Book Review: The Starfish and the Spider ~ by Ransom

The Starfish and the Spider

This is not one of my typical book reviews.  The ideas introduced in this book are important in ways the authors do not discuss.  Enough time has passed since the book was written that the topics have developed beyond what is presented.  I think the authors drew some incomplete or inaccurate conclusions from their research.  The most important lessons from the book are between the lines and several questions came to mind that the book did not address.

This article is divided into two primary parts; the review proper and my response.  If you have already read the book and do not want a refresher I encourage you to skip to my response as I cover a lot of material that the book does not adequately address.

Because of the importance I place on the material and its implications this book review will be longer than usual.  I hope you find the shift in perspective the ideas provide to be as valuable as I did.

Introduction

The Starfish and the Spider: The Unstoppable Power of Leaderless Organizations, by Ori Brafman and Rod Beckstrom, is a book about the behavior of distributed systems, how that behavior differs from that of centralized systems, the consequences of conflict between the two, and the nature of hybrid systems that combine attributes of each.

The book is fairly short and easy to read.  It comes across as a business book and does have value in that direction but the information it conveys is broadly applicable.

The Starfish and the Spider is named after two animals that exemplify the two ideas discussed.  While spiders and starfish share a certain superficial similarity due to their leggy appearance their actual function is strikingly different.  Spiders are centralized organisms; the loss of a limb is permanent, the loss of the head is fatal.  Starfish on the other hand have no central nervous system; like their major organs their neural structure is distributed throughout their bodies.  Actions are decided upon collectively by the entire body.  Lost limbs are grown back.  In some species severed legs will grow into new starfish.

The two animals are used as a shorthand throughout the book to describe centralized and decentralized systems.  I will do the same.

Human organizations take many forms and roles.  Their design determines how they act and respond to events.  System design elements are centralized or decentralized and work together to determine behavior.  The organization as a whole may be strongly centralized, decentralized, or a combination of the two.

Centralized, "spider" organizations are easy for us to grasp.  We see them everywhere.  Distributed, "starfish" organizations are often less intuitive and difficult to discern.  The two types of systems have different behaviors and respond differently to conflict.

Two examples of conflicts between spider & starfish pairs of human organizations are presented at the beginning of the book; the music industry in its conflict with fileshare services in the late 90s and early 00s and the Spanish war with the Apache tribes during the Conquistador era.

In the first example the music industry exemplifies the spider organization: centralized, wealthy, and in firm control of its destiny.  The fileshare services demonstrate starfish organizations in action: nimble, scrappy, and a general nuisance.  The filesharing services enabled people to share and download music without paying for it.  When the industry saw its profits impacted they took Napster to court and won.

Such a decisive legal victory had the appearance of solving the industry's problem but the big corporations found themselves battling a succession of adversaries increasingly difficult to stomp down.  Napster was rendered vulnerable by its centralized servers.  Successive organizations reduced their vulnerabilities until the eventual development of peer-to-peer filesharing left the music industry with no centralized entity to attack.

In the second example the Spanish Conquistadores exemplify the spider organization and the Apache tribes of the North American southwest exemplify starfish organizations.  Enriched and embolded by their victories over the gold-rich empires of Central America the Spanish turned northward and attempted to bring the simpler Apache tribes under Spanish control.

It did not work as planned.  The Spanish attacked and burned the villages.  The Apaches became nomadic hillsmen.  The Spanish engaged the bands in battle and killed their leaders.  More leaders arose to take their place.  For all the power and technical sophistication of the Spanish empire they were unable to defeat the simple Apache, who retained their independence for centuries after the collapse of Spanish imperial presence in the new world.

These two examples are used throughout the book to illustrate the behaviors of centralized organizations, their decentralized counterparts, and what happens when the two types of organizations come into conflict.

Principles of Decentralization

Drawing from these and other examples the book develops the Principles of Decentralization.  These principles describe the behaviors of spider and starfish organizations and the result of antagonism between them.

