Proof of Lemmingism

3 views
Skip to first unread message

Mark

unread,
Aug 10, 2006, 7:03:09 AM8/10/06
to top-politics
Hey everyone,
here is proof of lemminism among the masses:

http://www.livescience.com/othernews/060124_political_decisions.html
-------------------------------------------------------------------------------------------------
Democrats and Republicans Both Adept at Ignoring Facts, Study Finds
By LiveScience Staff
posted: 24 January 2006
10:03 am ET

Democrats and Republicans alike are adept at making decisions without
letting the facts get in the way, a new study shows.

And they get quite a rush from ignoring information that's contrary to
their point of view.

Researchers asked staunch party members from both sides to evaluate
information that threatened their preferred candidate prior to the 2004
Presidential election. The subjects' brains were monitored while they
pondered.

The results were announced today.

"We did not see any increased activation of the parts of the brain
normally engaged during reasoning," said Drew Westen, director of
clinical psychology at Emory University. "What we saw instead was a
network of emotion circuits lighting up, including circuits
hypothesized to be involved in regulating emotion, and circuits known
to be involved in resolving conflicts."

Bias on both sides

The test subjects on both sides of the political aisle reached totally
biased conclusions by ignoring information that could not rationally be
discounted, Westen and his colleagues say.

Then, with their minds made up, brain activity ceased in the areas that
deal with negative emotions such as disgust. But activity spiked in the
circuits involved in reward, a response similar to what addicts
experience when they get a fix, Westen explained.

The study points to a total lack of reason in political
decision-making.

"None of the circuits involved in conscious reasoning were particularly
engaged," Westen said. "Essentially, it appears as if partisans twirl
the cognitive kaleidoscope until they get the conclusions they want,
and then they get massively reinforced for it, with the elimination of
negative emotional states and activation of positive ones."

Notably absent were any increases in activation of the dorsolateral
prefrontal cortex, the part of the brain most associated with
reasoning.

The tests involved pairs of statements by the candidates, President
George W. Bush and Senator John Kerry, that clearly contradicted each
other. The test subjects were asked to consider and rate the
discrepancy. Then they were presented with another statement that might
explain away the contradiction. The scenario was repeated several times
for each candidate.

The brain imaging revealed a consistent pattern. Both Republicans and
Democrats consistently denied obvious contradictions for their own
candidate but detected contradictions in the opposing candidate.

"The result is that partisan beliefs are calcified, and the person can
learn very little from new data," Westen said.
---------------------------------------------------------------------------------------------------------------

MG

unread,
Aug 30, 2006, 3:30:58 AM8/30/06
to top-politics
Hello, I'm back but very busy running for elections. Will attend
Swedish TV3 on thursday..
To the science...:
This report is very good proof of that to most people, it is more
important WHO is saying something, than WHAT he says.
My solution for this is to de-identify the arguments as much as
possible and to have as many voices as possible, your is to promote
certain speakers as better than others?
When there are enough voices in the room, this stupid behaviour of not
reasoning to some voices will become impratical. It is only possible to
withstand as long as there are 2 or 3 easy identidied individuals to
hack on.

Mark

unread,
Aug 31, 2006, 12:20:48 AM8/31/06
to top-politics

>mG: This report is very good proof of that to most people, it is more

> important WHO is saying something, than WHAT he says.

-M: My main point from the beginning.

>mG: My solution for this is to de-identify the arguments as much as
> possible and to have as many voices as possible,...

-M: Yes, a multiplicity of lemming voices where the non-lemmings are
drowned out.
How is this a solution?

>mG: your is to promote certain speakers as better than others?

-M: Yes, better for political decision making.

shanti
Mark, Seattle WA USA

Abd ul-Rahman Lomax

unread,
Aug 31, 2006, 11:48:56 AM8/31/06
to top-po...@googlegroups.com
At 12:20 AM 8/31/2006, Mark wrote:
> >mG: My solution for this is to de-identify the arguments as much as
> > possible and to have as many voices as possible,...
>
>-M: Yes, a multiplicity of lemming voices where the non-lemmings are
>drowned out.
>How is this a solution?
>
> >mG: your is to promote certain speakers as better than others?
>
>-M: Yes, better for political decision making.

Mark and I agree on this. It is possible to do it in a way that
depersonalizes it, there is discussion going on on the Election
Methods list about this at this time, but Mark's method -- about
which I know little -- and the methods proposed by BeyondPolitics.org
both involve adding weight to certain "speakers."

Delegable Proxy accomplishes that by setting up a network based on
the assignment of trust by general members, though the naming of
proxies by members.

FA/DP (Free Association with Delegable Proxy) is generally conceived
to allow direct voting by all members; however, it is expected that
most members in a large organization will, instead, allow their votes
to be cast, routinely, by proxies. Further, allowing direct vote is
not the same as allowing every member to directly address the
assembly. There *must* be filtering systems in place. In an FA/DP
organization, a high-level meeting (mailing list, web site,
face-to-face) that has reached a scale where full participation is
impossible, one option for limiting participation to keep traffic
manageable, to keep the noise level down, is to restrict full
participation by those who enjoy broad trust. DP measures this by
considering proxies delegable and by thus concentrating proxies as
necessary to meet the standards of any particular meeting.

The FA context must be understood: FAs don't exercise power. The
kinds of political action, direct-legislative systems that some are
working on cannot be FAs. These power control systems could use DP,
but, in my view, we have insufficient experience to, in the
near-term, attempt to run them with DP or other similar or automated
system, and, besides, the existing oligarchy will strongly resist it,
we can predict. And until we have mass communication systems capable
of reaching broad consensus, we will not be able to overcome this obstacle.

But FAs can begin now. They are purely communications networks,
really, though communication can drive control where consensus can be
obtained through it. FAs don't have to address the security concerns
that can loom quite large with actual power control systems.

In particular, where votes move leveraged assets, the action points
are vulnerable to corruption. Where the assets remain distributed
among the members, corruption becomes much more difficult. There is
still the problem of media corruption, but DP distributes media
control among the members. Frauds may attempt to simulate memberships
and thus gain "rank" allowing them privileged access, but in an FA,
meetings may make their own rules. A meeting may decide to disallow
access by any member, for any reason whatever. I presume standard
Robert's Rules or an on-line equivalent, but meetings are completely
free to make their own rules. In an FA, meetings are rigorously
independent. The central organization merely facilitates connections.

FAs, in my experience, can bring people together beyond the
artificial boundaries that normally separate us. Combining the FA
experience (what I call FA is not terribly uncommon, just not much in
the public eye) with Delegable Proxy should make the
consensus-building possible in small FA groups accessible to large
groups, without imposing top-down oligarchical structure.

FA/DP does create a kind of elite, and thus it can be called
"elitist," but this is misleading. The elite in an FA depends
continuously on the successful maintenance of trust in it by the
general membership. Proxies have no term of office, they serve, in
the language of employment, at the will of the employer. The client.
The member.

Given that it is extremely simple, involves no significant expense,
and, in the FA context, it about as safe as could be imagined, I'd
say it is worth a trial.

"If we want to save the world, it has to be easy."

Because if it is not easy, it is *not* going to happen. If it is
hard, it takes centralized power, and centralized power, we have seen
all too often, has a way of institutionalizing itself and those who
hold key positions in it.

FA/DP is libertarian/anarchist, but, and this is a crucial
distinction, it is not based on any ideology, it does not promote, in
itself, libertarian/anarchist forms of government. An FA, quite
specifically, cannot be a government. But it can advise governments,
just as it can advise citizens as to how to apply their power. It
does so, not by centralization of advice, but by massive
decentralization of advice. Proxies, in the language of one Swedish
writer, are "advisors." To emphasize the bidirectional nature of
proxies, it has been proposed to call them Advisor/Proxies (AP).

Mark

unread,
Sep 1, 2006, 5:59:51 PM9/1/06
to top-politics

> Mark wrote:
> > >mG: My solution for this is to de-identify the arguments as much as
> > > possible and to have as many voices as possible,...

> >-M: Yes, a multiplicity of lemming voices where the non-lemmings are
> >drowned out. How is this a solution?

> > >mG: your is to promote certain speakers as better than others?

> >-M: Yes, better for political decision making.

>L: Mark and I agree on this. It is possible to do it in a way that
> depersonalizes it,...

-M: Why (attempt to) depersonalize it? What does this do?

>L:...there is discussion going on on the Election


> Methods list about this at this time, but Mark's method -- about
> which I know little -- and the methods proposed by BeyondPolitics.org
> both involve adding weight to certain "speakers."

-M: SD2 (the umbrella system) gives weight to generalists.
SD2-Smartocracy(system with my prefered defaults):
1. Measures popular consent, and uses popularity thresholds to maintain
issues within deliberation if need be.
2. Weights specialists on specific issues.
3. Defaults to generalists if specialists aren't selected by the voter.

I challange anyone to tell me why this isn't ideal.

>L: [...]There *must* be filtering systems in place.

-M: :-)

>L: But FAs can begin now. They are purely communications networks,


> really, though communication can drive control where consensus can be
> obtained through it.

-M: There is either control or not.
If there is no control, its just a boring FA.
If there is control, then the *former* FA can now render
*institutional* opinions.
It is now an *entity* with life of its own.

>L:[...]security concerns [...]where votes move leveraged assets, the action points are vulnerable to corruption. Where the assets remain distributed among the members, corruption becomes much more difficult.

-M: The best of both worlds is to have:
1. Popular consent measured, and decisive over deliberation.
2. Have the positions of leverage very fluid - those of high rank can
lose rank quickly, if need be.

>L:[...]with Delegable Proxy should make the consensus-building possible in small FA groups accessible to large groups, without imposing top-down oligarchical structure.

-M: Centralization is what *works* at being *effective*.
Do you want *ineffectiveness*?

>L: FA/DP does create a kind of elite, and thus it can be called


> "elitist," but this is misleading. The elite in an FA depends
> continuously on the successful maintenance of trust in it by the
> general membership.

-M: As with SD2.

>L: Proxies have no term of office, they serve, in the language of employment, at the will of the employer. The client. The member.

-M: Sounds linear. What about loops?
By contrast, with SD2, people are differentiated by rank.

>L: Given that it is extremely simple, involves no significant expense,


> and, in the FA context, it about as safe as could be imagined, I'd
> say it is worth a trial.

-M: Why try something that already appears outdated?

>L: "If we want to save the world, it has to be easy."


> Because if it is not easy, it is *not* going to happen. If it is

> hard, it takes centralized power,...

-M: What other option is there? Centralization is what works, and any
effective change done by decentralized systems will be much harder.

>L:...and centralized power, we have seen all too often, has a way of institutionalizing itself and those who hold key positions in it.

-M: Yes, but is this *innately* so, or is this just an inductive
argument?

>L: FA/DP [...] But it can advise governments, just as it can advise citizens as to how to apply their power.

-M: No it can't, because it has no voice of its own.

>L: It does so, not by centralization of advice, but by massive
> decentralization of advice.[...]

-M: The *measure* of advice is itself *centralization*.
If the consent is measured by *counting* the opinions, this is the
*centrality* algorithm of *in-degree*. If the advice is measured with
*PageRank* under SD2 constraints, this consent is called SD2.

Lomax, please study social network analysis.

Abd ul-Rahman Lomax

unread,
Sep 2, 2006, 2:34:44 AM9/2/06
to top-po...@googlegroups.com
At 05:59 PM 9/1/2006, Mark wrote:
>-M: There is either control or not.
>If there is no control, its just a boring FA.
>If there is control, then the *former* FA can now render
>*institutional* opinions.
>It is now an *entity* with life of its own.

Yes. And it also has just gained a bias, it is attached. And attachment blinds.

One who thinks FAs are boring simply has not experienced them. I'm
not speaking about *formal* FAs, not simply the default "free
associations" that take place every day. I'm speaking about
organizations which have institutionalized the freedom from position
and attached opinion, and, indeed, they are the least boring places
I've ever spent time in. They changed my life.

Does this mean that FAs won't control anything? Technically, yes. But
FAs make it possible for people to effectively control their own
lives and to cooperate with others in this project. FAs do not
inhibit *in the least* the freedom of action of the individuals who
participate in them. Rather, they enhance that freedom. As I've
mentioned, FAs don't act regarding controversy, but FA *members* can
and do. We call a group of FA members with some controversial agenda
an "FA caucus." Against, strictly speaking, an FA caucus won't
collect money *itself,* but its members can and will set up outside
organizations to do this.

An organization like MoveOn.org could be an example of an outside
organization that an FA could and would set up. It's a PAC, it is
oligarchically controlled. An FA caucus-initiated PAC might still
have a DP control structure, but it might be DP based on
contribution, for example. You contribute funds to the PAC, you have
so much control of it. It is your money which you are controlling.
But aside from a centralized treasury, there is also the MoveOn-like
distributed contribution of funds, where many people make small
contributions to some project which has been determined, in MoveOn,
by the oligarchy. In an FA caucus PAC, there might be such an
oligarchy, or there might be something more distributed.

But the difference here, the big difference, is that all this takes
place in a connected context, where broader consensus becomes
possible. If one FA PAC is, say, working to eject Republicans from
office, and another is working to elect Republicans, they are working
at cross-purposes, and the effort they put in and the funds they
spend are, to some degree at least, cancelling each other out. There
is a strong financial incentive for these two groups to find
consensus. As defined, through the conditions of oppositional
politics, this might seem difficult. But it gets easier if one
considers issues instead of parties which are clearly defined in
opposition to each other. Issues do not create natural opposition.
Rather, people have various concerns which lead them to take, as a
start, opposing positions. But, in fact, consensus solutions can
often be found. FAs create strong financial and energy incentives for
finding them. Otherwise groups which oppose each other simply waste
most of their effort, as they cancel each other out.


> >L:[...]security concerns [...]where votes move leveraged assets,
> the action points are vulnerable to corruption. Where the assets
> remain distributed among the members, corruption becomes much more difficult.
>
>-M: The best of both worlds is to have:
>1. Popular consent measured, and decisive over deliberation.
>2. Have the positions of leverage very fluid - those of high rank can
>lose rank quickly, if need be.

Stricly, in an FA, nobody has higher rank than another; however, the
relationships of trust -- which can change in a flash -- do create an
effective hierarchy. Some people have wider access, because they are
more widely trusted. Not only do I not see any way around this, but
it is as it should be.

Procedurally, it is important that proxies not serve for terms. The
common-law proxy is revocable at will by the client.

> >L:[...]with Delegable Proxy should make the consensus-building
> possible in small FA groups accessible to large groups, without
> imposing top-down oligarchical structure.
>
>-M: Centralization is what *works* at being *effective*.
>Do you want *ineffectiveness*?

No. We may not have a clear definition of "centralization." Is the
human nervous system centralized? Is there some small set of cells
with authority to act? How does intelligence get involved, is there a
central committee?

Socialism was based on central planning. It could not compete with
the decentralized market, which was -- and is -- far more effective
and efficient than central planning as generally conceived. It is
*distributed* planning.

But it is not magic. The market does not automatically always find
the optimum solutions. Still, the point is the decentralization is a
major marker of intelligence. When it depends on a center, it is
defective, generally, compared to distributed processing of
information and analysis.


> >L: FA/DP does create a kind of elite, and thus it can be called
> > "elitist," but this is misleading. The elite in an FA depends
> > continuously on the successful maintenance of trust in it by the
> > general membership.
>
>-M: As with SD2.
> >L: Proxies have no term of office, they serve, in the language of
> employment, at the will of the employer. The client. The member.
>
>-M: Sounds linear.

It is, that is, there is a linear connection between any client and
the highest active proxy. A chain. There are, to be sure, plenty of
cross-links, but there is a clear and single chain, a strong
connection of mutual responsibility and trust. I'm a bit suspicious
of "alternate proxies" for this reason. But, with proper caution, I
think they can be implemented. I just don't want to dilute the
responsibility, that primary connection that the proxy-client
relationship represents.

> What about loops?

What about loops? As I think we have agreed, loops are necessary,
they will exist. Small loops of people not well-connected can leave
those people, well, out of the loop :-)

But it's a Free Association. The FA is not going to tell people whom
to choose as a proxy, nor is it even going to require such a choice.
But people gain access and privileges by naming a proxy, naturally.

I assume that anyone can take a proxy list and analyze it, perhaps
together with voting records, and determine who is and who is not
represented, and can notify these people. For those who are
anonymous, the system can be set up to do this. Systemic
communications can be dangerous, but with proper safeguards....

(Existing nonprofits which have voting members and which allow
proxies, such as the California State Automobile Association,
maintain control on behalf of the oligarchy through the device of
sending proxy solicitations to members. Far too many members simply
sign them without having the foggiest idea of the implications. And,
besides, there being no parallel FA, as we would suggest, they would
not be able to accomplish anything anyway. A lawyer with substantial
resources tried to run a proxy fight, he failed. Those automatic
proxy solicitations, going to every member are quite powerful to
maintain the power of those who have the ability to send them.)


>By contrast, with SD2, people are differentiated by rank.

If the rank is an attribute of the relationships, this is a bit
circular. People in an FA/DP organization *may* according to specific
meeting rules have differing privileges. One method of defining the
privileges would be by what I've called "proxy rank." People with
sufficient proxy rank gain the right to directly address the
assembly, to enter motions, etc. Generally, however, except for what
are known under Robert's Rules as Questions of Privilege, which
involve the personal rights of participants in a meeting, all members
may be allowed to vote on any motion. That they are allowed to do so,
however, does not mean that more than a small percentage will
actually do it. Only those who are following the meeting and who have
a special concern are likely to directly vote. Most votes will be
cast by proxy, I'd predict, in large organizations. In small and
young organizations, most votes will be personally cast, unless
somehow we get full-blown DP ab initio, and even then when membership
is new, people are much more likely to be directly active, and when
the org is small, it is much more practical.


> >L: Given that it is extremely simple, involves no significant expense,
> > and, in the FA context, it about as safe as could be imagined, I'd
> > say it is worth a trial.
>
>-M: Why try something that already appears outdated?

If it has never been done, how can it be "outdated"?

But, hey, if you've got the Cat's Meow DP system, it could be used
for power structures, why not?

(Well, I could tell you why not, but I do believe you've got the
right to try. Indeed, I expect a multiplicity of such efforts, and
may the best one win.... I think that the FA model will prove, in the
end, to be the most *intelligent*, on theoretical grounds. Again, it
has to do with what I wrote above, attachment. All that time spent
studying Buddhism was not *entirely* wasted!)


> >L: "If we want to save the world, it has to be easy."
> > Because if it is not easy, it is *not* going to happen. If it is
> > hard, it takes centralized power,...
>
>-M: What other option is there? Centralization is what works, and any
>effective change done by decentralized systems will be much harder.

Your assertion, with which I do not agree. Centralization is, in
fact, the "outdated" method. It worked better than what came before,
which was informal direct democracy. But highly centralized systems,
everywhere, are losing ground in favor of more democratic forms,
mostly market systems and representative democracy. (The former is
decentralized and the latter largely centralized; technically, DP is
a hybrid. It creates a kind of centralization, but the centralization
is not fixed.)

> >L:...and centralized power, we have seen all too often, has a way
> of institutionalizing itself and those who hold key positions in it.
>
>-M: Yes, but is this *innately* so, or is this just an inductive
>argument?

I'd say that it is innate. However, if the centralized power is
continually dependent upon decentralized support, if the centralized
power collapses without that maintained support, the matter is different.

Indeed, it is *always* true that centralized power is dependent upon
decentralized support, except that, generally, the decentralized
support has no significant alternative. And I could go on about this,
but it is getting quite late.


> >L: FA/DP [...] But it can advise governments, just as it can
> advise citizens as to how to apply their power.
>
>-M: No it can't, because it has no voice of its own.

This is thoroughly incorrect. It has a voice. It may, for example,
issue press releases with poll results. Those are NPOV, in wikipedia
terms. It cannot itself, *as an organization* advocate a
controversial position, but any caucus may. And what do you think it
would mean if I had gone to, say, George W. Bush and said that I
represented fourteen million people, a caucus within a political
FA/DP organization, and those I represent have pledged to give $50
each -- indeed, they have *deposited* $50 each -- to a fund dedicated
to removing him from office, unless he gives due consideration to our
recommendation that he find an acceptable alternative to invading
Iraq. "Let's talk, Mr. Bush, while we gaze at this here $700 million
check which I would just love to tear up." Do you think it would have
gotten his attention?

Sure the FA doesn't take advocacy positions. But it brings together
the people who *may* and who *will* take such positions. So I
represent fourteen million people, and anyone can look at the records
of the FA and see that the opposing caucus enjoyed one million
supporters. Mr. Bush could decide, of course, to accept their $50
million to fight us.... his choice.

Or, perhaps, one imagines that Mr. Bush would be more likely to
respond to a few million people going out into the streets, waving
signs begging him to stop. What do you think? What moves power?

What prevented the peace movement in the US from stopping the war is
that, instead of creating an organization which could effectively
coordinate the collective power of its supporters, they wasted their
energy (about one billion dollars worth, more than enough to buy a
U.S. Presidential election) in an emotional display that went
nowhere. No organization capable of accomplishing the task was formed.

There is a reason why these people did not send in a check instead of
going into the streets. It is not that they don't have the money. It
is that they don't have an organization they can trust. Organizations
which collect money end up spending it according to the wishes of the
oligarchs. Sometimes that is just what people want, I'm not
condemning oligarchy. But it is limiting. What if we had
organizations that, by structure, we could trust? Indeed, what if we
had *governments* we could trust? But to get there with governments
is *much* more difficult than getting there with voluntary peer
organizations. It is quite hard enough to accomplish the latter,
under present conditions.

I'm working to change those conditions, mostly to awaken people to
the possibility that they are not nearly as powerless as they have
imagined. They are wealthy, but that wealth is inaccessible to them
because they are asleep.

> >L: It does so, not by centralization of advice, but by massive
> > decentralization of advice.[...]
>
>-M: The *measure* of advice is itself *centralization*.
>If the consent is measured by *counting* the opinions, this is the
>*centrality* algorithm of *in-degree*. If the advice is measured with
>*PageRank* under SD2 constraints, this consent is called SD2.
>
>Lomax, please study social network analysis.

Sorry. That's what I've been doing for quite a few years, though not
formally through standard texts, etc. Which is why I found what had
been neglected in the standard texts. If someone has something clear
to contribute to this discussion based on standard analysis, great.
But it had better be cogent, not merely an arrogant dismissal based
on shallow thinking and a ready assumption of better understanding.

Humans act ("make decisions") based on a distributed state of nervous
system activity. There is no one place where decisions are made.

Sure, were I twenty or thirty or even maybe forty or fifty years old,
I'd undertake new studies. But I'm not. I don't have time.

illegale

unread,
Sep 2, 2006, 8:50:58 PM9/2/06
to top-politics
Hey Magnus!

