Realization, description, and documentation of the system ​

In a systemic approach, it is very important to be adequate in thinking: to understand whether we are speaking about the physical reality of our world, whether we are tied to it, or whether we are simply fantasizing about some abstract world of ideas that may not correspond to reality. If we want to reliably change the physical world according to our intentions, if we are talking about human activity, then we need to somehow tie all our reasoning to the physical world, ultimately dealing with physical reality^[1].

The adequacy of systemic thinking is ensured by the fact that when we talk about a system, we mean system realization (the same root as real, literally it's about existence in reality, reality), usually at the time of operation, when the system is working/operating, interacting with the surrounding world.

A system is always understood as a concrete realization of the system in the physical world—an individual, unique physical object. For example, it's the company Apple (all its buildings, facilities, and employees of the company at the moment when they work there), a fuel pump with serial number #12345, installed on engine #5678, the performance (dancers' work) of the dance "Baroness" on the stage of the Ust-Uryupinsk theater in the evening of October 24, 2029 (the realization of the system in the future will suit us).

How can we know that a specific embodied system exists in the physical world? We will consider that in the physical world there are (including in the present, past, and future—time is not relevant here!) only those objects that occupy some place, volume of some peculiar form in our space-time, accepting Einstein's idea of the existence of the physical world in a four-dimensional space-time. There is some piece of space-time that we call system/system and pay attention to it from its environment, moreover, the system has a boundary/ies passing in space-time. Since there is space-time here, it doesn't matter where in space (here or in the neighboring galaxy) and when in time (for example, the system will exist for five minutes, but in a hundred years from "now").

Thus, a specific system has a certain length in space (that is, size—length, width, height, radius) and in time (that is, there is a moment when it started to exist, and a moment when it will cease to exist).

And this system also has changing states. Systems change as they interact with the systems around them, which are also in the physical world and have interacting subsystems within them. For example, "door::system is open", "door::system is closed", "door::system is broken"—these are all different states of a door. States are extremely important because different methods can lead to different states of a system. For example, "open door number four at 3:15 pm on March 24"::work can move door number four from the "door closed" state to the "door open" state (for example, "open with a key, turn the handle, and push with your hand"::"work method"), or "door broken" (for example, "a proper kick near the lock area"::"work method").

Systemic (representing different parts of the world as systems from the systemic approach) consideration of the world is physical/real. The concept of "system" comes from physics^[2]. Physics selects mathematical objects to express the properties of objects in the surrounding physical world that behave similarly in thought experiments. Semantics is a transdisciplinary field that links physics with its physical objects (subjects), concepts as mathematical/ideal objects, and signs as patterns on some carrier that represent objects and concepts (the discipline dealing with signs is semiotics). You have already studied the basics of semantics in the "Modeling and Consistency" course, in our "Systems Thinking" course, the application of semantics to the systemic approach will be revealed, and in the "Intellect Stack" course, many more details about semantics, physics, mathematics, semiotics, and their connection will be provided.

Thus, we clearly distinguish (and this difference is the subject of study of the transdisciplinary field of semantics):

• system realization (representing systems), occupying volume in physical space-time, as a physics subject.
• system description (often also called system definitions, do not confuse with "dictionary definitions"!), as an ideal/mathematical/mental/abstract/conceptual object. The behavior of descriptions as ideal objects is studied by mathematics.
• document with a system description, as the expression of the description of a system on a physical information carrier. Methods of expression (patterns on a physical carrier) are studied in semiotics, the "science of signs."

In the physical world, there are some non-zero size areas of space, and in mathematics—mathematical points, in semiotics—patterns on a carrier, signs/symbols. Of course, there are also distributed representations, which are studied in representation learning methods (we translate learning in AI methods as cognition). However, we will not delve into how intuitive fast thinking S1 is organized, which uses distributed representations. We will work with local/symbolic representations, dealt with by slow thinking S2. Errors are much harder to find in distributed representations—literally they are not localized. However, errors are easier to find in symbolic representations. This is why engineering challenges are more reliably solved in compact local representations with few signs, not in distributed representations (for example, neural networks—living in human brains or computer systems in AI).

The behaviors of mathematical objects and their corresponding physical objects always differ, therefore, physics, in the course of its development, replaces one mathematical representation with another that behaves more like physical objects. For example, the sum of the angles of a triangle near Earth is slightly more than 180° not because of the Earth's sphericity, but because of the curvature of space-time due to Earth's mass. This is why it would be impossible to implement GPS navigation relying on Euclidean geometry in the physical calculations of an engineering project. Thus, physics replaced Euclidean geometry with Riemannian geometry in the theory of relativity.