Principle 1: When attacked, a decentralized organization tends to become even more open and decentralized -- (Ransom: attacks are generally focused at the centralized points, so survivors learn to do without)

Principle 2: It's easy to mistake starfish for spiders -- (Ransom: people who are used to seeing the world in terms of centralized systems will often view decentralized systems as centralized; after all they often have the same parts put together differently)

Principle 3: An open system doesn't have central intelligence; the intelligence is spread throughout the system -- (Ransom: decisions are made not at the center but at the edges...and everything is an edge)

Principle 4: Open systems can easily mutate -- (Ransom: many units acting in concert permits many experiments to be run in parallel with improvements propagating among peers)

Principle 5: The decentralized organization sneaks up on you -- (Ransom: decentralized organizations often don't intend to do what they end up doing, it's just the result of their behavior; they generally have neither the interest nor the awareness to advertise the threat they pose to their centralized rivals in ways that the rivals are familiar with or even capable of processing)

Principle 6: As industries become decentralized, overall profits decrease

Principle 7: Put people into an open system and they'll automatically want to contribute

Principle 8: When attacked, centralized organizations tend to become even more centralized -- (Ransom: attacks are generally focused at the centralized points, so those who shore up their critical infrastructure are more likely to survive)

As will be seen in my response to the book some of these principles are incomplete or wrong.  Still, they are a good starting point for examining the world of decentralized organizations.

Legs

According to the authors' research decentralized systems have five elements that allow them to operate and grow.  These are described as "Legs".

Leg 1: Circles.  

Most starfish organizations are built on a foundation of peer groups -- circles.  Each circle is independent and autonomous, executing the organization's functions on its own.  The members of the circles are equals and participate voluntarily.  This is not to say that there are no specialists or that no-one is more important.  Members adopt specialized roles to further their own interests in the function of the circle.

Most of us are participants in circles whether we know it or not.  A regular poker group, a shooting club, dweebs hanging out in the comments section of a blog are all examples.

The voluntary nature of participation is critical to the function of the circle.  Coercion requires centralization.  Voluntary participation is regulated by the norms and rules of the circle.  Coercion maintains order by pain or profit.  Volunteerism maintains order by social goods such as approval and companionship.

Small circles such as Alcoholics Anonymous see high participation and reciprocation.  Large circles such as city-wide Craigslist categories see low commitment, freeloaders, and trolls.  Effective systems handle this gracefully.

Leg 2: The Catalyst

The catalyst is a person who brings people together into circles.  Where before were people with common interests is now a group of people acting on those common interests.

The catalyst is good at getting people enthused and bringing them together into valuable, persistant community.  The catalyst is a pioneer who builds starfish organizations and then leaves to do it again elsewhere.

Leg 3: Ideology

Belief is what motivates people to join and participate in starfish organizations.  Belief can be as large as religion or as specific as always holding a tea-cup with the right hand.  The organization provides a venue for participants to live out their beliefs and the beliefs in turn form the norms and rules that govern the organization's operation.

The norms and rules do not need to be explicit or even understood.  Ideology can be emergent based on the incentives & available actions.  Let's call it "belief-by-action."  System design determines what can be done and how.  The design of the system is a functional ideology in its own right.

Leg 4: The Preexisting Network

Starfish organizations are often build atop existing networks.  Existing networks contain people with common perspectives ready to be brought together and established lines of communication with which to evangelize successful developments.

People are more likely to join a group if they already share its values and know some of the people inside.  Existing networks are a low-cost vector of propagation for starfish organizations.

Big networks have the most potential participants for starfish growth but are often based on shallow interests and weak connections.  Small networks tend to be tighter and more enthusiastic but offer less potential growth.  Large networks of motivated people offer the greatest growth potential.

Leg 5: The Champion

While the catalyst builds circles and leaves, the champion stays with the circle and develops its potential.

Any group will see different rates of member participation.  The champion will be among the busiest.  The champion ensures members are cared for, recruits new participants, and carries the torch for organization values.

There may be more than one champion in a circle but successful circles always have champions.

I don't know if the authors chose five "legs" to match the starfish motif or if their research led to that result but it works.  There are probably additional features of successful starfish organizations but these five are a good start.

