Kill la Kill: cybernetics for kids or how the layers of a viable social system are interconnected through the limbic systems of their participants

Although KlK is still ongoing as of January 2014, its central theme is utterly interesting, and the already disclosed picture allows to discuss it in great detail. In this show the girl named Satsuki Kiryuin manages to build a viable and successful social system on the base of her school, but the system has strictly autocratic nature. Ryuko Matoi is Satsuki's rival backed by a secret rebel organization, she tries to ruin this system in retaliation for the murder of her father by a minion of the Kiryuin family.
Below we will look at the formal structure of power from the cybernetic point of view, understand why democracy is a form of soft autocracy, how the liberal values without the rule of law may lead to caste society, and how the liberal values enrooted by the system without any thoughts about civil responsibility may hinder the process of the creation of civil society.

The term "viable system", a system which is able to survive in a constantly and unpredictably changing environment, is coined by Stafford Beer, a prominent cybernetic theorist who developed the viable system model for economical applications as a model of a firm at 1960-70s. At the present time the cybernetics is decayed and specialized over the multiple disciplines including the control theory, the elements of AI such as expert systems, cognitive systems and so on, although Beer's works are deeply fundamental and are still actual.
The limbic system is a part of human brain which is responsible for basic instincts and emotions such as fear, pain, pleasure or reward. In Beer's theory, a system should utilize the equivalent of the limbic systems of its subsystems to avoid the "manual control" in the fields in which only the subsystems have a specialization.

Certain points of view presented below may be shocking or unacceptable for some readers, although this particular text does not intentionally contain any hidden subtexts.


[ep. 8] Rinne-Do Junior High School implements a poorly-built social system, it is mired in corruption and chaos; a group of students forces another student to jump off the roof using high social status of their patents to cover the incident.

Ira Gamagoori, the Student Council President; he is fond of order and discipline but isn't able to change anything, because he is the only such person in the whole school.

[ep. 9] A freshman Satsuki Kiryuin knows a way how to transform the academy into a viable social system to test the social model which may then even be able to conquer the world by force; this model is discussed in detail at the "Systemic Aspect" section below. In the context of the story, Kiryuin's name probably has a connotation of spring or purity of intentions, the name Satsuki may literally mean "May", but it uses the kanji that may literally mean "swamp". She considers the students of the academy as the ones who are preoccupied solely by their own interests, which is the primary source of corruption, and thinks that strict autocracy is necessary to make people to take system's and each other interests into the account.

Satsuki offers Ira to become her right hand man, but Ira resists her offer at first in hope to defeat Satsuki and prevent her plan. Being defeated by Satsuki, Ira becomes the first member of her team. The disproportion in sizes of the characters must be an artistic device which is used by artists since the Renaissance and shows the importance or power of a character in a given scene.

[ep. 15] Satsuki's psychedelic team (The Elite Four) represents the second layer in the chain of responsibility after Satsuki herself (who takes the role of the chief warlord) and consist of all the necessary components of a viable social system, such as law enforcement, feedback data analysis, tactical planning and close air support.

[ep. 10] Ryuko Matoi is a master of the red scissor blade who is blinded by rage and desperately trying to avenge her father; it seems, that she has no other goals. The name Ryuko may bear a connotation of a daughter of dragon, although in contrary to Kiryuin's name, it uses the kanji which may literally mean "flow". She wears a 100% life-fiber uniform named Senketsu which feeds on Ryuko's blood, has its own mind and gives her superhuman features.

[ep. 4] Similar life-fiber clothing with lesser amount of life-fiber is used by the Kiryuin family as the backbone element of the structure of Honnouji academy's social system. The clothing or its wearers may have a connotation of bureaucracy, but not in the sense of "the body of non-elective governmental officials", because in Kiryuin's system bureaucracy is selected and disciplined along with the rest of population through tournaments of "natural elections" in the best traditions of social Darwinism. An official with the higher rank has a greater amount of life-fiber in the clothing.

[ep. 5] The food, the authors put a great emphasis on food. This can not be made just accidentally.

