Georg Ivanovas From Autism to Humanism - systems theory in medicine

4. Systemic Basics

previous -- home -- content -- next


4.11 States and the nature of change

Synergetics is a concept of the German physicist Hermann Haken modelling spontaneous developments of order (Haken/Haken-Krell 1989: 6), respectively the emergence of phenomena in non-linear systems (Haken/Haken-Krell 1994: 41). It is a combination of recursive logic with chaos theory. It is here presented at some length as it reveals a lot of characteristics also important for medical purpose.

A simple example is the heating up of a fluid (e.g., silicon oil). As long as the difference of temperature between top and bottom is still small, no movement is observed. When the difference rises, the warmer and specifically lighter fluid moves upward, whereas the colder and specifically heavier fluid sinks downwards. This is no irregular process. The fluid develops a typical pattern with a sort of rotating cylinders. (Haken/Haken-Krell, 1994: 19-20)

The fluid is in an relatively stable balance of which can be said that it is the best way to transport the heat to the surface. It is the best way because it is in balance.

Therefore, the fluid is either at rest (little difference of temperature) or has a certain order. The order can be described by a mathematical parameter of order (q). In the words of Haken this parameter ‘enslaves’ the whole fluid.

In an analogical picture, Haken compares the behaviour of the parameter of order (q) with a ball in a hilly country. If the balance is disturbed, the ball rolls down again to the minimum (Haken/Haken-Krell, 1994: 23).

If there is only one optimum (minimum) for the present task, the valley is very narrow. Every intervention into the system (stirring in the vessel with silicon oil) will have no lasting effect. The same order as before will emerge, and, in the analogical picture, the ball will fall again to the bottom.

If the temperature is changed, the order of the fluid will change, as well. Some cylinders will grow more quickly, others more slowly. Some will survive, others not. It is Bertalanffy’s struggle between parts (chap. 4.7). In the analogical picture this means that the valley becomes more flat. If the temperature exceeds a certain limit the hilly country develops spontaneously two minima, i.e., two different orders might develop. The system has to decide to follow the one or the other (Haken/Haken-Krell, 1994: 25).

Illustrations after Haken 1994: 19-25

Heated silicon oil in a circular bowl has a longitudinal pattern that might have any direction. In this regard it is multi-stable. The direction of this pattern might develop spontaneously or is induced from outside through stirring in the fluid (Haken/Haken-Krell, 1994: 30).

If, through stirring, an already existing directional pattern is disturbed, this does not necessarily change the pattern. It might, however, change the direction. When enough cylinders remain undistorted they are able to reproduce the pre-existent order (Haken/Haken-Krell, 1994l: 28).

This is, in very short words, the concept of synergetics. The analogy of the hilly landscape is a description of free energy (Haken/Haken-Krell, 1989: 16-17) and represents nothing else but the attractors in their basins (Haken/Haken-Krell, 1989: 50-51).

The change of a parameter (temperature) might lead to an instability. The system leaves its old state and strives for a new one. Some collective movements serve as principles of order. They might increase and enslave other parts of the system (Haken/Haken-Krell, 1989: 27). By this the parameter of order (q) is comparable with the operator of the recursive functions (chap. 4.2), as states are always the result of recursive processes. The emergent states can be monostable, bistable and multistable.

Bi-stability, a very common phenomenon in physiology, “is the tendency for a system's output to be drawn toward either one or the other of two stable states. For example, in excitable cells such as neurons, depolarization elicits sodium entry, which in turn elicits depolarization—a positive feedback loop. As a result, large depolarizations drive neurons to fully discharge their membrane potential, whereas small depolarizations decay back to a resting state. Thus, the neuron tends strongly toward one or the other of these two states. The stability of each state brings with it a sort of intrinsic robustness— i.e., once a cell is in one state, it takes a fairly large disturbance to move it into the other.“ (Lander 2004).

Most physiological phenomena are based on a bi-stable processes on a cellular level, such as nerval excitation or receptor reaction. Lander could demonstrate that in Drosophila melanogaster bi-stability on a cellular level is also able to organize genetic expression.

Haken showed that also phenomena of perception might be explained in these terms. Rubin’s vase or Necker’s cube are typical examples for a bistable visual state.

An example for multi-stable states is the bacterium Eschericha coli. Although the bacterium has 7 genes and more than 1000 metabolic regulations and is able to live in many different environments (15,580 different environments had been tested in computer simulation), the bacterium actually exists in only about 5 different metabolic states (Barrett et al 2005).

