[EM] Trees by Proxy

[Abd ul-Rahman Lomax]

At 03:47 AM 3/21/2007, Dave Ketchum wrote:
>On Tue, 20 Mar 2007 23:52:45 -0400 Abd ul-Rahman Lomax wrote:
>Voters in a village give proxies to elders in their 
>village.  Village elders give proxies to town trustees in their town.
>
>While holding one effective voter proxy directly makes one an elder, 
>and holding one indirectly is enough to make one a town trustee, 
>there have to be limits on quantity of active elders or 
>trustees.  Anyone holding too few can combine with others to create 
>one active member (much like proxy, but I avoid the word on purpose):
>      Very few, and no participation.
>      Few more, and can vote in government, but not be more active 
> in meetings, etc.
>      Enough to make one a dictator, and voting power is limited to 
> prevent this.
>
>While village and town give the ideas framework, more levels make 
>sense, and sizes could be tailored as makes sense.

Here is what I'm suggesting be looked at: if every member of a 
society, who cares to participate, names a proxy, and if proxies are 
considered delegable, one can take a proxy list and analyze it to 
determine a hierarchy of connections. There will be some loops, but 
fewer than one might expect.

Neglecting loops -- or presuming that they are resolved in some way, 
which happens when any member of the loop selects a proxy outside the 
loop, thus integrating the whole loop with a larger structure -- such 
a system can be seen to devolve into representation in a group of any 
desired size. Including, perhaps, down to one person who is a 
superproxy. More likely, since we do not wish to coerce proxy choices 
in any way, it is essential that they be totally free, subject only 
to mutual consent between proxy and client, there will appear what we 
call "natural caucuses." How many natural caucuses will appear we 
don't know, but there might not be that many of significant size. 
Naming a proxy doesn't necessarily signify full approval of 
everything the proxy does; where a representative body has a limited 
size, a caucus which isn't of sufficient size -- or otherwise 
qualified -- for direct participation in the body can name a proxy 
from among those who will have participation rights, or may join with 
other such unrepresented caucuses to create a common participating proxy.

I suggest that such bodies, I often simply call them "meetings," may 
make their own rules. These rules define who has participation 
rights, among other things. The rules may include, I'm noting, that 
all extended members have the right to vote, but may limit "floor 
rights," i.e., the right to address the assembled meeting or to enter motions.

I'm describing a self-organized structure that is created from the 
unconstrained choices of members. Many people thinking about DP for 
the first time start to imagine this or that rule which should be 
imposed. A common one is limiting the number of proxies which can be 
held by a single individual.

I can understand this in the governmental applications of delegable 
proxy, but it is utterly unnecessary in the Free Association context. 
Part of what the FA context is about is to serve as a test bed for 
delegable proxy.

Within the FA context, if all members wish to directly or indirectly 
agree to allow a single person to represent them, what's the harm? 
Or, more accurately, who is to say that this is harmful? Trying to 
set a rule is an a priori restraint on the freedom of the members. 
The fact is that a single person in a large FA actually functioning 
to represent everyone is only likely to happen in some kind of 
emergency. Otherwise, there will be plenty of people participating, 
and when people participate directly, their proxy assignment is not active.

In the village representation problem, process within the village 
would result in a small number of people, ordinarily, who, 
collectively, represented the entire village. If it is practical for 
all of them to participate at a higher level, fine. If not, then 
somebody, somewhere, will have to compromise or won't be represented. 
And that is what would happen: the system does not pretend that 
someone is represented who has not chosen an active representative 
(or who is not directly participating). There will always be some 
unrepresented people, for various reasons.

Suppose the voting power of U.S. Congresspeople depended on the votes 
that they actually received.... instead of them having power 
proportional to the district populations they represent. Might 
encourage high turnout, don't you think?