Long time no see. :-)

MG wrote:
> Hello, I'm back but very busy running for elections. Will attend
> Swedish TV3 on thursday..


Really? You might send us some report about whole thing! BTW, today I
came from socialist youth conference where i talked about Internet &
democracy, gonna write some findings to top-politics.


> To the science...:
> This report is very good proof of that to most people, it is more
> important WHO is saying something, than WHAT he says.

That is actually needed IMO for perserving society. Concretely, if some
bump aproaches to you and says "tiki miki swiki", youll look at him and
go away. If that was told by someone you highly apreciate and you think
he is serious about that, you will start thinking about the whole point
maybe for several days even.

And the problem is that we do not have enough time to analyse any
thought of any person, though we have some time that we have to use the
best way can. This WHO is about credibility, very usefull thing for
openning of your thought process.

All in all, this phenomenon enables us to share usefull infomration
much faster than if we did not have such mechanism.


> My solution for this is to de-identify the arguments as much as
> possible and to have as many voices as possible, your is to promote
> certain speakers as better than others?


That is the fact. Some speakers are better than others. Same as some
lawyers are better than others.


> When there are enough voices in the room, this stupid behaviour of not
> reasoning to some voices will become impratical. It is only possible to
> withstand as long as there are 2 or 3 easy identidied individuals to
> hack on.

More people, more thoughts, more powerfull info that is not
personalised, more decentralisation. Yet, people should care for their
reputation IMO, that is about responsibility to themselves, which is
rather good thing. If you have mechanisms that enable this reputation
be used for good, than you have very powerfull principle in use.

ATB,
Gale

Mark

unread,
Sep 3, 2006, 4:39:31 PM9/3/06
to top-politics

Abd ul-Rahman Lomax wrote:
> At 05:59 PM 9/1/2006, Mark wrote:
> >-M: There is either control or not.
> >If there is no control, its just a boring FA.
> >If there is control, then the *former* FA can now render
> >*institutional* opinions. It is now an *entity* with life of its own.

>L: Yes. And it also has just gained a bias, it is attached. And attachment blinds.

-M: Yes, it is biased. But in the case of SD2-S, there could be
opposing biases, such as the popular opinion being opposed to the
expert opinion

>L One who thinks FAs are boring simply has not experienced them.

-M: They are boring for political purposes because they are
inneffective.
However they are great for spiritual purposes. I have experienced both.

>L: I'm not speaking about *formal* FAs, not simply the default "free


> associations" that take place every day. I'm speaking about
> organizations which have institutionalized the freedom from position
> and attached opinion, and, indeed, they are the least boring places
> I've ever spent time in. They changed my life.

-M: Was it the effectiveness that changed you, or was it the experience
of community?
I suspect that it was the latter. If so, then get a love-circle. And
leave politics to the experts.

> > >L:[...]security concerns [...]where votes move leveraged assets,
> > the action points are vulnerable to corruption. Where the assets
> > remain distributed among the members, corruption becomes much more difficult.
> >
> >-M: The best of both worlds is to have:
> >1. Popular consent measured, and decisive over deliberation.
> >2. Have the positions of leverage very fluid - those of high rank can
> >lose rank quickly, if need be.

>L: Stricly, in an FA, nobody has higher rank than another; however, the


> relationships of trust -- which can change in a flash -- do create an
> effective hierarchy.

-M: But it is a hierarchy based on popularity, not a deep trust. It is
too immediate and direct. :-(

> >-M: Centralization is what *works* at being *effective*.
> >Do you want *ineffectiveness*?

>L: No. We may not have a clear definition of "centralization." Is the
> human nervous system centralized?

-M: Yes, as evidenced by the potential for decisive action.

>L: Is there some small set of cells with authority to act?

-M: Yes, the ones employed for the task at hand.

>L: How does intelligence get involved, is there a central committee?

-M: There are competing memes, and some win dominence.
This is centralization that has emerged from decentralized and
distributed processes.

>L: Socialism was based on central planning. It could not compete with


> the decentralized market, which was -- and is -- far more effective
> and efficient than central planning as generally conceived. It is
> *distributed* planning.

-M: Both systems are centralized, decentralized and distributed.
The problem with commie-nism is that it more entrenched.
What is needed is unentrenched centralization that decentralize where
need be.
SD2-S is an example of this.

>L: But it is not magic. The market does not automatically always find


> the optimum solutions. Still, the point is the decentralization is a
> major marker of intelligence. When it depends on a center, it is
> defective, generally, compared to distributed processing of
> information and analysis.

-M: Again, centralization isn't the problem, it is entrenchment.
[I'll answer the rest, but I have to run.]

Abd ul-Rahman Lomax

unread,
Sep 5, 2006, 10:43:43 PM9/5/06
to top-po...@googlegroups.com
At 04:39 PM 9/3/2006, Mark wrote:
> >L One who thinks FAs are boring simply has not experienced them.
>
>-M: They are boring for political purposes because they are
>inneffective.
>However they are great for spiritual purposes. I have experienced both.

Great.

I'm a Muslim which, among other things, means that I don't accept the
division of the world into spiritual and political realms. There is one world.

Non-DP FAs are quite effective in politics *if* the scale is small.
They are ineffective on a large scale if they have no means of
forming action groups on the necessary scale. Always, where
large-scale action has been needed, FAs have devolved into standard
power structures, or they remained all talk and no walk.

DP is the missing element. I think FA/DP organizations will be *far*
from boring. But we won't know until it is tried!

>-M: Was it the effectiveness that changed you, or was it the experience
>of community?
>I suspect that it was the latter. If so, then get a love-circle. And
>leave politics to the experts.

The community and the effectiveness cannot be separated. FAs are
effective, in their realm, because of the community. And the
communities grow because they are effective.

But the application of FA traditions to the political realm has never
been tried, as far as I know. And without DP, I'd expect it to fail.
I.e., to be ineffective.

(You can consider many discussion groups to be, effectively, FAs.
Such as this group. But without an organizing, filtering structure,
this group cannot grow to the necessary scale to be truly effective,
beyond its function of allowing a relatively small number of people
to communicate and cooperate.)

As to leaving politics to experts, are there any around here? Any
that I might be willing to vote for, so I could, indeed, "leave it to" them?

Let's put it this way. It ain't you. Bye.

pether....@gmail.com

unread,
Sep 6, 2006, 7:30:15 AM9/6/06
to top-politics
Hello,

Haven't been following this group for a while, but good to hear you
guys are actually running for this election.

Been busy with my day job the last few months but doing some progress
on the direct democracy portal.

Best regards
Pether

Mark

unread,
Sep 6, 2006, 5:53:21 PM9/6/06
to top-politics

> > >L One who thinks FAs are boring simply has not experienced them.

> >-M: They are boring for political purposes because they are
> >inneffective. However they are great for spiritual purposes. I have experienced both.

>L: Great. I'm a Muslim which, among other things, means that I don't accept the


> division of the world into spiritual and political realms. There is one world.

-M: OK, there is one world, and the way in which we *think* about this
world should be unified. However we still have different tools for how
we *interact* with this world.

With spirituality, people can interact with each other in a highly
egalitarian environment, and this works. But with politics, hierarchy
and quality filtering is in order.

>L: Non-DP FAs are quite effective in politics *if* the scale is small.


> They are ineffective on a large scale if they have no means of
> forming action groups on the necessary scale. Always, where
> large-scale action has been needed, FAs have devolved into standard
> power structures, or they remained all talk and no walk.
> DP is the missing element. I think FA/DP organizations will be *far*
> from boring. But we won't know until it is tried!

-M: There is FA, there is FA/DP and there is institutionalized DP.

Institutionalized DP is the only one here with its own voice, therefore
it is the only one capable of *walk* - the others are just *talk*,
therefore boring(for political purposes).

> >-M: Was it the effectiveness that changed you, or was it the experience
> >of community?
> >I suspect that it was the latter. If so, then get a love-circle. And
> >leave politics to the experts.

>L: The community and the effectiveness cannot be separated. FAs are


> effective, in their realm, because of the community. And the
> communities grow because they are effective.
> But the application of FA traditions to the political realm has never
> been tried, as far as I know. And without DP, I'd expect it to fail.
> I.e., to be ineffective.

-M: Wouldn't *politics+ DP* = political DP?
Where does FA come into play?
Does this look different than political DP?

>L: (You can consider many discussion groups to be, effectively, FAs.


> Such as this group. But without an organizing, filtering structure,
> this group cannot grow to the necessary scale to be truly effective,
> beyond its function of allowing a relatively small number of people
> to communicate and cooperate.)

-M: OK, filtering structure.

>L: As to leaving politics to experts, are there any around here?

-M: I think I am pretty good, but my discourse has been away from
ideology(except for how it relates to voting).

>L: Any that I might be willing to vote for, so I could, indeed, "leave it to" them?

-M: My system recommends at least two trustees. Maybe Marcus or Gale
could be your other trustee.

>L: Let's put it this way. It ain't you. Bye.

-M: This is your option, but keep in mind that the issues here are
currently for system design. Lets stay focused. The issues are *how to*
filter experts, and *why* filter experts, not *who* those experts are.

illegale

unread,
Sep 6, 2006, 10:42:46 PM9/6/06
to top-politics
Hey Pether!

Do you have some expectation of when are you going to release user
friendly product of your development?

Maybe we might help you with testing and commenting it?

ATB,
Gale

pether....@gmail.com

unread,
Sep 7, 2006, 8:26:24 AM9/7/06
to top-politics
Been in pre-alpha stage for quit some time now but are finally getting
happy with the
current technologies and infrastructure work I have done..

Before I did complicate things by treating the applications as
separate modules but now I refactored everything into a more simple
application. Also doing test driven development now since before I did
the mistake of adding/changing the application but never maintained old
functionality. Where before I didn't have any test for my application I
now have a suite of automated unit/integration & acceptance test..

Been talking to much before so don't want to make any promises, but
plan to make an alpha release the next few months that will bundle the
application together with jboss & hsql (java database) so people have a
simple way of testing the application..

At some point before that I will put an alpha demo version on
ww.directdemocracyparty.net.

Currently busy with my day job working with the same set of
technologies doing extreme programming and TDD so been forced to work
(and learn) instead of slacking of having to much fun.

Current functionality is almost more limited then before,
Register User
Login
View Organisation
View Activity
View Members
View Online Members
View Elections
View View Elected Officials
Create Political Party

As always updating www.directdemocracyportal.org whenever I do update
the code, but currently have some code that I haven't checked in..

Waiting for the final release of Spring 2.0 and Hibernate 3.2 and the
ambition is the the first alpha version would be a small showcase how
these technologies can be used in combination with testing tools like
HtmlUnit, DbUnit, JMock & built using Maven.

Since it wasn't finished for the Swedish election this year I now have
another 4 years until next deadline :)

Good to see that everybody is still making some progress and will post
a link as soon as I sorted out a server to host the application to get
some comments.

Best regards
Pether

Abd ul-Rahman Lomax

unread,
Sep 6, 2006, 11:57:48 PM9/6/06
to top-po...@googlegroups.com
It's a bit of a shame. Mark does have an unusual understanding of the
issues involved with DP. But, as it has been pointed out, his
behavior alienates.

In a previous post, Mark wrote, referring to me:


> >-M: Was it the effectiveness that changed you, or was it the experience
> >of community?
> >I suspect that it was the latter. If so, then get a love-circle. And
> >leave politics to the experts.

I then responded


>L: As to leaving politics to experts, are there any around here?

> >L: Let's put it this way. It ain't you. Bye.

At 05:53 PM 9/6/2006, Mark wrote, in response to this:

>-M: This is your option, but keep in mind that the issues here are
>currently for system design. Lets stay focused. The issues are *how to*
>filter experts, and *why* filter experts, not *who* those experts are.

It was Mark who had made the reference to the identity of experts,
or, more specifically, to the identity of an alleged non-expert, me.
This kind of writing continually distracts from what he correctly
identifies as the proper focus of this group. And then he blames
everyone else for the distractions.

I think it is enough for me to point this out once. Not more.


Mark

unread,
Sep 7, 2006, 4:25:17 PM9/7/06
to top-politics

Abd ul-Rahman Lomax wrote:
> It's a bit of a shame. Mark does have an unusual understanding of the
> issues involved with DP. But, as it has been pointed out, his
> behavior alienates.
>
> In a previous post, Mark wrote, referring to me:
> > >-M: Was it the effectiveness that changed you, or was it the experience
> > >of community?
> > >I suspect that it was the latter. If so, then get a love-circle. And
> > >leave politics to the experts.
>
> I then responded
> >L: As to leaving politics to experts, are there any around here?
>
> > >L: Let's put it this way. It ain't you. Bye.
>
> Mark wrote, in response to this:
>
> >-M: This is your option, but keep in mind that the issues here are
> >currently for system design. Lets stay focused. The issues are *how to*
> >filter experts, and *why* filter experts, not *who* those experts are.

> It was Mark who had made the reference to the identity of experts,
> or, more specifically, to the identity of an alleged non-expert, me.

-M: Lomax,
When I said "leave politics to the experts"
I meant: "leave politics to the experts"
meaning that us system designers should build systems that find the
experts.
I don't know if these systems would select you or not, and it would be
presumptous of me to assume either way.

>L: This kind of writing continually distracts from what he correctly


> identifies as the proper focus of this group. And then he blames
> everyone else for the distractions. I think it is enough for me to point this out once. Not more.

-M: I can see how this misunderstanding occured, because when I said
"If so, then get a love-circle. And leave politics to the experts" it
was as if I was establishing a *mutually exclusive dichotomy*, which
people often do. I need to be more careful, and I appologize for any
hard feelings.

Lomax, I do like you, and I do hope that we continue our exchanges. :-)

Abd ul-Rahman Lomax

unread,
Sep 9, 2006, 1:03:11 PM9/9/06
to top-po...@googlegroups.com
At 04:25 PM 9/7/2006, you wrote:
>-M: Lomax,
>When I said "leave politics to the experts"
>I meant: "leave politics to the experts"
>meaning that us system designers should build systems that find the
>experts.

Of course. That is exactly what DP is designed to do. It does it
through the concentration of trust. Experts should not necessarily be
the ones actually making decisions, however. I don't want a "medical
expert," for example, deciding what treatments should be available to
me. I want such experts to *participate*. Not to control.

There may be a field in which I am generally not an expert, but I
happen to know a great deal about some narrow slice of this field. So
I'd leave most decisions in the area to my proxy, assuming that my
proxy will either participate directly or will similarly pass on the
responsibility, and that bringing in expert testimony will be a part
of this process, but there may be a particular question that I
understand, in my opinion, better than my proxy. So I want to vote
directly on that. I shouldn't have to ask my proxy for permission.

I'd like to be able to participate in the discussion, also. But that
can't be available to me by default, for reasons that have been well discussed.

It is also important to note that my non-participation in a topic may
not be due to lack of qualifications. It may be due to lack of time.
Voting is fast, compared to actually writing comments on a topic.

Generally, the decision of whether or not to vote directly or to
leave the voting to a proxy (whether or not a specialist proxy)
should remain firmly in the hands of the voter. Proxies are servants,
not governors.

But the right to speak to an assembly is another matter. That right
is something which is socially assigned, through some means.

Thus we distinguish between the right to vote and the right to speak
at high levels. FA/DP organizations, we conceive, will be direct
democracies, but on an unlimited scale. Yes, unlimited, at least as
far as the present population of the earth is concerned....

FA/DP is an organizational concept, not a specific implementation.
For an organization to claim to be FA/DP, it need not implement all
details, because some details are not significant with small and new
organizations. DP itself is an example. The FA characteristics,
however, would mean that the organization is open to the
establishment of a DP structure, and bringing such a structure to the
point of official recognition (as part of the internal
decision-making process) by the FA should not be difficult if it is
following the FA traditions, and if the structure is actually needed.

illegale

unread,
Sep 9, 2006, 6:12:27 PM9/9/06
to top-politics
Hey Pether.

Who is your targeted public for tool you are developing actually? Do
you know somebody in person, or you suppose what would fictive
character actually need?

ATB,
Gale

pether....@gmail.com

unread,
Sep 10, 2006, 7:05:19 AM9/10/06
to top-politics
As always in open source people get what they pay for, so are really
developing it because I want to see a transfer from representative
democracy to direct democracy and at the same time improve my skill
levels as a developer.

Probably mentioned it before but plan to host one instance of the
application at www.directdemocracyparty.net once it's get to a state
where I'm happy with it.

Pether

illegale

unread,
Sep 12, 2006, 5:24:15 PM9/12/06
to top-politics
Thank you Pether for giving me answers I was looking for.

ATB,
Gale

Mark

unread,
Sep 12, 2006, 5:44:30 PM9/12/06
to top-politics

> >When I said "leave politics to the experts"
> >I meant: "leave politics to the experts"
> >meaning that us system designers should build systems that find the
> >experts.

>L: Of course. That is exactly what DP is designed to do. It does it


> through the concentration of trust. Experts should not necessarily be
> the ones actually making decisions, however. I don't want a "medical
> expert," for example, deciding what treatments should be available to
> me. I want such experts to *participate*. Not to control.

-M: OK, with SD2-S the generalists/trustees would control the forum(not
the issues)
where the expert specialists/delegates would be participating.

>L: There may be a field in which I am generally not an expert, but I


> happen to know a great deal about some narrow slice of this field. So
> I'd leave most decisions in the area to my proxy, assuming that my
> proxy will either participate directly or will similarly pass on the
> responsibility, and that bringing in expert testimony will be a part
> of this process, but there may be a particular question that I
> understand, in my opinion, better than my proxy. So I want to vote
> directly on that. I shouldn't have to ask my proxy for permission.

-M: OK, since SD2-S is DP, voting directly on an issue is an option,
but with your prefered defaults, how does it count?

My system, assuming that the the defaults haven't been overrided by the
voters, only 40% popular approval is needed as long as the expert
opinion is 60% in favor.

>L: I'd like to be able to participate in the discussion, also. But that


> can't be available to me by default, for reasons that have been well discussed.
> It is also important to note that my non-participation in a topic may
> not be due to lack of qualifications. It may be due to lack of time.

-M: This is why I would make politics 24/7. This can be done with
web-based technology.

>L: Voting is fast, compared to actually writing comments on a topic.


> Generally, the decision of whether or not to vote directly or to
> leave the voting to a proxy (whether or not a specialist proxy)
> should remain firmly in the hands of the voter. Proxies are servants,
> not governors.

-M: Instead of :"...decision of whether or not to vote directly *or* to
leave the voting to a proxy..."
Did you mean?: "...decision of whether or not to vote directly *and/or*
to leave the voting to (a) prox(y)(ies)..."

This does make a difference.

>L:[...]The FA characteristics, however, would mean that the organization is open to the establishment of a DP structure,[...]

-M: How? If the FA can render no institutional opinion, how can it
decide on DP?
This is why FA talk annoys me.

Abd ul-Rahman Lomax

unread,
Sep 13, 2006, 1:19:01 AM9/13/06
to top-po...@googlegroups.com
At 05:44 PM 9/12/2006, Mark wrote:

>-M: OK, since SD2-S is DP, voting directly on an issue is an option,
>but with your prefered defaults, how does it count?
>
>My system, assuming that the the defaults haven't been overrided by the
>voters, only 40% popular approval is needed as long as the expert
>opinion is 60% in favor.

FA/DP regulates actual power behavior through the members.
Controversial actions are taken by caucuses, or even by individual
members, collectively or individually, not by the organization, so
the organization does not ordinarily need to determine what
threshhold of approval is necessary. It simply reports poll results,
and, perhaps, provides a few analytical tools, such as expanding a
poll with DP.

> >L: I'd like to be able to participate in the discussion, also. But that
> > can't be available to me by default, for reasons that have been
> well discussed.
> > It is also important to note that my non-participation in a topic may
> > not be due to lack of qualifications. It may be due to lack of time.
>
>-M: This is why I would make politics 24/7. This can be done with
>web-based technology.

24/7. What a joke. Every minute counts for me, it is difficult to
find any time at all. Let me guess that Mark does not have small
children. (But, yes, I agree with the desirability of "making
politics 24/7." It does increase potential participation. It is just
that this is not enough. Proxy representation is really necessary.)

>-M: How? If the FA can render no institutional opinion, how can it
>decide on DP?
>This is why FA talk annoys me.

Well, if it is that easy and simple, Mark's objection, then we are almost home.

FAs do need to make internal decisions. Where do we meet? Do we serve
tea as well as coffee? How long should the meeting last?

And what software should we use for our web site? Should we use proxy
voting in our ordinary decisions?

FAs make these decisions, in my experience, using pretty much
standard democratic process, Robert's Rules if they are large enough,
with a bias toward seeking and finding consensus, which FAs value
greatly. DP would be implemented, if it is going to be implemented as
part of an official structure, by an ordinary vote. That's right,
proxies would not be counted in this vote. Or they might be. It is
quite likely that the difference would not be major. But if it is,
those who want to have a larger and more powerful organization will
form one, counting the proxies, and the rest can continue to
collectively twiddle their thumbs.... FAs don't force anyone to do
anything. And they splinter at the drop of a hat. But they tend to
splinter in a way that simply makes the organization more diverse;
since they don't own property, there is little to fight over.

If our theory is correct, organizations which use DP will be more
efficient and will be able to leverage more power. So they will tend
to grow and to be more effective.

This is how FAs would come to use DP.

Got it?

Mark

unread,
Sep 13, 2006, 6:57:09 AM9/13/06
to top-politics

> >-M: OK, since SD2-S is DP, voting directly on an issue is an option,
> >but with your prefered defaults, how does it count?
> >My system, assuming that the the defaults haven't been overrided by the
> >voters, only 40% popular approval is needed as long as the expert
> >opinion is 60% in favor.

>L: [...]It simply reports poll results, and, perhaps, provides a few analytical tools, such as expanding a poll with DP.

-M: Sounds weak. :-(

> > >L: I'd like to be able to participate in the discussion, also. But that
> > > can't be available to me by default, for reasons that have been
> > well discussed. It is also important to note that my non-participation in a topic may not be due to lack of qualifications. It may be due to lack of time.
> >
> >-M: This is why I would make politics 24/7. This can be done with
> >web-based technology.

> L:[...](But, yes, I agree with the desirability of "making politics 24/7." It does increase potential participation. It is just that this is not enough. Proxy representation is really necessary.)

-M: We can't force participation, only make it available to those with
the time.
The rest can be proxied.

> >-M: How? If the FA can render no institutional opinion, how can it
> >decide on DP? This is why FA talk annoys me.

>L: [...]FAs make these decisions, in my experience, using pretty much
> standard democratic process, ...

-M: 'standard democratic process' - this is either binding or its not.
If it is binding, then it is an institutional voice. If not, then its
just play.

>L: [...] If our theory is correct, organizations which use DP will be more


> efficient and will be able to leverage more power. So they will tend
> to grow and to be more effective. This is how FAs would come to use DP.
> Got it?

