Evolution

Physical and informational fields exert an influence upon matter and its structure.

The optimization process aiming at the reduction of free energy has to arrive at decisions. The direction of the steepest potential drop has to be determined. This basic task holds good for physical and cybernetic systems in equal measure. For cybernetic systems in the process of evolution, the additional task of adaption to an informational field comes into play.

This decision process requires information about the spatial structure, about potential differences. When taking a straight course in the sphere of influence of physical fields, this decision process is very simple: the system follows the direction of effect of physical fields. Structure and energy are not yet separated from each other. After the separation of structure and energy, as it is realized in cybernetic systems, the motion of the physical system in the direction of the steepest potential drop takes on a new quality. The direction of motion is determined by structures existing separate from the energy sources. As such, a motion in the direction of the steepest potential drop doesn't necessarily have to take place. New factors intervene into this simple optimization process. It is not the local energy distribution which supplies the optimization criterion but the structure exerting an influence. The physical criterion of the steepest potential drop continues to be valid. It is merely superimposed by a new criterion determining an essential part of the further evolution. See Fig. on pg. and pg.

Hence, the separation into energy- and information sources is the prerequisite for the evolution of animated nature. A flow of information transfers structure onto matter adapting to that structure. As the physical field is inevitably accompanied by a flow of information, it itself determines the changes of the target system in the overall physical system.

As matter under the influence of a field can also be the source of physical processes, order can be accumulated across several stations which is an additional prerequisite for evolutionary processes. Matter therefore is subject to changes caused by the flow of information acting upon it, with the changes accumulating steadily.

In the following, the cause of changes, the source of information, shall be characterized as evolutionary criterion. The affected system adapts to the evolutionary criterion in an optimization process, resulting in the overall physical system heading towards a lower level of energy. Cybernetic systems will, to the degree that it is possible to them, adapt to a criterion. Structure and energy are going separate ways. The entropy can increase faster or slower.

The physical system always takes the path of the steepest local potential drop. In doing so, it can happen that the system is moving upwards, relative to a field exerting an influence upon it; thus, it is building up field energy rather than reducing it. However, this energy always originates from another field. The total potential always drops, and the entropy of the overall system rises steadily. As field effects can be initiated via amplifiers which then depend upon the structure of weak fields, a system can be moved by a field which does not possess a momentarily real space/time structure, but the structure of which reproduces stored information. Hence, the source structure is located in a farther spatial/temporal distance. Information transports this physical field/space structure, thus possessing a field character itself. With respect to the effect on a cybernetic system, we can justifiably talk about an informational field.

The mechanism of adapting to a criterion also dominates the evolution by natural selection, because the preferred systems are those which, in the sense of the criterion, have already experienced the steepest potential drop, and/or which are most in accordance with the evolutionary criterion. Thus, the overall system also follows the evolutionary path of least resistance and/or steepest potential drop. The evolutionary mechanism of the genome shows parallels to the law of entropy governing the physical world. The task is to enter this trend of changes, induced in the object by the information of the evolutionary criterion, into a general evolutionary theory, and to make this theory available to as many fields of knowledge as possible. Perhaps we will be able one day to describe evolutionary processes in biology, technology, and society with one single theory.

Obviously, one thing is common to all evolutionary processes: an object changes under the influence of a source of information. General laws and principles seem to be at the basis of the large trend of these changes.

First of all, let us take a look at the evolutionary processes effecting the transport of information from the evolutionary criterion to the object.

The first aspect to mention would be natural selection. See Fig. on pg.

It facilitates the "self start" of an evolution as it only requires a passive system and can do without any additional intelligent system. The evolutionary criterion determines the survival of system copies slightly deviating from one another. Natural selection optimizes the system in the sense of the criterion. Whatever is adapted best to the evolutionary criterion has the best chances of survival and/or the best opportunities for procreation. At the same time, the object is completely passive. It cannot move actively in the state space in which the optimization takes place. Motion in the state space comes into existence only by means of deviations occurring during the process of copying the single systems. Motivity in the state space by means of deviations during the process of copying is a prerequisite for evolution coming into existence according to these processes. If no potential difference, no criterion exists, no directed motion will gain momentum. The systems would diverge.

Another evolutionary process shall be termed here as "construction". See Fig. on pg.

