[EM] [MG] Helping the Pirate Party to vanish

[Abd ul-Rahman Lomax]

At 05:31 AM 3/18/2013, Paul Nollen wrote:

>>-----Original Message----- From: Kristofer Munsterhjelm
>>Sent: Monday, March 18, 2013 11:00 AM
>>To: Paul Nollen
>>Cc: Metagovernment Project ; 
>>Meinungsfindungstool ; Votorola ; Comunicación 
>>; Election Methods ; AG Liquid Democracy
>>Subject: Re: [MG] [EM] Helping the Pirate Party to vanish
>>
>>On 03/18/2013 09:58 AM, Paul Nollen wrote:
>>>Liquid democracy is tested for many years in every big (and small)
>>>corporation. It is unthinkable that shareholders have the obligation to
>>>give their voice for more than one General assembly to anyone. Every
>>>shareholder can vote for himself or appoint a representative at his
>>>choice only for that dedicated General Assembly. This system of "liquid
>>>democracy" is proven over many years all around the world.
>>>It is only in politics that voters are forced to give a mandate for many
>>>years for decissions unknown.
>>
>>The way I understand liquid democracy (and correct me if I'm wrong),
>>people can give their votes to proxies, and these proxies can in turn
>>give their votes to other proxies;
>
>- Multiple proxies is, or has to be, an 
>option.  There is nowhare any rule that liquid 
>democracy has to allow indirect proxy.

If I say that the tail of a dog is a "leg," how many legs does the dog have?

The standard answer to the riddle is being, 
"Four. Because I say a tail is a leg doesn't make it one."

I had been working on delegable proxy for long 
before the usage "liquid democracy" came into 
being, but it was published before I published 
anything, in the 1990s, as I recall. The concept 
was independently invented in at least a 
hafl-dozen places around the world, over roughly 
a decade. Delegability is essential to the *new* 
concept. Yes, it's based on old concepts, and in 
some cases, traditional proxies could be 
delegable in some sense, but that was not expected or reliable.

The current Demoex parliamentary rep has a blog 
and is writing a book, in which he suggests that 
Wikipedia is a fine example of democracy. It is, 
in fact, a fine example of how direct democracy 
can fail. In any case, currently, if you load the 
link, you are redirected to the article on 
deletaged voting. However, that is the result of 
an anonymous edit -- and nobody who has a clue was paying attention.

http://en.wikipedia.org/wiki/Liquid_democracy

Until a few months ago, and almost four years, the page redirected to

http://en.wikipedia.org/wiki/Proxy_voting#Delegated_voting

There was originally an article on Liquid 
democracy itself. It was moved to the title of 
"Delegable proxy" in 2008. That page was deleted, see
http://en.wikipedia.org/wiki/Wikipedia:Articles_for_deletion/Delegable_proxy

The content, though, was moved to 
http://wiki.electorama.com/wiki/Delegable_proxy 
It was never properly reformatted for that wiki.

If you want a clue as to what *really* happened 
back then, see 
http://en.wikipedia.org/wiki/Wikipedia:Delegable_proxy. 
This proposal was merely for an experiment, 
setting up a proxy table system, it proposed no 
changes in policy and did not implement proxy 
voting. But it would have made proxy analysis 
possible. The rejection was violent, almost 
visceral. The proposer was rapidly blocked; he'd 
been a long-term editor who had a habit of 
changing names, but I checked at the time. His 
previous incarnation had been an editor in good 
standing, for years. He was indefinitely blocked 
for offenses that might have raised an eyebrow 
for an ordinary editor, and it all came down very 
rapidly once he proposed delegable proxy structures.

I don't frequently use the term "liquid 
democracy," but ... here are some old pages.

http://www.communitywiki.org/LiquidDemocracy 
seems to be from 2004, but the page history 
indicates an earlier version existed in 2003.
http://www.communitywiki.org/en?action=rc;all=1;from=1;showedit=1;rcidonly=LiquidDemocracy

Soemthing from 2003, specifies delegation:

http://ming.tv/flemming2.php/_d10/_v10/__show_day/_w2003-05-15#000010-000797

This seemed to be from 1999, and it has a dead 
link to a page by James Green-Armytage, who I 
consider one of the various independent inventors of delegable proxy.
http://maparent.ca/Why_online_tools_for_public_deliberation_

This page doesn't discuss delegation. Apparently 
the concept of ordinary proxy representation is 
so radical for some people they don't understand 
it's been around for centuries. I now think that 
this page was not from 1999, but later.

However, here is a link to a paper written later by James G-A:

http://www.econ.ucsb.edu/~armytage/voting/proxy.htm#introduction

He explicitly discusses delegation, and he calls 
the method "delegable proxy." I think he may have 
gotten that usage from me, but may have been 
working on the concepts as "liquid democracy" or 
proxy voting long before we had contact. I'm not 
certain. I've also called it "fractal democracy," 
and some people like that name, because the 
organizational structure is a fractal. A large DP 
structure will self-resemble at all scales, at least roughly.

