Some systems, known as self-maintaining systems, are thought to have both maintenance sub-systems and adaptive mechanisms. The maintenance sub-system sustains the relationship between the other sub-systems and holds the entire system together. The adaptive mechanisms promote changes to inputs, outputs, and processes, to keep the system in equilibrium with its environment. Living things, for example, are self-maintaining, but not exclusively so. People also create self-maintaining machines, computer programmes, etc.
Self-maintenance and adaptation are carried out through a process of feedback. Information on inputs, processes and outputs are passed to the controlling sub-system. The latter then processes this information, and issues commands, again in the form of information, to subsystems engaged in accepting inputs, in processing them, and in delivering outputs. For control to be successful important aspects of the latter must appear as a white box to the former, i.e., must be known by it. This existence of controlling and subordinate systems is known as requisite hierarchy.
Another principle, requisite variety, applies to the operation of controlling sub-systems. This principle was discovered by W. Ross Ashby and is also known as the First Law of Cybernetics. It holds that the degree of control of a system is proportional to the amount of information available. Variety refers to the number of states of a system. If a controlling sub-system can recognise all possible states, then it has full knowledge of the systems behaviour and can therefore issue appropriate instructions. If it does not have knowledge of all possible states, uncertainty arises. Ashby believed that “When the variety or complexity of the environment exceeds the capacity of a system (natural or artificial) the environment will dominate and ultimately destroy that system.”
Such systems are known as self-maintaining systems because they perform these operations autonomously and without any assistance from their environment. However, they can use a large part of their inputs in self-maintenance as opposed to producing outputs. The boundaries of systems which are not self-maintaining are defined by the observer. However, self-maintaining systems define their own boundaries. In a living system, such as a bacterial cell or a multicellular organism, this property is known as autopoiesis.
Systems with higher levels of organisation can display purposive behaviour or agency. That is, they have choices available to them, and produce an end result after a period of time. Systems with purposive behaviour can “extract” inputs from other systems in their environment, or can “exchange” them for their own outputs. Without adaptation, a system can become unsustainable. It may, for example extract inputs at a rate greater than its environment can produce them, or it may produce outputs at a rate greater than its environment can process them.
The internal organisation of a system can increase in complexity without being guided or managed by an outside source. This is known as self-organisation and relies on four main ingredients. They are: positive and negative feedback; a balance between the exploitation of existing opportunities and the exploration of new ones; and multiple interactions. The latter are not merely one-way causal inputs, but also, two-way output/input relationships with other systems in the environment. More information on self-organisation can be found here: https://psychology.fandom.com/wiki/Self-organization
The reliability of a system can be increased through redundancy. That is, the duplication of critical components. One important redundancy is known as redundancy of potential command. This principle was first identified by the American neurophysiologist, Warren McCulloch, in the 1950s. When studying the transmission of signals between the brain and the nervous system, it was found that two identical signals from the same source were being delivered by a primary channel and an auxiliary channel. From this, McCulloch developed the principle that knowledge, i.e., correct information, constitutes authority. This will be explored further in my next post.
To these principles, I would add the two variational principles described in my earlier article on decision making and behaviour. That is, pressing needs and the efficient use of resources. A self-maintaining system may have several functions but limited resources. So, it is necessary to prioritise its processes and attend to the most pressing needs first. It must also employ its resources as efficiently as possible to maximise the benefits of its processes and outputs. Together, these variational principles help to maintain the system and contribute to the likelihood of its continued existence.