A higher-level cybernetic system changes the object in the sense of the evolutionary criterion. The criterion can, but doesn't have to be, a component part of this higher system. If the cybernetic system does not produce the changes by itself, but if it makes a selection from different individual systems, a natural genetic selection is the case. Again, the criterion can be a component part of the system selected but doesn't have to be. Hence, what is required for the evolutionary process is:

1. a criterion providing the required information;
2. an optimization process, a mechanism moving the object in the state space in a targeted or random (stochastic) way, changing it;
3. an evolutionary process transporting information from the criterion to the object.

Some additional evolutionary processes are to be mentioned. Self-organization requires a higher-level system which already possesses a model of the environmental structure which it can orient its active adaption to, directed at itself and its well-being. See Fig. on pg.

The system must be capable of moving actively in the state space. It must be in a position of being able to change its characteristics, its behavior.

1. In order to test the direction in the state space, the capability of active motion/change is sufficient, as well as the existence of sensors reporting back the information of the criterion.
2. Part of the oriented self-organization is an environmental model, a model of the criterion which makes an orientation in the sense of a projection (advance calculation) possible.

A system which has acquired a model of its environment, albeit fractional and incomplete, is already organized fairly high. As the system wasnīt active from the beginning we can say that the criterion has made itself independent, because the system is, in its entirety, its product as the information originated from the criterion exclusively.

The criterion became the system; function by function, it has reproduced itself into the cybernetic system. What used to be a requirement in the criterion is being realized in the system by means of functions. See Fig. on pg.

With the aid of evolutionary processes, the dead passive order of nature has transformed into a cybernetic system. Its activities, however, are not confined to itself.

In the course of the action, the environment turns into the object. The system undertakes to change its environment, stochastically via sensor feedback or in an oriented manner - according to its evolutionary status. Now the cybernetic system itself constitutes the criterion. It determines how its environment has to look. See Fig. on pg.

Function by function, hierarchy level by hierarchy level, the criterion copies itself into an existing cybernetic system, starts to become consciously aware of itself, and changes its environment, all this being a consequence of field effects - evolution without confrontation with the increase in entropy! See Fig. on pg.

Each single condition the criterion imposes upon the object follows the evolution towards higher levels, as rapidity of adaption is a general requirement. The system is subject to variation. Natural selection picks suitable systems. In the state space we can observe a motion within the dimension corresponding to this requirement. The systems capable of free mobility within this dimension have a significant advantage with respect to selection. The requirement is adopted as a function by the control system. The system was able to speed up the rapidity of adaption and to gain more freedom. In this way, the long-term evolution of a system turns into active short-term adaption, into a mobility of the system.

I cannot identify a significant difference between long-term evolution and short-term motion. Obviously, both are subject to the same principles of motion and/or evolutionary processes. The higher-level evolutionary processes are more efficient, that is, they provide the same progress with less material expenses in a shorter amount of time. Consequently, an explanation would have to be found as to why the rapidity of evolution in nature seems to accelerate despite an increasing complexity of the systems. Thus, it should be suspected that evolution according to the principles of natural selection exclusively would result in a deceleration of the rapidity of evolution.

The formation of evolutionary processes thus strives for faster methods and, as we shall see, for methods getting along with as little information from the environment as possible, that is, methods oriented to a model. These are even faster and require less input information for the same performance. Slow evolution thus turns into free mobility. The fact that we can move freely, respond to stimuli, and pick up information from the environment is due to the fact that the evolutionary processes themselves are naturally subject to certain evolutionary pressures as well. See Fig. on pg.

As we already could see from the description of the evolutionary processes, each process, except for one, has basically two options at its disposal to realize the directed motion of the object in the state space:

1. selection of random motions;
2. directed motion via orientation to the environment and/or a model.

With respect to natural selection, understandably there is no possibility of optimization via orientation. Feasible in this case is only the slow method of natural selection bringing about the least flow of information, but which is capable of a "self start" as soon as a system is able to replicate itself with slight deviations, and/or is able to move in the state space with random steps, capable of retaining feedback information from the environment.