I don't know if I could find the first mention of 
liquid democracy. Many old pages are gone, and 
there is no way that I know of to search the 
Internet Archive. But I did find this page in the 
Archive, from a dead link elsewhere: 
http://web.archive.org/web/20070930030223/http://wiki.uniteddiversity.com/tiki-index.php?page=LiquidDemocracy

This, which may have been from 2004, though it 
was captured later, refers to "chained answer 
recommendation." I.e. the author is thinking of 
what I call the Free Association concept, and has 
the idea that this is for advice The "chained" 
means that the designations are delegable. The 
structure is a fractal. Another way to say it, 
mathematically, is that it is a directed graph.

Many, many sources I saw in this recent search 
clearly refer to delegation. Some merely refer to 
proxy representation. But proxy representation is 
not a new idea at all, in fact, that was Paul's 
point, except that, instead of simply noting that 
proxy representation is an old concept, he 
claimed that "liquid democracy" was old. I'm 
going to make one attempt to find a truly old mention.

This page version was last modified in 2002:
http://web.archive.org/web/20020225024957/http://twistedmatrix.com/users/jh.twistd/python/moin.cgi/LiquidDemocracy 
and it links to

http://web.archive.org/web/20020826094730/http://twistedmatrix.com/users/jh.twistd/python/moin.cgi/LiquidDemocracyVotingSystem

I really do think this guy may have been the 
first to use the term "liquid democracy." He 
calls the system scalable, but the way he 
describes it doesn't take advantage of the 
scalability of delegable proxy, and, in fact, he 
tries to address the problems that he can see would arise.

 From my perspective, looking back at a decade of 
massive discussions of delegable proxy, he's got 
some aspects of the Free Association concept, but 
only a primitive idea of how to actually make the 
structure scalable. Without having any present 
need, he's already trying to write code. He's 
proposing complex solutions to problems that don't exist yet.

So ... what is "liquid democracy." I'd noticed 
images along the way. Here are some, all returns 
on a search for images for "liquid democracy."

http://www.karl.aegee.org/lf/ This one uses, 
primitively, what I'd use to graphically display 
a proxy table, concentric circles. Initially, all 
clients are arranged around a first circle. When 
a client names a proxy, the proxy is moved to a 
second smaller circle, opposite the client. If 
another client names that proxy, the client is 
moved around the circle to be next to the first. 
If a proxy and client name each other, they are 
moved to that inner cricle, adjacent. If another 
person names the first client as a proxy, that 
ooriginal client moves inward to the first inside 
circle and then the original proxy moves to the third, and so on.

When the graph is complete, the circle represents 
what I've called the "proxy rank." Essentially, 
it is the number of people the proxy represents 
if nobody else shows up. So the outer circle is 
circle 1. The positions of the individuals are 
arranged around the circle so that proxy-client 
lines don't cross. So as information is 
transferred from client to proxy, it is 
concentrating, and when it is transferred from 
proxy to client, it is spreading.

So, one problem. Loops. A two-person loop is easy 
to display, but a larger loop could complicate 
the graph. I'm not solving that here.

http://agsm.fsk.uni-heidelberg.de/index.php/Datei:LD-Grafik.png 
shows only one delegation.
http://www.ruprecht.de/no_cache/nachrichten/archive/2010/june/19/article/demokratie-20/ 
shows multiple layers of delegation.
http://commons.wikimedia.org/wiki/File:Democracy.svg 
comes from the Common Good Bank, a local organization here.
http://www.tageswoche.ch/de/2012_15/schweiz/414125/die-noch-direktere-demokratie.htm

*On the other hand*
http://aktivdemokrati.se/wp-content/uploads/2010/12/enghumans.jpg 
does not show delegation. Delegation, by the way, 
doesn't necessarily become important until the 
organization becomes large and traffic starts to 
overwhelm proxies. That's when proxies would 
start to recommend people move their assignments 
to a client of the proxy. (or elsewhere, but that 
would be for other reasons .... i.e., they don't 
want to communicate with the person. Proxy 
assignments should be *mutually* agreed upon.

The original coiner of the term, if I've 
identified this person correctly above, didn't 
think to this level, but did realize that there 
could arise a problem of excessive traffic, he 
handles it by having the proxy start to charge 
for "advice." Indeed, that may happen. But such a 
highly trusted individual would suggest that some 
of her clients take on some of her proxies, and 
might suggest a small charge as standard, and the 
clients would keep some and pass some on to her. 
Yes. This could become an occupation.

These are -- must be -- *voluntary relationships* 
and can be covered by special and personal 
agreements. The system doesn't care about that, 
but, to be scalable, it *must* allow delegation. 
That's what makes the structure into a fractal, 
like the human brain. Can you imagine what 
thinking would be like if every neuron could not 
pass a signal on through other neutrons? Basically, it would not exist.

Here is how I'll resolve this. Liquid democracy 
proposals and discussions clear include, 
routinely, transitive delegation of 
representation. Obviously, a particular structure 
at a particular time may not be using delegation. 
However, if a structure does not use delegation, 
it's not anything new, it's simply direct 
discussion, ordinary democratic process, or is a 
step up with recognizing proxy representation. 
The original liquid democracy proposal did not 
contemplate delegation, which is why it got 
complex for the proposer, he claimed it was 
scalable but missed how delegation would make 
that actually possible and practical.