Time and again we see the described principles in everyday medical practice. Stable states can often be observed. An example of my practice: A mid-aged women developed an anaphylactic rush after she had a bath in the cold sea for nearly one hour. From that day on she always developed this rush, whenever she took a bath in cold water, even if it was only for a very short time. Similarly many pathological states just show up and remain. Adverse reactions of drugs, especially of the delayed type, mainly express that way. Often this happens in times of stress, when the person is heated up, just like silicon oil.

A classical example of heating up are the exercise ECG or the oral glucose tolerance test. These tests reveal hidden tendencies. In systemic psychotherapy it is often necessary to set the family under stress in order to reveal the pathological pattern of interaction. When everything is fine, no abnormal behaviour might be visible.

Pathological states of later life often show first symptoms quite early. 4100 normotensive black and white men and women were exposed to distressing tasks (cold pressure, star tracing, and video game tasks). The larger the blood pressure responses were to each of the 3 tasks, the earlier hypertension occurred (Matthews et al 2004). In the words of the model it could be said that the valley of hypertension was very small in the beginning and the pattern of hypertension developed only when a lot of energy was administered. This pattern changed during lifetime with the hypertensive valley becoming deeper.

Important is, however, that these changes of state are rather quick. The parameter of order (q) changes from one minimum to another just as seen in the ambiguous pictures. Change is no linear or gradual development. Not even processes like the climate change. In the history of the earth there have been two different states of CO2 levels, one around 190 parts per million during the glaciations, and the other around 280 ppm during the interglacials. No other concentrations persisted for long (Pearce 2003). There was either glacation or conditions similar to ours. The transition from one state to the other might not have been as quickly as in the movie The day after tomorrow. But ecosystems rarely undergo gradual changes. Mostly the change is abrupt and sometimes irreversible (Rietkerk et al 2004).

This indicates another characteristic of states: if a state is attained, it is not changed without an outer or inner stimulus. This is even true if another state would be as likely or more likely for the defined system. Even when a transition from one state to another takes place, the system does not react immediately to the stimulus, as the parameter of order first has to leave the valley. This phenomenon is called “hysteresis”.

Hysteresis is a term in physics that literally means to be late. It describes systems that do not directly follow the forces applied to them, but react slowly, or don't return completely to their original state: that is, systems whose states depend on their immediate history. For instance if you push on a piece of putty it will assume a new shape, and when you remove your hand it will not return to its original shape, or at least not entirely. Hysteresis phenomena does not only show up in magnetical and ferromagnetical materials, but are present in the elastic and electromagnetic behavior of materials, in which a lag occurs between the application and the removal of a force or field and its subsequent effect. Electric hysteresis occurs when applying a varying electric field, and elastic hysteresis occurs in response to a varying force. Although the hysteresis loop depends on the material property behaviours, there is no complete theoretical description that explains the physical phenomenon. Hysteresis was initially considered to be a dirty, unwanted, phenomena of materials. But its behaviour is now considered to be of very great importance in technology, and the property is for example used when constructing permanent memory” (Wkipedia 3.8.04)

In medicine, hysteresis is known in a lot of contexts. In antidepressive therapy a single dose of an antidepressant drug is able to boost neurotransmitter levels to the defined level, but it takes weeks to bring relief to a patient (Farley 2004).

By no means do changes from one state to another follow a linear relation of cause and effect. Haken explains this with the laser (Haken/Haken-Krell, 1994: 38-39). Laser waves consist of one frequency and are synchronized. In order to build up the typical laser wave, a certain critical energy is necessary. Below this critical point only spontaneous emissions occur. When the critical point is reached waves become synchronized. It is a recursive process between two mirrors where the parameter of order enslaves the spontaneous emissions (Haken/Haken-Krell, 1989: 56). However, if energy is raised beyond another critical point, a totally new behaviour can be observed. The laser emits flashes of light. This means that there are different, distinct states that are attained through a gradual change of the influencing parameter (administered energy). Such points in the change of the behaviour is called ‘phase transition’ in physics. A superconducter might become a superisolater only by a slight change of temperature (Vinokur et la 2008). Some believe that the brain activity shows an according behaviour with a spontaneous activity in rest and a phase transitions towards certain patterns during action (Cowan 2008). It is rather probable that other physiological or biochemical process work the same way, e.g., the immune system.

*

The synergetic model shows that the usual idea of a linear relationship between drug administration and response is not supported by scientific findings. Even the opposite, the dose related reverse effect (Kratky 2003: 83-84) is seen in physics. This phenomenon is called hormesis (chap. 6.10). “Hormesis, a dose-response relationship phenomenon characterized by low-dose stimulation and high-dose inhibition, has been frequently observed in properly designed studies and is broadly generalizable as being independent of chemical/physical agent, biological model, and endpoint measured” (Calabrese/Baldwin 2003). The implications of this effect are manifold (Kaiser 2003, Calabrese 2008). A medical example for this is the protein alpha-Synuclein which might contribute to Parkinson’s disease in a higher concentration and protect against Parkinson’s disease in a lower concentration (Chandra et al. 2005).