Combating Starfish

While decentralized organizations can be a serious challenge for centralized organizations to combat they are not invulnerable.

The conventional methods of defeating rivals do not work for spider organizations attempting to bring down starfish.  As discussed above attacks on decentralized organizations tend to result in even more decentralization among surviving elements without addressing the threat.

So what does work?  The book provides two strategies for attacking starfish organizations.

Changing Ideology

Starfish organizations use ideology to maintain cohesion.  Whether the ideology is individually received or structurally emergent it coordinates behavior among adherents without the need for a command-and-control structure.  Changing people's ideology may be an effective way to stop those organizations.

Changing the beliefs of organization members will reduce or remove their incentive to participate.  If they don't believe in the organization's mission and aren't getting their own needs met, why keep going?

Changing the beliefs of the population from which the organization recruits will starve it of growth material and isolate it from its social context, creating disincentives to participate.

Changing the design of the organization through means such as architecture or technological shifts change the structural beliefs of the organization, possibly putting it at odds with the beliefs of the members or making participation uncomfortable or prohibitive.

One example of this is the use of medical aide in tribal areas of the Middle East to combat insurgent groups.  The insurgents cannot match or exceed the services provided and have far more difficulty convincing the populace that the foreign powers are enemies.  The fact that aide is so often wrapped up in exploitative NGOs is another matter entirely.

Centralize the Starfish

Starfish organizations use their decentralization to avoid mortal threats, dancing around adverseries while less decentralized organizations would be threatened into submission.  Centralizing these organizations may be an effective way to rein them in.

The Spanish were unable to bring the Apache tribes under their control.  When the Spanish left the Mexican empire tried their hand and failed.  It was not until the early twentieth century that the Americans succeeded where their predecessors did not.  How did they do it?

The Apache tribes were defined by leaders called Nant'ans.  Nant'ans had no coercive power over their peers but lead by example.  If others were inspired to follow they would have a following.  If their leadership failed their followers would look for inspiration elsewhere.

Killing Nant'ans did not work.  When one fell another would appear.  The Americans won their war with the Apaches by centralizing their society so it would be easier to control.  They gave the Nant'ans cattle.

Cattle were a concrete form of wealth that gave their owners economic power.  Nant'ans with cattle were now able to direct the behavior of other Apaches.  Cattle were easy for the U.S. government to give & control.  By changing the power dynamics of the culture in a way the U.S. government found favorable the Apache tribes were centralized, controlled, and tied down to their assigned territories.

Decentralize Yourself

Decentralized organizations can be more effective than centralized ones at attacking decentralized rivals.

Voluntary decentralization is not an easy decision.  Centralization facilitates the creation and flow of physical wealth to be exchanged in the larger economy while decentralized systems rely on more intrinsic, ideological motives and the direction creation or acquisition of desired ends.  The process is not a reskinning but a change in the nature of the structure all the way down to the incentives of individual people.  The people who work well with the one system may not do so with the other.

Centralized systems are better at conventional wealth creation & accrual while decentralized systems have more intrinsic, ideological motivation.  Centralized systems create trade goods to be exchanged with the larger network for what actors want while decentralized systems more directly create or acquire what actors want.

Decentralization represents a shift to an entirely different type of system with a loss of optionality for the decision makers.  Decentralization requires a change in fundamental operations if not in motives.  This is a reason why involuntary decentralization is easier; the opportunity cost has been reduced by the loss of centralized nodes.  Hybrid systems are a workaround in some circumstances.

The Hybrid Organization

Hybrid organizations combine elements of centralization and decentralization to use the strong points of each, usually profit potential and self-organization.

Two types of hybrid organizations are discussed.

The first is the centralized company that decentralizes the user experience.  Ebay is the example used for this category.  Ebay has a centralized payment system with defined institutions & visibly enforced rules to handle money while buyer reviews & open-to-anyone selling allow users to self-organize and inform each other.  The tight money management makes users confident that they won't lose financial assets while the peer networking enables them to discover & share valuable information that a centralized system could not economically provide.