-M: I think that organizations should be started with DP.
All the FA looks like play.

Abd ul-Rahman Lomax

unread,
Sep 13, 2006, 11:44:25 AM9/13/06
to top-po...@googlegroups.com
At 06:57 AM 9/13/2006, Mark wrote:


> > >-M: OK, since SD2-S is DP, voting directly on an issue is an option,
> > >but with your prefered defaults, how does it count?
> > >My system, assuming that the the defaults haven't been overrided by the
> > >voters, only 40% popular approval is needed as long as the expert
> > >opinion is 60% in favor.
>
> >L: [...]It simply reports poll results, and, perhaps, provides a
> few analytical tools, such as expanding a poll with DP.
>
>-M: Sounds weak. :-(

It is information about the state of agreement of the FA community,
and, as such, would be useful in two major ways: a caucus
contemplating action can know if sufficient groundwork has been done
so that proceeding is not going to be a waste of energy, and others
may take advice from it. For example, one function of FAs would be to
advise public officials, including elected officials. Advise, not control.

How powerful is advice? Depends, doesn't it?

Yes, the FA *is* weak. Just as your brain is weak unless your muscles
decide to respond as you desire. The weakness, in the case of an FA,
is part of the design. It is, in fact, essential. Without it, the FA
becomes yet another oppressive government, or at least that is the risk.

>-M: We can't force participation, only make it available to those with
>the time.
>The rest can be proxied.

Yes. A system may automatically assign proxies, as I think you
suggest, but there is a danger in this, actually, more than one
danger. I think those dangers would require attention, and a pure
software solution would have to be generally agreed to be
incorruptible. I do think it can be done. AA meetings often have an
officer called the "greeter," who makes a point of noticing new
people and welcoming and talking with them.

If the assignment process is TOP, and enjoys general agreement, it
should be fine. But assigned proxies are weak links and I'd probably
want to devalue them in expanding votes.

Remember, in an FA, how to analyze poll results is really up to the
analyst, who is or is engaged by those who will use the results.


> > >-M: How? If the FA can render no institutional opinion, how can it
> > >decide on DP? This is why FA talk annoys me.
>
> >L: [...]FAs make these decisions, in my experience, using pretty much
> > standard democratic process, ...
>
>-M: 'standard democratic process' - this is either binding or its not.
>If it is binding, then it is an institutional voice. If not, then its
>just play.

It is binding, that is, it makes irrevocable decisions, in a sense,
but they only bind the FA, not its members, who can, if they wish,
recreate the FA in a new form, not bound by the decision, in a flash.
Once again, the implications of FAs not holding property must be
noticed. Institutional inertia, of the kind that can be so
frustrating in dealing with nonprofits, becomes an issue because the
central organization *does* own property, as well as, often,
controlling access to and between the members. FAs are *different*,
very different.


> >L: [...] If our theory is correct, organizations which use DP will be more
> > efficient and will be able to leverage more power. So they will tend
> > to grow and to be more effective. This is how FAs would come to use DP.
> > Got it?
>
>-M: I think that organizations should be started with DP.
>All the FA looks like play.

To you. To people whose lives have been saved by an FA, and there are
many, it is more than play. It is, indeed, deadly serious. If you can
imagine deadly serious with an advanced sense of humor, at the same
time, you might have a sense of what it can be like.

My point about DP is that, with an awakened membership, it is trivial
to convert from standard proxy to DP. In an FA, the difference is not
terribly important, and since poll analysis is not a crucial central
function, anyone who wants to use the already-existing proxy list and
consider proxies delegable is completely free to do so.

But there is no harm in having DP from the beginning, except that it
complicates explaining the organization to people who just want to
talk about their Ford, and perhaps act as a customer lobby with the
company. My realization was that emphasizing DP at the beginning made
the tablet a bit too large to swallow at first sight. So, in fact,
I'd build it into the system, but not necessarily into the internal
decision-making network (because there are people who *will* object
to it without understanding it. Many.)

Were I active in one of the FAs based on the AA traditions, I expect
that I could arrange the implementation of DP pretty quickly. I might
do it. Depends. It's a huge time commitment.... (If I had worked out
the implications of DP and the FA applications, say ten years
earlier, I think we would already have some clear demonstration
models. I was in a position to do it.)


Mark

unread,
Sep 13, 2006, 6:50:20 PM9/13/06
to top-politics
Lomax wrote:
>Mark wrote:
> > > >-M: OK, since SD2-S is DP, voting directly on an issue is an option,
> > > >but with your prefered defaults, how does it count?
> > > >My system, assuming that the the defaults haven't been overrided by the
> > > >voters, only 40% popular approval is needed as long as the expert
> > > >opinion is 60% in favor.
> >
> > >L: [...]It simply reports poll results, and, perhaps, provides a
> > few analytical tools, such as expanding a poll with DP.
> >
> >-M: Sounds weak. :-(

>L: It is information about the state of agreement of the FA community,


> and, as such, would be useful in two major ways: a caucus
> contemplating action can know if sufficient groundwork has been done
> so that proceeding is not going to be a waste of energy, and others
> may take advice from it. For example, one function of FAs would be to
> advise public officials, including elected officials. Advise, not control.
> How powerful is advice? Depends, doesn't it?

-M: Who's flipp'n advice? The FA can't render an institutional opinion,
so the advice would be just of several people. So? Why do they need a
flipp'n FA?

The Council on Foreign Relations, because it is a 501(c)3
tax-exempt(under American federal IRC code), registered as an
'educational organization', it can't be a 'political advocacy' group.
It gets around this limitation by forming 'Disappearing Task Forces'.
The DTF renders a political opinion in the name of the DTF - while the
CFR simply serves as a *forum* for the formation of DTFs.

Notice how they nail down their organizational procedures in a binding
way? This is how the enemy works, and if we want to be powerful like
them, then we should organize ourselves atleast as well as them.

No flipp'n MUSHY FAs!

>L: Yes, the FA *is* weak. Just as your brain is weak unless your muscles


> decide to respond as you desire. The weakness, in the case of an FA,
> is part of the design. It is, in fact, essential. Without it, the FA
> becomes yet another oppressive government, or at least that is the risk.

-M: Coercive organization of society is manditory. There is no other
option, so this isn't a variable. So this now means that any coercive
components of society should be organized with as much legitimacy and
as little corruption, as possible.

> >-M: We can't force participation, only make it available to those with
> >the time. The rest can be proxied.

>L: Yes. A system may automatically assign proxies, as I think you
> suggest,...

-M: There is this default, but people are requested to avoid this by
voting for at least two general-trustees.

>L:...but there is a danger in this, actually, more than one danger. I think those dangers would require attention, and a pure software solution would have to be generally agreed to be incorruptible.

-M: My algorithm would never give default voting rank to the top guy,
only to those of lower rank in rank distribution gaps. This is
counter-entrenchment and incorruptible.

>L: I do think it can be done. AA meetings often have an officer called the "greeter," who makes a point of noticing new people and welcoming and talking with them. If the assignment process is TOP, and enjoys general agreement, it should be fine.

-M: I think so.

>L: But assigned proxies are weak links and I'd probably want to devalue them in expanding votes.

-M: I would call them *intentionally unstable* links, ones intended to
have temporary inflated rank for the purpose of having the network
ranks not fall into an entrenchment, such as splitting into an
'in-group' and 'out-group'.

If the rank distributions are smooth, there are no identifyable
clusters that could demarkate such an 'in-group'.

>L: Remember, in an FA, how to analyze poll results is really up to the


> analyst, who is or is engaged by those who will use the results.

-M: Who-the-flip is an analyst? What makes one better than another?
Would everyone have their own pet analyst?
By contrast, an actual institution would have its analyst selection
procedures nailed down. Then something could actually get done.

[...]


> >-M: I think that organizations should be started with DP.
> >All the FA looks like play.

>L: To you. To people whose lives have been saved by an FA, and there are


> many, it is more than play. It is, indeed, deadly serious. If you can
> imagine deadly serious with an advanced sense of humor, at the same
> time, you might have a sense of what it can be like.

-M: OK, I have seen FAs for community purposes, and they can be
impressive.
But politics and other forms of collective *action* are a different
domain.

[...]
>L: [...] My realization was that emphasizing DP at the beginning made


> the tablet a bit too large to swallow at first sight.

-M: Then don't use it for board selection, just commitee selection.

>L: So, in fact, I'd build it into the system, but not necessarily into the internal


> decision-making network (because there are people who *will* object

> to it without understanding it. Many.)[...]

-M: Agreed. Inertia is easier than progress.

Abd ul-Rahman Lomax

unread,
Sep 14, 2006, 12:17:09 AM9/14/06
to top-po...@googlegroups.com
At 06:50 PM 9/13/2006, Mark wrote:

>-M: Who's flipp'n advice? The FA can't render an institutional opinion,
>so the advice would be just of several people. So? Why do they need a
>flipp'n FA?

My, my.... must have touched a nerve!

The advice might indeed be just that of several people, or it might
be the consensus of millions. It might be the consensus of everyone
in the FA except for one. And thus the entire FA (except perhaps that
one) would move in coordination according to that consensus. Weak?
Not unless the members were *very* weak.

The FA exists to broaden understanding and consensus.

(There is an unresolved question in the definition of FA. FAs don't
take controversial positions. But what if every member agrees? My
view is that it is better if the FA avoids controversy even there is
even one reasonable objection from anyone who might become a member,
if the FA does not adopt the position. If an FA is large enough, I'd
think, it would become less important to consider theoretical outside
objections.)

>The Council on Foreign Relations, because it is a 501(c)3
>tax-exempt(under American federal IRC code), registered as an
>'educational organization', it can't be a 'political advocacy' group.
>It gets around this limitation by forming 'Disappearing Task Forces'.
>The DTF renders a political opinion in the name of the DTF - while the
>CFR simply serves as a *forum* for the formation of DTFs.

Yes. Whether or not this is legitimate depends on who does the
formation. If the CFR forms the DTF, it is responsible for it. If
members of the CFR form the DTF, if the CFR never endorses the DTF,
it could be quite legitimate. Belonging to an organization should not
make one unable to form other organizations.

>Notice how they nail down their organizational procedures in a binding
>way? This is how the enemy works, and if we want to be powerful like
>them, then we should organize ourselves atleast as well as them.

Brilliant.

First of all, they are not "enemies," per se. They are self-centered
and self-motivated, moving within a system in ways that pursue their
own advantage. If we duplicate their procedures, we will discover
that we are just like them, and the system we create will simply
perpetuate the problems, just with new faces. In other words, by
opposing them, we become them. We need something new, something that
takes us outside the existing system. And, if I'm correct, this will
enable us to use the existing system. It is vulnerable to control by
special interest groups. I'm just trying to form a special interest
group of the whole, through what I'll call "the vanguard of the
whole," but, unlike a prior claimant to that, this SIG specifically
avoids crystallizing itself, it seeks, instead, to incorporate and
reflect all elements of society, so that it becomes, not just another
pretender at representing the people, but a true superconsciousness
built from the individual awarenesses of the people, able to act
through them directly and thus to control, as necessary, power structures.

>No flipp'n MUSHY FAs!

Would seem to be a waste, wouldn't it? Look, you think an FA is
useless, don't join it! Simple! But let me suggest this: if you see a
large FA coming at you, don't stand in its way. It will be polite and
careful. To a degree.

> >L: Yes, the FA *is* weak. Just as your brain is weak unless your muscles
> > decide to respond as you desire. The weakness, in the case of an FA,
> > is part of the design. It is, in fact, essential. Without it, the FA
> > becomes yet another oppressive government, or at least that is the risk.
>
>-M: Coercive organization of society is manditory.

too bad. Let me just say that I'd prefer to keep this man's hands as
far away as possible from the actual nuts and bolts of a democratic
system. Coercion may be necessary (and thus "mandatory") under some
circumstances. But an organization rooted in coercion is tyranny, not
democracy.

> >L: Remember, in an FA, how to analyze poll results is really up to the
> > analyst, who is or is engaged by those who will use the results.
>
>-M: Who-the-flip is an analyst? What makes one better than another?

Nothing. Who said that any analyst was better than another? I want to
spend my money, I hire or consult the best analyst available to me.
Or I'm the analyst myself. I'm responsible, it is my effort and my money.

Mark, I really think you have missed the concept, almost entirely.

>Would everyone have their own pet analyst?

Sure. It's called a proxy.

>By contrast, an actual institution would have its analyst selection
>procedures nailed down. Then something could actually get done.

The trains would run on time, crime and deviation would be
eliminated, and disagreement would disappear. One big happy society,
having properly selected Big Brother through the best available
analyst selection procedures. Anyone who disagrees obviously is an
enemy of society, desiring to corrupt the pure analysis of the great System.

>-M: OK, I have seen FAs for community purposes, and they can be
>impressive.
>But politics and other forms of collective *action* are a different
>domain.

FAs engage in collective action. It is just that we have not seen
this extend into politics. I've been expressing the theory of how to
do so. Take it or leave it. I'm certainly not soliciting Mark's
participation. I'd find it, I'm afraid, a bit scary. But he's free to
do so, if he wishes. I'd be inconsistent if I tried to keep him away.

>[...]
> >L: [...] My realization was that emphasizing DP at the beginning made
> > the tablet a bit too large to swallow at first sight.
>
>-M: Then don't use it for board selection, just commitee selection.

The point is missed entirely, but it is too late, morning comes early.

Mark

unread,
Sep 16, 2006, 6:30:32 PM9/16/06
to top-politics

> >-M: Who's flipp'n advice? The FA can't render an institutional opinion,
> >so the advice would be just of several people. So? Why do they need a
> >flipp'n FA?

>L: My, my.... must have touched a nerve!

-M: Yes, something I keep seeing among visionaries is what seems like a
'counter-culture' influence, a "ineffectiveness is a virtue" kind of
approach.

>L: [...]FAs don't take controversial positions.[...]

-M: Then they are weak.
By contrast, an SD2-S group would encourage controversy.

> >The Council on Foreign Relations, because it is a 501(c)3
> >tax-exempt(under American federal IRC code), registered as an
> >'educational organization', it can't be a 'political advocacy' group.
> >It gets around this limitation by forming 'Disappearing Task Forces'.
> >The DTF renders a political opinion in the name of the DTF - while the
> >CFR simply serves as a *forum* for the formation of DTFs.

>L: Yes. Whether or not this is legitimate depends on who does the


> formation. If the CFR forms the DTF, it is responsible for it. If
> members of the CFR form the DTF, if the CFR never endorses the DTF,
> it could be quite legitimate. Belonging to an organization should not
> make one unable to form other organizations.

-M: OK, this is already how it works.

> >Notice how they nail down their organizational procedures in a binding
> >way? This is how the enemy works, and if we want to be powerful like
> >them, then we should organize ourselves atleast as well as them.

>L: Brilliant.

-M: Thanks. :-)

>L: First of all, they are not "enemies," per se.

-M: WTF? Imperialists of the Round Table network aren't our enemy?
http://www.nexusmagazine.com/articles/Rhodes&SecretS3.html

>L: They are self-centered and self-motivated, moving within a system in ways that pursue their own advantage.

-M: Yes, this is how *Empire* ususally works.

>L: If we duplicate their procedures, we will discover that we are just like them, and the system we create will simply perpetuate the problems, just with new faces. In other words, by opposing them, we become them.

-M: NO! The only difference is that we be democratic-republicans!
Case in point, the American Revolution.
There is little structural difference between the British and
Americans.

>L: We need something new, something that takes us outside the existing system.

-M: No, the system exists because it works.
What we need is just an enhanced form of democratic-republicanism, like
SD2-S.

>L:[...] I'm just trying to form a special interest group of the whole, through what I'll call "the vanguard of the whole," but, unlike a prior claimant to that, this SIG specifically avoids crystallizing itself, it seeks, instead, to incorporate and reflect all elements of society, so that it becomes, not just another pretender at representing the people, but a true superconsciousness built from the individual awarenesses of the people, able to act through them directly and thus to control, as necessary, power structures.

-M: Its a non-institution institution, that is hierarchally opposed to
hierachy.

> >No flipp'n MUSHY FAs!

>L: Would seem to be a waste, wouldn't it? Look, you think an FA is


> useless, don't join it! Simple! But let me suggest this: if you see a
> large FA coming at you, don't stand in its way. It will be polite and
> careful. To a degree.

-M: Why should I care about an FA?
How is it going to oppose me if it has no voice?
Likewise, how is it going to fight the enemy?

> > >L: Yes, the FA *is* weak. Just as your brain is weak unless your muscles
> > > decide to respond as you desire. The weakness, in the case of an FA,
> > > is part of the design. It is, in fact, essential. Without it, the FA
> > > becomes yet another oppressive government, or at least that is the risk.

> >-M: Coercive organization of society is manditory.

>L: too bad. Let me just say that I'd prefer to keep this man's hands as


> far away as possible from the actual nuts and bolts of a democratic
> system. Coercion may be necessary (and thus "mandatory") under some
> circumstances. But an organization rooted in coercion is tyranny, not
> democracy.

-M: Organizations are rooted in coercion.
If I stole from someone's church or mosque, shouldn't I go to jail?
How is this tyranny? This seems fair even if the organization isn't
democratic.

[...]


> >By contrast, an actual institution would have its analyst selection
> >procedures nailed down. Then something could actually get done.

>L: The trains would run on time, crime and deviation would be


> eliminated, and disagreement would disappear. One big happy society,
> having properly selected Big Brother through the best available
> analyst selection procedures. Anyone who disagrees obviously is an
> enemy of society, desiring to corrupt the pure analysis of the great System.

-M: Anyone who disagrees should:
1. Change their proxies, and/or
2. Offer a better algorithm, and/or
3. Voice their opinion

Got a better idea?(Your vision above is very trippy and Orwellian.)

> >-M: OK, I have seen FAs for community purposes, and they can be
> >impressive. But politics and other forms of collective *action* are a different
> >domain.

>L: FAs engage in collective action. It is just that we have not seen


> this extend into politics. I've been expressing the theory of how to
> do so. Take it or leave it. I'm certainly not soliciting Mark's
> participation. I'd find it, I'm afraid, a bit scary.

-M: Me, scary?
I am a lovable guy with moderate opinions.

>L: But he's free to do so, if he wishes. I'd be inconsistent if I tried to keep him away.

-M: OK, and you are welcome in any SD2-S forum.

Abd ul-Rahman Lomax

unread,
Sep 17, 2006, 11:55:48 AM9/17/06
to top-politics
Obviously, there is something that I need to explain.

At 06:30 PM 9/16/2006, Mark wrote:

> > >-M: Who's flipp'n advice? The FA can't render an institutional opinion,
> > >so the advice would be just of several people. So? Why do they need a
> > >flipp'n FA?
>
> >L: My, my.... must have touched a nerve!
>
>-M: Yes, something I keep seeing among visionaries is what seems like a
>'counter-culture' influence, a "ineffectiveness is a virtue" kind of
>approach.

Except that my position is that ineffective is far from a virtue. So
if Mark is seeing this in me, I'd suggest that his vision is like
that of a lemming, i.e., it only sees what its history has trained it to see.


> >L: [...]FAs don't take controversial positions.[...]
>
>-M: Then they are weak.
>By contrast, an SD2-S group would encourage controversy.

First of all, FAs *engage* people where controversies exist. They
encourage consensus, but not the dead consensus of "you'd better
agree with us or you're history, or irrelevent," but *real*
consensus, that is, broad agreement among the membership forged from
a thorough examination of a controversy. When possible. If not, it
leaves the factions unmolested and free to act independently.

Now, suppose I have a circle of friends. Some of those friends are
dreamers and couldn't start a fire if they had a match. Others are
powerful, influential, wealthy. Now, would you say that my circle is "weak'?

What Mark seems to be butting his head up against is a wall, an
assumption that a weak organization cannot accomplish anything. What
a weak organization can do is to bring factions together in a neutral
space. If it is weak, it has no power to force this, but it may
enable it. However, the factions vary in strength. (We call them
"caucuses." No faction, alone, can control the society, unless it is
so large that we no longer call it a faction, we call it the
majority. Or the general consensus, to take it even further.

The FA allows caucuses to measure their relative strength. This is
why the analysis is distributed, in the default proposals we are
making. Anyone can analyze polls using proxy lists. And what we
pointed out was that there is always incentive for a caucus to extend
agreement on an issue. Do the math. Suppose that there are two sides
to an issue, and each side has equal resources per member, available
for political use. A simple majority will result in a net effect
close to zero. As the majority ratio increases, the subtractive
effect decreases, and the net resources climb. A two-thirds majority
is one-third as effective as a full consensus.

At a certain point, the broad agreement is sufficient that the actors
decide to act. Who decides what level of consensus is necessary? *The
actors do, those whose resources will be dedicated to the cause.*

The FA is weak, in itself, as a single organization, but it is not
weak in terms of its overall effect. It is a consensus organization,
generally, and traditionally consensus organizations become weaker
the larger they are, because of the difficulty of finding consensus.
However, in this case, the organization's very weakness is a necessary feature:

The FA has no power to prevent a caucus from acting whenever the
caucus decides to act. It has no sequestered member resources, as
dues, fees, accumulated property, bequests, etc. If if were not weak,
it would not function with the power that we expect.

That is, the overall function, in context, including the individual
actions of its members, is powerful. It, in itself, is like a brain
without the power to act. But together with its members, it is an
intelligent organism with the full resources of its members behind
it, potentially.

Given that Mark has clearly not understood this, the rest of this
discussion is premature, and I've got a wife and children begging --
or demanding -- that I go out the door with them.

Abd ul-Rahman Lomax

unread,
Sep 17, 2006, 11:04:58 PM9/17/06
to top-po...@googlegroups.com
At 06:30 PM 9/16/2006, Mark wrote:
> >L: First of all, they are not "enemies," per se.
>
>-M: WTF? Imperialists of the Round Table network aren't our enemy?
>http://www.nexusmagazine.com/articles/Rhodes&SecretS3.html
>
> >L: They are self-centered and self-motivated, moving within a
> system in ways that pursue their own advantage.
>
>-M: Yes, this is how *Empire* ususally works.

Yes. And if you oppose them, you become them. Under a new name.

Not opposing them does not mean being ignorant. It does not mean
failing to use the power that we have in the service of higher goals.
And, indeed, we have power. More than them. However, if we organize
to oppose them, as we grow, they will, for sure, notice. And the
revolution, if it is to come, will come too soon. It will come at a
time when the only way it can succeed is to reproduce the same
coercive methods. Or worse.

No, what I'm proposing is that *we* organize. And "we" includes
"them." Our organization will not oppose them. It will not seek their
destruction, nor even their removal from power. It just might,
however, make them irrelevant. Without harming them.

Being an oligarch can be extremely dangerous. It can result in the
extermination of your entire family.