As the normal medical approach does not refer to states, it is unable to say anything about the conditions under which a state arises. Although textbooks and reviews list a lot of parameters as etiological factors, they are often an epidemiological patchwork of unrelated facts. A review of schizophrenia mentions genetic factors, environmental factors as prenatal and perinatal events, social class, family structure, but also pathophysiological factors such as alterations of the brain anatomy, blood flow or neurotransmitters (Mueser/McGurk 2004). All this does not contribute to a coherent picture of development and change. This is a pity as schizophrenia and even more bipolar disease are classical examples of bi- and multistable processes.

States and their changes have been explicitly observed in family therapy. The presence or absence of certain family members may totally alter the behaviour of a patient within seconds (Minuchin/Fishman 1981).

We all know such changes through the change of the environment. We observe it in ourselves and in others. One moment we feel full of energy, the next moment we feel exhausted, or a self-confident person might become a helpless child being together with his parents.

In psychotherapy synergetics has been defined as complex ‘(psycho-) physiological’ phenomena (Perlitz et al 2004) or as distinct affective and cognitive states of processing and experiencing (Beierle/Schiepek 2002). A report of a psychotherapy structured according to the principles of synergetics shall be analysed more in detail as it reveals some basic problems of medical perception.

Beierle and Schiepek analysed the tapes of a therapy of a young women and mother who suffered from functional problems of stomach, heart and circulation combined with fears and a lack of self-confidence. In the treatment the principles of ‘brief therapy’ were used (Beierle/ Schiepek 2002).

The authors defined six states:

  1. reports and looks for help
  2. suffers and moans
  3. resigns, has given up
  4. feels ‘real’ anger, resists
  5. works therapeutically
  6. self-confident and active (desired state)

Of course these definitions are highly arbitrary and not convincing for a reader but might have been coherent for the observer. The description (the map) is always poorer than the territory, That is, the state of a system can only be assessed according to the parameters taken into account by an observer. Thus, every kind of diagnosis and description leaves out many aspects. As a consequence, change can only be described according to the defined state which is still much better than using only surrogate-parameters.

Beierle and Schiepek did this in observing how often and how long the patient remained in one state during the therapeutic sessions. Then they measured the changes during therapy which lasted 13 hours. By that, the progress of the therapy could be documented and at the end of the therapy she remained mostly in state 6. This process was accompanied by a lot of positive changes in her life.

This development is, of course, pleasing. It is a result aimed for by most cognitive and behavioural strategies. But according to Ashby’s Law of requisite variety respectively to von Foerster’s ethical imperative (chap. 4.7) a more flexible pattern would be desirable. The description of this therapy does not indicate a larger variety, i. e. the emergence of new states. It was the shift from one state to another. Under stable circumstances this might be a good result, but there is no evidence what happens in a changing environment. I personally believe that a patient with such a good development shows new states. May be this was not the case with this patient, or not mentioned in the paper, or not observed by the authors, or did not show up in the therapeutic session.

This leads to a core problem of all therapy. If a state is defied, very often it is aimed to pin the patient down to the ‘desired’ state, something Bateson called ‘conscious purpose’ (chap. 6.1). Although this seems to be an effective therapy, it is true only under certain circumstances in a certain frame of time. As it represents the change of the operand, not of the operator (chap. 4.2), long-term effects might not be not touched.

These reflections show a limitation of the synergetic model. States can only be defined empirically and depend therefore on the subjectivity and theory of the observer. Due to their non-trivial nature the emergence of new states is hardly predictable by theory (Schiepek et al 2002).

Furthermore, change can only be described in terms of different shapes of valleys, how they emerge and vanish. A change on the meta-level is not foreseen. Thus, second-order learning cannot be modelled..

Another term that might lead to some confusion in this context is robustness. Robustness can be seen as the depth of a valley, the tendency of the system to remain in a certain state.

The robustness… is a result of the fact that the desired pattern is a stable steady state. In a system of ordinary differential equations… such states correspond to stable fixed points. These are generic features of such systems; small changes in parameters or initial conditions will not change them qualitatively” (Ignola 2004).

However, when robustness is defined in this way it cannot be distinguished from rigidity. Both are represented by a deep valley. But in terms of health, or in the judgement of the whole system, the two are basically different (chap. 6.4).

This weakness of the synergetic model is a typical example for the weakness of all physical models when applied to the principles of the living. Although valuable, they have to be used with some caution.


previous -- home -- content -- next