This capability of order retention, of recording and storing particular states of order, is a basic requirement for an evolution which, after all, is nothing else than the storage of order information flowing in from the criterion. The criterion is stored in the object. It produces an image of itself in the system subject to evolution. Motion via orientation represents an evolution towards higher states resulting in the faster attainment of a more beneficial condition with less input information. Whereas natural selection requires years for a certain plant species to populate a favorable habitat, an animal in contrast requires only a few hours. In addition, the animal has completely different ways and means at its disposal:

1. complex environmental model;
2. highly developed sensors;
3. free mobility in space.

These factors are contingent upon one another, therefore they can only evolve mutually, parallel to and backed up by each other. If an important requirement for one factor is missing so that it canīt evolve, it does not result in evolutionary pressure for the other factors.

Due to their very primitive environmental model, the possibilities of motion for plants are poorly evolved. They are limited to differences in concentration of growing agents and turgor (osmotic pressure) changes. Light, gravitation, or stimulation by touch trigger directed growth or motion. Simultaneously, the growing agents are sensor, logical connection, and actuator; as such, they represent a very simple and efficient system. Wherever light is destroying growing agents, the plant shows a slower rate of growth, thus bending towards the light. In this context, an impression of the surrounding space, an environmental model, has already been stored as information. Motions due to turgor changes, e.g. with Mimosa and Venus flytraps, are already significantly faster. Their mechanism is more complex as well. Sensor and actuator are separated from one another. Stimulus conduction to the actuator has to take place. The entire system is only functional on the conditions it has evolved under; in other words, it has internalized and modeled these conditions.

Without a system equivalent to a nervous system, storage of a more complex environmental model doesn't seem to be possible. On the other hand, it is evident that an absence of sensors would not allow the build-up of a current environmental model. Only very old functions and/or structures such as, for example, the three-dimensionality of the space surrounding us seem to be available to higher developed animals prior to any other experience. These have already advanced, by natural selection, into the permanent storage device called genome.

With the aid of the optimization process of orientation, the result is a significantly higher flow rate of information. A higher speed of evolution is attained, culminating in the system being able to directly react to the environment and even being able change it. It is not necessary to scan the space with the trial-and-error method, nor is a slow, wear-and-tear natural selection required which optimizes the system as such, and where each individual merely represents an experimental trial. Orientation can take place directly in space by means of sensors, or it draws on an acquired environmental model in its further evolution which represents stored experience, providing more information for the system than the environment can at the moment.

Therefore, the process starts with systems copying themselves with slight deviations, and which thus can be subject to natural selection, that is, evolution. At the outset of evolution the system is passive. However, there is not just an evolutionary pressure to adapt to certain environmental conditions, but there is also a pressure in the direction of the ability to adapt and react accordingly in general. Thus it follows that evolutionary processes are also subject to evolutionary pressures.

The functions acquired by the control system in the course of evolution are, in the final analysis, adaption functions. The control system bestows the system with adaptability. More and more, it takes over functions from the criterion. First are the basic and most elementary environmental demands impressing their structure upon the system and exerting the strongest pressure upon it. Each single function the system control is capable of executing is subject to the pressure to adapt. In the course of evolution, it integrates itself into a control hierarchy in which higher-ranking functions are acting upon subordinate functions like command variables. All larger systems represent such hierarchical structures. Either the demands of the criterion itself are structured in such a way that the system is only an image of the criterion, the demands of which it has taken over step by step as system functions, or demands such as reliability and performance efficiency enforce the formation of hierarchical structures.

The evolutionary pressure results in an advancement of evolutionary processes and optimization processes. See Fig. on pg.

The question about the flow of information can be answered in five different ways:

1. natural selection of systems conforming to the criterion » genetic selection;
2. active information reception via self- organization » self-organization;
3. active transfer to the environment » action;
4. transfer via the cybernetic system to the passive object » construction;
5. conflict/contest between active cybernetic systems » battle.

The last two processes can also be considered to be a special version of the third one.

Optimization is possible by:

1. natural selection;
2. orientation to the criterion;
3. orientation to a theory.

The evolution of individual system functions to higher states, and thus the evolution of information reception, takes more or less place according to Fig. on pg. in the appendix. It must be kept in mind though that the classification applied here can only be an auxiliary means for a clear layout of the process described.

Evolution is a reciprocal process with one advancement acting as a prerequisite for the next one. It might never be possible to exactly analyze the evolutionary path of large systems; however, we should be able to describe the evolutionary path of an individual system function.