The second type of hybrid organization is the centralized company that decentralizes parts of its own organization.  General Electric's operations under CEO Jack Welch is the example used for this category.  A monolithic entity when Jack came on board, it was broken up into business units that were each required to keep their own books.  Internal products and resources used by one unit had to be purchased from the others at full market price, creating visibility and accountability where there had been none.  Unit leaders were given profit requirements and wide latitude in their means of achieving them.

The sweet spot between centralization & decentralization for a hybrid organization is neither stationary nor obvious.  Its location must be narrowed down based on what does not work and tracked as it changes with technical & societal changes.  There are some rules of thumb:

- Systems based on sharing information will have the sweet spot towards decentralization

- Systems based on anonymity will have the sweet spot towards decentralization

- Systems that benefit from security & accountability will have the sweet spot more towards centralization

Note that the "sweet spot" as used in the book is mostly based on profit maximization.  Organizations with other motives will measure this differently.

The New World

Modern technologies have created a still mostly-unexplored world where decentralization will lead to disruption and innovation.  The authors discuss some new paradigms that break from the old ways of doing things.  Some of these are summaries of previous points while others are new observations.

Diseconomies of scale: being small and nimble can be more of an advantage than ever before.  Big organizations are optimized for doing things a certain way over and over.  Small ones can try new things at low cost and adapt quickly to the results.

Network effect: people are already able to connect more now than ever before in history.  Turning those potential connections into active networks is now fast & inexpensive and the benefits of doing so can be quickly evangelized across abundant existing channels.

The power of chaos: the inability of existing enforcement structures to keep up with decentralization means that people can be more experimental at a lower cost than ever before.

Knowledge at the edge: knowledge is generated at the active edge of an organization, not the insulated center.  Organizations with lots of edge and a structure that allows that knowledge to freely propagate & be acted upon will learn & innovate quickly.

Everyone wants to contribute: people generally enjoy being helpful.  Systems that allow people to do so at a minimal opportunity cost enable behavior with low financial burden.

Beware the hydra response: attacking decentralized organizations will tend to multiply them.

Catalysts rule: catalysts inspire the growth and development of self-motivated systems.  The new era of decentralization radically increases the reach & availability of catalysts.

The values are the organization: decentralized systems are structured by self-reinforcing beliefs rather than the command-and-control structure of centralized systems.  Most successful starfish organizations were started with what were at the time radical ideologies.

Measure, monitor, and manage: the health of a decentralized organization is not determined in reports or spreadsheets but by the activity of circles, retention of members, growth & mutation of the network, and movement towards or away from decentralization.

Flatten or be flattened: since starfish will tend to outcompete spiders organizations would be well-advised to pursue a hybrid model to use the power of decentralization while retaining the financial core that enables the production to continue.

Ransom's Criticism

As I said at the beginning the ideas introduced in this book are important in ways the authors do not discuss and some of their conclusions are incomplete or inaccurate.  Up to this point I have attempted to present their ideas with a minimum of editorialization or translation into my own views.  The rest of the article will attempt to address the wrong assumptions in the book and the corrections, consequences, and questions that occurred to me.

Western Civilization

One big assumption the authors did not address is that the events & systems in the book take place in an environment that is now in contention.

Network Effect

A fact mentioned briefly in the book is that decentralized organizations, especially circles, form most readily from existing looser associations.

Western Civilization has ideological underpinnings that affect decentralization; a shared Christian history and the idea that people should be generally free to make their own associations.  Western Civ has a high population of white people that especially in the US is disinterested in the differences between them.  Western Civ has little in the way of tribalism.  All this forms a sprawling low-level network effect inside the civilization that enables ideas to spread with low friction & civilization members to have many avenues of further self-organization.

As Western Civilization loses its coherence due to ideological shifts, importation of dissimilar peoples, and rising tribalism due to both these factors, the vast ocean of weak connections waiting to be developed becomes fragmented and polarized.  Our potential network effect is reduced and the available feedstock for forming tighter connections fractures.

Benign Neglect

The decentralized networks in the book were generally left alone so long as they did not impact finances or engage in terrorism.

In times of abundance people are more willing to experiment and are more indulgent of others doing so.

