Georg Ivanovas From Autism to Humanism - systems theory in medicine

4. Systemic Basics

previous -- home -- content -- next

4.1 Signal and information

Information is a widely used term in modern society which has even been called the ‘era of information’. But there is much dispute on the question, what information really is. There are lot of different definitions and many a discussion fails, because there is no clarity what is actually talked about. A main problem is the confusion between signal and information. Many scholars advocate a concept of information more congruent with the notion of signal.

First, I would like to suggest three possible definitions of ‘information’. They shall demonstrate the difficulty with this term. Simultaneously they shall exemplify how ‘information’ is understood and used in this thesis.

  1. Information is the difference that makes a difference.
    Bateson’s definition (Bateson, 1972: 453) has been widely accepted, but he himself was cautious about it. He stated that a letter that has not been written can be more an information than a written letter.
  2. Information is the interpretation of a signal. A signal is a disturbance in the universe in which the hearer who experiences a disturbance of the universe is embedded.
    Von Foerster’s definition (von Foerster/Bröcker: 344, my translation) has the advantage of showing a relation between a signal and its information but it lacks the ‘information’ when something expected does not happen.
  3. Information is the smallest defined recursive unit.
    In self-organization information is embedded in a circular exchange of signals. The meaning is attributed to the semantic context (Jantsch 1979: 88). Out of this context an observer defines a unit, calling it information. However, also this definition is not able to overlook relational patterns, e.g., a signal that becomes an information only after a long period of time. It has also the disadvantage to be quite difficult.

The classic concept of information theory goes back to Shannon and Weaver who worked in the Bell Laboratories, researching telephone communication. What they called information theory should better be called “signal theory” (von Foerster/Bröcker: 343).

The classical concept of the transmission of a signal is

A sender sends a signal that is transformed by a first transformer, then transferred through a channel, retransformed by a second transformer and finally reaching the receiver.

This model fits quite well for telephone communication. The sound waves are transformed into electrical impulses, send through a cable or through the air and then are retransformed into sound waves. Everything that disturbs this process is called “noise”. So if I call my friend and my children put on their music, or the telephone crackles because it has fallen down too often, or my friend’s companion asks without interruption “whom are you talking with?”, all this is noise. And as a result we miss each other, because the time of the appointment was not clear.

More problematic is this concept if I write a letter to my aunt. I write (transformer 1), the post (channel) transports the letter to my aunt and she reads it (transformer 2). Here we have many possibilities of noise. I write awfully, the postman does not pay attention and raindrops fall onto the letter. And, after all, my aunts does not see very well. What does she read?

The signal changes through outer influences:

What of all this is the information? In classical information theory information is what reaches the receiver. This makes sense. If we want a satellite to go into the orbit of Saturn but it takes its way to Pluto, something went wrong. And the one responsible is not excused by explaining to the investigation committee that he had sent the right information, but unfortunately an eruption of the sun altered the signal. The committee will insist that he had not done his job correctly by not foreseeing that kind of noise and that the loss of 15 million Euro is his fault.

This notion of information corresponds to von Foerster’s hermeneutical principle which says: The hearer not the speaker determines the meaning of an utterance. However, this does not only mean that information is receiving a signal. It comprises, as well, its interpretation.

Therefore an informational program has information if and only if there is someone who hears or listens and for whom it is information. Out of five people, seeing the same program one might receive an information and four not. Von Foerster maintained: “The world contains no information” (von Foerster/Poerksen 2002: 95). This is not at all in line with our normal concept of information, where information is something that is sent.

To show the complexity of the issue I often present two photographs, both showing the same vulture (Gypus fulvus) connected with the question:

Which picture contains more information?

There are 4 possibilities:

  1. picture one has more information
  2. picture two has more information
  3. both pictures have the same information
  4. both pictures have no information

Mostly answer two is preferred. But if we analyse the possibilities we find that the correctness of the answer depends on our criteria. If taxonomy is the main concern, answer two has to be chosen. Answer one is adequate if the normal human view is of interest. If we look at it formally (seize, number of pixels), we have to choose answer three. Those who are bored by the game or don’t like to decide find answer four appropriate.

What shall be demonstrated by that is

  • the question of information is an undecidable question
  • information might depend more on the interests of the observer than on the presented data
  • there is no general information, information arises according to criteria

To make a simple medical example. 20 IU of insulin are a well-defined signal. But its information depends on the ability of the organism to react, so it is a different information for a normal person and for a diabetic.

