Federalism as a way to scale CZ and avoid bureaucracy

[Larry Sanger]

I learned this morning that one topic discussed at the 2007 "Wikimania" was how to deal with Wikipedia's burgeoning bureaucracy.  This is, essentially, a problem of "scaling the community," or how to keep governance sane and simple in the light of expansion.  Of course, we could always stay small, but I now seriously doubt this.  I think we're going to see our activity double within the next six months, and increase even more after that.

So I am throwing an idea out there just for us to kick around.  Let me begin with a definite (albeit vague) proposal, not because I want to get behind this proposal (it's just what comes to mind now), but because I think it might help us focus our thoughts.  I think that this could turn out to be one of the most important problems for us to solve, so--please think creatively!

We could reconceive of Citizendium as a federalized project, with virtually all day-to-day editorial decisions and oversight, of the sort I now do, being made at the workgroup level (with leadership from the workgroup's Chief Editor).  As each workgroup kicks off as an independent unit, it gains its own Constabulary, and its other apparatus, such as its homepage and mailing list, become much more heavily used.  Moreover, each workgroup gains its own advisory committee, if for no other reason than to be attractive to experts within the field.

This sort of full-fledged workgroup is activated when numbers warrant, according to a policy pre-established and as easy to implement as possible.  A workgroup becomes "active" only when there are N active editors (defined in some particular way).  I envision a workgroup-building procedure that involves the active editors we have on board getting in touch with people in their personal networks, to build an advisory board, and then the advisory board puts its stamp of approval on the selection of the Chief Editor.

Let me take a step back now.  The problems with bureaucracy are familiar: getting decisions takes too long; that's because procedures are too complex or involve bottlenecks; the decisions are also sometimes just silly, or wrong; and that's because the rules are actually not complex enough to handle every little exception; interaction with bureaucrats are frequently unpleasant and impersonal.

Bureaucracy is the poor solution of central decisionmakers to the "problem" of growth.  It simply is not appropriate for bottom-up projects like CZ.  The more appropriate solution is along the lines of what I suggest above: an ever-growing federation of small, self-contained organizations, each one as nimble as CZ is now, and each able to act quickly and sensibly (and with personality and "heart"!).

Perhaps federalization will create too many different rule sets, however.  But not necessarily.  I doubt that we need to have to have long, elaborated sets of rules for different subjects.  The basic rules could be the same everywhere; there will be a few local variants on certain topics, and there will be special templates and rules for special sets of articles, standard examples being articles about countries and about species.  But a great advantage of a federalized system of the sort envisioned is that it will always be very clear who's in charge: the workgroup.

Another potential problem is that independent and more greatly empowered workgroups will be far more aggressive in securing their own rights, and it will be difficult to get them to work together on articles.  This reminds me of something in ''The Federalist Papers''--if I remember right, it was one of Hamilton's argument for a relatively strong central government (which is what we got) instead of a mere federation of completely sovereign states.  The idea is that the "central government" of CZ will be able to step in and manage any inter-workgroup disputes, and because it will have this authority, that will ultimately make it easier for workgroups to work together.  The other solution is to have exactly one workgroup for each topic, and in some cases, to have different articles on the same topic for different workgroups.  I can imagine [[God (religion)]] and [[God (philosophy)]], or [[Language (linguistics)]] and [[Language (philosophy)]].  Actually, we have left our policy on these issues relatively unsettled.  I am inclined to think that we currently file articles under too many workgroups, and this might cause a problem later, when and if workgroups do become more empowered.

[Matthias Röder]

Thank you, Larry, for this stimulating post. Generally I am in favour of the federalist structure you propose. It will make day to day work easier and having a workgroup editor-in-chief will speed up the productivity of each workgroup.

We should realize however, that moving away from a centralized editorial system towards a federalist one will have one major consequence: Citizendium will gradually change its character from a "general encyclopedia" to a collection of "specialized subject encyclopedias (or better put, dictionaries)."