The FA/DP revolution will not destroy anything. It leaves existing
structures in place. One way of looking at it is that it calls the
bluff of the oligarchs. The oligarchs generally claim to have the
best interests of society at heart. (This claim is very old, it is
the divine right of kings.) Okay, so if we organize to create a
neutral intelligence, one capable of understanding, better than any
individual, what is true benefit to society, that intelligence will
not remove the oligarchs from power, it will leave them to exercise
their benevolent function. There will still be traditional leaders.
FA/DP, remember, is not a power structure. It is a communcation &
intelligence structure. It does not replace existing systems, it does
not, in itself, oppose them. And I do not think it would generate
action to oppose them *unless* clear intelligence led to that
conclusion. Such action would be far less likely to cause the kind of
destruction we have seen from revolutions in the past. It would be
far more likely to find consensus solutions. It would be more likely
to discover ways to gently and gradually change. Shock, in general,
is not good for society.

Indeed, this is an argument frequently used by the oligarchs as why
they need to hold onto power. It is to stave off chaos and mass disruption.

I'm very serious about this. I have the Communist Party of China in
mind, but it applies to most other oligarchical structures. The
people who effectively control nonprofit organizations frequently
justify this as being necessary, because they are the ones, they
assert -- and believe -- who know what's best.

> >L: If we duplicate their procedures, we will discover that we are
> just like them, and the system we create will simply perpetuate the
> problems, just with new faces. In other words, by opposing them, we
> become them.
>
>-M: NO! The only difference is that we be democratic-republicans!
>Case in point, the American Revolution.
>There is little structural difference between the British and
>Americans.

It's not a good example, actually. The American Revolution did not
make deep structural changes. It did change the way the king is
chosen, which certainly opens up possibilities.

> >L: We need something new, something that takes us outside the
> existing system.
>
>-M: No, the system exists because it works.

*Of course it exists because it works.* Once again, Mark hasn't
recognized what is being suggested.

>What we need is just an enhanced form of democratic-republicanism, like
>SD2-S.

This is like saying that, if you have pneumonia, you need fluids.
Yes, you need fluids. But you may need much more, if you are to truly
recover. Mark, you have a false hope. You think that this tweak will
untie the knot.

It might be a *piece* of what will untie it. If it is a good power
structure, it might, for example, be used by caucuses. Indeed, if it
is flexible enough, it might be used in an FA. I don't know enough about it.

But, remember, and I know this has been missed, caucuses in an FA can
be quite the same as traditional organizations; indeed they can *be*
traditional organizations.

Consider that an FA/DP organization composed primarily of Democrats
arises, and starts to gain some traction. It is open membership. It
has to be to be an FA. A group of Republicans notice this and decide
to join. Why not? Joining doesn't provide funding to the Democratic
FA, it does not signify approval, it only *connects.* So these
Republicans not only join, but they encourage all Republicans to
join. Indeed, so many Republicans join that there are more
Republicans in the FA than Democrats. What happens?

Oh, I should mention. All it takes is one Republican to join and be
active. The rest can join and can name, directly or indirectly, that
one as their proxy. They don't have to personally slog through all
the progressive garbage. As they might see it....

Can these Republicans torpedo the Democratic initiatives. No, they
can't. All they can do is to present arguments in a neutral forum,
one that is able to boil down issues by confining the discussion to a
relatively small number of participants. Every meeting has its own
rules. A large FA will have *many* meetings, it may have more than
one meeting that could be called "top level." If the Republicans
really are just one person representing many who have signed up but
aren't active, well, suddenly there is one Republican to talk to,
massively trusted by Republicans. And this person is talking with
Democrats. But, in fact, there will be more than one. There will be
many. Nevertheless, at the top levels, where meetings have more
stringent participation requirements -- they *must*, or the noise
will make them truly useless -- there will be a relative few, and
collectively they will represent almost all the Republicans. The same
situation will exist for the Democrats. The FA has become a mirror of
society, and a place to rapidly test for consensus.

I think it would transform society. But we have not yet even seen one
FA/DP organization. There are plenty of FAs, but the combination with
DP, which makes the organization scalable, has never been tried. It
is very difficult to try it, because organizations of a size where DP
becomes critically useful already have developed methods of dealing
with the problem of scale, and these methods are conservative, that
is, they preserve themselves. Rare would be the organization which
would give them up in order to return power to the membership.

There are only a few critical points where the FA/DP concepts can be
introduced. Sooner or later, one of the seeds we are sending out will sprout.

> >No flipp'n MUSHY FAs!

> >L: Would seem to be a waste, wouldn't it? Look, you think an FA is
> > useless, don't join it! Simple! But let me suggest this: if you see a
> > large FA coming at you, don't stand in its way. It will be polite and
> > careful. To a degree.
>
>-M: Why should I care about an FA?
>How is it going to oppose me if it has no voice?
>Likewise, how is it going to fight the enemy?

You have thought that because the FA, in itself, has no voice, that
the entire dharma, the entire phenomenon, has no voice. No, the
entire phenomenon has *many* voices, many hands, more money than you
could dream of, more connections, more power. When it is not united,
it is weak. That's designed in!

But when it is united, there is nothing that could stop it.

If you think an FA/DP organization is about to make a decision that
will harm you, I'd highly suggest making sure that your arguments and
your needs are considered! And you can do that by joining and finding
*anyone* already participating in the organization who will listen to
you long enough to understand them.

>-M: Organizations are rooted in coercion.

As I wrote, keep this man away from the tools of democracy.

>If I stole from someone's church or mosque, shouldn't I go to jail?
>How is this tyranny? This seems fair even if the organization isn't
>democratic.

Nobody has said that coercion is tyranny, per se. Unjust coercion is
tyranny. And coercion, undertaken through fixed rules by an
organization, tends to be unjust at times. Coercion is dangerous.
Indeed, it may be necessary, I'm not a Libertarian, as such. But so
to might some dangerous drug be necessary to cure an illness. It does
not mean that one may routinely use it without hazard.

Coercion alienates. It divides society into the coerced and the
coercers. The legitimate motto of the police is "Protect and Serve."
But for some segments of society, this has become "Harass and
Humiliate." Police, practically by definition, have the power of
coercion. That power is quite dangerous, it requires careful
supervision, which is often lacking. And it is much more convenient
to simply make a behavior unlawful and punish the offenders than it
is to discover the root causes of an undesirable behavior and
eliminate or redirect them.

Lao Tsu knew this thousands of years ago....


>[...]
> > >By contrast, an actual institution would have its analyst selection
> > >procedures nailed down. Then something could actually get done.
>
> >L: The trains would run on time, crime and deviation would be
> > eliminated, and disagreement would disappear. One big happy society,
> > having properly selected Big Brother through the best available
> > analyst selection procedures. Anyone who disagrees obviously is an
> > enemy of society, desiring to corrupt the pure analysis of the
> great System.
>
>-M: Anyone who disagrees should:
>1. Change their proxies, and/or
>2. Offer a better algorithm, and/or
>3. Voice their opinion
>
>Got a better idea?(Your vision above is very trippy and Orwellian.)

What an amazing coincidence!


> > >-M: OK, I have seen FAs for community purposes, and they can be
> > >impressive. But politics and other forms of collective *action*
> are a different
> > >domain.
>
> >L: FAs engage in collective action. It is just that we have not seen
> > this extend into politics. I've been expressing the theory of how to
> > do so. Take it or leave it. I'm certainly not soliciting Mark's
> > participation. I'd find it, I'm afraid, a bit scary.
>
>-M: Me, scary?
>I am a lovable guy with moderate opinions.

Actually, so were quite a number of those who assisted Adolf Hitler.
I'm not making any accusations here, but the arguments being advanced
are eerily familiar. Impatience with democratic process, readiness to
coerce....

> >L: But he's free to do so, if he wishes. I'd be inconsistent if I
> tried to keep him away.
>
>-M: OK, and you are welcome in any SD2-S forum.

And who is Mark that he can make me welcome? Will he run all these forums?

He might design the forum such that anyone can join. But joining is
not welcome, per se. Welcoming is something that *people* do.
Software can only imitate it. And imitation welcome is far more
bitter than imitation sugar.

"Thanks for sending us input, it is valuable to us. We welcome
comments." But then there is no reply that actually shows that anyone
read the contribution.

Mark

unread,
Sep 21, 2006, 4:46:36 PM9/21/06
to top-politics

Abd ul-Rahman Lomax wrote:
> At 06:30 PM 9/16/2006, Mark wrote:
> > >L: First of all, they are not "enemies," per se.
> >
> >-M: WTF? Imperialists of the Round Table network aren't our enemy?
> >http://www.nexusmagazine.com/articles/Rhodes&SecretS3.html
> >
> > >L: They are self-centered and self-motivated, moving within a
> > system in ways that pursue their own advantage.
> >
> >-M: Yes, this is how *Empire* ususally works.

>L: Yes. And if you oppose them, you become them. Under a new name.


> Not opposing them does not mean being ignorant. It does not mean
> failing to use the power that we have in the service of higher goals.
> And, indeed, we have power. More than them. However, if we organize
> to oppose them, as we grow, they will, for sure, notice. And the
> revolution, if it is to come, will come too soon. It will come at a
> time when the only way it can succeed is to reproduce the same
> coercive methods. Or worse. No, what I'm proposing is that *we* organize. And "we" includes "them." Our organization will not oppose them. It will not seek their
> destruction, nor even their removal from power. It just might, however, make them irrelevant. Without harming them.

-M: How do you make them irrelevant without removing them from power?

>L: Being an oligarch can be extremely dangerous. It can result in the


> extermination of your entire family.

-M: My hero, Machiavelli, made this clear.

>L: The FA/DP revolution will not destroy anything. It leaves existing
> structures in place.

-M: So does my approach. You aren't making a comparison.

>L: One way of looking at it is that it calls the


> bluff of the oligarchs. The oligarchs generally claim to have the
> best interests of society at heart. (This claim is very old, it is
> the divine right of kings.) Okay, so if we organize to create a
> neutral intelligence, one capable of understanding, better than any
> individual, what is true benefit to society, that intelligence will
> not remove the oligarchs from power, it will leave them to exercise
> their benevolent function. There will still be traditional leaders.

-M: If the assholes stay, the result is Assholism. :-(

>L: FA/DP, remember, is not a power structure. It is a communcation &


> intelligence structure. It does not replace existing systems, it does
> not, in itself, oppose them. And I do not think it would generate
> action to oppose them *unless* clear intelligence led to that
> conclusion. Such action would be far less likely to cause the kind of
> destruction we have seen from revolutions in the past. It would be
> far more likely to find consensus solutions. It would be more likely
> to discover ways to gently and gradually change. Shock, in general,
> is not good for society. Indeed, this is an argument frequently used by the oligarchs as why they need to hold onto power. It is to stave off chaos and mass disruption.

-M: Fuck'em. They need to be replaced with trustworthy leadership.

>L: I'm very serious about this. I have the Communist Party of China in


> mind, but it applies to most other oligarchical structures. The
> people who effectively control nonprofit organizations frequently
> justify this as being necessary, because they are the ones, they
> assert -- and believe -- who know what's best.

-M: If they are using the best leadership selection methods available,
then they probably *do* know what is best.

> > >L: If we duplicate their procedures, we will discover that we are
> > just like them, and the system we create will simply perpetuate the
> > problems, just with new faces. In other words, by opposing them, we
> > become them.
> >
> >-M: NO! The only difference is that we be democratic-republicans!
> >Case in point, the American Revolution. There is little structural difference between the British and Americans.

>L: It's not a good example, actually. The American Revolution did not


> make deep structural changes. It did change the way the king is
> chosen, which certainly opens up possibilities.

-M: My point exactly. The Revolution was quite effective, despite
little structural differences.

> > >L: We need something new, something that takes us outside the
> > existing system.
> >
> >-M: No, the system exists because it works.

>L: *Of course it exists because it works.* Once again, Mark hasn't


> recognized what is being suggested.

-M: You are suggesting a communication network with no collective
voice.

> >What we need is just an enhanced form of democratic-republicanism, like
> >SD2-S.

>L: This is like saying that, if you have pneumonia, you need fluids.


> Yes, you need fluids. But you may need much more, if you are to truly
> recover. Mark, you have a false hope. You think that this tweak will
> untie the knot. It might be a *piece* of what will untie it. If it is a good power
> structure, it might, for example, be used by caucuses. Indeed, if it

> is flexible enough, it might be used in an FA. I don't know enough about it.[...]

-M: Do you understand the concept of Markov Chain convergence?

> > >No flipp'n MUSHY FAs!
>
> > >L: Would seem to be a waste, wouldn't it? Look, you think an FA is
> > > useless, don't join it! Simple! But let me suggest this: if you see a
> > > large FA coming at you, don't stand in its way. It will be polite and
> > > careful. To a degree.
> >
> >-M: Why should I care about an FA?
> >How is it going to oppose me if it has no voice?
> >Likewise, how is it going to fight the enemy?
>
> You have thought that because the FA, in itself, has no voice, that
> the entire dharma, the entire phenomenon, has no voice. No, the
> entire phenomenon has *many* voices, many hands, more money than you
> could dream of, more connections, more power. When it is not united,
> it is weak. That's designed in!

-M: You are designing in non-unity and weakness. :-(

>L: But when it is united, there is nothing that could stop it.

-M: This is why SD2-S us united under a centrality algorithm.

> If you think an FA/DP organization is about to make a decision that
> will harm you, I'd highly suggest making sure that your arguments and
> your needs are considered! And you can do that by joining and finding
> *anyone* already participating in the organization who will listen to
> you long enough to understand them.
>
> >-M: Organizations are rooted in coercion.

>L: As I wrote, keep this man away from the tools of democracy.

-M: I don't *like* coercion - I am a peaceful and loving man, I simply
know how things innately work.

Abd ul-Rahman Lomax

unread,
Sep 21, 2006, 11:14:43 PM9/21/06
to top-po...@googlegroups.com
At 04:46 PM 9/21/2006, Mark wrote:

> > No, what I'm proposing is that *we* organize. And "we" includes
> "them." Our organization will not oppose them. It will not seek their
> > destruction, nor even their removal from power. It just might,
> however, make them irrelevant. Without harming them.
>
>-M: How do you make them irrelevant without removing them from power?

I didn't say that they would not be removed from power. I wrote that
"our organization will not [seek that]." And, in fact, it might not
be necessary. Depends on how they respond. What FA/DP organization
could do, potentially, is to make their power far less significant.
Even tyrants are powerless if people, collectively, decide not to
follow them. There have been some striking examples in recent
history: Iran and the fall of the Shah, the Philippines. When nearly
all the people went into the streets, the army simply joined them.
There was actually little bloodshed. At first. In Iran, things did
not go so well *after* the revolution, pointing out the danger of
overthrowing a government when what will replace it is untried.

> >L: One way of looking at it is that it calls the
> > bluff of the oligarchs. The oligarchs generally claim to have the
> > best interests of society at heart. (This claim is very old, it is
> > the divine right of kings.) Okay, so if we organize to create a
> > neutral intelligence, one capable of understanding, better than any
> > individual, what is true benefit to society, that intelligence will
> > not remove the oligarchs from power, it will leave them to exercise
> > their benevolent function. There will still be traditional leaders.
>
>-M: If the assholes stay, the result is Assholism. :-(

Of course, the incompetent and vicious won't be allowed to remain in
positions of leadership.... What FA/DP is likely to do is to make it
all quite visible. FA/DP creates an intelligence network, bottom-up.
If I didn't think that the FA principles and other aspects of the
system would be protective, I'd think it quite sinister. Big Brother
on steroids. But it wouldn't be Big Brother, it would be more like Big Family.

> >L: I'm very serious about this. I have the Communist Party of China in
> > mind, but it applies to most other oligarchical structures. The
> > people who effectively control nonprofit organizations frequently
> > justify this as being necessary, because they are the ones, they
> > assert -- and believe -- who know what's best.
>
>-M: If they are using the best leadership selection methods available,
>then they probably *do* know what is best.

Not necessarily. The "best" available can leave the leadership
seriously out of touch with the rank and file membership. Leadership
selection still gives us leaders who are only individuals. If the
overall system can collect the intelligence of the members, then it
will know better than even the best individuals.

> > >-M: No, the system exists because it works.
>
> >L: *Of course it exists because it works.* Once again, Mark hasn't
> > recognized what is being suggested.
>
>-M: You are suggesting a communication network with no collective
>voice.

Not quite. There *is* a collective voice, but it is not expressed
through the FA/DP structure *as a whole*. However, the individual
members, including high-level proxies, still have voices, and,
further, they have a way of knowing with reasonable ease whether or
not their views are isolated or represent broad agreement or even consensus.

*Parts* of the network are used to coordinate the expression of these
voices. There is not *one* voice, in that sense there is "no
collective voice." But there may be one voice, the collective voice
of a large caucus, which is very strong, very clear, and with great
power behind it, and other voices that are weak. FA/DP does not
suppress and hide dissent.

>-M: Do you understand the concept of Markov Chain convergence?

No, at least not under that name. Should I?

> > You have thought that because the FA, in itself, has no voice, that
> > the entire dharma, the entire phenomenon, has no voice. No, the
> > entire phenomenon has *many* voices, many hands, more money than you
> > could dream of, more connections, more power. When it is not united,
> > it is weak. That's designed in!
>
>-M: You are designing in non-unity and weakness. :-(

No, I'm designing in weakness in the absence of unity. That is, the
strength of a position taken by members of the organization varies
with the measure of the degree of consensus enjoyed.


> >L: But when it is united, there is nothing that could stop it.
>
>-M: This is why SD2-S us united under a centrality algorithm.

*I don't want to create unity, artificially.* In other words, I very
specifically don't want to create an unstoppable organization, unless
that unstoppability is a direct result of consensus.


> > If you think an FA/DP organization is about to make a decision that
> > will harm you, I'd highly suggest making sure that your arguments and
> > your needs are considered! And you can do that by joining and finding
> > *anyone* already participating in the organization who will listen to
> > you long enough to understand them.
> >
> > >-M: Organizations are rooted in coercion.
>
> >L: As I wrote, keep this man away from the tools of democracy.
>
>-M: I don't *like* coercion - I am a peaceful and loving man, I simply
>know how things innately work.

Organizations are not rooted in coercion. They are rooted in
cooperation. At least most modern and successful organizations are.
There are *aspects* of coercion which afflict most, but they are
*not* at the root.

Top-down organizations are generally more coercive than others.

Mark

unread,
Sep 22, 2006, 2:52:51 PM9/22/06
to top-politics

> > > No, what I'm proposing is that *we* organize. And "we" includes
> > "them." Our organization will not oppose them. It will not seek their
> > > destruction, nor even their removal from power. It just might,
> > however, make them irrelevant. Without harming them.

> >-M: How do you make them irrelevant without removing them from power?

>L: I didn't say that they would not be removed from power. I wrote that


> "our organization will not [seek that]." And, in fact, it might not
> be necessary. Depends on how they respond. What FA/DP organization
> could do, potentially, is to make their power far less significant.
> Even tyrants are powerless if people, collectively, decide not to
> follow them. There have been some striking examples in recent
> history: Iran and the fall of the Shah, the Philippines. When nearly
> all the people went into the streets, the army simply joined them.
> There was actually little bloodshed. At first. In Iran, things did
> not go so well *after* the revolution, pointing out the danger of
> overthrowing a government when what will replace it is untried.

-M: OK. This is what I want from SD2-S.

> > >L: One way of looking at it is that it calls the
> > > bluff of the oligarchs. The oligarchs generally claim to have the
> > > best interests of society at heart. (This claim is very old, it is
> > > the divine right of kings.) Okay, so if we organize to create a
> > > neutral intelligence, one capable of understanding, better than any
> > > individual, what is true benefit to society, that intelligence will
> > > not remove the oligarchs from power, it will leave them to exercise
> > > their benevolent function. There will still be traditional leaders.

> >-M: If the assholes stay, the result is Assholism. :-(

>L: Of course, the incompetent and vicious won't be allowed to remain in


> positions of leadership.... What FA/DP is likely to do is to make it
> all quite visible. FA/DP creates an intelligence network, bottom-up.
> If I didn't think that the FA principles and other aspects of the
> system would be protective, I'd think it quite sinister. Big Brother
> on steroids. But it wouldn't be Big Brother, it would be more like Big Family.

-M: Same with SD2-S, I just think that my algorithms are better.

> > >L: I'm very serious about this. I have the Communist Party of China in
> > > mind, but it applies to most other oligarchical structures. The
> > > people who effectively control nonprofit organizations frequently
> > > justify this as being necessary, because they are the ones, they
> > > assert -- and believe -- who know what's best.
> >
> >-M: If they are using the best leadership selection methods available,
> >then they probably *do* know what is best.

>L: Not necessarily. The "best" available can leave the leadership


> seriously out of touch with the rank and file membership.

-M: Then they aren't the *best*. I will repeat:
*If they are using the best leadership selection methods available,


then they probably *do* know what is best."

>L: Leadership selection still gives us leaders who are only individuals. If the


> overall system can collect the intelligence of the members, then it
> will know better than even the best individuals.

-M: The *best* leaders will gather the collective intelligence or they
aren't the *best*.

> > > >-M: No, the system exists because it works.
> >
> > >L: *Of course it exists because it works.* Once again, Mark hasn't
> > > recognized what is being suggested.
> >
> >-M: You are suggesting a communication network with no collective
> >voice.

>L: Not quite. There *is* a collective voice, but it is not expressed


> through the FA/DP structure *as a whole*. However, the individual
> members, including high-level proxies, still have voices, and,
> further, they have a way of knowing with reasonable ease whether or
> not their views are isolated or represent broad agreement or even consensus.
> *Parts* of the network are used to coordinate the expression of these
> voices. There is not *one* voice, in that sense there is "no
> collective voice." But there may be one voice, the collective voice
> of a large caucus, which is very strong, very clear, and with great
> power behind it, and other voices that are weak. FA/DP does not
> suppress and hide dissent.

-M: A measureable communication network.

> >-M: Do you understand the concept of Markov Chain convergence?

>L: No, at least not under that name. Should I?

-M: Its what allows for *Augmented Democracy*, in which there is *peer
selection* instead of popular selection, in a democratic environment.
This allows the best leaders to select the best leaders, with no
predetermination of what a *best leader* is.

> > > You have thought that because the FA, in itself, has no voice, that
> > > the entire dharma, the entire phenomenon, has no voice. No, the
> > > entire phenomenon has *many* voices, many hands, more money than you
> > > could dream of, more connections, more power. When it is not united,
> > > it is weak. That's designed in!
> >
> >-M: You are designing in non-unity and weakness. :-(
>
> No, I'm designing in weakness in the absence of unity. That is, the
> strength of a position taken by members of the organization varies
> with the measure of the degree of consensus enjoyed.

-M: This sounds too populistic.