The lack of distinction between the signal and its information has led to difficulties in the assessment of adverse drug reactions. Initially adverse drug reaction was classified in “type A reactions, dose dependent and predictable from the known pharmacology of the drug, and type B reactions, not dose dependent and unpredictable. This classification is simple; it helps drug regulation because prelicensing studies can reveal type A reactions, and it predicts that dose titration will reduce the risk of some reactions. However, it is sometimes difficult or impossible to assign a reaction to one type. For example, dose dependent (type A) nausea and vomiting due to erythromycin could also be classified as type B because it is not pharmacologically predictable. Furthermore, other types of adverse reactions are not comfortably classified by the system. For example, osteoporosis from corticosteroids depends not only on dose but also on duration of treatment. And some reactions, such as asthma from adrenoceptor antagonists, do not occur in all patients. The classification has gradually been extended to other alphabetically labelled types, including type C (dose and time dependent (chronic) reactions), type D (delayed reactions), type E (withdrawal reactions), and type F (failure of therapy). These modifications have mitigated some of the difficulties of the classification system but have introduced others” (Aronson/Ferner 2003). The authors, therefore, add patient susceptibility as a factor to dose relatedness and timing, creating a three dimensional model. This is, one must be clear, nothing than a rough map of an unknown territory. Due to the nonlinear nature of physiology nothing can be said why the receiver changes the way of perceiving the signal in time, thus altering the information.

Type E (withdrawal reactions) are of a different logical type I would like to call it the engine driver's type of information. This type of information has to be considered carefully in order to understand the meaning of information.

An engine driver of a train has to push a certain button every some minutes. If he does not the train will automatically be stopped. Here the information is to receive no signal. It is a change of the a pattern, a ‘difference’ in Bateson’s terms.

This kind of information is crucial in order to understand organisation, as organisation highly depends on the engine driver’s type of information. This is most obvious in social organisations. From the ‘good morning’ that is an information if and only if it is not said, until to the last ‘good bye’ communication is characterized by an exchange of signals not meant as information but to uphold a structure (Bateson 1972: 9-13, Ivanovas 2003). In couples it was estimated that about two thirds of conversation is not to exchange information but a kind of “relation-changing” (Soskin/John 1963). In such a case the description of the signal is insufficient. Only a broader view reveals its sense. Something that is true as well for the description of metabolism.

In a machine as in a train the process is clearly defined. It is a standardized reaction leading, if missing, to a standardized reaction. This is only possible when sender and receiver are structurally coupled, if there is an agreement on the meaning of the signal. In the case of the engine driver an observer will be able to understand how sender and receiver are interconnected, because the reaction is linear and immediate. However, the observer might not even realize that the pressing of a button has an important function for the journey, if he never experienced the stopping of the train. How often in the history of medicine doctors saw something as unnecessary and were surprised when the train suddenly stopped after they had removed parts of the body (tonsils, thymus, spleen) or suppressed inner regulations. The same is probably true for a lot of the so-called ‘junk genes’. Some scientists believe that “one in 200 of our human genes can be inactivated with no detectable effect on our health.” (Wellcome Trust Sanger Institute 2009; Yngvadottir et al 2009).

In certain cases it might even be impossible for an observer to understand the procedure. When the management does not congratulate the employee to his/her birthday, this can be an important information changing the behaviour of the employee. An observer not customized with this tradition will not understand why all of the sudden the employee changes his behaviour. The same might be true in physiology. There might be a lot of reactions triggered by not receiving a certain signal. But we do not understand this and may be we never will, because it happens so rarely. As current research only lately is able to overview and conceptualise circadian rhythms, it would be arrogant to suppose that such rare events (every year or every 20 years) do not happen. Genomic contributions to the adult onset of neurodegenerative diseases suggest that such long time scales exist for physiological processes (Kennedy et al 2003). We have here a non-trivial situation where something that does not exist and can’t be measured may change the behaviour of a system.

For understanding the information in a social context – something that has been called conversation theory (Pask 1975) – it has to be understood that the ‘agreement’ on how to interpret the signal (the legend, chap. 3.8) is never part of the content. How to code and decode a signal (or the missing of a signal) is laid down in the frame. There must be (or have been) a connection between sender and receiver. Otherwise a signal does not become an information. This relation is not static, but dynamic.

To send a signal meant as an information the sender must have already some information about the receiver.

Even in the army, with a highly standardized structure, the sergeant giving the instruction to his soldiers to turn right and to start marching has to know that there are soldiers (otherwise he will be put into a mental asylum) or that there is no abyss on the right (1). Social information only exists as an exchange of information. It is a recursive process.

The same is true in the medical context. If the relation between sender and receiver is shattered we have severe states of disease.


(1) This refers to an example of von Foerster who maintained that in the military command the notion of signal and information are identical (Foerster/Bröcker 2002: 345). In opposite to his opinion I claim that also in this context information has a recursive character which exceeds the linear idea of the signal.

previous -- home -- content -- next