Traditionally, the correspondence of descriptions as ideal/mathematical objects/concepts, expressed/presented as patterns/repetitions on some physical objects (documents), and the objects/systems they represent/deal with, is studied in semantics. However, semantics can also deal with the description of descriptions (for instance, you can describe the mathematical language used to describe a physical object). But this chain of descriptions always ends with the description of real physical objects; otherwise, we cannot agree on what all these descriptions mean without grounding/confirmation with what they describe.

Semantics is sometimes considered a branch of informatics/computer science along with algorithms; you can learn more about this, for example, in David Deutsch's book "The Beginning of Infinity: Explanations that Transform the World," where he explores the relationships between physics, mathematics, and computer science. More explanations can be found in the "Intellect Stack" course. In systems thinking, it is important that all descriptions in the world are ultimately descriptions of systems because we represent the entire physical world as embodiments of interacting systems. The caveat "ultimately" is that sometimes one description describes another.

Descriptions as mental/ideal/conceptual/mathematical/abstract objects do not have a place in the real world. It cannot be said that the mathematical/mental object "number 300" determining the height in meters (the realization of the system "Eiffel Tower") is located somewhere in the real world and has its own length-width-height. If you point to this "number 300" and say that it exists and has its own volume—you are pointing not to the number itself, but to the document as an information carrier that encodes this number with its patterns/forms as forms of material (patterns/forms of paint particles, or the transparency of the material, or something else). Therefore, the place in the physical world is occupied not by "300" as a number or a part of the description of the Eiffel Tower (number 300 describes the Eiffel Tower), but by the material object—a piece of documentation of the Eiffel Tower, i.e., the carrier/document with the information/description written on it. Documentation is the abstract/ideal/mathematical/conceptual description of the system written on some physical information carrier. One or several documents are usually referred to as the documentation of the system.

In documentation, concepts/concepts/mathematical/abstract objects are encoded by signs on a medium, representing some (ultimately since there are sometimes descriptions of descriptions) different properties of physical objects. Once again, we repeat this triad with different words: sign, concept, signified.

In systems thinking, you will need a solid understanding of the basics of semantics (for example, semantics is taught in the "Modeling and Consistency" course). It will be a pity if you poorly understand this material and will confuse a car as the target system with a photograph of the car as a document providing a description of the car. A photograph is material, but its materiality is important only as an opportunity to carry a description of a physical object—the materiality of information carriers should not confuse us—this is not the materiality of an embodiment of the represented system.

People are ultimately interested in embodiments of the system (houses, airplanes, managerial skills, gaming sessions, money), while descriptions and system documentation (information models, paper sheets with writing, images on a screen) are of interest only to the extent that without them making the embodiment of the system is difficult, especially when it comes to systems created by many people:

• When people do not even think about creating a system, but write another "concept." It is necessary to check whether the concept describes something existing in reality, not just the creation of documentation and avoids describing the embodiment of the system. For example, "regional development concept": is this a utopia? What are the tools for implementing the concept? After all, a "region" is an absolutely material object. Who and what should be done (perform work, having the resources and the authority to allocate resources) to translate the region from the state of "was" to "has become" according to the concept? Is the "concept" accurately describing the physical object "region" that will exist in the future?

• When developers of project documentation believe that their target system is the documentation. Engineers from the design bureau or design institute often confuse the realization of the system with the moments when it has already been put into operation with the information model of such a system, which they "fantasized" with no chance of realization/implementation/manufacturing. No, the target system is always the embodiment of the system; without the embodiment of the system, there is no point in dealing with its description (the result of thinking, a mental object), or the documentation of that description. Without the intention to make the system, it is about "artistic descriptions"; they can contain any mistakes, and there won't even be a desire to check them. You need to ask the designer what will be the result of his work. If the designer of, for example, a building, answers "building,” then it's okay. If such a designer answers "information model of the building"—then it's alarming since the designer's thought clearly does not reach the realization of the system and is limited to the description and documentation of the system.