This change, will create some unavoidable tension between the various workgroups, for instance when one topic can be approached by various workgroups at the same time. One element that will be of crucial importance here, is the communication between the various workgroups. Will workgroup editors-in-chief be able to find proper solutions for this type of tensions? Another problem might be that the individual workgroup policies will be so different from each other, that a unified appearance of Citizendium is no longer guaranteed.

In all these cases I think it is necessary to have a strong Citizendium editor-in-chief who has the last word and is responsible for the overall direction of the project.

Making this project more federalist is good as long as we will be able to navigate a reasonable course that lies somewhere inbetween "general encyclopedia" and "specialized subject dictionary".

All the best,
Matthias

[Martin Baldwin-Edwards]

Larry, on the contrary I am opposed to the focus on academically conventional workgroups. In my view, some of the most exciting and important innovations in social science are interdisciplinary, or derived from a migration of ideas between disciplines, or even are multidiscipllnary research endeavours. That tradition-bound universities are stuck with these constraints is a problem, which favours and penalizes various types of researcher and research: it would be foolish to imitate a structure which is embedded in conventional learning, but prescribes CZ development along very conservative lines. One of the real achievements of Wikipedia has been to challenge those lines, albeit with other rather visible deficits.

If you really feel the need for a federalized bureaucracy, why not larger divisions? I would suggest the folllowing:

Natural Sciences
Arts and Music
Classics and Humanities
Social Sciences
Sports

These would make more sense than workgroups, in that much of the valuable cross-disciplinary work would still be possible. Still, I don't know where History would go: it is not strictly a social science, yet is so important for high-quality work. It is also important for all the other broad disciplines...

Postscript: I see that we have already some "Areas" for workgroups, which probably are better thought out than my divisions.

[John Thompson]

Interesting idea and worthy of discussion, Larry. To me, it appears to be another in an ongoing, albeit disconnected, discussion of what CZ is or should be, and how. It would seem that we're discussing the merits of the "Goldilocks Syndrome" in which we're attempting to find that "just right" spot between between too top down (bureaucracy gone wild) and too wide open (aka Wikipedia).

[MikeJohnson]

I really like the idea. Breaking a large-scale project down into many small-scale, local contexts seems like a win.

If we do cede a lot of power/responsibility to workgroups (a good idea, I think), I do wonder what the checks-and-balances on workgroups will be-- the editorial council, certainly, and the editor-in-chief and judiciary in more limited contexts, but... perhaps there's a place for an ombudsman in all of this? Perhaps as part of the judiciary branch (though perhaps not, if part of an ombudsman's job is to be an independent analyst/advocate)?

Just today, Jason Calacanis mentioned something along these lines for his start-up, and though it's a different situation, I think having an ombudsman would be a good, ongoing 'sanity check' on our governance.
http://www.calacanis.com/2007/08/15/mahalo-ombudsman

[Robert_W_King]

Independantly I just want to state my concerns:

- Splitting up a large body into a collective of little ones can lead to powergrabs and rule setting on a micromanaged scale; maybe a form of "tribalism" we want to avoid.

- Additionally, I believe that it might be more difficult to set and maintain standards for every group.

I have to agree with Martin that maybe larger divisions are better.  Perhaps even the current form of Workgroups seems to be working "fine".

I am just really worried that we'll have an irreversible snowball effect.  Maybe we should consider keeping our current position until it becomes absolutely necessary to refine the structure, the "don't fix it until it's broken" philosophy.

[Greg Woodhouse]

I am just really worried that we'll have an irreversible snowball effect.  Maybe we should consider keeping our current position until it becomes absolutely necessary to refine the structure, the "don't fix it until it's broken" philosophy.

The problem is that the current system is broken. For Citizendium to do well in mathematics and the sciences, it is important to have editorial standards set by people with some understanding of these subjects and some investment in seeing Citizendium excel in these areas. Right now, with some exceptions, what we have is heavily biased toward the humanities and social sciences. Providing those authors and editors working in the sciences with more ability to make decisions in these areas could certainly help develop a more balanced editorial policy, not to mention help Citizendium to grow.

