Skip to content
Create an account for full access.

Not all systems are called systems

All the different definitions of a system converge on the idea that a system as a whole consists of interacting parts, which in their interaction give rise to emergence/system effect, i.e. new (emergent) properties appear in the system as a whole during the interaction of its parts, properties that the individual parts do not possess. Roughly speaking, the gears and springs in a mechanical wall clock have certain properties (for example, precision and durability for gears, elasticity for springs), but the clock itself as a whole composed of these gear and spring parts exhibits completely different properties (for example, time accuracy, dimensions). Similarly, the interior of a house, as a whole made up of clocks, walls, furniture, household utensils, and decorations, exhibits yet another set of properties (for example, ease of passage, lighting). Emergence is the appearance of new properties, such as "time accuracy" emerges when attention is shifted from the parts of the clock to the whole clock, and ease of passage emerges when attention is shifted from the clock to the interior as a whole. This is important for division of labor: one group of people usually deals with the accuracy of gear manufacturing (we use "people" for simplicity, it would be more correct to say "agents") with a high degree of skill in the method of precise manufacturing of metal parts, another group deals with the accuracy of the timekeeping, and yet another group, the convenience of movement within an interior, is handled by artists and interior designers.

While nuances may vary, the assembly of parts (we are speaking not of dividing into parts but of assembly, the relation of composition), the interaction of assembled parts, and the resulting emergence/system effect can be found in all schools of systems thinking. In different schools of systems thinking, there are two traditions for interpreting the relations between parts and wholes: engineering, where parts can only be physical objects, and general-philosophical, where parts can be of any nature - physical or mental objects.

Our interpretation in the course is - engineering-based, material/physical, where we talk about the parts of the system as physical objects in space-time. We roughly adhere to the provisions of the engineering ontological standard ISO 15926-2:2003. A part can also be a role-based object, however only when the role of this part-object is played by some constructive object or even a group of constructives. A part can be a space location - but in physical space, not "mental"/mathematical.

We do not recommend considering processes as parts of a system::behavior, but both the individual parts of a system and the system as a whole inherently exhibit behavior: changes in the state of parts and the system as a whole as they interact. The behavior of gears in clocks - they rotate and transmit motion, the behavior of springs - they compress and extend, storing and releasing energy, the behavior of clocks - they "tick" and show time, the behavior of an interior - it displays convenience for anthropomorphic agents (humans and robots) and is well-lit during both day and night. These changes in states (the assembly of the system from parts during creation, changes in parts during use) constitute "behavior," so we always link the consideration of processes to the physical world: processes are represented by parts involved in these changes of state. If one wants to depict the process of "an apple falling," it must be represented by the apple, the section of the Earth below the apple, and the change in state of the apple (its relative speed with respect to the Earth and the distance between the apple and the Earth during vertical fall). It's all physical.

In the physical/material context, for abstract/ideal/mathematical objects (classes, types, sets, etc.), there are no parts, as these parts are not physical: they do not occupy volume in the physical world, so it is unclear what they are made of. For example, it is impossible to control whether all "mathematical molecules" as parts are included in the number of the "mathematical cell" or the "mathematical organism," as a part cannot be interpreted as part of the volume of space-time occupied by the whole.

We reject systems thinking using the concept of a "part" for non-physical/mathematical/mental objects (for example, descriptions: discussions about parts of pictures, parts of texts), as it becomes entirely unclear what "interaction" results in a system effect when the parts are non-physical objects. Here, people often talk about an algorithm that performs calculations involving several non-physical objects simultaneously - but these objects themselves do not interact; the discussion shifts to the computational process with its algorithm and operational methods! This is constructivism in mathematics, a shift to operations of creating objects instead of discussing the relationships between objects. In our version of systems thinking, we do not follow this approach towards constructive mathematics but adhere to the engineering tradition and exclude mathematical objects from consideration as full-fledged systems.

Continued in the next message.