As ideological & demographic changes occur so do the location & scope of perceived zero-sum situations.  Existential rivalries strongly motivate us to preemptively disadvantage opponents.  Cultural rules that we had always assumed would be true are now in question and their potential loss must be hedged against.

Thus decentralized systems must not only unite more divergent members they must also endure more outside hostility.

Because circles are internally organized by positive incentives rather than negative incentives it is a question how they handle negative incentives from the outside.  There is precedent with organizations such as radical animal rights organizations and terror organizations.  How that spreads to the larger world, and how far, remains to be seen.

The Nature of Centralization and Decentralization

The book does not answer all the questions about centralization & decentralization that it brought to mind.

Decentralization: Absolute or Relative?

Is centralization or decentralization an absolute or relative measure?  Is an organization truly decentralized or just in comparison to something else?

If centralization / decentralization is absolute is there a "continental divide" in the exact middle?

If it is relative how do the many axes of variation compare to each other?  If one organization is decentralized in one way and its rival in another what determines which is the decentralized half of the contest?

I expect the answer is a mixture of the two.  Centralization and decentralization have objective structural differences while the behavioral differences are relative.  The structure determines what behaviors are possible.  The behaviors determine what actually happens.

The Nature of Losses

The authors make a big deal over the war between centralized music industry titans and the spunky decentralized pirates. "Cut off one of the hydra's heads and two more will take its place."  The problem with this example is that two different scales of operation are being treated as the same.  On the one hand are the music corporations.  On the other is the swarm of piracy programmers and downloaders.  One side constitutes organizations and the other is composed of the people who form organizations.  These are not two different types of organizations.  These are two different things.

While the piracy faction was successful as a faction its component organizations were often less successful.  Comparing the centralized and decentralized organizations in this conflict shows that the pirates suffered far more entity fatalities than the music corporations.  Pirates were shut down & put into prison while corporations merely lost revenue and consolidated.  Making a true comparison shows the decentralized faction as less attractive than presented.

The Spanish / Apache example shows the same properties.  While the Apaches outlasted the Conquistadors I expect that far more Apache villages were burned and Apache people were killed than their Spanish counterparts.

Centralization Over Scale

The book glosses over the question of difference over scale.

Everything contains something else and is contained by something else.  An industry is a world-wide network of producers and consumers that gracefully gains & loses participants -- decentralized.  An individual firm is managed by a board of directors and an executive structure -- centralized.  The firm employs thousands of people who can shift roles and employers -- decentralized.  An individual human body has a central nervous system and billions of cells that rely on the collective function to survive -- centralized.  The human race is decentralized across the world but centralized on one planet.

Any one thing may be centralized or decentralized but it is not isolated; it exists on a scale and the other levels on that scale may have different levels of centralization.  Subsystems of a centralized system may be decentralized.  Subsystems of a decentralized system may be centralized.  Events that are fatal to entities on one level of scale may be beneficial to entities on a higher or lower level of scale.

Specific Points of Disagreement

So much for the general disagreements I have with the material.  Here are some specific points of disagreement usually originating in the authors' being restricted by a conventional viewpoint.

Principles of Decentralization

The eight principles of decentralization were generally good but some correction is needed.

Principle 1: "When attacked, a decentralized organization tends to become even more open and decentralized."

Not exactly.  The patterns of the defending systems will shift towards decentralization but the specific organization targeted is probably going to be crushed.

Principle 3: "An open system doesn't have central intelligence; the intelligence is spread throughout the system."

I am going to flip this around because the authors are speaking from the frame that centralized systems are normal.  They aren't.  Decentralized systems are normal.  Intelligence everywhere is natural.  Centralized systems deliberately design away functional intelligence & decision-making capabilities from the bulk of the structure leaving only that which is necessary for a command-and-control system.

I am writing in terms of human organizations here.  Naturally centralized systems such as complex organisms are another matter.

Principle 4: "Open systems can easily mutate." As I already wrote it is the survivors that get the advantage.  Members & entities of open systems can be more vulnerable than their closed counterparts.