[Martin Baldwin-Edwards]

Actually, Greg, we are very weak in the social sciences -- both in terms of "live"editors and also approved articles. In what way is there an "unbalanced" editorial policy? I do not understand this point.

[Ivo Janecka]

I learned this morning that one topic discussed at the 2007 "Wikimania" was how to deal with Wikipedia's burgeoning bureaucracy.  This is, essentially, a problem of "scaling the community," or how to keep governance sane and simple in the light of expansion.  Of course, we could always stay small, but I now seriously doubt this.  I think we're going to see our activity double within the next six months, and increase even more after that.

So I am throwing an idea out there just for us to kick around.  Let me begin with a definite (albeit vague) proposal, not because I want to get behind this proposal (it's just what comes to mind now), but because I think it might help us focus our thoughts.  I think that this could turn out to be one of the most important problems for us to solve, so--please think creatively!

We could reconceive of Citizendium as a federalized project, with virtually all day-to-day editorial decisions and oversight, of the sort I now do, being made at the workgroup level (with leadership from the workgroup's Chief Editor).  As each workgroup kicks off as an independent unit, it gains its own Constabulary, and its other apparatus, such as its homepage and mailing list, become much more heavily used.  Moreover, each workgroup gains its own advisory committee, if for no other reason than to be attractive to experts within the field.

This sort of full-fledged workgroup is activated when numbers warrant, according to a policy pre-established and as easy to implement as possible.  A workgroup becomes "active" only when there are N active editors (defined in some particular way).  I envision a workgroup-building procedure that involves the active editors we have on board getting in touch with people in their personal networks, to build an advisory board, and then the advisory board puts its stamp of approval on the selection of the Chief Editor.

Let me take a step back now.  The problems with bureaucracy are familiar: getting decisions takes too long; that's because procedures are too complex or involve bottlenecks; the decisions are also sometimes just silly, or wrong; and that's because the rules are actually not complex enough to handle every little exception; interaction with bureaucrats are frequently unpleasant and impersonal.

Bureaucracy is the poor solution of central decisionmakers to the "problem" of growth.  It simply is not appropriate for bottom-up projects like CZ.  The more appropriate solution is along the lines of what I suggest above: an ever-growing federation of small, self-contained organizations, each one as nimble as CZ is now, and each able to act quickly and sensibly (and with personality and "heart"!).

Perhaps federalization will create too many different rule sets, however.  But not necessarily.  I doubt that we need to have to have long, elaborated sets of rules for different subjects.  The basic rules could be the same everywhere; there will be a few local variants on certain topics, and there will be special templates and rules for special sets of articles, standard examples being articles about countries and about species.  But a great advantage of a federalized system of the sort envisioned is that it will always be very clear who's in charge: the workgroup.

Another potential problem is that independent and more greatly empowered workgroups will be far more aggressive in securing their own rights, and it will be difficult to get them to work together on articles.  This reminds me of something in ''The Federalist Papers''--if I remember right, it was one of Hamilton's argument for a relatively strong central government (which is what we got) instead of a mere federation of completely sovereign states.  The idea is that the "central government" of CZ will be able to step in and manage any inter-workgroup disputes, and because it will have this authority, that will ultimately make it easier for workgroups to work together.  The other solution is to have exactly one workgroup for each topic, and in some cases, to have different articles on the same topic for different workgroups.  I can imagine [[God (religion)]] and [[God (philosophy)]], or [[Language (linguistics)]] and [[Language (philosophy)]].  Actually, we have left our policy on these issues relatively unsettled.  I am inclined to think that we currently file articles under too many workgroups, and this might cause a problem later, when and if workgroups do become more empowered.