> > >L: But when it is united, there is nothing that could stop it.
> >

> >-M: This is why SD2-S is united under a centrality algorithm.

>L: *I don't want to create unity, artificially.* In other words, I very


> specifically don't want to create an unstoppable organization, unless
> that unstoppability is a direct result of consensus.

-M: Centrality algorithms are what measure consensus. Did you know
this?
I will suggest that PageRank is the least arbitrary(artificial) of
centrality algorithms.

> > > If you think an FA/DP organization is about to make a decision that
> > > will harm you, I'd highly suggest making sure that your arguments and
> > > your needs are considered! And you can do that by joining and finding
> > > *anyone* already participating in the organization who will listen to
> > > you long enough to understand them.
> > >
> > > >-M: Organizations are rooted in coercion.
> >
> > >L: As I wrote, keep this man away from the tools of democracy.
> >
> >-M: I don't *like* coercion - I am a peaceful and loving man, I simply
> >know how things innately work.

>L: Organizations are not rooted in coercion. They are rooted in
> cooperation.

-M: Organizations are rooted in both coercion and cooperation. An
organization is usually a court recognized entity, as such, is backed
by the guns of government.
And guess what happens if an organization doesn't pay its taxes. It
gets sanctioned by the government, who can potentially gather
collections by force, if need be.

>L: At least most modern and successful organizations are. There are *aspects* of coercion which afflict most, but they are *not* at the root.

-M: Try gathering a group of people together as a collective without
giving them a government issued entity status: "Just *understand* that
we are a collective" they say," and just *understand* that this is our
treasury."

Lomax, without government(coercive) recognition of agreements,
organizations would be unstable.

>L: Top-down organizations are generally more coercive than others.


Nobody has said that coercion is tyranny, per se. Unjust coercion is
tyranny.

-M: OK.

>L: And coercion, undertaken through fixed rules by an


organization, tends to be unjust at times. Coercion is dangerous.
Indeed, it may be necessary, I'm not a Libertarian, as such. But so
to might some dangerous drug be necessary to cure an illness. It does
not mean that one may routinely use it without hazard. Coercion
alienates. It divides society into the coerced and the coercers. The
legitimate motto of the police is "Protect and Serve." But for some
segments of society, this has become "Harass and
Humiliate." Police, practically by definition, have the power of
coercion. That power is quite dangerous, it requires careful
supervision, which is often lacking. And it is much more convenient to
simply make a behavior unlawful and punish the offenders than it
is to discover the root causes of an undesirable behavior and eliminate
or redirect them.
Lao Tsu knew this thousands of years ago....

-M: OK. I think that we should steer conversation toward centrality
algorithms.

>[...]
> > >By contrast, an actual institution would have its analyst selection
> > >procedures nailed down. Then something could actually get done.

> >L: The trains would run on time, crime and deviation would be
> > eliminated, and disagreement would disappear. One big happy society,
> > having properly selected Big Brother through the best available
> > analyst selection procedures. Anyone who disagrees obviously is an
> > enemy of society, desiring to corrupt the pure analysis of the
> great System.

>-M: Anyone who disagrees should:
>1. Change their proxies, and/or
>2. Offer a better algorithm, and/or
>3. Voice their opinion
>Got a better idea?(Your vision above is very trippy and Orwellian.)

>L: What an amazing coincidence!

-M: I am only suggesting democracy. Just democracy with a better
centrality algorithm, then you change it into a 1984 fantasy. :-(

> > >-M: OK, I have seen FAs for community purposes, and they can be
> > >impressive. But politics and other forms of collective *action*
> are a different domain.

> >L: FAs engage in collective action. It is just that we have not seen
> > this extend into politics. I've been expressing the theory of how to
> > do so. Take it or leave it. I'm certainly not soliciting Mark's
> > participation. I'd find it, I'm afraid, a bit scary.

>-M: Me, scary? I am a lovable guy with moderate opinions.

>L: Actually, so were quite a number of those who assisted Adolf Hitler.

-M: Hay Caramba! :-(

>L: I'm not making any accusations here, but the arguments being advanced


are eerily familiar. Impatience with democratic process, readiness to
coerce....

-M: No! Democracy already is coercive.
I am only suggesting changing the centrality algorithm.

> >L: But he's free to do so, if he wishes. I'd be inconsistent if I
> tried to keep him away.

>-M: OK, and you are welcome in any SD2-S forum.

>L: And who is Mark that he can make me welcome? Will he run all these forums?

-M: I will run the algorithms, and the algorithms will be open source.

>L: He might design the forum such that anyone can join. But joining is


not welcome, per se. Welcoming is something that *people* do. Software
can only imitate it. And imitation welcome is far more bitter than
imitation sugar.
"Thanks for sending us input, it is valuable to us. We welcome
comments." But then there is no reply that actually shows that anyone
read the contribution.

-M: Strawman argument.

Abd ul-Rahman Lomax

unread,
Sep 22, 2006, 11:56:49 PM9/22/06
to top-po...@googlegroups.com
At 02:52 PM 9/22/2006, Mark wrote:
>-M: Centrality algorithms are what measure consensus. Did you know
>this?
>I will suggest that PageRank is the least arbitrary(artificial) of
>centrality algorithms.

If this is the case, then presumably those leading caucuses will use
that algorithm. However, while I'm not a student of Machiavelli, I'd
suggest that the measure of consensus used by caucus leaders might be
more pragmatic. That is, it might be more on the line of "How much
money and votes and energy can our opposition gather? How much damage
will it cause if we fight instead of cooperating?"

What I've been suggesting is that within an FA/DP context, there is
strong motivation to find broad consensus before acting. It is *not*
a requirement. There is no program saying "Act now, consensus has
reached level IV," unless, of course, a caucus decides to obtain and
use such a program. Because power is not centralized in the FA/DP
organization, but remains distributed (or, to be sure, concentrated
in other organizations involving members, which organizations may
have any structure whatever, including classic top-down,
my-way-or-the-highway leadership), it is not necessary, for FA/DP
purposes, to determine a complex common decision algorithm, though
the org may use practically any. The DP structure makes it much less
important, because an organization that gets it wrong will simply be ignored.

Present organizations tend to be gatekeepers to the membership (which
is necessary for various reasons) which means that members
dissatisfied with leadership often have few practical alternatives.
"If you don't like it, go start your own Sierra Club!"

If the Sierra Club were DP, a significant caucus could do exactly
that, practically overnight. But there would still be the substantial
assets held by the original organization. Members who have worked to
build those assets might be quite reluctant to leave them behind.

If it were FA/DP, there would be no substantial collected assets. The
assets remain with the members, but may still be expended as needed,
for they will be contributed as needed. (Technically, the assets may
go, as national AA donations go, to a corporation which uses them as
its board decides, but this corporation itself keeps little reserve,
keeping it heavily dependent upon the continued active support of
those who fund it.)


> > > > If you think an FA/DP organization is about to make a decision that
> > > > will harm you, I'd highly suggest making sure that your arguments and
> > > > your needs are considered! And you can do that by joining and finding
> > > > *anyone* already participating in the organization who will listen to
> > > > you long enough to understand them.
> > > >
> > > > >-M: Organizations are rooted in coercion.
> > >
> > > >L: As I wrote, keep this man away from the tools of democracy.
> > >
> > >-M: I don't *like* coercion - I am a peaceful and loving man, I simply
> > >know how things innately work.
>
> >L: Organizations are not rooted in coercion. They are rooted in
> > cooperation.
>
>-M: Organizations are rooted in both coercion and cooperation. An
>organization is usually a court recognized entity, as such, is backed
>by the guns of government.

"An organization is usually a court recognized entity." Some
organizations are. Others can be if they need to be. An
unincorporated association can have standing, sometimes, in court.
But this does not mean that it is rooted in coercion. It means that
it may call upon coercion if it needs to. It may never need to.

There are essentially two kinds of organizations. One kind is a
control structure created by an oligarchy to develop, extend, and
maintain its power. Very often such organizations devolve to a single
person as strong leader. Does this involve coercion? Not necessarily.
If I hire an employee, is providing instruction to that employee, in
a context where it is understood that failure to perform may lead to
dismissal, coercive?

Possibly. It would depend on details of conduct, I'd suggest. If
dismissal is punitive, yes, it would be coercive. But dismissal could
simply mean "I can't afford to keep this person on the payroll, since
I'm not getting any work from him."

It is much more straightforward to consider such work a trade. I give
you A in return for you giving me B. Is this coercive? It might be.
But, and this is my point, normally it is not. Normally, in most
societies, such trades are voluntary, neither party experiences them
as coercive, *so they are not coercive.*

And such trades, including more complex cooperation, are the
foundation of society. Substantial coercion is not the daily
experience of *most* of us, I'd suggest, in healthy societies. But
people on the fringes, disadvantaged, or, indeed, oppressed, may
certainly see it differently. On the other hand, there are ancient
habits of coercion that many of us take for granted. Taxes are
generally coercive. Voluntary payments for services are not. An
agency which collects voluntary payments may have some protective
service that prevents theft. We could say that private property is
inherently coercive, which might be the theory behind what Mark has
written. But we don't ordinarily consider it coercive if I exercise
private property rights in a non-offensive way, and we consider theft
a form of coercion. (And thus the use of power to prevent theft is
not a primary coercion, but is necessary to *prevent* coercion.)

>And guess what happens if an organization doesn't pay its taxes. It
>gets sanctioned by the government, who can potentially gather
>collections by force, if need be.

Sure. If it has anything to be taxed. Or any assets to seize.

> >L: At least most modern and successful organizations are. There
> are *aspects* of coercion which afflict most, but they are *not* at the root.
>
>-M: Try gathering a group of people together as a collective without
>giving them a government issued entity status: "Just *understand* that
>we are a collective" they say," and just *understand* that this is our
>treasury."

Aha! "Treasury." What's in that treasury? Bottle caps? If me and my
friends collect bottle caps and put them in our treasury, there is no
taxable activity and no need to obtain government issued entity
status. And we could do down a whole alley of thought regarding the
nature of money. Do we really need this here?

Suffice it to say that handling money brings us into the province
that governments typically consider theirs to rule. If I want to form
a voting collective, where we organize to vote in concert, there is
no requirement to register the organization. There may even be a
common treasury if it is handled in certain ways, and, in particular,
if the sums collected are relatively small, and they are directly
expended for the purpose, which is a purpose not for profit, and they
do not benefit any individual, except in trade. (I.e., if the
treasury buys coffee, it benefits the coffee vendor.) This is done
all the time! "Unincorporated Association" is what it is called,
legally. Legally, the association has no existence apart from its
members. If there is a treasurer, the treasurer holds funds in trust.
An unincorporated association is similar, legally, to a partnership
of the members. A treasurer who handles the funds of the association
might end up getting taxed if they take certain actions. For example,
if they hire an employee. But unincorporated associations don't
generally do that....

>Lomax, without government(coercive) recognition of agreements,
>organizations would be unstable.

No, some would be unstable. Recognition of agreements, however, is
not only a governmental function, it is an interpersonal function,
part of the social glue. We can hardly call it "coercion" if I'm
expected to keep a promise I made....

It's my opinion that large FA/DP organizations would be quite stable,
and, by definition, they would not be founded on coercion, not even
the coercion involved in the inviolability of contracts. They do not
depend on the performance of specific agreements, the kind of
agreements that could be enforced in a court.

(But, again, power is coordinated through caucuses, which may use
standard legal forms, and which would thus be legal entities engaging
in contracts, etc. Typically, if the caucus uses a corporate form for
this purpose, the members involved are independent of and are not
liable for the acts of the corporation, unless they personally
controlled them.)

>-M: No! Democracy already is coercive.
>I am only suggesting changing the centrality algorithm.

If by "democracy" we mean "government by the people," yes, there is
an unavoidable coercive aspect to *government*. If I *govern* you, it
means I control you. If I were merely guiding or advising you, I
would not use the word "govern." Control is coercion.

Control and coercion are distinct from voluntary cooperation. The
latter might look like control, sometimes, but it is not.

Mark

unread,
Sep 23, 2006, 4:09:04 PM9/23/06
to top-politics

> >-M: Centrality algorithms are what measure consensus. Did you know
> >this? I will suggest that PageRank is the least arbitrary(artificial) of
> >centrality algorithms.

>L: If this is the case, then presumably those leading caucuses will use
> that algorithm.

-M: 'Leading' - by what *measure*?
Again, centrality algorithms are what measure consensus.

I am for starting with the theoreticly most advanced algorithms.

>L: However, while I'm not a student of Machiavelli, I'd suggest that the measure of consensus used by caucus leaders might be more pragmatic. That is, it might be more on the line of "How much money and votes and energy can our opposition gather? How much damage will it cause if we fight instead of cooperating?"

-M: Machiavelli didn't know about SNA.

>L: [...] any structure whatever, including classic top-down,


> my-way-or-the-highway leadership), it is not necessary, for FA/DP
> purposes, to determine a complex common decision algorithm, though
> the org may use practically any. The DP structure makes it much less
> important, because an organization that gets it wrong will simply be ignored.

-M: With SD2-S, I am trying to do it right from the beginning.

>L:[...] If it were FA/DP, there would be no substantial collected assets. [...]

-M: Then it is weak. :-(

> > > > > If you think an FA/DP organization is about to make a decision that
> > > > > will harm you, I'd highly suggest making sure that your arguments and
> > > > > your needs are considered! And you can do that by joining and finding
> > > > > *anyone* already participating in the organization who will listen to
> > > > > you long enough to understand them.
> > > > >
> > > > > >-M: Organizations are rooted in coercion.
> > > >
> > > > >L: As I wrote, keep this man away from the tools of democracy.
> > > >
> > > >-M: I don't *like* coercion - I am a peaceful and loving man, I simply
> > > >know how things innately work.
> >
> > >L: Organizations are not rooted in coercion. They are rooted in
> > > cooperation.
> >
> >-M: Organizations are rooted in both coercion and cooperation. An
> >organization is usually a court recognized entity, as such, is backed
> >by the guns of government.

>L: "An organization is usually a court recognized entity." Some


> organizations are. Others can be if they need to be. An
> unincorporated association can have standing, sometimes, in court.
> But this does not mean that it is rooted in coercion. It means that
> it may call upon coercion if it needs to. It may never need to.

-M: The threat of coercion is always present.

>L: There are essentially two kinds of organizations. One kind is a


> control structure created by an oligarchy to develop, extend, and
> maintain its power. Very often such organizations devolve to a single
> person as strong leader. Does this involve coercion? Not necessarily.
> If I hire an employee, is providing instruction to that employee, in
> a context where it is understood that failure to perform may lead to
> dismissal, coercive? Possibly. It would depend on details of conduct, I'd suggest. If
> dismissal is punitive, yes, it would be coercive. But dismissal could
> simply mean "I can't afford to keep this person on the payroll, since
> I'm not getting any work from him." It is much more straightforward to consider such work a trade. I give you A in return for you giving me B. Is this coercive?

-M: If A was found to be defective and/or fraudulent, and you didn't
want to return B, I may want to take you to court.

>L: It might be. But, and this is my point, normally it is not. Normally, in most


> societies, such trades are voluntary, neither party experiences them
> as coercive, *so they are not coercive.*

-M: There is coercive potential.

> [...] Voluntary payments for services are not. An agency which collects voluntary payments may have some protective service that prevents theft. We could say that private property is inherently coercive, which might be the theory behind what Mark has


> written. But we don't ordinarily consider it coercive if I exercise private property rights in a non-offensive way, and we consider theft a form of coercion. (And thus the use of power to prevent theft is not a primary coercion, but is necessary to *prevent* coercion.)

-M: The counter-coercion to protect against theft is itself coercion.

> >And guess what happens if an organization doesn't pay its taxes. It
> >gets sanctioned by the government, who can potentially gather
> >collections by force, if need be.

>L: Sure. If it has anything to be taxed. Or any assets to seize.

> > >L: At least most modern and successful organizations are. There
> > are *aspects* of coercion which afflict most, but they are *not* at the root.
> >
> >-M: Try gathering a group of people together as a collective without
> >giving them a government issued entity status: "Just *understand* that
> >we are a collective" they say," and just *understand* that this is our
> >treasury."

>L: [...] "Unincorporated Association" is what it is called,


> legally. Legally, the association has no existence apart from its
> members. If there is a treasurer, the treasurer holds funds in trust.
> An unincorporated association is similar, legally, to a partnership
> of the members. A treasurer who handles the funds of the association
> might end up getting taxed if they take certain actions. For example,
> if they hire an employee. But unincorporated associations don't
> generally do that....

-M: This is still an entity, it just isn't separate from the members as
with a corporation.
The point being that it is still recognized by the government.

> >Lomax, without government(coercive) recognition of agreements,
> >organizations would be unstable.

>L: [...] It's my opinion that large FA/DP organizations would be quite stable,


> and, by definition, they would not be founded on coercion, not even
> the coercion involved in the inviolability of contracts. They do not
> depend on the performance of specific agreements, the kind of
> agreements that could be enforced in a court.

-M: What if someone wanted to doctor the poll results?
What protects against fraud?
Answer: the threat of coercion.
Someone who fucks-up, gets fined and/or goes to jail.

>L: (But, again, power is coordinated through caucuses, which may use


> standard legal forms, and which would thus be legal entities engaging
> in contracts, etc. Typically, if the caucus uses a corporate form for
> this purpose, the members involved are independent of and are not
> liable for the acts of the corporation, unless they personally
> controlled them.)

-M: OK, FAs are a communication network that can generate REAL
orgnizations.

> >-M: No! Democracy already is coercive.
> >I am only suggesting changing the centrality algorithm.

>L: If by "democracy" we mean "government by the people," yes, there is


> an unavoidable coercive aspect to *government*. If I *govern* you, it
> means I control you. If I were merely guiding or advising you, I
> would not use the word "govern." Control is coercion.
> Control and coercion are distinct from voluntary cooperation. The
> latter might look like control, sometimes, but it is not.

-M: But voluntary cooperation is still protected with the threat of
coercion.

Abd ul-Rahman Lomax

unread,
Sep 24, 2006, 11:40:25 PM9/24/06
to top-po...@googlegroups.com
At 04:09 PM 9/23/2006, Mark wrote:


> > >-M: Centrality algorithms are what measure consensus. Did you know
> > >this? I will suggest that PageRank is the least arbitrary(artificial) of
> > >centrality algorithms.
>
> >L: If this is the case, then presumably those leading caucuses will use
> > that algorithm.
>
>-M: 'Leading' - by what *measure*?
> Again, centrality algorithms are what measure consensus.

Yes. "Leading" was used in context which has been deleted. Without
looking back, I think it meant caucuses which have sufficient support
that they are considering taking action, and the context was a large
FA/DP organization concerned with major issues.

>I am for starting with the theoreticly most advanced algorithms.

I'm for starting with the simplest. If I've got a group of twenty
people, I don't want to use a complex algorithm. At what point is
more sophisticated analysis appropriate? I have no fixed idea. I've
been involved with Range Voting, which is actually quite
sophisticated; it estimates, if people vote sincerely, the social
utility of various courses of action.

> >L: However, while I'm not a student of Machiavelli, I'd suggest
> that the measure of consensus used by caucus leaders might be more
> pragmatic. That is, it might be more on the line of "How much money
> and votes and energy can our opposition gather? How much damage
> will it cause if we fight instead of cooperating?"
>
>-M: Machiavelli didn't know about SNA.

I'm in good company then. Neither do I. And if I'm starting a
political organization, I'm not sure I want to start having to
explain to them centrality algorithms and "SNA."

K.I.S.S. Know what that means, Mark?

> >L:[...] If it were FA/DP, there would be no substantial collected
> assets. [...]
>
>-M: Then it is weak. :-(

Mark continues to assert this. "There would be no substantial
collected assets" refers to the FA/DP organization itself. However,
caucuses within the organization can form organizations which collect
assets, and, historically, they do. That is, there *are* substantial
collected assets, but the FA/DP organization is not entangled with
them. Gaining control of the FA/DP organization itself does not gain
control of the assets, unless by "gaining control" we mean actually
convincing the members who contribute the assets to join the relevant cause.

> >L: "An organization is usually a court recognized entity." Some
> > organizations are. Others can be if they need to be. An
> > unincorporated association can have standing, sometimes, in court.
> > But this does not mean that it is rooted in coercion. It means that
> > it may call upon coercion if it needs to. It may never need to.
>
>-M: The threat of coercion is always present.

No. The existence of a theoretical possibility does not mean the
existence of any threat. Threat, in this context, would have to mean
that someone is threatening, and if there is no such person, there is
no threat. If you come into my house and take something of mine, is
there a threat that I will use coercion to get it back? Not if you
were trusted by me!

That, under various circumstances, the situation would shift and I
would use coercive means does not mean that the threat exists prior
to those changes, only that the possibility exists. The threat arises
when the situation shifts.


> >L: There are essentially two kinds of organizations. One kind is a
> > control structure created by an oligarchy to develop, extend, and
> > maintain its power. Very often such organizations devolve to a single
> > person as strong leader. Does this involve coercion? Not necessarily.
> > If I hire an employee, is providing instruction to that employee, in
> > a context where it is understood that failure to perform may lead to
> > dismissal, coercive? Possibly. It would depend on details of
> conduct, I'd suggest. If
> > dismissal is punitive, yes, it would be coercive. But dismissal could
> > simply mean "I can't afford to keep this person on the payroll, since
> > I'm not getting any work from him." It is much more
> straightforward to consider such work a trade. I give you A in
> return for you giving me B. Is this coercive?
>
>-M: If A was found to be defective and/or fraudulent, and you didn't
>want to return B, I may want to take you to court.

You might. Someone else might not. Thus you might threaten, and
another might not.

> >L: It might be. But, and this is my point, normally it is not.
> Normally, in most
> > societies, such trades are voluntary, neither party experiences them
> > as coercive, *so they are not coercive.*
>
>-M: There is coercive potential.

Coercion does not exist when the parties voluntarily participate.
That it *could* exist under different circumstances does not mean it
exists absent those circumstances.

>(And thus the use of power to prevent theft is not a primary
>coercion, but is necessary to *prevent* coercion.)
>
>-M: The counter-coercion to protect against theft is itself coercion.

That's right. That's why I used the term "primary," to refer to an
initiating coercion, aggression, and to distinguish it from coercion
which becomes necessary to counter coercion. I'll note that there is
controversy over this without going into it.

>-M: What if someone wanted to doctor the poll results?
>What protects against fraud?
>Answer: the threat of coercion.
>Someone who fucks-up, gets fined and/or goes to jail.