[ep. 15] The Nudist Beach organization: the ideological opponents of the life-fiber clothing produced by the Kiryuin family at Revocs Corporation. They consider the clothing a extraterrestrial life-form which resides at the top of the food chain, feeds on humans and may subjugate them, so they intend to destroy Kiryuin's fiber manufacture using Ryuko's abilities.

The Success of Kiryuin's rule

Being ruled by Satsuki, Honnouji academy (formerly known as Rinne-Do) has schemed a plot against the Nudist Beach organization. During the tri-city schools raid trip Satsuki's team had overthrown the regimes in the schools which are supported by the anti-fiber rebels. After that she had finally succeeded by destroying the most of the members of Nudist Beach. The reasons that have led the other regimes to the defeat allow to understand why Satsuki's system is viable, and why others are not.

[ep. 14] Kami-Kobe High School looks like a plain military corporation, but it has failed because of reliance on the brute force without any coordination.

The social system of Abekamo Academy uses religion and mysticism for organization and confrontation, although it has failed against Kiryuin's team backup plan, because Abekamo Academy had no backup plans.

[ep. 15] Naniwa Kinman High School is organized due to loyalty based on money and uses money as a weapon. But no one from the academy except its leader wishes to fight Kiryuin from an explicitly disadvantaged position, because money or reward alone do not provide any solidarity. Despite that the leader had a backup plan, he failed due to the lack of training and mastery.

Satsuki celebrates her victory with the sacramental phrase: "It's not the money that rules men. It's fear."
So, we see that Kiryuin's system is successful and viable because it is well organized and disciplined; it uses self-audit along with strategical and tactical coordination of its own behavior; its participants are competent and fearless because Satsuki uses a kind of strict social "natural selection" to choose her elite without any doubts about human rights. And the immediate question arises: is her system a well-built one? The answer is: temporarily by the systematic aspect. But how about the personal aspect? It's obvious that nothing but fear eventually will lead people to the deep stress. As we see below, a well-built social system uses equivalents of both fear and reward to govern its subsystems in the chain of responsibility.

The Systemic Aspect

Let's take a closer look at Kiryuin's system from the cybernetic point of view and find out, why the process of proper social control has formally nothing to do with a political system or the liberal values.


A viable system must adapt itself to the constantly changing environment, or it may be destroyed by the changes that it can not react to. Nature uses such ability of the phenomenon of genetic mutations and implements such mechanism in the form of dominant and recessive gene alleles, although these mechanisms are completely probabilistic.
A rational player bestowed with memory and cognitive abilities may implement heuristic algorithms to adapt to environment. Such a system takes some strategy, which is chosen not arbitrarily but according to the conditions of the situation, and remembers the result of its application to the environment. Then it changes the strategy and compares the result of the new strategy with the old one according to some criteria. By repeating this operation it creates a new strategy and stops after a certain amount of steps if it can not find a better strategy until the environment changes. However, a properly organized viable system should produce a spontaneous strategy with some low probability to not to fall into the trap of overspecialization.
It's possible to successfully automate this process using artificial expert systems if the field of the application is sufficiently narrow and well-formalized, although its narrowness along with the need of formalization imply that decision-making expert systems will never be able to replace human mind completely.

Kiryuin's system is able to use self-audit and coordinate itself in any extent to maintain good adaptability, because it is strictly autocratic and thus has unidirectional top-down chain of responsibility which does not take people's opinion into account. This is the fundamental flaw of strict autocracy from the personal but not from the systemic point of view. It also opens the door for corruption if the chain of responsibility is too weak.

