autocracy).

What's the solution? Is membership? To be honest, I don't think so. A vote, by its very nature, implies winners and losers, and counting noses. Give the size and nature of our community, I honestly can't imagine any rational scheme for membership that wouldn't allow a particular faction to "stuff the ballot box" to the detriment of others -- especially on a issue that differentially impacts various populations.

Frankly, I believe the only viable option is to pursue consensus, by adopting a more formal, transparent, feedback-oriented communication process (a la #2, with a fallback to #1). Anything less, IMHO, would tear the community apart

I think one of the problems is that the original board was too busy, or didn't have the right skills, to maintain ongoing involvement. Some organizational knowledge was lost, and the new Board (IMHO) didn't understand what all was at stake. A healthier mechanism (again, IMHO) would be for the current Board of Directors (who're doing the actual work) to be required to consult with an "emeritus" Board to validate policies before presenting them to the public. And similarly to consult with the public before implementing those policies.

In particular, I would recommend that, in order to increase their perceived legitimacy, the OSI:

I. Suspend the current policy on anti-proliferation impacting license approval; not as a "bad" idea, merely as "poorly formed"

II. Develop a public process for how new policies will be drafted, aired, and implemented
 (including things like a Slashdot posting and public comment period)

III. Establish a Board of Advisors, consisting of diverse, high-profile, well-respected community "elders"

IV. Better publicize the membership of the Board of Directors, including the criteria and process for their election

I realize that may be a lot of work, but frankly I don't see how the OSI Board can hope to simultaneously:
 a) pursue an activist agenda, and
 b) maintain legitimacy
if they do NOT do something like this.

Anyway that's my two cents -- perhaps this will get a reaction. :-) I welcome your feedback.

Best wishes,
Ernest N. Prabhakar, Ph.D.
[speaking only for myself]

Posted: Sat - October 1, 2005 at 02:14 PM        

©