Wikipedia broke down because not only did it not 
allow proxy representation, it did not understand 
the need to manage traffic, so discussions becoem 
unwieldy if deep and inadequately considered if 
kept to sound-bite dimensions. In frustration, 
blocking and banning were used to manage traffic.

>>whereas in corporations, the proxying
>>happens only once. That is, either you vote or you say "X will vote for me".
>>
>>(I imagine that if indirect proxying had been possible in corporations,
>>we'd heard more about cycles. But again, I could be wrong. I've never
>>been to a general assembly.)
>>
>>Also, for corporations in general, there are arguments that the real
>>power resides as much with the board as the assembly. This is given as
>>an explanation for the weird incentive structures that often appear,
>>with enormous bonuses given to executives even when the companies
>>struggle or fail. The recent Swiss Minder referendum can be seen in that
>>light.
>>
>>In effect, that argument goes that the board can more effectively
>>exercise power than the general assembly can counter it, so the real
>>decisions are made by the board. If this is true, then having a more
>>corporate system in the political domain may not be such a good thing.
>
>- I only stated that "liquid democracy" existed 
>as a system already for a long time.

Only if by liquid democracy you mean ordinary 
proxy. Mostly the term is used for something new. 
Now, this much is true. The advice-chains that 
have been mentioned in liquid democracy 
organizations are *very* old, ancient. It's how 
human society actually works, but liquid 
democracy and delegable proxy formalize it, and 
much becomes possible with the formalization, including scalability.

>>On the other hand, perhaps the boards are powerful only because the
>>general assembly meets so infrequently. As an imperfect analogy: "you
>>can accomplish lots of things even in a prison if the guards only look
>>on you once every year". And if *that*'s the reason, then liquid
>>democracy could work in a political setting - at least for advisory
>>purposes.
>
>- That depend of the corporation ofcourse. I can 
>imagine that in a cooperative like Mondragon the 
>situation is different than in, let's say for example, General Motors.

Well, roles are filled in different ways. 
However, if General Moters were to run an ESOP 
and if the ESOP came to own a majority of the 
stock, the operations might converge.

>- if people are not granted decisive power it 
>doesn't interest me, but that is my personal opinion of course.

The people *have* decisive power, so who is going to "grant" it to them?

Here is what I suspect you have not considered. I 
think you have in mind some way in which "the 
people" make some collective decision, some 
*mechanism.* For example, they cast votes, and a 
clerk certifies the vote, and an executive goes 
out and acts based on the vote. But how does the 
clerk certify the vote? Well, in a standard 
deliberative body, used in power structures, 
questions are confined to Yes/No answers, and 
proceed through a process that allows unlimited 
amendment, and eventually the rules allow a vote 
on the question, and a majority is required for a 
decision, or else the whole process must start 
over. And that's how deliberative bodies ordinarily work, all over the world.

However, when the organization increases in size, 
the necessary discussions get rapidly more 
cumbersome, even impossibly so. Above a fairly 
small number, organizations start to delegate 
early deliberation to committees. And when the 
business of the body gets complex, officers start 
to control it. Instead of considering this an 
abuse of power, as proponents of direct democracy 
think, it's simply a means of handling the 
traffic distributing the load. Simply returning 
to direct democracy does not address the problem, it recreates it.

In very small organizations, with training, 
people can often find 100% consensus. It's work, 
but it can also be exhilarating. On the other 
hand, doing this week in and week out for years, 
as in cohousing communities that have tried it, 
is still exhausting, and fewer and fewer people 
participate. There is an obvious solution, proxy 
representation. But that's often rejected in 
these communities because they think of proxy 
voting as *directed voting.* It certainly could 
be, but a directed voter cannot participate in 
the process that can lead to consensus.

As organizations become very large, they 
problably must abandon any *demand* for complete 
consensus, and even modest organizations that 
require fulll consensus end up, in my experience, 
seeing that bring minority rule, if the status 
quo favors a stubborn minority. I've seen it happen.

So I've concluded that the majority does have the 
*right* of decision, but that maximizing 
consensus remains always desirable. Hence the 
importance of structures that can efficiently 
*negotiate* consensus on a large scale. That's 
delegable proxy. Other approaches can do it, but 
DP is *so simple* that it deserves testing. I've 
seen it work on a small scale, but it's 
intrinsically scalable. It *should* work for the 
entire population of the earth, it merely takes 
longer for messages to travel through the 
network, a few days per traverse. (System 
messages, centrally broadcast, of course, can be 
practically instantaneous, but there are very 
good reasons to use those only in an emergency. 
Central broadcast means, intrinsically, central 
control. So we want to limit the usage of that.)

>>I also note that vote-buying would make less sense in a corporation
>>since voting power is given by wealth, in terms of numbers of shares
>>held, in the first place. And then I wonder if there have been instances
>>of vote coercion on shareholders.
>
>- yes, in corporations it is not "one man one vote" but that is another issue.

It's simply one-share, one-vote, and there are 
corporations where members are limited to a single share.