The Problem of Complexity

You might have noticed that a system comprised of the large amount of free entities may become quite chaotic being left for itself, of course, if this is not the case of a hive, ant colony or a military parade. You might say that a system would magically order and coordinate itself if it is preprogrammed with the liberal values and civil responsibility. Yes, it will do, but as we see below, only when it does nothing or performs some simple actions directed towards some situational goal that does not require complex coordination.
In some specific situations, for example in distributed computing, a hive or a decentralized (purely flat) system may be a primary answer for complexity, but every such system has one interesting feature. Because every element participates in the process of control, every element of such system should be deprived of freedom by strict programming and completely subjugated to the interests of the system itself; you know that worker ants do not reproduce. As the result, a real-world hive may be very effective in a narrow set of situations, but it has a limited extent of adaptability in a limited set of preprogrammed tasks. Moreover, the intention of such system to extend or shrink its capabilities through self-reprogramming would affect behavior of every element of the system. But we value freedom and personal inviolability, aren't we?
To free ourselves from the process of control we delegate it to some person or organ which (ideally) should become the embodiment of the interests of the system as a whole and can perform the task of coordination, playing the role of a brain of the system. But as we saw earlier, it can do nothing being left alone, because it faces the problem of complexity again and just physically unable to perform all the decisions related to all elements of a complex system. In addition, it also has problems with the control of their behavior even if it possesses a heavily armored vehicle.

Every system engineer, as well as every Roman emperor, knows that the best solution of the problem of complexity is an explicit or implicit hierarchy. But only system engineers know that hierarchy allows to reduce the complexity or variability of the control process to the amount proportional to the logarithm of the number of system elements (the base of the logarithm depends on the amount of child nodes in the levels of the resulting hierarchy tree). For example, search in an unsorted array of 65535 elements or recognition of 65535 symbols may require about 65535 (216) tests or sensors, but search of an element in a balanced and sorted binary tree (a regular tree with two child nodes) or recognition of a symbol with the hierarchically organized set of sensors will require just about 16 tests or sensors. A good sorting or search algorithm may also implicitly implement hierarchy through recursion.
So, the controlling organ builds a set of hierarchies bounded by the chain of responsibility to solve the problem of complexity in the processes of decision making, law enforcement and so on. This grants freedom of the control process to the most of system elements.

Kiryuin's system implements a five-layer hierarchy. The bureaucracy represented by decision-makers and law-enforces wear the star-ranked life-fiber uniform which enhances their abilities according to the amount of the stars. Kiryuin applies a severe social selection procedure to the entire social system to keep the bureaucracy competent.

Satsuki Kiryuin (100% life-fiber uniform)
Student Council (The Elite Four, *** goku-uniform)
Club Presidents (** goku-uniform)
Club Members (* goku-uniform)
Other Students (casual uniform)


The essential control process implies that the controller constantly monitors the system output, and if the output exceeds a certain threshold, the controller makes corrections to the system input to stabilize it.
Every libertarian dreams of the situation when the system is distancing itself from society and gently pushing it from the way to a suddenly opened manhole or a technological catastrophe. The society is a very good coordinator by itself, says a libertarian; freedom of the control process that you granted to me is not enough, so, dear system, please do not disturb me with your control procedures. Let's assume, that this is true. The paradox is that the system should have very precise knowledge about society's velocity towards the manhole to push it aside gently enough, so the libertarian might even not notice this. A well-built social system actually should act mostly as the classical controller. To satisfy the libertarian, the system should perform precise self-audit by descending the hierarchies of sensors into society, and when a libertarian sees a massive hierarchy of sensors which hangs from his ceiling, he becomes outraged and tells the system to go away completely. But if the system will leave, the next time he may need to get the society out of the manhole.
You might say, that there is no need in sensors if society is preprogrammed with liberal values and civil responsibility, and you are almost right, this should decrease the amount of sensors, because this just will make society more stable, so a well-built system is interested in this. But you still need a controller in any form along with its sensors and control procedures to stabilize society's fluctuations and to maintain its survivability in the external environment. As we saw earlier, society is a good self-coordinator only when there is no need in coordination, because it can not form a strict chain of responsibility; this should be one of the obvious features of the liberal values. Even a hive is not a good self-coordinator, because only the controller may obtain the most complete informational picture (but not necessarily a correct one) through the feedback hierarchy, and because it directly participates in the process of control, it is able to efficiently correct society's trajectory using the actuator hierarchy. The morals is that efficient feedback or coordination is impossible without a hierarchy based on the strict chain of responsibility.
There is an interesting problem, though: the capacity of a feedback or actuator hierarchy should be an adequate answer to complexity. Because complexity may change, a hierarchy should be able to figure out that it needs to shrink or grow by observing changes in its own data throughput. But in the real world hierarchies may have their own interests, because they comprised of people and not of ethereal expert systems. Because of this they might not want to shrink. Although the development of expert systems may reduce need in the personnel that participates in governmental hierarchies, is it possible to efficiently fight with their interests in the context of liberal values? Some believe that only by using civil responsibility at the independent judicial system. Kiryuin solves this problem with the strict autocracy which enforces the responsibility.