Principle 6: "As industries become decentralized, overall profits decrease."

The authors are limited by a conventional view of "profit:" hard assets, real estate, money in the bank -- especially in corporate accounts.  This is insufficient.  Profit is any benefit from an action.  It need not be financial.  It need not be financeable.  A distributed system may be more profitable than its centralized counterpart if the total benefit to all participants is measured though the act of measuring may not be possible.

Spiders are not optimized to accrue wealth, they are optimized to concentrate specific types of wealth to specific nodes.  Centralized systems generally focus on creating wealth that is easy to count to the decimal point and easy to exchange.  Total utility across a system is more broadly distributed and takes non-financial forms.  As an example consider that the rise of career women has created an explosion of GDP and a collapse of net human experience.

Spiders specialize in either-or wealth.  While capitalism incentivizes providing services for money the spider does so to acquire money which cannot be in the possession of both producer and consumer simultaneously.  Spiders focus on money, objects, and material-heavy services.  Spiders require audited numbers to sustain trust so that the system can operate.

Starfish specialize in both-and wealth.  Both-and wealth does is not exchanged so much as shared.  Starfish specialize in information, social goods & movements, relationships.  Starfish require trust in the ruleset so that the system can operate.

Principle 7: "Put people into an open system they'll automatically want to contribute."

That's nice.  It's also bogus.  The authors only looked at distributed systems where this is true because distributed systems that don't win contributors stop existing.  It's a survivorship bias issue.

Not everyone wants to contribute and not every contributor puts in the same amount of work.  There are super contributors and there are moochers.  How many people write wiki articles compared to the number of people who read them?  For that matter, how many of those readers ever contribute financially to wiki projects?

As mentioned in the book small groups have more equal dynamics while larger groups have more freeloaders.

Freeloaders are part of life.  Any successful open system must be able to function with freeloaders.

Technological Shifts

The shift towards decentralization is in some ways a return to normal.  The twentieth century was defined by technological developments that centralized power in both government and industry.  The information & transportation technology allowed central actors to monopolize markets that had before been filled by a variety of smaller entities.  Those smaller entities were not generally decentralized but they were compared to the organizations that forced them out.

The one-to-many systems of the twentieth century led to the development of the many-to-many technologies we see in the early twenty-first century.  The new stuff does not necessarily replace the old but just exists on top of it.  We can see this layering in depth: the decentralized internet was formed on the centralized technology of the phone networks, the "big nodes" of the internet, such as YouTube, are centralized entities on the decentralized internet, and the content & comments on YouTube are decentralized.

One big difference between the current day and the "old days" is in the nature of the networks.  With easier communication & travel we can form associations based on very specific traits.  When travel & communication was more expensive connections were more geographically determined.  The fact that networks tend to be built on existing networks means that those neighbors had existing relationships while our associations are more likely to start among strangers with shared interests.

Group Size Issues

I was interested to learn that groups of a size larger than 14 tend to break down.  I have seen this number elsewhere.  The book Brave New War discusses how insurgent organizations have a big challenge growing from small groups to larger groups.  I unfortunately do not have the book on hand but as I recall there is a "death zone" for group sizes between teens-size and seventy-or-so members.  Growing larger than teen size results in loss of connections and dysfunction while groups smaller than seventy struggle to develop the specialists necessary to keep things coordinated.

As I recall one solution to the problem was the franchise model such as al-Qaida employs.  Circles adopt publicly-available rulesets & affiliations with existing social proof.  Rather than controlling its affiliates the central organization will send advisors to franchises needing their experience or specialized knowledge.

Belief

The book often talks about the importance of ideology for forming starfish organizations but forgoes the opportunity to discuss it in depth.

An important aspect of ideology seen in some examples but not discussed is the importance of fervor.

- Organizations that believe more fervently have an advantage over those that don't  

- Members of an existing network that believe more fervently will form a sub-network

Holding strong beliefs on a subject makes people more willing to commit for it and sacrifice for it.  Holding strong beliefs leads people to build stronger bonds and more active relationships.  People with strong beliefs will be accommodated by the indifferent.