Energy consumption and assimilation is one of the fundamental problems a living organism has to solve. This example can serve to clarify a few points. Living organisms incapable of motion are completely dependent upon luck and coincidence when it comes to their habitat and the question of sufficient food supply at all times in that location. Certainly, these are not favorable conditions. That is why plants invented the active motion towards light by accelerating growth in parts of the plant not subjected to light. The ensuing motion towards light proves the existence of a primitive model of space. Mobile protozoa are able to master problems of space three-dimensionality and varying food distribution with the following models or mini programs:

"slow down motion whenever food is available" or "move in reverse direction whenever food becomes scarce".

In this case, a function is already accessing the function of motion on a lower hierarchy level. Prerequisites for this simple form of self-organization:

• simple sensor
• primitive environmental model
• locomotor system

The improvement of these functions is still left to natural selection as long as no learning aptitude exists, and as long as sensor and locomotor system cannot be optimized on a conscious level.

If the sensor provides a better image of the food supply than just signaling its existence, and if the control system is able to identify it and trigger correctly corresponding motions in its direction in space, then a living organism is already able to chase after the object. In doing so, it increases the evolutionary pressure on this object, resulting in an intensified "evolutionary arms race" between robber and prey. Thereafter, natural selection unearths the most versatile forms such as, for example, the hooked tentacles of deep-sea fish.

In the meantime, the transition to self-organization continues. Hunting strategies are geared to the different types of prey. Obviously, the behavior of an animal is more flexible than its anatomy. Significantly faster learning processes are evolving. Thus it becomes clear that self-organizational processes in individuals can be primarily recognized in behavioral changes. A self-organization of sensors and/or the locomotor system is still out of question at this point. It is not until the emergence of man that attempts are undertaken to compensate for these deficiencies with: animals of prey, mounts, long-range weapons, binoculars, vehicles, etc.

Of course, vehicles were not designed for hunting in particular as the evolutionary criterion is very complex. Anyone considering this too exhausting and tedious should resort to other methods of supplying food, for example by fencing livestock, setting traps, growing crop, etc.

That in itself is already action in the sense of directing a personally acquired criterion towards the environment. Man, as a product of the environment, changes this environment according to his own criterion and a model acquired.

But this is not the end of the flagpole; further advancements are possible. If mortal man intervenes into the evolutionary path of any given animal species for the purpose of selective breeding - for example if it tastes too bland to him - then the future evolutionary path for this species is determined by means of construction via selection, because a cybernetic system projects its criterion into the object, adapting it to itself. From the viewpoint of this do-gooder, it is just action, that is, environmental adaption via selection.

Targeted genetic manipulations already point towards the construction of a model and towards optimization via orientation.

The emerging system is an image of the criterion providing information. The buildup of its control hierarchy should thus reflect the demands of the criterion on the system. A close look into the system interior would simultaneously constitute a look at the evolutionary history of the system, and it should also reveal the conditions that existed during its inception. In biological systems we can recognize a hierarchical buildup: the formerly self-contained cell, tissues and organs. The behavior of the living organism is hierarchically organized as well. Younger higher-level functions can fall back on older, more elementary ones. Such a consistent hierarchical buildup, with the system even being able to replace failing functions and thus repair any potential damages, is largely missing in today's technical systems. Maybe we will always be able to see through artificially, that is, constructively created systems and recognize that they haven't come into existence on a "self-start" basis. Given this, the possibility should exist to take a look at unknown systems and to be able to recognize the ways and means of their origin/formation.

In essence, the pressure the system and its evolution is subjected to is characterized by the following two factors, free space and speed of information reception.

1. The system expands its free space by consistently taking over adaptive functions into its control system. This fact is easily understood. A system penetrating a state space - in which it was previously moved only passively via selection - with motion and orientation, has a considerable cutting edge, a significant advantage, over its competitors. It can react to changes in the state space, it can tap more food supplies, it is less prey to enemies, etc.

   Any dimension in which only a shift via selection was possible before, and which the system learns to master, becomes a newly acquired degree of freedom. This strive for freedom extends beyond motion in the state space all the way to its mastery via active change, that is action. This urge towards autonomy is the basic requirement of any evolutionary criterion - in fact, of evolution as such.