• When developers of the source code of programs (describing the realization of the system) also consider that their responsibilities end when they have delivered the documentation (source code with any number of errors) to the version storage system (carrier—a computer's external memory storage). No, their target system must be a functioning program (changing its state! Program variables change their state in the computer's memory! This is a physical process!), not the source code on any carriers. Moreover, it is necessary to verify whether a working program is needed, or a working program in combination with people who can do useful work with the program—then you will also need to consider the "software in people's minds," which also needs to be developed and then manufactured (someone should be worried about the skill of the people working with the embodiment of the software at the time of operating the program, not its source code).

• The manager who developed the plan of events for a project (a description of jobs existing in the form of project documentation in the project management software). The embodiment of the system for the project (set of works) will be the organization with the necessary resources and authorities, performing the work according to the plan, and not the plan itself. Moreover, the target system, most likely, will be the system resulting from the project of these works, not the system-organization conducting the works! The compiler of the plan should primarily think about the system that will result from the execution of the plan!

• ..and so on. It is necessary to pay attention to the differences for the system embodiment (physical), the description (non-physical), and the system documentation (physical as an information carrier, not the described system) to prevent common mistakes of people playing the roles of designers in system creation projects:

  • When people do not even think about creating a system, but write another "concept." It is necessary to check whether the concept describes something existing in reality, not just the creation of documentation and avoids describing the embodiment of the system. For example, "regional development concept": is this a utopia? What are the tools for implementing the concept? After all, a "region" is an absolutely material object. Who and what should be done (perform work, having the resources and the authority to allocate resources) to translate the region from the state of "was" to "has become" according to the concept? Is the "concept" accurately describing the physical object "region" that will exist in the future?

  • When developers of project documentation believe that their target system is the documentation. Engineers from the design bureau or design institute often confuse the realization of the system with the moments when it has already been put into operation with the information model of such a system, which they "fantasized" with no chance of realization/implementation/manufacturing. No, the target system is always the embodiment of the system; without the embodiment of the system, there is no point in dealing with its description (the result of thinking, a mental object), or the documentation of that description.

  • When developers of the source code of programs (describing the realization of the system) also consider that their responsibilities end when they have delivered the documentation (source code with any number of errors) to the version storage system (carrier—a computer's external memory storage). No, their target system must be a functioning program (changing its state! Program variables change their state in the computer's memory! This is a physical process!), not the source code on any carriers. Moreover, it is necessary to verify whether a working program is needed, or a working program in combination with people who can do useful work with the program—then you will also need to consider the "software in people's minds," which also needs to be developed and then manufactured (someone should be worried about the skill of the people working with the embodiment of the software at the time of operating the program, not its source code).

  • When some companies make a big deal about "science-based" strategies, but they are not grounded to evidence or reality—This translates into aiming at making decisions that are not data-driven or efficient.

The design project of an atomic power station is ultimately the realization of a working atomic power station, not the description of the power station, even documented in the form of an information model of the station in a dozen computers. The result of a choreographer's work is ultimately a dancer's performance (executed at a specific time in a specific physical place in the universe), not the description of choreography (movement sequences) in ideas in the minds of the dancers, or even less a document (paper or electronic) with a description of the choreography. This is despite the fact that the designer themselves does not build atomic power stations but only describes them, and the original meaning of a choreographer is a describer of a dance performance (from ancient Greek χορεία—round dance, chorus + γράφω—write. The primary meaning of choreography is not composing or setting dance performances, but the art of recording dancing as an observed behavior of a dancer^[3], in simple terms, it was not a "composer," but a "scribe").

A map is not the territory. People walk not on the map but on the territory. The map is just a description of the territory, which can be presented in different documents (electronic or paper, or plastic) and this is indeed true for all descriptions/documents, not only for geographical maps. People look at a geographical map but walk on the ground.

A cocktail map is not cocktails, it" not to be drunk. The map reflects the conceptual/mental/conceptual/mathematical/abstract structure of the world of physical cocktails, and the information on it is a description of the system. From a cocktail map, you can look at it, and you can't pour it into a glass; maps only show what is in the territories. You can't drink a cocktail map, but cocktails can surely be enjoyed.

And indeed the cocktails documented on the map—the systems (embodiments of the system: products, items, businesses, people, equipment), they occupy space in space-time, can be knocked on, can be pointed at. Cocktails can even be drunk, but you can't drink map pictures. And that's why the information model of an atomic electrical station does not generate electricity. The description of a dance performance on paper or a computer doesn't evoke the same emotions in the dancer and the audience as the actual dance.

    Text translated successfully to en.