Where Are They Now? and Other Examples

Some of the examples provided by the book have changed significantly since publishing and are worth a new look.  Some other examples came to mind that may be of value.

iTunes

iTunes represents a successful hybrid strategy for the music industry.  It provides security features that build trust, a legal source of digital music that allays ethical concerns, and a large library that makes it a good first stop.  Its size and features create a low opportunity cost and high reputation.

Piracy still exists but iTunes offers demonstrated value to both customers and producers.

Wikipedia

Wikipedia has moved in the opposite direction from iTunes in losing its value and thus dominance.  The increasing convergence of the Wikipedia contribution team is a good example of how decentralized organizations wage war with each other.

The contributors at the beginning of Wikipedia's life were left-leaning but genuinely enthused about the project itself.  They enjoyed writing stuff.  SJWs came in and made changes that the originals did not care about or did not care about enough to successfully push back against.

As SJWs continued to enter the team they formed a group within the group defined by its stronger beliefs.  This tighter network eventually dominated the larger, less-committed network and forced them to either comply or drop out.

Bitcoin

The best-known and best-funded form of cryptocurrency, Bitcoin is an interesting exploration of how a centralized system largely subverted a decentralized system.

The decentralized nature of cryptocurrency is its main selling point.  The fact that it decentralized a feature strongly identified with centralized systems, money, is both an accomplishment and the source of its partial downfall.

In the early days of its life cycle Bitcoin was used by enthusiasts as a proof of concept.  As less-fervent users adopted it Bitcoin quickly transitioned from a cool idea to a means of exchange to an opportunity to speculate.  The fact we can now hold Bitcoin in ETFs demonstrates that what was intended to be an alterative to the existing financial system was instead subverted by it.  What was once a way to escape the dollar is now a means to make a quick buck.  How did that happen?

Bitcoin did not change, the environment around it did.  The financial system's embrace of Bitcoin changed the incentives around its use and brought in a new kind of participant that swamped the old, changing the nature of the system.  How much of that was intentional and how much was accidental I cannot say but it is undeniable that Bitcoin is now an appendage of the system it was advertised as competing against.

The Manosphere

Five years ago the manosphere was growing quickly around a handful of successful, visible websites.  There were many smaller blogs maintained by enthusiasts but nearly everyone orbited the big players.  ReturnOfKings.com even made enough money for its owner to live on.

When the big sites shut down due to deplatforming, changing political winds, or the destructive behavior of their managers the participants went their many separate ways.  Today the manosphere is a far more decentralized set of small enthusiast blogs and members have no unifying clearinghouses.  To my knowledge no-one is making much money.

The manosphere was attacked by several threats.  The big players mostly failed and the members dispersed into a less centralized structure.  Old timers look back favorably on the good old days but not only are there more members today there is also more development of the ideas.  A field started by pickup artists has ripened into the study of building families.

Jerry Pournelle's Iron Law of Bureaucracy

Quoted from Jerry Pournelle's website (https://www.jerrypournelle.com/reports/jerryp/iron.html):

Pournelle's Iron Law of Bureaucracy states that in any bureaucratic organization there will be two kinds of people:

First, there will be those who are devoted to the goals of the organization. Examples are dedicated classroom teachers in an educational bureaucracy, many of the engineers and launch technicians and scientists at NASA, even some agricultural scientists and advisors in the former Soviet Union collective farming administration.

Secondly, there will be those dedicated to the organization itself. Examples are many of the administrators in the education system, many professors of education, many teachers union officials, much of the NASA headquarters staff, etc.

The Iron Law states that in every case the second group will gain and keep control of the organization. It will write the rules, and control promotions within the organization.

While several dynamics are at play here it is a good demonstration of the power of ideological fervor.  In this case the ideology is contained in the structure of the system rather than the minds of the participants.

In bureaucracy the ideology is the primacy of the bureaucratic structure.  The structure is designed to reward those who feed it.  People who serve the stated purpose of the structure are less effective than those who serve the structure itself.  Whatever the actual beliefs of the staff is those who act like they believe in the enlargement of the structure are rewarded by it.  The process naturally accrues people with that belief-by-action and pushes to the periphery those without it, resulting in a feedback loop that is only checked by a loss of inputs or a failure to maintain structural integrity.