Larry, you posed a question: “…how to deal with Wikipedia's burgeoning bureaucracy.  This is, essentially, a problem of ‘scaling the community,’ or how to keep governance sane and simple in the light of expansion. “

Systems Science provides guidelines to survivability of an ever-expanding group of living entities in an open system format, and on any scale. Citizendium can be considered in this light. The system’s principles emphasize that a primary focus within such an organization be on the relationships among its components. The guiding light of operations needs to be the desire to optimize the entire system instead of maximizing any of the individual components. In addition, the system needs to be engaged in internal self-organization facilitated by a common language as well as self-adaptation to its external environment. The main goal of the entire system is the creation of its emergence, something which none of the components could have accomplished on its own; the whole then is greater than the sum of its parts.
Couple of other points: A well-functioning system needs to have some sort of external boundary as either extreme-boundaryless system or totalitarian system-are detrimental to the functioning of the systems.  An optimal system boundary must be a dynamic one based on biologic principles of semipermeability. This fact implies that such a boundary functions in a dynamic and system-enhancing selectivity of transport in and out of the system; simply stated, not everything gets in and not everything gets out.
The other point, essential for the long-term functioning of a balanced system, is how the system prioritizes its evolution.  For example, the survival of the fittest, often useful in early stages of a system’s development, must, at a certain critical growth inflection, be proactively changed to a focus on system’s diversity. Otherwise, the system will become very unstable, prone to mutations, and eventually follows an entropic implosion.
You suggestion of federalization of Citizendium, to achieve a healthy governance, is pointing in this direction. Incorporating systems’ principles could provide a theoretical framework for success of the organization.
Ivo

[Martin Baldwin-Edwards]

Excuse me, IPJ, but you have an illegitimate name and can only post on non-member boards. WE have a real-names policy on CZ, which you must observe, along with a link to a biography on your CZ user page.

[Larry Sanger]

I think that's Ivo Janecka--Dr. Janecka, can you click on Profile; then under Modify Profile, in the left column, click on Account Related Settings, and put your name under Name; then, also under Modify Profile, click on Forum Profile Information, scroll down to Signature (near the bottom) and write in the URL of your CZ user page (which is: http://en.citizendium.org/wiki/User:Ivo_Janecka ).

Also, you're also one of our early recruits who still doesn't have a bio on his wiki user page!  Please give us one soon.

[Thomas Simmons]

On 16/8/07 21:00, "Larry Sanger" <sanger-lists@citizendium.org> wrote:

> We could reconceive of Citizendium as a federalized project, with
> virtually all day-to-day editorial decisions and oversight, of the sort
> I now do, being made at the workgroup level (with leadership from the
> workgroup's Chief Editor).  As each workgroup kicks off as an
> independent unit, it gains its own Constabulary, and its other
> apparatus, such as its homepage and mailing list, become much more
> heavily used.  Moreover, each workgroup gains its own advisory
> committee, if for no other reason than to be attractive to experts
> within the field.


I would support this. Each group could have a Constable on a central Constabulary, coordinating with other group constables and basically insure that we don't end up with a plethora of interpretations while the increasing load of work will be spread to those with motivated interest in their area.

The discipline/field dependant advisory committee idea would be optimal in that each field or discipline will have some specific ideas and practices (about how things are written, presented, who is making headway, that sort of thing) and adapt readily to their norms.

[Gareth Leng]

It seems to me that we should first be pragmatic, and that federalism should evolve as needed for particular workgroups, some might be ready for this now and want it, while others may be far from ready or not see benefits now and should stay within the main group. There are now many important articles that fall between worgroups or indeed outside them, and it would be a great shame if any prematurely built structure were to leave them even more in the cold than they are now.

I think it is important that CZ should become what its members want it to become, and if different groups need or want greater devolution then it would be wrong to try to suppress that.

[Ivo Janecka]

I think that's Ivo Janecka--Dr. Janecka, can you click on Profile; then under Modify Profile, in the left column, click on Account Related Settings, and put your name under Name; then, also under Modify Profile, click on Forum Profile Information, scroll down to Signature (near the bottom) and write in the URL of your CZ user page (which is: http://en.citizendium.org/wiki/User:Ivo_Janecka ).

Also, you're also one of our early recruits who still doesn't have a bio on his wiki user page!  Please give us one soon.

I have updated my profile.
Ivo

[Trevor Walker]

I think Ivo is right
Trevor