2. Evolution means gaining freedom - and time! In the case of natural selection, the speed of information reception is the lowest. The more the system evolves, the more it resorts to higher-level evolutionary processes. It is therefore in a position to pick up information faster, in short, to speed up its rate of evolution. Indeed, the speed of evolution in living organisms seems to have increased in the course of natural history. Living organisms move from passive information reception to active information reception, to the development of a theory, a model of the environment. Gaining information, freedom, viability, and speed could be considered as basic requirements of a general evolutionary criterion. In order to increase adaptability, the transition to orientation is necessary. The amount of information increases. A theory of the environment is formed which constitutes the prerequisite for action. This theory represents the assimilated criterion the system was subjected to. It encompasses its entire reservoir of functions assimilated and will, at a later point, confront the environment again as working principle of the system. This environment doesn't just consist of dead matter. The other systems watching this system will refer to all his reactions as its working principle. The criterion under which the system came into existence confronts the environment as its principle. Therefore, whoever stipulates the requirements a system has to conform to, in the course of its evolution, is responsible as well for the way that system will behave towards its environment later on!

As already mentioned, the theory is also subject to a hierarchical classification. Elementary functions settle on lower levels. The higher the levels examined are, the more general and abstract are their input and output quantities. The advancement of the environmental theory heightens the degree of abstraction of the higher control levels. The evolutionary conditions of the criterion are conceived in increasingly abstract terms. See Fig. on pg.

Abstraction permits the system to reduce storage efforts for the environmental theory. Different functions can resort to a lower-level function which can thus be utilized in multiple ways. This is directly connected to the hierarchization of the control system. Abstraction permits the system to make statements about environmental reactions for which there are no experiential data available at that time. This in turn allows for an extrapolation beyond the immediate sphere of experience.

With these advantages, abstraction constitutes a pivotal progress in the evolution of a cybernetic system. It becomes possible to sift universally valid information out of the environmental information. In doing so, the most important basis for higher-level cognitive processes, for example optical and acoustic identification, has been created.

In the further course of evolution, the system will not only register its environment and abstract the information received by it so as to conceive the nature of this environment; in addition, the integration of system/environment relationships will increasingly gain in significance. It is certainly an advantage to be able to reflect on the consequences of one's own activities already at a relatively low level. An integration of the consequences of an activity into the theory already starts at a low level; that is inseparably connected with action. At this time, we are of course still dealing with unconscious behavior. However, in the course of advancement to higher levels, most notably within the scope of increasing theory abstraction, an abstraction of system/environment relationships ensues, leading to the evolution of conscious awareness on the part of the system.

Via system/environment relationships, the system starts to conceive itself. Only at a later point it will learn to conceive itself as part of the environment/world.

Every cybernetic system physically capable of constructing an information processing mechanism enabling the system to go into action, and which is capable of abstraction, that is, a gradual buildup of higher system levels processing more general information, will eventually obtain conscious awareness. It is the inevitable consequence of reflecting on system/environment relationships.

Hence, in the evolutionary history of man there has not been a sudden transition to today's conscious awareness. Much rather, it is a presently existing awareness having evolved gradually, and which will evolve further in the future with increasing velocity. For this reason, it would be utterly nonsensical to completely deny animals at a higher evolutionary level a conscious awareness. Their awareness does not seem to differ so significantly from ours. Analogous evolution of awareness within animals of different tribes only proves the determinant function of the evolutionary criterion.

In addition, I am convinced that there is no reason why it shouldn't be possible for machines and societies processing information to be able to attain such a high evolutionary level. In my view, the correlations presented here are universally valid as they are only oriented to the flow of information.

It would be very important, both for natural and social sciences, to examine whether it is possible to formulate a general evolutionary theory on the basis of cybernetics. Based on the physical nature of order, the information, it would be a comprehensive theory which could interconnect physics, cybernetics, biology, ethology (behavior research), and the social sciences. It could represent an extension and enhancement of cybernetics as the philosophy of the natural sciences.

Analogy examinations between different cybernetic systems like biological, technical, and social ones, could lead to such a theory and an activation of the sciences involved for mutual benefit which would facilitate approaches in the field of natural philosophy and, most notably, which would enable mankind to master its social problems in due time.