Decentralization and Fourth Generation Warfare

A proper treatment of 4GW is beyond the scope of this article.  Suffice to say fourth generation warfare is conflict where loyalties are not aligned with the State but instead revert to more immediate, organic systems.  Tribe, family, religion, community, these have claims on their participants' identities where the fickle and tenuous mechanism of the State does not.

The preexistent networks & incentives provided by the natural identities will lead to the growth of decentralized organizations.  Without the order provided by a uniform legal system conflict will be resolved by other means and the decentralized organizations will be a part of that with all the behaviors discussed above.

A Few Words on Spiders

The focus on decentralized organizations in the book left some aspects of centralized systems unstated.  Here are my thoughts.

Spiders have generally obvious edges and are aware of their territory.  Where are the edges of a starfish?  Starfish may not be aware of their own existence.

Spiders have obvious edges because they are unnatural -- they show up clearly against the starfish background.  Spider activities are activities people go and do.  Starfish activities are activities people just do.

Errata

Here is a list of ideas & questions I was not able to develop or answer.  Perhaps you will find value in them.

Starfish are defined by rules not hierarchies.  Society is a starfish.  Functional rules provide substrate for hierarchies to build upon.  Why do ruleset contests lead to more coercive hierarchies?  Low-coercion hierarchies can rely on compliance with existing social rulesets to direct behavior in valuable directions.  High-coercion hierarchies can establish functional if not voluntary rulesets.  Why does social conflict result in more authoritarian hierarchies?  Is this due to loss of amiable hierarchies that could push back against aggressors?  Do authoritarian hierarchies that take advantage of social conflict tend to originate from inside or outside the conflict zone?

Chaotic/disrupted/rapidly-changing environments depress viability of forming (some?) spiders.  Probably disrupt formation of starfish as well.  People want to avoid too much risk.  Existing systems with proven success in adverse environments reduce risk.

What is the most decentralized system possible?  Society itself?

Conclusion

Well this was a heck of an article.  I read the book three times or so.  The book contains some information I did not address, such as the attributes of catalysts and list of questions to help determine the decentralization of a system, but the most important stuff is here.  If you found it valuable I suggest you purchase the book and read it yourself.  The use of stories to reinforce memorization of important information works for me and the book is full of them.

Addition: A Third Method of Attacking Starfish

A third method of attacking starfish is discussed in its own post but the basics are as follows:

Centralized organizations use controlled methods of communication to ensure that only approved messages are transmitted only to and from command nodes.

Distributed organizations do not use communication to send commands and monitor progress; they communicate to keep up with events, learn from other nodes, and form consensus.

The more loosely-connected the distributed organization the more open of a communication system it tends to use.  Exceptionally sprawling, loosely-connected organizations may rely entirely on environmental or opposition messaging to fill the role of communication.

The tendency of Distributed, "Starfish" organizations to use open and uncontrolled communication presents the third method of attacking Starfish.

Adversaries can use these communication systems to send signals to which Starfish respond by behaving against their own interests.

A partial list of such signals & tactics includes:

• Disproportionate emphasis on real problems that demoralizes members of the target distributed organization, reducing their motivation to act
• False threats that waste time & energy to guard against, require resources spent filtering out static, and possibly create rifts or civil wars as parts of the distributed organization disagree on the reality of the threat
• False victories that create complacency and reward unproductive behavior thus training the distributed organization to develop structural inertia against its true interests
• Forming large communication clearinghouses, usually raised to prominence by massive capitalization rather than organic growth through proven utility, that disincentivise members of distributed organizations from forming many lateral connections with other members & groups, reduce spontaneous mass behaviors, and obscure the true number of members from each other
• Unreliable reliability that trains distributed organizations to rely too heavily on controlled information sources which then become unreliable or absent at critical points

These methods promote waste, friction, incorrect emphasis, low trust, internal conflict, and fragmentation, all of which reduce a distributed organization's ability to act out its agenda.