Satsuki monitors situation at the LCD-table. The obsolete monochrome LCD-technology and the design of the table resemble the artifacts in the control room of the Cybersin system - the application of Beer's ideas to Allende's Chilean command economy which allowed to test a semi-decentralized form of it.

 The chain of responsibility and the leak of control

If you participate in a hierarchy, you may delegate some work to its lower levels to solve the problem of complexity in a recursive manner. If the lower levels are comprised from groups of competent specialists or expert systems, they may create a very complex strategy that you would never be able to comprehend (this problem takes a fundamental nature in the case of expert systems). Moreover, different levels may have different specializations, and there is a situation when the levels aren't able to communicate with each other: for example, a scriptwriter, director and producer may use languages of ideas, aesthetics and profitability to speak about their common work and use different criteria for its evaluation.
    Level 1/Language 1         [S1]
                               / \ 
    Level 2/Language 2       [S2][S3]
                             / \   \
    Level 3/Language 3     [S4][S5][S6]
How would you solve this problem? There are two obvious extremes: you may say that the strategy is good or wrong according to your criteria and the better one is still necessary, or you may do the work by yourself. The first one trains the underlying system to work in the domain of the language of your level, although this process is cumbersome by itself and resembles the training of a dog: you must call to her limbic system using scolding or sugar, because the dog does not understand your words. 
Stafford Beer calls "algedonic" a subsystem that allows a generalized expert system to find out is the obtained strategy a good or wrong one. Although the algedonic subsystem has just only one binary input ("good" or "wrong"), it allows a system which remembers the quality of its strategies to find the direction to a good one. Of course, this would work only inside a chain of responsibility, so it's impossible to create a reliable control hierarchy at once by taking random people from the streets, this fundamentally makes the society not a good self-coordinator.
The second extreme represents the leak of control to the lower levels of hierarchy, and you may easily imagine its consequences, although there may be no other choices, if there is no chain of responsibility or a subsystem does not have enough competency.

It seems that Kiryuin uses only fear at the lower levels of her hierarchy, and this should be the main drawback of her system, because only fear which enforces responsibility does not allow to find a way to a good strategy. But this may be not necessary, if her social system is intended for use as food by the fiber life form; it represents a steady pyramid, where more powerful social layers are sorted and placed to the top by the enforced "natural social selection" and oppress/discipline the bottom ones. 
Yeah, this resembles the 100-year-old political cartoon of American capitalist caste society pyramid, but Kiryuin sorts her bureaucracy by competency.

In theory, liberal values should help to overcome hierarchical segregation through universal suffrage and formal equality of all members of society, including also the members of the control hierarchies who ideally should be chosen by competency. It's obvious, that it's practically impossible to avoid some sort of hierarchical segregation in any society, because it's principally impossible to avoid hierarchy in the process of control. So, as we saw in the example of hive, the tradeoff between a hierarchical or flat system is the tradeoff between the relative personal freedom and relative personal prosperity; it's also obvious, that a capitalist system offers more personal freedom. But the problem of such segregation becomes aggravated in the presence of overly egocentric establishment and absence of the rule of law.