No, not in an FA/DP organization, unless they break some specific
law. It's quite possible to join FA/DP organizations, to create sock
puppets. However, an army of sock puppets has no more power than the
master of the puppets, and no more ability to convince. FA/DP
organizations do *not* make decisions based on majority vote. If
there is an organization with 1000 members, and 600 of them are sock
puppets controlled by only one, what we really have is 400 members.
Who have the assets of 400 members. A caucus of 399 members opposed
by 601 has just as much power as the entire organization minus one.
Would they be fooled by the sock puppets? I don't think so. They
would look at the behavior and arguments of the one they actually
communicate with, the proxy representing the 601. And they'd see that
this just doesn't make sense. So they would disregard that proxy in
determining whether or not the time was ripe for action. They are
completely free to do that because they are *not* following some
mechanistic algorithm which can be fooled by manipulation of the
data-gathering mechanisms. You can fool google, that is, you can
distort page rank by taking certain actions.

The protection against fraud in an FA/DP organization is the freedom
of the members. Members *can* decide, also, to validate memberships.
But that can be expensive. I think it will be simpler to just look at
what makes sense, in the deliberations and polls. And the proof is in
the pudding. If a broad FA/DP organization has a caucus that
disregards another, thinking it to be a sock puppet army, so it goes
ahead and attempts to act, it will find itself opposed with real force.

> >L: (But, again, power is coordinated through caucuses, which may use
> > standard legal forms, and which would thus be legal entities engaging
> > in contracts, etc. Typically, if the caucus uses a corporate form for
> > this purpose, the members involved are independent of and are not
> > liable for the acts of the corporation, unless they personally
> > controlled them.)
>
>-M: OK, FAs are a communication network that can generate REAL
>orgnizations.

By Mark's definition of "organization," yes. However, these "real"
organizations tend to function differently than organizations that
are not linked so closely to such a communications network.

Mark

unread,
Sep 27, 2006, 1:17:24 PM9/27/06
to top-politics

> > > >-M: Centrality algorithms are what measure consensus. Did you know
> > > >this? I will suggest that PageRank is the least arbitrary(artificial) of
> > > >centrality algorithms.
> >
> > >L: If this is the case, then presumably those leading caucuses will use
> > > that algorithm.
> >
> >-M: 'Leading' - by what *measure*?
> > Again, centrality algorithms are what measure consensus.

>L: Yes. "Leading" was used in context which has been deleted. Without


> looking back, I think it meant caucuses which have sufficient support
> that they are considering taking action, and the context was a large
> FA/DP organization concerned with major issues.

-M: 'taking action' - as decided by which centrality algorithm and
decision threshholds?
Again, we are back to centrality algorithms.

> >I am for starting with the theoreticly most advanced algorithms.

>L: I'm for starting with the simplest. If I've got a group of twenty


> people, I don't want to use a complex algorithm. At what point is
> more sophisticated analysis appropriate? I have no fixed idea. I've
> been involved with Range Voting, which is actually quite
> sophisticated; it estimates, if people vote sincerely, the social
> utility of various courses of action.

-M: 'fixed idea' - sure, we shouldn't have fixed ideas, but we can
still have favorite algorithm(s) by which to try. A decision has to be
made.

> > >L: However, while I'm not a student of Machiavelli, I'd suggest
> > that the measure of consensus used by caucus leaders might be more
> > pragmatic. That is, it might be more on the line of "How much money
> > and votes and energy can our opposition gather? How much damage
> > will it cause if we fight instead of cooperating?"
> >
> >-M: Machiavelli didn't know about SNA.

>L: I'm in good company then. Neither do I. And if I'm starting a


> political organization, I'm not sure I want to start having to
> explain to them centrality algorithms and "SNA."

-M: Then don't explain it. People will choose to use it or not.

>L: K.I.S.S. Know what that means, Mark?

-M: *Keep It Simple Stupid* - yes, exactly!
This is why the voter would have a one page data input field with
convienient defaults.
The PROGRAM would do the work, not the voter.
Complexity is for my programmer friends and myself.

[...]


> >-M: The threat of coercion is always present.

>L: No. The existence of a theoretical possibility does not mean the


> existence of any threat. Threat, in this context, would have to mean
> that someone is threatening, and if there is no such person, there is
> no threat. If you come into my house and take something of mine, is
> there a threat that I will use coercion to get it back? Not if you
> were trusted by me!

-M: OK, but now there is a greater vulnerability.

>L: That, under various circumstances, the situation would shift and I


> would use coercive means does not mean that the threat exists prior
> to those changes, only that the possibility exists. The threat arises
> when the situation shifts.

-M: OK.

> > >L: There are essentially two kinds of organizations. One kind is a
> > > control structure created by an oligarchy to develop, extend, and
> > > maintain its power. Very often such organizations devolve to a single
> > > person as strong leader. Does this involve coercion? Not necessarily.
> > > If I hire an employee, is providing instruction to that employee, in
> > > a context where it is understood that failure to perform may lead to
> > > dismissal, coercive? Possibly. It would depend on details of
> > conduct, I'd suggest. If
> > > dismissal is punitive, yes, it would be coercive. But dismissal could
> > > simply mean "I can't afford to keep this person on the payroll, since
> > > I'm not getting any work from him." It is much more
> > straightforward to consider such work a trade. I give you A in
> > return for you giving me B. Is this coercive?
> >
> >-M: If A was found to be defective and/or fraudulent, and you didn't
> >want to return B, I may want to take you to court.

>L: You might. Someone else might not. Thus you might threaten, and
> another might not.

-M: The point being that the *potential* exists.

> > >L: It might be. But, and this is my point, normally it is not.
> > Normally, in most societies, such trades are voluntary, neither party experiences them as coercive, *so they are not coercive.*
> >
> >-M: There is coercive potential.

>L: Coercion does not exist when the parties voluntarily participate.


> That it *could* exist under different circumstances does not mean it
> exists absent those circumstances.

-M: The point being that potential coercion protects the environment of
voluntary participation.

> >(And thus the use of power to prevent theft is not a primary
> >coercion, but is necessary to *prevent* coercion.)
> >
> >-M: The counter-coercion to protect against theft is itself coercion.
>
> That's right. That's why I used the term "primary," to refer to an
> initiating coercion, aggression, and to distinguish it from coercion
> which becomes necessary to counter coercion. I'll note that there is
> controversy over this without going into it.
>
> >-M: What if someone wanted to doctor the poll results?
> >What protects against fraud?
> >Answer: the threat of coercion.
> >Someone who fucks-up, gets fined and/or goes to jail.

>L: No, not in an FA/DP organization, unless they break some specific
> law.

-M: 'break some specific law' - this is what I mean by 'fucking-up'.

>L: It's quite possible to join FA/DP organizations, to create sock


> puppets. However, an army of sock puppets has no more power than the
> master of the puppets, and no more ability to convince.

-M: '...army of sock puppets..." - LOL! Colorful.

>L: [...] So they would disregard that proxy in determining whether or not the time was ripe for action. They are completely free to do that because they are *not* following some mechanistic algorithm which can be fooled by manipulation of the


> data-gathering mechanisms. You can fool google, that is, you can
> distort page rank by taking certain actions.

-M: Google scans the open-web - this is VERY different than the
analysis of a closed network. The ONLY way my proposed system can be
scammed is with blatant and obvious fraud. With my system, there is
even the potential to have *paper ballots*.

>L: The protection against fraud in an FA/DP organization is the freedom


> of the members. Members *can* decide, also, to validate memberships.
> But that can be expensive. I think it will be simpler to just look at
> what makes sense, in the deliberations and polls. And the proof is in
> the pudding. If a broad FA/DP organization has a caucus that
> disregards another, thinking it to be a sock puppet army, so it goes
> ahead and attempts to act, it will find itself opposed with real force.

[...]

-M: 'polls', 'caucuses' - these are all measurements, hence centrality
algorithms.
Since I am already decided on a centrality algorithm and decision
threshholds, I am already atleast a step ahed of you. Why do you want
to walk behind me?

illegale

unread,
Sep 27, 2006, 4:21:10 PM9/27/06
to top-politics

Mark wrote:
> > > > >-M: Centrality algorithms are what measure consensus. Did you know
> > > > >this? I will suggest that PageRank is the least arbitrary(artificial) of
> > > > >centrality algorithms.
> > >
> > > >L: If this is the case, then presumably those leading caucuses will use
> > > > that algorithm.
> > >
> > >-M: 'Leading' - by what *measure*?
> > > Again, centrality algorithms are what measure consensus.
>
> >L: Yes. "Leading" was used in context which has been deleted. Without
> > looking back, I think it meant caucuses which have sufficient support
> > that they are considering taking action, and the context was a large
> > FA/DP organization concerned with major issues.
>
> -M: 'taking action' - as decided by which centrality algorithm and
> decision threshholds?
> Again, we are back to centrality algorithms.

By reading your messages, person could easily start thinking that one
could not even go to piss without previously established centrality
algorithms. Yet, I find it be far from reallity.

ATB,
Gale

MG

unread,
Sep 27, 2006, 4:53:26 PM9/27/06
to top-politics
Hello!

The elections in Sweden 060917 went so so..
We got 83 votes in total..out of around 6000.000.. 1% would have given
us state payed voting bills in the coming 2 elections and over 3%:
millions in state funding.
The TV3 show was great though and the number of members has rocketed
since then, about 1 or 2 new every week now. It is obvious that media
expo is still everything even with the internet in place.
Our next action is to trying already now to unite all other small
parties who all failed the magical 4% limit for entering the swedish
parliament.
The thing is that they could all be present under the AD-umbrella with
full possibilities to raise opinion even for very small issues.
And if all new small parties where a part of a bigger one, not so many
would consider a vote for us in the next elections in 2010 to be in
risk of being wasted due to the 4% barrier.
To be continued..

To the report-this thread..:
I think that basically I agree on that there is a need for a
newcomer/not experiensed citizen of being able to judge whom to listen
to and whom not. At least in the beginning.
This could be made via points from other forum users on the postings
made.
Also the SD2 system with active voting for the best speakers can be
useful.
BUT, and Echarp and others agree on this: there is no need to take away
the final say/vote for the particulary issue because of that!
Why are you, Lomax and Mark so afraid of letting the people decide
directly once they have the correct information?

MG

unread,
Sep 27, 2006, 4:58:52 PM9/27/06
to top-politics
Happy to read that you are making progress, Pether!
You have a customer in AD as you know..

Abd ul-Rahman Lomax

unread,
Sep 27, 2006, 11:57:15 PM9/27/06
to top-po...@googlegroups.com
At 01:17 PM 9/27/2006, Mark wrote:
> >L: K.I.S.S. Know what that means, Mark?
>
>-M: *Keep It Simple Stupid* - yes, exactly!
>This is why the voter would have a one page data input field with
>convienient defaults.
>The PROGRAM would do the work, not the voter.
>Complexity is for my programmer friends and myself.

Mark, it seems that you are unclear on the concept of what politics
is, at least politics in a democracy.

>Since I am already decided on a centrality algorithm and decision
>threshholds, I am already atleast a step ahed of you. Why do you want
>to walk behind me?

Because I'm walking a narrower path, my guess. I'm making decisions
relevant to what I need to do, today, now. I trust that those who
come after me will see the situation better than I, so I'm not making
decisions for them. I'm just trying to anticipate what they will need
to be able to get to that point, the point where they can see better than I.

In other words, I'm not inventing -- or adopting -- a "centrality
algorithm" because I consider it premature. If Mark invents or has
found a good one, fine. I'm sure that the process I'm laying the
groundwork for will fully consider it.

Now, how will these future analysts decide what decision algorithm to
use? Will they adopt Mark's proposal simply because he claims it's best?

If they can correctly decide, with any reliability, *they already
have the means to make the best decisions.* If they cannot so decide,
then they would be adopting the algorithm, if they adopt it, on faith
or intuition. Now, based on what we have read here, would we have
faith in Mark's proposals?

However, I do have a kind of "centrality algorithm." I do consider
that a considered consensus is the best possible decision for a group
to make. If such a consensus cannot be obtained, and to the degree to
which it cannot be approached, the likelihood increases that the
proposed options are not the best.

So a good deal of my work is about how to find consensus. In truly
large groups.

If you have 100% agreement, there is no need for a complex algorithm!
If you have enough agreement that the disagreement remaining is moot,
in that those who disagree have been fully heard and, from the
record, one can conclude that something other than reason -- or even
intuition -- is the cause of the expression of dissent, and those who
disagree do not have significant power to impede the group decision,
then, again, the decision is relatively simple.

So the system of interest

(1) Insures that all available inputs are considered as thoroughly as possible.
(2) Actively solicits input from specialists or anyone with special knowledge.
(3) Facilitates deliberation in forums or assemblies protected from
noise but open to all considered inputs.
(4) Measures the degree of agreement to proposals, dynamically and
relatively continuously.
(5) Makes the formation of caucuses easy.
(6) Provides tools for analysis of consensus with respect to such
factors as the identity or other characteristics of the participants.
(7) Does not inhibit caucuses from independently acting, except as
they may voluntarily refrain based on available information, as with
(4) and (6).

Abd ul-Rahman Lomax

unread,
Sep 28, 2006, 12:02:32 AM9/28/06
to top-po...@googlegroups.com
At 04:21 PM 9/27/2006, illegale wrote:
>By reading your messages, person could easily start thinking that one
>could not even go to piss without previously established centrality
>algorithms. Yet, I find it be far from reallity.

Actually, there are, I'm sure, decision-making algorithms that we use
to decide when to undertake such activities.

Probably the simplest of these is a kind of majority voting.... If my
mind is active on something, I'll forget to take care of these bodily
functions until they actually become urgent. However, if I'm brought
"down to earth" by a need to focus on, say, what an individual is
saying, and the understanding process requires me to become aware of
my body sense (It's called "being present), I'll notice that I, in
fact, have an urgent need. It takes over my thinking....

There are, perhaps, "action potentials," perhaps these are skandhas,
and when the vector sum is sufficient to move us, we move....

Abd ul-Rahman Lomax

unread,
Sep 28, 2006, 12:18:37 AM9/28/06
to top-po...@googlegroups.com
At 04:53 PM 9/27/2006, MG wrote:
>Why are you, Lomax and Mark so afraid of letting the people decide
>directly once they have the correct information?

I'm not afraid of that. In what I've proposed for actual
decision-making methods, direct vote is allowed. But also proxy
voting is allowed. So people can decide whether or not to decide.
This is an increased level of freedom, it is not in any way a
restriction, a failure to "let the people decide."

Rather, systems as MG would support, perhaps, are coercive: they
*force* people to decide or allow others, *not* trusted, to decide.
Proxy voting allows one to vote directly (in all the systems I
contemplate) or to delegate the right and responsibility.

It is what people of wealth and power demand. (That's why it is
standard practice in business.) Why not allow it for everyone, in government?

In order to make optimal decisions about the thousands of matters
which affect me, in government, I'd have to have more information
than I could possibly review and consider in the available time. At
some point I have to trust someone.

But if I *wish* to vote directly on any of these matters, I should be
allowed to. Only when this is possible will we really have
implemented true democracy.

This is really old hat, that is, the problem of scale in democracy.
Where I depart from traditional analysis is that the latter generally
concludes that direct democracy is impractical. And the problem is
not voting, it is deliberation. Direct voting is easy, always has
been. But *deliberation* takes time. *Lots* of time. And the
necessary time, absent a good solution not contemplated by the
traditional political scientists, expands as the square of the number
of participants.

Even traditional deliberative bodies recognize this, so deliberation
is typically chunked and assigned to committees, a device for
reducing the number of participants.

I agree with Mark in one sense: most people are not prepared to make
optimal decisions. Some think that opacity is the problem. I don't
agree. Opacity is *a* problem, not *the* problem. Even with complete
transparency, the problem of scale remains.

*However*, all people, to my mind, have the *right* to decide whether
or not to delegate consent. Who am I to say that you are not
competent to make *this* decision? Or vice-versa? By what intrinsic
right can I claim that I will decide better than you?

Let me say that, quite frequently, when I'm thoroughly knowledgeable
about a subject, and I see one person claiming that another is
incompetent about it, I can tell that the incompetence is in the accuser.

A dog has a right to decide whether or not he consents. And if a dog
has that right, certainly so do all people.

*Consent* is the foundation of democracy. Societies could be measured
and judged by the degree to which the members of the society consent
to its activities. If everyone wants to allow a single individual to
be king, to make all decisions, that is a democracy *so long as the
consent remains.*

MG

unread,
Sep 28, 2006, 9:11:18 AM9/28/06
to top-politics
OK, so then I think we agree on the balance between DD and DP/RP!
Please also have a look at:
http://groups.google.com/group/top-politics/browse_frm/thread/e7c77e6d814a3d6d/#

Mark

unread,
Sep 28, 2006, 1:17:59 PM9/28/06
to top-politics
Lomax wrote:
>illegale wrote:
> >By reading your messages, person could easily start thinking that one
> >could not even go to piss without previously established centrality
> >algorithms. Yet, I find it be far from reallity.

>L: Actually, there are, I'm sure, decision-making algorithms that we use to decide when to undertake such activities. Probably the simplest of these is a kind of majority voting.... If my mind is active on something, I'll forget to take care of these bodily functions until they actually become urgent. However, if I'm brought "down to earth" by a need to focus on, say, what an individual is saying, and the understanding process requires me to become aware of my body sense (It's called "being present), I'll notice that I, in fact, have an urgent need. It takes over my thinking.... There are, perhaps, "action potentials," perhaps these are skandhas, and when the vector sum is sufficient to move us, we move....

-M: Yes, I suspect that even individuals have the equivalents of
centrality algorithms by which to prioritize activities. Groups
certainly need them.

Mark

unread,
Sep 28, 2006, 3:19:48 PM9/28/06
to top-politics
Lomax wrote:
> Mark wrote:
> > >L: K.I.S.S. Know what that means, Mark?
> >
> >-M: *Keep It Simple Stupid* - yes, exactly!
> >This is why the voter would have a one page data input field with
> >convienient defaults. The PROGRAM would do the work, not the voter.
> >Complexity is for my programmer friends and myself.

>L: Mark, it seems that you are unclear on the concept of what politics


> is, at least politics in a democracy.

-M: I am a systems designer for *collective action*.
If this is too advanced for politics - fine, someday people will be
ready.

> >Since I am already decided on a centrality algorithm and decision
> >threshholds, I am already atleast a step ahed of you. Why do you want
> >to walk behind me?

>L: [...] In other words, I'm not inventing -- or adopting -- a "centrality algorithm" because I consider it premature. If Mark invents or has found a good one, fine. I'm sure that the process I'm laying the groundwork for will fully consider it. Now, how will these future analysts decide what decision algorithm to use? Will they adopt Mark's proposal simply because he claims it's best? If they can correctly decide, with any reliability, *they already have the means to make the best decisions.* If they cannot so decide, then they would be adopting the algorithm,...

-M: I am designing a *testable* system, so my 'claims' become *moot* -
a non-issue.

>L:..if they adopt it, on faith or intuition. Now, based on what we have read here, would we have faith in Mark's proposals?

-M: What other options do you all have?
I have debated against any competing systems.

>L: However, I do have a kind of "centrality algorithm."

-M: You have centrality algorithm *guidlines*. You have nailed nothing.

>L: I do consider that a considered consensus is the best possible decision for a group to make. If such a consensus cannot be obtained, and to the degree to which it cannot be approached, the likelihood increases that the proposed options are not the best. So a good deal of my work is about how to find consensus. In truly large groups.

-M: How can you find something that you seem to refuse to define?

>L: If you have 100% agreement, there is no need for a complex algorithm!

-M: Then you have popular opinion, and nothing controversial.
Nothing that will challenge the entrenchment-monkeys. :-(

>L: [...] So the system of interest


> (1) Insures that all available inputs are considered as thoroughly as possible.

-M: Scratch this.
Systems are *people*, and people can't be forced to consider all
available inputs.
(But I do like this as an *ideal*. And SD2-S does *encourage* the
consideration of inputs.)

>L: (2) Actively solicits input from specialists or anyone with special knowledge.

-M: SD2-S would encourage this by *identifying* those with specialized
knowledge and capabilities.
But again, the players can't be *forced*.

>L: (3) Facilitates deliberation in forums or assemblies protected from noise but open to all considered inputs.

-M: Yes, SD2-S has hierarchal quality filtering so there is noise
control.

>L: (4) Measures the degree of agreement to proposals, dynamically and
> relatively continuously.

-M: Yes, I am working on how this will work.
Each proposal will have several measurements:
1. Rank of top endorser
2. Average rank of endorser
3. Median rank of endorser
4. PageRank of proposal
5. Popular consent of proposal

>L: (5) Makes the formation of caucuses easy.

-M: Yes, each issue will have identifyable specialists.

>L: (6) Provides tools for analysis of consensus with respect to such


> factors as the identity or other characteristics of the participants.

-M: The participants have 2 charcteristics:
1. their voting inputs
2. their endorsements by others

>L: (7) Does not inhibit caucuses from independently acting, except as


> they may voluntarily refrain based on available information, as with
> (4) and (6).

-M: *independent action* - each vote is a specialist vote, so
independent action is built in.

Lomax, are you happy with this?

Abd ul-Rahman Lomax

unread,
Sep 28, 2006, 9:43:18 PM9/28/06
to top-po...@googlegroups.com
At 01:17 PM 9/28/2006, Mark wrote:
>-M: Yes, I suspect that even individuals have the equivalents of
>centrality algorithms by which to prioritize activities. Groups
>certainly need them.

And, of course, they already have them.

So along comes Henry Hacker and invents a new algorithm, which he
claims is superior. It might even seem that this is supported by the
math, but something is easily overlooked: "Superior" needs
definition. Further, what we need is heuristics as a science, not
merely as theory, and, historically, it has been a very hard lesson
to discover that great theories can make for terrible consequences in
real life.

To prove that a new algorithm actually does improve outcomes
(assuming that "better outcome" can be defined) requires *testing*
it, and a simulation is probably not adequate, because people will
behave differently if they know that a process is going to control
outcomes than if it is merely an experiment without immediate consequences.

So, once again we come up against the basic question: we have a
utopian concept, how do we get from here to there?

I've been writing about this for quite some time now, so I won't
repeat it, except to mention that how to get from here to there,
safely, is very much a part of the concept I've been describing.

Mark

unread,
Sep 29, 2006, 3:05:57 PM9/29/06
to top-politics

Lomax wrote:
> Mark wrote:
> >-M: Yes, I suspect that even individuals have the equivalents of
> >centrality algorithms by which to prioritize activities. Groups
> >certainly need them.

>L: And, of course, they already have them. So along comes Henry Hacker and invents a new algorithm, which he claims is superior. It might even seem that this is supported by the math, but something is easily overlooked: "Superior" needs definition.

-M: In science, a superior model is one that is more predictive of
future data than competing models.
Likewise a superior centrality algorithm is one that is more predictive
of the future outputs of competing algorithms than vice-versa.

Right now the main competing centrality algorithm to SD2-S is
*in-degree*.
I think that SD2-S would be faster in finding leaders selected by
*in-degree* than vice-versa.