Why representative democracy is also fundamentally flawed

Although modern technology allows to relatively easily implement the direct democracy, and this may be especially useful in some situations, no one would seriously discuss it as a primary political system because, yes, you already know, because society is not a good self-coordinator.
Representative democracy is better than the direct one because it provides a set of controlling organs with the corresponding control strategies and chains of responsibility; in addition, it has the bottom-top algedonic feedback channel which makes system's chain of responsibility bidirectional. This allows to perform a some sort of painless communication between the different layers which talk on the languages of the systemic and personal aspects, although every system periodically faces the language of emotions at the protest actions.
In an ideal case, people should choose a strategy according to their interests, but you know, that the controlling organs may have their own interests, and in reality controlling strategies choose people through attractive promises. So, there is generally no strategy in people's choice of strategy, which shows again that society is not a good self-coordinator.
You might say, that this may work as a heuristic algorithm in which people evaluate a strategy post-factum, and yes, this may work, but only if the system is protected against the traps of chaotic fragmentation and overspecialization, so you might need to limit (traditionally or legally) the amount or possible types of strategies and put a limit on their relative size along with the time of application.
As the strict autocracy may give priority to the interests of the system without any attempts to find any balance (as Kiryuin's system does), some flawed strategy of the representative democracy may give priority to its promises. This may ruin the system, because system's interests may fall out of strategy's ones. So, the competition of strategies does not guarantee viability. As Kirhyin's system shows, a system may be viable only if there is competency-based competition among the bureaucracy inside the strict chains of responsibility.

The Personal Aspect

As we saw earlier, the democracy is actually a form of soft autocracy and from the systemic point of view in the general sense the proper process of social control has nothing to do with liberal values: its key aspects are the intelligence, capacity, responsiveness and responsibility of the controller. A properly built social system is also interested in civil responsibility of society. If there is no such thing, Kiryuin's type of strict autocracy is probably our only choice if we want to immediately enforce discipline, select the competent bureaucracy and reduce corruption, but there are some problems with social freedoms. We all value freedom, and of course, a certain set of liberal values may reflect society's wish to make the system to take its interests into account. And it's good if a bottom-top algedonic feedback channel works keeping the system functioning in the domain of society's interests. But what we should do when the lack of the balance between interests of the system and society casts the society into the Brownian chaos or leads the system to corruption?
Some say that society should help in the development of an independent juridical system which then should lessen the corruption and thus the amount of various non-governmental advocacy organizations. But if this is too complex goal to achieve, there is something other that we should consider as well. Civil responsibility is a thing that we probably should actually value, remembering at the same time that the society is not a good self-coordinator and never will be. How would you cultivate it in people? You already know about the language of emotions which is understood by every human being, and you may use it if a person does not behave civilly and responsively when he or she should do it. But if you are disciplined enough, you probably would not want to participate in a system which talks to you on the language of instincts and emotions or prioritizes its own interests. Probably, you would choose a system which prefers the language of reason. A worthy choice, is there such a political system? So, we also have to defend our interests, what else can we productively do? But if you don't know what exactly you're defending, you are a part of a something that is not a good self-coordinator.
The understanding of when it's reasonable to defend our interests against the systemic ones, of what exactly it's necessary to do to defend them along with the estimation of our real chances is a far more complex task than the stimulation of system's algedonic channel at protest actions. You probably need to understand the limits of the flat self-coordination that society by itself can provide and also where your sphere of interests crosses or concides with the interests of others and, of course, with the interests of the governmental control hierarchies. You will never be able to get rid of them, you know, the society needs them because it is not a good self-coordinator, and any hierarchy will also always tend to protect its own interests besides the interests of its activity. Even a non-governmental hierarchy which tells you that it fights for your rights and prosperity against interests of the governmental hierarchies is not an exception, even if it actually does its work. You know, the hierarchies are comprised not of ethereal cybernetic systems but of people endowed with some status, and it's naively to think that a hierarchy will reject its own interests when it crosses the boundary which separates it from the system, especially if it explicitly intends to do this.

So, it turns out, that civil society is a society that is programmed with civil responsibility and is able to constructively protect its own interests if the government does not take them into account, although it simply might to not do this due to some objective reasons. But if society is heavily fragmented, there would be many multiply-crossing spheres of interests, so the civil society also should be homogenous. A system may use a certain set of liberal values to enforce homogeneity, but if the system does not program the society with civil responsibility, such a society would not be able to create flat chains needed for self-coordination and thus would not be able to protect anything. Some communities may have some traditional values which allow them to create such chains. But does a liberal have a right to completely reprogram them according to its own criteria if he respects freedom, the right of self-identification and personal inviolability? A quite interesting question.

No comments:

Post a Comment