>L: Further, what we need is heuristics as a science, not merely as theory, and, historically, it has been a very hard lesson to discover that great theories can make for terrible consequences in real life. To prove that a new algorithm actually does improve outcomes (assuming that "better outcome" can be defined) requires *testing* it, and a simulation is probably not adequate, because people will behave differently if they know that a process is going to control outcomes than if it is merely an experiment without immediate consequences. So, once again we come up against the basic question: we have a utopian concept, how do we get from here to there?[...]

-M: The system must be built so that it can be tested.

Mark

unread,
Sep 29, 2006, 4:10:11 PM9/29/06
to top-politics
Lomax wrote:

> MG wrote:
> >Also the SD2 system with active voting for the best speakers can be
useful. BUT, and Echarp and others agree on this: there is no need to
take away
the final say/vote for the particulary issue because of that!

-M: How does this relate to theory and facts?

>>mG: Why are you, Lomax and Mark so afraid of letting the people decide directly...

-M: Because the people fuck-up too much, and I have given examples.

>>mG:...once they have the correct information?

-M: Who do you trust to give the lemmings correct information?
If you trust them, then why not have them make the decisions?

>L: I'm not afraid of that. In what I've proposed for actual decision-making methods, direct vote is allowed.

-M: SD2-S has an input field for direct votes, and direct votes can put
issues into deliberation if a threshhold isn't met.

>L: But also proxy voting is allowed. So people can decide whether or not to decide. This is an increased level of freedom, it is not in any way a restriction, a failure to "let the people decide."

-M: SD2-S is also democratic, so nowhere do the people *not decide* -
its just that they don't directly decide. *ALL* political power is an
extention of their consent.

>L: Rather, systems as MG would support, perhaps, are coercive: they *force* people to decide or allow others, *not* trusted, to decide.

-M: Rock on, Lomax!

>L: Proxy voting allows one to vote directly (in all the systems I contemplate) or to delegate the right and responsibility. It is what people of wealth and power demand. (That's why it is


> standard practice in business.) Why not allow it for everyone, in government?

-M: Agreed, this is why SD2-S is proxy based.

>L: In order to make optimal decisions about the thousands of matters


> which affect me, in government, I'd have to have more information
> than I could possibly review and consider in the available time. At
> some point I have to trust someone.

-M: Yes, I have been trying to tell them this.

>L: But if I *wish* to vote directly on any of these matters, I should be


> allowed to. Only when this is possible will we really have
> implemented true democracy.

-M: But above, did you complain?: "...they *force* people to decide or


allow others, *not* trusted, to decide."

If you have a vote that is counted *directly* without winning the trust
of others, then isn't this the *force* that you describe above?

By contrast, SD2-S does measure the popular/direct vote, but it just
isn't used to make decisions - this way the unearned *right to force*
never comes into play.

>L: [...]it is deliberation. Direct voting is easy, always has


> been. But *deliberation* takes time. *Lots* of time. And the
> necessary time, absent a good solution not contemplated by the
> traditional political scientists, expands as the square of the number
> of participants.

-M: It doesn't take lots of time if the people with ready answers can
be found.

>L: Even traditional deliberative bodies recognize this, so deliberation is typically chunked and assigned to committees, a device for reducing the number of participants. I agree with Mark in one sense: most people are not prepared to make optimal decisions.

-M: :-)

>L: Some think that opacity is the problem. I don't agree. Opacity is *a* problem, not *the* problem. Even with complete transparency, the problem of scale remains. *However*, all people, to my mind, have the *right* to decide whether or not to delegate consent.

-M: But as you said above "At some point I have to trust someone", this
is true not just for decisions, but also the administration of those
decisions.

This makes delegation *manditory* whether or not you want to delegate.
This is why I have the algorithm decide your representitives *IF, and
only IF* you do not provide representitives.

>L: Who am I to say that you are not competent to make *this* decision? Or vice-versa? By what intrinsic right can I claim that I will decide better than you?

-M: You are a voter. You are the data input component of a democratic
system.
Its the job of a centrality algorithm to make sense of this data. I see
no other option.

>L: Let me say that, quite frequently, when I'm thoroughly knowledgeable


> about a subject, and I see one person claiming that another is
> incompetent about it, I can tell that the incompetence is in the accuser.
> A dog has a right to decide whether or not he consents. And if a dog
> has that right, certainly so do all people. *Consent* is the foundation of democracy. Societies could be measured and judged by the degree to which the members of the society consent to its activities. If everyone wants to allow a single individual to
> be king, to make all decisions, that is a democracy *so long as the
> consent remains.*

-M: Yes, and SD2-S is consent based. SD2-S's accumulative voting makes
it so that those with the most accumulated consent are given decision
making positions.

Abd ul-Rahman Lomax

unread,
Sep 30, 2006, 8:33:15 PM9/30/06
to top-po...@googlegroups.com
At 03:19 PM 9/28/2006, Mark wrote:

>Lomax wrote:
> > Mark wrote:
> > > >L: K.I.S.S. Know what that means, Mark?
> > >
> > >-M: *Keep It Simple Stupid* - yes, exactly!
> > >This is why the voter would have a one page data input field with
> > >convienient defaults. The PROGRAM would do the work, not the voter.
> > >Complexity is for my programmer friends and myself.

Sure. However, the ultimate simplicity is dictatorship. Very simple.

The problem here is that for such a system to be politically
feasible, the *voter* must trust what is going on "under the hood."
There is one way for that to happen, quite possibly: it's been vetted
by those the voter trusts. This is why I think that FA/DP comes
first, then more sophisticated systems for "keeping it simple" for the voter.

FA/DP is terminally simple: the essential action on the part of the
members is to name a person the members trust to handle association
business when

(1) They are too busy with other activities or are incapacitated.
(2) They are insufficiently informed about the subject *in their own opinion.*

FA/DP is "mere" communication. But it is more than simple
communication. It creates networks of trust. And these networks may
well work to implement, in the societies power structures, something
like what Mark is proposing. Or something even better, if that is possible.

Mark does not see that communication, and, in particular,
communication between trusted partners, is powerful. He thinks it is
weak. A failure of vision, is my summary.

> >L: Mark, it seems that you are unclear on the concept of what politics
> > is, at least politics in a democracy.
>
>-M: I am a systems designer for *collective action*.

Nobody hired you, especially not "Collective Action." So you don't
have design specifications. Rather, you are doing it on spec. It's
hazardous. If it's fun, fine!

>If this is too advanced for politics - fine, someday people will be
>ready.

It is getting ready that is precisely my project. I'm trying to get
people ready to discover and implement decision-making structures,
but what I've seen is that communications structures come first, if
we want intelligent decision-making.

Further, any new decision-making structure is going to face
*structural* resistance from those who are attached to the status
quo, and, generally, these are those who have, under the status quo,
excess power. Why, pray tell, has the archaic appendix of the
Electoral College persisted for more than 200 years? The Electoral
College itself is quite a good idea, but without a reasonable method
of choosing the electors, it is quirky and dangerous; the political
parties quickly realized what they could do to manipulate it,
state-by-state, and almost all the states fell into line. And it is
terrifically hard to change. There is an initiative in California
that just might succeed; it actually is similar to a suggestion I
made on the Election Methods list before I'd ever heard about the
initiative.... The trick is to avoid tipping the balance against
those who are making the decision to change the system.

(My proposal, instead of awarding all the state's electors to the
party that wins the national vote, as happens with the California
initiative if a sufficient number of states ratify the compact that
it creates, would award proportional electors based on the proportion
of the vote in the state, similarly if the compact is sufficiently
ratified. My version would re-establish the intended function of the
Electoral College, undoing the damage done two hundred years ago; the
California version essentially replaces the College with national
popular vote. The advantage of my version would be that a
deliberative body is re-created, with proportional representation,
and, incidentally, ameliorating the "spoiler effect" of third-party
candidates, the California version is simpler, but is vulnerable to
the bugaboo of a close national election requiring recounts
everywhere. I'm not quite sure why the latter would be so
terrible.... but that's a common excuse advanced against getting rid
of the College.... and for retaining the all-or-nothing wackiness
that the political parties inflicted on us.)


> > >Since I am already decided on a centrality algorithm and decision
> > >threshholds, I am already atleast a step ahed of you. Why do you want
> > >to walk behind me?

I don't. I haven't decided to follow you at all. I'm not the one to
choose a "centrality algorithm." Those with the power to make changes
will choose it. And I'm quite sure, if your ideas are sound, that
they will be considered. Well considered. Indeed, the whole FA/DP
concept was developed by me as a method of making it easy for ideas
to be thoroughly considered. The world does *not* beat a path to the
inventor of a better mousetrap. Unless he hires a patent attorney and
a marketing firm. And then it is a matter of the skill of those, not
of the quality of the mousetrap. There is a better way!

>-M: I am designing a *testable* system, so my 'claims' become *moot* -
>a non-issue.

If you can get people to test it. Look at how difficult it is to get
secure, verifiable voting. It's a trivial problem. For
computer-assisted voting, the technology of scanning paper ballots
has been around for many decades. It was used when I was in high
school. That's almost fifty years ago. All it takes is a scanner and
a PC. The equipment is lying about, actually being junked all the
time. The software could easily be public-source, it is not at all
difficult or complicated. The paper ballots exists for validation
when necessary, yet they can be scanned locally and the results
transmitted immediately.

But instead, we have governments buying complicated and vulnerable
"voting machines." Why?

Well, the paper lobby doesn't see this as a major increase in the use
of paper. The "voting machine" lobby sees money in what they are
doing, so they hired that marketing firm. (i.e., the Republican Party :-)

Now, who is going to promote (read, spend big money on) your system?
The system itself is not expensive, I'm sure. But getting it
implemented, unless something new appears on the scene, is.

And that something new is exactly what I'm working on, which you
think "weak." Fine. If you don't know where your bread is buttered,
someone else will eat it.

> >L:..if they adopt it, on faith or intuition. Now, based on what we
> have read here, would we have faith in Mark's proposals?
>
>-M: What other options do you all have?
>I have debated against any competing systems.

Beta was a better format than VHS.

> >L: However, I do have a kind of "centrality algorithm."
>
>-M: You have centrality algorithm *guidlines*. You have nailed nothing.

Right. It does not ordinarily involve hammers and nails. People,
indeed, have known how to make decisions since long before either of
those, anyway.

> >L: I do consider that a considered consensus is the best possible
> decision for a group to make. If such a consensus cannot be
> obtained, and to the degree to which it cannot be approached, the
> likelihood increases that the proposed options are not the best. So
> a good deal of my work is about how to find consensus. In truly large groups.
>
>-M: How can you find something that you seem to refuse to define?

I'm not the one who is going to find it. Those who are ready to
implement it will define it and find it.

> >L: If you have 100% agreement, there is no need for a complex algorithm!
>
>-M: Then you have popular opinion, and nothing controversial.
>Nothing that will challenge the entrenchment-monkeys. :-(

I'm not challenging anyone. I'm facilitating consensus, at least
that's the goal. How do you find consensus in large groups, Mark? Any clue?

Now, you might be able to see the light of day with your system if
you can convince critical leaders, present systems are vulnerable to
that. (FA/DP systems do have entry points of increased sensitivity,
but they are, theoretically, far more open to input from any source.)
The problem is that these critical leaders are *very* difficult to
approach with new ideas. After all, they are the ones who enjoy
amplified influence under the status quo.

Try it for a few years, Mark. See what you think then.

> >L: [...] So the system of interest
> > (1) Insures that all available inputs are considered as
> thoroughly as possible.
>
>-M: Scratch this.
>Systems are *people*, and people can't be forced to consider all
>available inputs.

First of all, individuals can't. Neither can systems be forced, but a
system may be interested, collectively, in seeing that all ideas are
sufficiently considered. FA/DP is designed to maximize consideration.
A wacky idea gets a response at the very entry point, between two
people who have a reasonable level of trust between them. Call it
"rapport." Yes, there will be some people who are unable to form such
relationships, and they are going to have some trouble getting their
ideas considered. However, and this is important, in an FA/DP system,
all that loner has to do is convince *one* person who is better
connected, and the idea goes forth.

Mark's response was ... thoroughly off. He missed the "as thoroughly
as possible."

>(But I do like this as an *ideal*. And SD2-S does *encourage* the
>consideration of inputs.)

By whom?

> >L: (2) Actively solicits input from specialists or anyone with
> special knowledge.
>
>-M: SD2-S would encourage this by *identifying* those with specialized
>knowledge and capabilities.
>But again, the players can't be *forced*.

Of course not. Nobody can be forced to do *anything* in FA/DP. The
organization doesn't collect power, so it can't move power that I
contributed against my wishes. However, it can certainly advise. And
advice from someone *I chose* for trustworthiness? Quite likely to be
followed, I'd suggest.

> >L: (5) Makes the formation of caucuses easy.
>
>-M: Yes, each issue will have identifyable specialists.

A "caucus" is not about "specialists." A caucus is, formally or
informally, a group of people in agreement on some issue or platform.
In an FA/DP political organization, there would be, for example, I'd
expect, a Democratic caucus, a Republican caucus, a Libertarian caucus, etc.

A proxy together with his constituency is what we call a "natural
caucus." It is essentially a group of people advised, directly or
indirectly, by a single individual, and thus it has a certain natural
unity. And, thus, a certain natural power.

[some further descriptions of how Mark's system satisfies some of the
criteria mentioned,]


>Lomax, are you happy with this?

No. I did not see a description of how the system satisfies those
criteria, what was written was far too sketchy. The answer to (5),
above, shows this. A specialist is, as I mentioned, it, a servant. I
think that Mark interpreted the term as a leader. A natural caucus
does have a leader, the top proxy in it.

Abd ul-Rahman Lomax

unread,
Sep 30, 2006, 10:33:34 PM9/30/06
to top-politics

That is an entire thread. At what should we look?

MG

unread,
Oct 2, 2006, 9:44:41 AM10/2/06
to top-politics

-My answer to you..;-)

Mark

unread,
Oct 3, 2006, 3:57:50 PM10/3/06
to top-politics
Lomax wrote:
> Mark wrote:
> >Lomax wrote:
> > > Mark wrote:
> > > > >L: K.I.S.S. Know what that means, Mark?
> > > >
> > > >-M: *Keep It Simple Stupid* - yes, exactly!
> > > >This is why the voter would have a one page data input field with
> > > >convienient defaults. The PROGRAM would do the work, not the voter.
> > > >Complexity is for my programmer friends and myself.

>L: The problem here is that for such a system to be politically


> feasible, the *voter* must trust what is going on "under the hood."

-M: The voter trusts by seeing it work.

>L: There is one way for that to happen, quite possibly: it's been vetted


> by those the voter trusts. This is why I think that FA/DP comes
> first, then more sophisticated systems for "keeping it simple" for the voter.

-M: SD2-S is also a DP system, so you aren't making a comparison.

>L: [...] Mark does not see that communication, and, in particular,


> communication between trusted partners, is powerful. He thinks it is
> weak. A failure of vision, is my summary.

-M: Yes, communication is already used everywhere where there is
failure.
So communication alone isn't the answer.

> > >L: Mark, it seems that you are unclear on the concept of what politics
> > > is, at least politics in a democracy.

> >-M: I am a systems designer for *collective action*.

>L: Nobody hired you, especially not "Collective Action." So you don't


> have design specifications. Rather, you are doing it on spec. It's
> hazardous. If it's fun, fine!

-M: I am working on design specs, and have submitted some for peer
review.
I challenge others to find a better system then SD2-S.

> >If this is too advanced for politics - fine, someday people will be
> >ready.

>L: It is getting ready that is precisely my project. I'm trying to get


> people ready to discover and implement decision-making structures,
> but what I've seen is that communications structures come first, if
> we want intelligent decision-making.

-M: You can't make a decision as a group without a centrality
algorithm, therefore communication alone won't give you 'intelligent
decision-making' because it won't give you *decisions*.

>L: Further, any new decision-making structure[...]

-M: 'decision-making structure' - which you don't have.

> > > >Since I am already decided on a centrality algorithm and decision
> > > >threshholds, I am already atleast a step ahed of you. Why do you want
> > > >to walk behind me?

>L: I don't. I haven't decided to follow you at all. I'm not the one to


> choose a "centrality algorithm." Those with the power to make changes

> will choose it.[...]

-M: Under what centrality algorithm will they choose the centrality
algorithm?

By contrast, I actually start with an *initial condition* that makes
sense.
I actually start with a centrality algorithm.

> >-M: I am designing a *testable* system, so my 'claims' become *moot* -
> >a non-issue.

>L: If you can get people to test it. [...]

-M: All I need is a website where people can submit their ideas in a
competitive format.
(gotta run.)

Message has been deleted

Mark

unread,
Oct 4, 2006, 2:32:12 PM10/4/06
to top-politics

Lomax wrote:
> Mark wrote:
> >Lomax wrote:
> > > Mark wrote:
[...]
>L: Now, who is going to promote (read, spend big money on) your system?

-M: Large organizations who want increases in efficiency and user
satisfaction.
Probably most.

>L: The system itself is not expensive, I'm sure. But getting it


> implemented, unless something new appears on the scene, is.
> And that something new is exactly what I'm working on, which you
> think "weak." Fine. If you don't know where your bread is buttered,
> someone else will eat it.

-M: It is weak because its a communication network and not a
decision-making network.

> > >L:..if they adopt it, on faith or intuition. Now, based on what we
> > have read here, would we have faith in Mark's proposals?

> >-M: What other options do you all have?
> >I have debated against any competing systems.

>L: Beta was a better format than VHS.

-M: Agreed. An entrenchment occured because people adapted to VHS
first.
My concern is that with your DP/FA, people will entrench themselves
into the more familiar counting/in-degree centrality algorithm for
decision making, instead of the theoreticly more advanced SD2-S.

> > >L: However, I do have a kind of "centrality algorithm."
> >
> >-M: You have centrality algorithm *guidlines*. You have nailed nothing.

>L: Right. It does not ordinarily involve hammers and nails. People,


> indeed, have known how to make decisions since long before either of
> those, anyway.

-M: Yes, and they have been quite expert in entrenching themselves into
mediocrity.

> > >L: I do consider that a considered consensus is the best possible
> > decision for a group to make. If such a consensus cannot be
> > obtained, and to the degree to which it cannot be approached, the
> > likelihood increases that the proposed options are not the best. So
> > a good deal of my work is about how to find consensus. In truly large groups.
> >
> >-M: How can you find something that you seem to refuse to define?

> I'm not the one who is going to find it. Those who are ready to
> implement it will define it and find it.

-M: Then they will probably entrench themselves.

> > >L: If you have 100% agreement, there is no need for a complex algorithm!
> >
> >-M: Then you have popular opinion, and nothing controversial.
> >Nothing that will challenge the entrenchment-monkeys. :-(

>L: I'm not challenging anyone. I'm facilitating consensus, at least


> that's the goal. How do you find consensus in large groups, Mark? Any clue?

-M: Give them the truth, even if it is controversial?
No, that doesn't work. Cram the truth down their throat?
Maybe I am clueless. Maybe so is everyone else. :-(

>L: Now, you might be able to see the light of day with your system if


> you can convince critical leaders, present systems are vulnerable to
> that. (FA/DP systems do have entry points of increased sensitivity,
> but they are, theoretically, far more open to input from any source.)

-M: Again, SD2-S is a DP system.
We are comparing FA/DP systems with SD2-S, correct?

>L: The problem is that these critical leaders are *very* difficult to


> approach with new ideas. After all, they are the ones who enjoy
> amplified influence under the status quo.

-M: I won't approach those entrenchment-monkeys.

>L: Try it for a few years, Mark. See what you think then.

-M: Maybe member-based nonprofits and cooperatives are best.

> > >L: [...] So the system of interest
> > > (1) Insures that all available inputs are considered as
> > thoroughly as possible.

> >-M: Scratch this. Systems are *people*, and people can't be forced to consider all
> >available inputs.

>L: First of all, individuals can't. Neither can systems be forced, but a


> system may be interested, collectively, in seeing that all ideas are
> sufficiently considered. FA/DP is designed to maximize consideration.
> A wacky idea gets a response at the very entry point, between two
> people who have a reasonable level of trust between them. Call it
> "rapport." Yes, there will be some people who are unable to form such
> relationships, and they are going to have some trouble getting their
> ideas considered. However, and this is important, in an FA/DP system,
> all that loner has to do is convince *one* person who is better
> connected, and the idea goes forth.

-M: OK, this increases the *likelihood* that a good idea will
proliferate.
But again, the individuals involved can't be forced.

>L: Mark's response was ... thoroughly off. He missed the[point] "as thoroughly
> as possible."

-M: I responded to two points, your communicated point and your
(apparently) intended point, so I don't think that I missed anything.

> >M: (But I do like this as an *ideal*. And SD2-S does *encourage* the
> >consideration of inputs.)

>L: By whom?

-M: Since SD2-S is a DP system, its the individuals involved. As you
said above:


"all that loner has to do is convince *one* person who is better
connected, and the idea goes forth."

DP creates rank hierarchies and quality filtering.

> > >L: (2) Actively solicits input from specialists or anyone with
> > special knowledge.
> >
> >-M: SD2-S would encourage this by *identifying* those with specialized
> >knowledge and capabilities.
> >But again, the players can't be *forced*.

>L: Of course not. Nobody can be forced to do *anything* in FA/DP. The


> organization doesn't collect power, so it can't move power that I
> contributed against my wishes. However, it can certainly advise. And
> advice from someone *I chose* for trustworthiness? Quite likely to be
> followed, I'd suggest.

-M: Advise from specific individuals, and collectively as measured by
an array of different centrality algorithms?:

----------------------
Issue X
Indegree: 57% yes
Top selectee:
Frank Popular: Yes

PageRank: 42% yes
Top selectee:
Jane Elite: No

3rd-degree-Markov: 40% Yes
Top selectee:
John Kissass: Yes [...]
-------------------------------
Lomax, would it look like this?
What is this supposed to do?

> > >L: (5) Makes the formation of caucuses easy.
> >
> >-M: Yes, each issue will have identifyable specialists.

>L: A "caucus" is not about "specialists." A caucus is, formally or


> informally, a group of people in agreement on some issue or platform.
> In an FA/DP political organization, there would be, for example, I'd
> expect, a Democratic caucus, a Republican caucus, a Libertarian caucus, etc.

-M: OK, you are using the term 'caucus' conventionally.
The trend in discussions has been toward specialist systems, and this
is what threw me off.

>L: A proxy together with his constituency is what we call a "natural


> caucus." It is essentially a group of people advised, directly or
> indirectly, by a single individual, and thus it has a certain natural
> unity. And, thus, a certain natural power.
> [some further descriptions of how Mark's system satisfies some of the
> criteria mentioned,]

> >M: Lomax, are you happy with this?

>L: No. I did not see a description of how the system satisfies those


> criteria, what was written was far too sketchy. The answer to (5),
> above, shows this. A specialist is, as I mentioned, it, a servant. I
> think that Mark interpreted the term as a leader. A natural caucus
> does have a leader, the top proxy in it.

-M: SD2-S has both generalists and specialists.
The generalists are the highest ranking general trustees - the
'top-proxies', as you describe.
They, in turn, organize the top-proxies for specific issues, for both
administrative and debate purposes. Didn't you already know this?

Abd ul-Rahman Lomax

unread,
Oct 4, 2006, 11:35:10 PM10/4/06
to top-po...@googlegroups.com
At 02:32 PM 10/4/2006, Mark wrote:
> >L: Now, who is going to promote (read, spend big money on) your system?
>
>-M: Large organizations who want increases in efficiency and user
>satisfaction.but
>Probably most.

For large numbers of organizations to adopt it, a few must try it
out. For a few to try it out, *one* must adopt it first.

Now, you might find a leader in an org who is looking for better
decision-making process. *However*, you will run into what I've
called the "Lomax effect," since I haven't seen anyone else
explicitly describing it, but it could also be called the
"persistence of power inequities effect."

Those who have excess power will act to preserve it. And, by the
conditions given, they have excess power and are thus relatively able
to succeed in preserving it.

A leader may be interested in better decision-making, *but* few will
be interested in this if it means that their excess power will
vanish. A good implementation of just about any variety of proxy
democracy *will* make leaders less relevant, at least some of them,
in ways that are difficult to predict. For quite a few reasons, some
of them even sound, these leaders don't think that their loss of
personal power would be a benefit to the organization. After all,
they are the ones who have the divine right ...

oops, time warp, sorry, I mean, "clearly understand the needs of the
organization and its members."

They are quite afraid that the lemmings will run the show, and will
be subject to corrupt influence.

It is remarkable to see this in organizations dedicated to reforming
democracy. I've seen it again and again; fear of democracy. So Center
for Voting and Democracy is oligarchical. Common Cause is
oligarchical. And I could go on and on. Key question, even in
organizations which nominally are controlled by members, who elect a
Board: is proxy voting allowed?

It is often *specifically* prohibited. Why? Well, for starters, the
existing leadership of the organization makes sure that the annual
meeting is held where it is convenient for them. So they have, I've
seen, a huge advantage at any such meeting....

> >L: The system itself is not expensive, I'm sure. But getting it
> > implemented, unless something new appears on the scene, is.
> > And that something new is exactly what I'm working on, which you
> > think "weak." Fine. If you don't know where your bread is buttered,
> > someone else will eat it.
>
>-M: It is weak because its a communication network and not a
>decision-making network.

Any communications network can become, when needed, a decision-making
network. FA/DP specifically, however, leaves the ultimate decisions
in the hands of the membership, who make the final call by deciding
whether or not to actually put muscle to the wheel of a particular cause.

Technically, however, any proxy is the leader of what we call a
"natural caucus," and can effectively make decisions for that caucus.
In a Free Association, this leader still must convince his clients.
But he or she is at the center of the natural caucus communications.
And he was chosen because he or she was trusted. Thus decision-making
in an FA boils down to decision-making *by a single individual,* who
can use whatever tools desired.

What is *not* done in an FA is to make a decision for the whole
organization, beyond operational details.

FA/DP uses the DP structure to develop a broad, deliberated
consensus. DP alone could be used in a power structure, as with more
traditional membership organizations. However, in an FA, the power is
left entirely with the members, who are the ultimate safeguard
against corruption. The proxies can decide whatever they want, but,
to get power behind it, they must convince their clients.

In AA, the Conference does *not* make decisions for AA World
Services, Inc. Except for board elections, and I'm not even sure
about that, they might merely make recommendations which AAWS
routinely accepts, it might not be legally binding. Rather, the
Conference attempts to seek and express a consensus. Generally, a
two-thirds vote of the Conference is considered *as if* it were
binding, but, legally, the Conference does not control AAWS, the
board members do. However, AA as a whole voluntarily maintains
on-going funding for AAWS. If AAWS improperly refused to implement a
Conference consensus, the Conference could recommend back to its
constituent regional intergroups to stop funding the national office.
It's never come to that, I think.


> >L: Beta was a better format than VHS.
>
>-M: Agreed. An entrenchment occured because people adapted to VHS
>first.
>My concern is that with your DP/FA, people will entrench themselves
>into the more familiar counting/in-degree centrality algorithm for
>decision making, instead of the theoreticly more advanced SD2-S.

FA/DP can use any decision-making algorithm the members choose. And
it is in the nature of FA/DP that Mark would have full opportunity to
present his proposals, to have them heard, *and answered*. Either
accepted and used, or not accepted *with an explanation.*

>-M: Yes, and they have been quite expert in entrenching themselves into
>mediocrity.

No, they have made remarkable strides. You have to consider where
they started, and you also have to consider the possibility that your
own knowledge is not the be-all and end-all of it. Perhaps those
lemmings are much better adapted than you think. The Marxists wanted
to dump what was, to them, clear exploitation of the workers.
However, some of this was functional in ways that the Marxists simply
did not understand, and when they *did* succeed in getting rid of
what they thought was the problem, they had a whole new host of
problems that the status quo, previously, was dealing with.

It was the collision of a beautiful theory with the real world....
Millions of years of evolution sometimes comes up with better than we
can imagine. At least initially. I think that if we *respect* the
status quo, something which comes with difficulty for Mark, we may be
better able to understand how it functions, and thus to understand
better how to tweak it without damage.

Give the wrong people an efficient decision-making system, and you
could have a real disaster on your hands....

> > > So
> > > a good deal of my work is about how to find consensus. In truly
> large groups.
> > >
> > >-M: How can you find something that you seem to refuse to define?
>
> > I'm not the one who is going to find it. Those who are ready to
> > implement it will define it and find it.
>
>-M: Then they will probably entrench themselves.

Mark is describing himself. What else could he be describing? "Those
who are ready to implement it (i.e., consensus) will define it." This
is just what Mark thinks he has done, he has found the best way.

But he is quite entrenched!

> >L: I'm not challenging anyone. I'm facilitating consensus, at least
> > that's the goal. How do you find consensus in large groups, Mark? Any clue?
>
>-M: Give them the truth, even if it is controversial?
>No, that doesn't work. Cram the truth down their throat?
>Maybe I am clueless. Maybe so is everyone else. :-(

No. There are people who know how to do it, people who make a living
at it, and some of them are very good. I've done this work, on
occasion, though never professionally. I chaired a national
conference, using some early versions of my ideas, with over a
hundred delegates in attendance. Nearly every vote was unanimous,
*because the groundwork had been done.*

(To a degree. In fact, the Conference, elated by its new-found
ability to actually function and make decisions, something which had
eluded it for years, went too quickly. It freaked the Board out, and
the Board acted to make sure that this wouldn't happen again! And
there were other structural problems that meant that, indeed, it did
not happen again. This was pre-DP. Had DP been in use, it would have
been a different story.)


>-M: Again, SD2-S is a DP system.
>We are comparing FA/DP systems with SD2-S, correct?

Not really. I'm explaining FA/DP. I don't understand enough about
SD2-S to make a comparison, except to bounce off impressions of a few details.

> >L: The problem is that these critical leaders are *very* difficult to
> > approach with new ideas. After all, they are the ones who enjoy
> > amplified influence under the status quo.
>
>-M: I won't approach those entrenchment-monkeys.

Then you have only one approach left: work with totally new
organizations. But how are you going to recruit one? After all, new
organizations start small and don't think that they need a
complicated decision-making process, or even a fairly simple one, for
that matter. (DP is *extremely* simple and it scales all the way up
from a few members to the population of the earth. In theory.)

> >L: Try it for a few years, Mark. See what you think then.
>
>-M: Maybe member-based nonprofits and cooperatives are best.

Quite possibly. But, once gain, I've been working on this problem.

See, if we had FA/DP organizations, it would be easy to bring in a
new idea like SD2-S. FAs don't have an entrenched power structure,
beyond what is natural conservatism. They don't have gatekeepers in
the way that traditional organizations do.

> > However, and this is important, in an FA/DP system,
> > all that loner has to do is convince *one* person who is better
> > connected, and the idea goes forth.
>
>-M: OK, this increases the *likelihood* that a good idea will
>proliferate.
>But again, the individuals involved can't be forced.

Right. They can't be forced, but, collectively, the new idea will be
considered unless the one proposing it is so socially dysfunctional
that nobody will talk to him or her. Which is pretty bad, there are
people who listen to just about every lunatic out there.

> >L: Mark's response was ... thoroughly off. He missed the[point]
> "as thoroughly
> > as possible."
>
>-M: I responded to two points, your communicated point and your
>(apparently) intended point, so I don't think that I missed anything.
>
> > >M: (But I do like this as an *ideal*. And SD2-S does *encourage* the
> > >consideration of inputs.)
>
> >L: By whom?
>
>-M: Since SD2-S is a DP system, its the individuals involved. As you
>said above:
>"all that loner has to do is convince *one* person who is better
>connected, and the idea goes forth."
>
>DP creates rank hierarchies and quality filtering.

Yes, we agree there. For power structures, which is what Mark is
working on, DP should be quite suitable. With power structures, there
are certain problems to be solved, such as identity validation, the
detection and interdiction of sock puppet armies, and so forth. FA/DP
avoids this by *postponing* it. My view is that once we have better
experience with DP, we will be more able to anticipate and deal with
the problems.

> >L: Of course not. Nobody can be forced to do *anything* in FA/DP. The
> > organization doesn't collect power, so it can't move power that I
> > contributed against my wishes. However, it can certainly advise. And
> > advice from someone *I chose* for trustworthiness? Quite likely to be
> > followed, I'd suggest.
>
>-M: Advise from specific individuals, and collectively as measured by
>an array of different centrality algorithms?:

The collective advice, from a political science perspective, is
"aggregative." This is not the foundation of democracy, which is
deliberation. Aggregation is needed for ultimate decision-making, in
a power structure, but aggregation is not a method of generating good
advice. It merely indicates that a particular proposal is popular.
Yes, more sophisticated systems can consider, in addition, *who* has
the opinions being collected, but, I'll tell you this, I do not trust
popular opinion. It is far too easily deceived.

It is crucial that, in a DP system, the proxy relationships be
*personal.* Otherwise there is no safeguard, and individuals will
exercise mass power based solely on popularity. This kind of mass
collection of power is necessary in standard representative
democracy, based on elections. But it is not necessary in DP.


>----------------------
>Issue X
>Indegree: 57% yes
>Top selectee:
>Frank Popular: Yes
>
>PageRank: 42% yes
>Top selectee:
>Jane Elite: No
>
>3rd-degree-Markov: 40% Yes
>Top selectee:
>John Kissass: Yes [...]
>-------------------------------
>Lomax, would it look like this?
>What is this supposed to do?

I don't understand what you've presented. Explain it to me in terms
that would be understood by a non-specialist.


> > > >L: (5) Makes the formation of caucuses easy.
> > >
> > >-M: Yes, each issue will have identifyable specialists.
>
> >L: A "caucus" is not about "specialists." A caucus is, formally or
> > informally, a group of people in agreement on some issue or platform.
> > In an FA/DP political organization, there would be, for example, I'd
> > expect, a Democratic caucus, a Republican caucus, a Libertarian
> caucus, etc.
>
>-M: OK, you are using the term 'caucus' conventionally.
>The trend in discussions has been toward specialist systems, and this
>is what threw me off.

Yes. I used the word in its ordinary usage.... a group of people who
agree on something for some purpose.


> >L: A proxy together with his constituency is what we call a "natural
> > caucus." It is essentially a group of people advised, directly or
> > indirectly, by a single individual, and thus it has a certain natural
> > unity. And, thus, a certain natural power.
> > [some further descriptions of how Mark's system satisfies some of the
> > criteria mentioned,]
> > >M: Lomax, are you happy with this?
>
> >L: No. I did not see a description of how the system satisfies those
> > criteria, what was written was far too sketchy. The answer to (5),
> > above, shows this. A specialist is, as I mentioned, it, a servant. I
> > think that Mark interpreted the term as a leader. A natural caucus
> > does have a leader, the top proxy in it.
>
>-M: SD2-S has both generalists and specialists.
>The generalists are the highest ranking general trustees - the
>'top-proxies', as you describe.
>They, in turn, organize the top-proxies for specific issues, for both
>administrative and debate purposes. Didn't you already know this?

Well, that is pretty much how I see a large FA/DP organization
functioning. The top proxies act to ensure that specialist input is
found. Indeed, they may *hire* specialists (or, more accurately, in
an FA, they may arrange, for example, for members to subscribe to a
particular media, thus funding that media organization. Or to donate
to a foundation, etc.)

Mark

unread,
Oct 5, 2006, 3:33:13 PM10/5/06
to top-politics

Lomax wrote:
> Mark wrote:
> > >L: Now, who is going to promote (read, spend big money on) your system?
> >
> >-M: Large organizations who want increases in efficiency and user
> >satisfaction[but]Probably most.

>L: For large numbers of organizations to adopt it, a few must try it


> out. For a few to try it out, *one* must adopt it first.
> Now, you might find a leader in an org who is looking for better
> decision-making process.

-M: Or have a new organization that uses the process.

>L: *However*, you will run into what I've called the "Lomax effect," since I haven't seen anyone else


> explicitly describing it, but it could also be called the "persistence of power inequities effect." Those who have excess power will act to preserve it. And, by the conditions given, they have excess power and are thus relatively able to succeed in preserving it. A leader may be interested in better decision-making, *but* few will be interested in this if it means that their excess power will vanish.

-M: Quigley's Law of Institutionalization:
http://www.nsu.ru/filf/pha/papers/wilqui.htm
---------------------------------------
[...]The law of institutionalization is to be understood as follows: a
social instrument originally set up in order to fulfill a certain
function, which lives and spreads just because it fulfills this social
function, ceases to fulfill that function and becomes a parasitic
"institution," through overspecialization, rigidification, or
misappropriation. The personnel who man the instrument may be so
independent or become such thoroughgoing specialists that there is no
one to see to the general function of the instrument. They may become
so comfortably well trained in fixed roles that they resist adapting
the instrument to new conditions. They may divert their energies and
the instrument's receipts to their own private benefit. All these
things tend to happen in time; and so every instrument tends to become
an institution.[...]
-----------------------------------------------------------

>L: A good implementation of just about any variety of proxy


> democracy *will* make leaders less relevant, at least some of them,
> in ways that are difficult to predict.

-M: OK

>L: For quite a few reasons, some of them even sound, these leaders don't think that their loss of


> personal power would be a benefit to the organization. After all,
> they are the ones who have the divine right ...
> oops, time warp, sorry, I mean, "clearly understand the needs of the
> organization and its members." They are quite afraid that the lemmings will run the show, and will
> be subject to corrupt influence.

-M: This is a concern about populistic-democracy.
I want to see DP for democratic-republicanism.

>L: It is remarkable to see this in organizations dedicated to reforming


> democracy. I've seen it again and again; fear of democracy. So Center
> for Voting and Democracy is oligarchical. Common Cause is
> oligarchical. And I could go on and on. Key question, even in
> organizations which nominally are controlled by members, who elect a
> Board: is proxy voting allowed?
> It is often *specifically* prohibited. Why? Well, for starters, the
> existing leadership of the organization makes sure that the annual
> meeting is held where it is convenient for them. So they have, I've
> seen, a huge advantage at any such meeting....

-M: This can inhibit particapatory democracy.

> > >L: The system itself is not expensive, I'm sure. But getting it
> > > implemented, unless something new appears on the scene, is.
> > > And that something new is exactly what I'm working on, which you
> > > think "weak." Fine. If you don't know where your bread is buttered,
> > > someone else will eat it.
> >
> >-M: It is weak because its a communication network and not a
> >decision-making network.

>L: Any communications network can become, when needed, a decision-making


> network. FA/DP specifically, however, leaves the ultimate decisions
> in the hands of the membership, who make the final call by deciding
> whether or not to actually put muscle to the wheel of a particular cause.

-M: The decision requires a centrality algorithm.
Then its no longer just a communication network, its now an entity.

>L: Technically, however, any proxy is the leader of what we call a


> "natural caucus," and can effectively make decisions for that caucus.
> In a Free Association, this leader still must convince his clients.
> But he or she is at the center of the natural caucus communications.
> And he was chosen because he or she was trusted. Thus decision-making
> in an FA boils down to decision-making *by a single individual,* who
> can use whatever tools desired.

-M: So? An individual can still make a decision without an FA.

>
> > >L: Beta was a better format than VHS.
> >
> >-M: Agreed. An entrenchment occured because people adapted to VHS
> >first. My concern is that with your DP/FA, people will entrench themselves
> >into the more familiar counting/in-degree centrality algorithm for
> >decision making, instead of the theoreticly more advanced SD2-S.

>L: FA/DP can use any decision-making algorithm the members choose.

-M: By what centrality algorithm would they choose a centrality
algorithm?
You haven't given a starting point.

>L: And it is in the nature of FA/DP that Mark would have full opportunity to


> present his proposals, to have them heard, *and answered*. Either
> accepted and used, or not accepted *with an explanation.*
>
> >-M: Yes, and they have been quite expert in entrenching themselves into
> >mediocrity.
>

>L: No, they have made remarkable strides. You have to consider where


> they started, and you also have to consider the possibility that your
> own knowledge is not the be-all and end-all of it. Perhaps those
> lemmings are much better adapted than you think.

-M: No, as evidenced by the fact that neocons control my country. :-(

>L: The Marxists [...] they had a whole new host of
> problems that the status quo, [...] It was the collision of a beautiful theory with the real world....

-M: Had people listened to Marx's competing economists like HC Cary,
this BS wouldn't have happened. It is the rightous that need to be
given power.

>L: Millions of years of evolution sometimes comes up with better than we


> can imagine. At least initially. I think that if we *respect* the
> status quo, something which comes with difficulty for Mark, we may be
> better able to understand how it functions, and thus to understand
> better how to tweak it without damage.

-M: The status-quo needs to fall.

>L: Give the wrong people an efficient decision-making system, and you


> could have a real disaster on your hands....

-M: This is interesting. What would happen if oligarchs were given
SD2-S?
I think that the mellowest of them would be selected.

> > > > So a good deal of my work is about how to find consensus. In truly
> > large groups.
> > > >
> > > >-M: How can you find something that you seem to refuse to define?
> >
> > > I'm not the one who is going to find it. Those who are ready to
> > > implement it will define it and find it.
> >
> >-M: Then they will probably entrench themselves.

>L: Mark is describing himself. What else could he be describing? "Those


> who are ready to implement it (i.e., consensus) will define it." This
> is just what Mark thinks he has done, he has found the best way.
> But he is quite entrenched!

-M: There are SNA programs with many c-algorithm options.
I have chosen a modified version of PageRank as my c-algorithm, and few
challenge this.
If I am entrenched, then I am not the only one.
Does anyone here have something better than SD2-S?
If not, then lets use SD2-S.

> > >L: I'm not challenging anyone. I'm facilitating consensus, at least
> > > that's the goal. How do you find consensus in large groups, Mark? Any clue?
> >
> >-M: Give them the truth, even if it is controversial?
> >No, that doesn't work. Cram the truth down their throat?
> >Maybe I am clueless. Maybe so is everyone else. :-(

>L: No. There are people who know how to do it, people who make a living


> at it, and some of them are very good. I've done this work, on
> occasion, though never professionally. I chaired a national
> conference, using some early versions of my ideas, with over a
> hundred delegates in attendance. Nearly every vote was unanimous,
> *because the groundwork had been done.*
> (To a degree. In fact, the Conference, elated by its new-found
> ability to actually function and make decisions, something which had
> eluded it for years, went too quickly. It freaked the Board out, and
> the Board acted to make sure that this wouldn't happen again! And
> there were other structural problems that meant that, indeed, it did
> not happen again. This was pre-DP. Had DP been in use, it would have
> been a different story.)

-M: OK, and I think SD2-S would choose talented people like this. :-)

[...]


> >-M: I won't approach those entrenchment-monkeys.
>
> Then you have only one approach left: work with totally new
> organizations. But how are you going to recruit one? After all, new
> organizations start small and don't think that they need a
> complicated decision-making process, or even a fairly simple one, for
> that matter. (DP is *extremely* simple and it scales all the way up
> from a few members to the population of the earth. In theory.)

-M: I won't recruit one, I will start one.

[...]

> >-M: Since SD2-S is a DP system, its the individuals involved. As you
> >said above:
> >"all that loner has to do is convince *one* person who is better
> >connected, and the idea goes forth."
> >
> >DP creates rank hierarchies and quality filtering.
>
> Yes, we agree there. For power structures, which is what Mark is
> working on, DP should be quite suitable. With power structures, there
> are certain problems to be solved, such as identity validation, the
> detection and interdiction of sock puppet armies, and so forth.

-M: OK.

> > >L: Of course not. Nobody can be forced to do *anything* in FA/DP. The
> > > organization doesn't collect power, so it can't move power that I
> > > contributed against my wishes. However, it can certainly advise. And
> > > advice from someone *I chose* for trustworthiness? Quite likely to be
> > > followed, I'd suggest.
> >
> >-M: Advise from specific individuals, and collectively as measured by
> >an array of different centrality algorithms?:
>
> The collective advice, from a political science perspective, is
> "aggregative." This is not the foundation of democracy, which is
> deliberation. Aggregation is needed for ultimate decision-making, in
> a power structure, but aggregation is not a method of generating good
> advice. It merely indicates that a particular proposal is popular.
> Yes, more sophisticated systems can consider, in addition, *who* has
> the opinions being collected, but, I'll tell you this, I do not trust
> popular opinion. It is far too easily deceived.
> It is crucial that, in a DP system, the proxy relationships be
> *personal.* Otherwise there is no safeguard, and individuals will
> exercise mass power based solely on popularity. This kind of mass
> collection of power is necessary in standard representative
> democracy, based on elections. But it is not necessary in DP.

-M: RD instead of DD. OK, we have covered this.

> >----------------------
> >Issue X
> >Indegree: 57% yes
> >Top selectee:
> >Frank Popular: Yes
> >
> >PageRank: 42% yes
> >Top selectee:
> >Jane Elite: No
> >
> >3rd-degree-Markov: 40% Yes
> >Top selectee:
> >John Kissass: Yes [...]
> >-------------------------------
> >Lomax, would it look like this?
> >What is this supposed to do?

>L: I don't understand what you've presented. Explain it to me in terms


> that would be understood by a non-specialist.

-M: The first term in each group is the centrality algorithm that is
measuring consent to the proposal.
The last group is the top selectee by that centrality algorithm.
If this is not what it would look like, what would it look like?

shanti
Mark, Seattle WA USA

[...]

Reply all
Reply to author
Forward
0 new messages