The Chaotic self.

"For the want of a horseshoe nail the horse was lost
for the want of a horse the rider was lost
for the want of a rider the message was lost
for the want of a message the battle was lost
for the want of a battle the war was lost
for the want of a war, the nation was lost
    ...all for the want of a horseshoe nail"
         - Nursery Rhyme from the US civil war.


- - - - - - - - - -  

INTRODUCTION - Theoretical Claim Being Made

The 'self' of a person can be understood as an emergent function from a non-linear dynamical or chaotic, system (Marks-Tarlow, 1999). The two core propositions presented in this section are that this non-linear system has a strange attractor that can be broadly grouped as the sympathetic arm (amygdalic based, more emotional, subjective) and the parasympathetic arm (pre-frontal context based, more conceptual, objective) of the Autonomic Nervous System (ANS), such that homo sapiens consciousness has arisen from this non-linear system. And secondly that this dynamical system has "self-similarity" with the dyad of a pairing bond and also with political alignment when groups of people make decisions.

This is a bold claim, and in this section the basis of scientific enquiry is examined first in light of the core methodological technique used in science, that of a 'linear' statistical approach and how this has shaped scientific enquiry itself. From here an overview of non-linear dynamical systems is presented which has made inroads into all fields of scientific enquiry, and how it can now be used to understand human psychology.

In the psychological exploration of non-linear systems, there is an overview of the philosophical problems of how the self can understand itself (a recursive loop that leads to the illusion of the self) and how this can be a stepping off point for asserting that a dynamical system is in operation. Then the clinical problems of how the subjective and objective organisation of experience interact, how the fields of Self-Psychology and Intersubjectivity attempt to deal with these problems, and the inherent flaw many other theories of psychology which, because of their linear nature, have largely been unable to solve the problems of severe disorders like psychosis or schizophrenia.

From here the neurological underpinnings human relationships are summarised and how this forms the basis of a system that must necessarily extend beyond the self of a single individual into the dyadic relationship. The argument that a human self is co-created from the many relationships they experience, from a social phenomenological narrative that is constructed, and how when a linear approach is applied to the study of relationships, it breaks down so readily into a masculine and feminine critic that it may very well be the same strange attractor of the self in an individual.

Following this is an extrapolation of these same ideas that look to how groups form and make decisions, specifically in light of the many divides that politics throws up like left-wing and right-wing 'beliefs', or liberal / conservative values. And finally this idea of a fractal relationship between self formation and group formation is presented such that it can be falsified by way of an examination of the self-similar nature of the fractal conceptualisation of the sympathetic and parasympathetic arm of the autonomic nervous system, the masculine and feminine aspects of a dyad and the broadly 'right-wing' and broadly 'left wing' components of a group decision making process. This theoretical fractal, if it exists, is as much a story of the development of consciousness in a person as it is an encapsulation of the evolution of consciousness in humans in general.

Remember to breath.

- - - - - - - - - -  

The Generalised Linear Model and those pesky extraneous variables.

The Generalised Linear Model (GLM) is a statistical approach which inherits the mechanistic thinking that science in predicated upon. The GLM when used with experimental methods is a powerful tool, but with correlational methods and to a lesser extent, the anecdotal case study, the GLM is confronted by problems of context. In the experimental method, there is an attempt to explain what one thing causes another (causality), through the use of test and control groups, which are subjected to different factors in the design. But because it is either impossible, too difficult, or unethical to set up a control / test experiment, there is a tendency in psychology to use observational methods of general associations (correlations), which can only be simple associations, and can not infer causality. From a dynamical systems perspective, linear modelling will only work at a scale where the small perturbations are masked by larger more general trends. This is both the strength and the weakness of the GLM, because a statistical measure such as the mean, can only refer to a population of individuals and not the specific individuals themselves. The mean, variance, skewness, kurtosis etc. are useful in sociological terms, but they cannot, by definition, account for all individual differences in psychological terms.

It was Karl Popper who originally challenged the observational-inductive philosophy of science stating that no theory is ever proven and in order to be scientifically valid, theory itself should be something that can be falsified. That is, theory needs to be stated such that it can be shown to be false. In Popper's view, every scientific theory has to have a hangman's noose around it's neck, and where every experiment or observation of that theory is a potential trapdoor releasing exercise. Put this way, who would want to expound a theory in the first place without some a-priori suggestion that it will stand the test of time. When scientific careers rely on publication and exposure in a marketplace of ideas, no-one would want to do this. As noble as the idea of falsifiability is, it would seems far too scary a proposition without some hint that the direction of the experiment won't ultimately pull the trap door that hangs theory. At best the most a scientist could hope to achieve is being permanently in an agnostic liminal space of not knowing.

In science, a lot of emphasis is placed on peer-reviewed published results. What is not noted when talking about deduction and falsifiability is that the researchers themselves have typically spent years examining the field, collecting data, reading up on theory, working in clinical/real world systems, and have built up an internal 'emotional' map of experience around the data they have taken in. This defies Popper, because it is a subjective and affective feeling about the direction in which they are heading. Often it is because of 'random' accidents that great discoveries happen, but it may also be that the researchers have a 'gut-instincts' approach about what to explore in the first place. This affective component in cognition is what Damasio calls a somatic marker (Cozolino, 2006; Damasio, 1994). According to Popper puritans, gut instincts should not exist.

But the dominant game in science is 'publish or perish'. The GLM has become the obsession of contemporary research in psychology, it is the life raft that numerically challenged psychology students cling to in order to gain the credibility of scientific rigour. The vast bulk of research in psychology is a one-way street of mechanistic thinking that inevitably becomes reductio ad absurdum. But in this psychologists are not alone, the whole of science is an example of the methodological tail wagging theoretical, empirical and clinical dog.

The earlier, larger, all encompassing theories (called meta-theories, although I think the use of the prefix is ill-advised) such as the psychoanalytic tradition (Freud), object-relations (Klein), behaviourism (Pavlov, Watson, Skinner) and phenomenology (Husserl) have given way to smaller, more specific areas of theoretical extrapolation like Cognitive Dissonance (Festinger), Attributions Theory (Heider, Kelley) and Attachement Theory (Bowlby, Ainsworth). These later theories have arisen from research with observable data. Much has been achieved in understanding the human condition, however it seems at least superficially that the mode of inquiry itself has resulted in a sort of painting oneself into a theoretical corner. All this research has been conducted using linear methods of enquiry. What's missing, in one simple word, is 'context'.

To be fair, contemporary psychological researchers are largely aware of the many problems they face. It gets taught as part of the core syllabus in psychology. There are financial constraints that limit the research that can be conducted, this reflects most obviously in sampling problems where a disproportionately large amount of research is conducted on first year psychology students. There are ethical problems in using humans in many experimental designs. But there are also very real physical limitations as well, the most striking of which is that demographic variables cannot be assigned experimentally, for example you can't test a person who is male, make them female, and then test them again. But so much of the literature will take into account basic demographic variables like gender, age and socio-economic status. So instead what often happens is that a cross-section of the population is studied at a single point in time. And even when researchers have the funds to gather data at multiple points in time (longitudinal data), these multiple time points are being examined in terms of their within and between sample differences, and will similarly fail to take into account simple context.

Context is hard to examine, because it often the very thing that the Generalised Linear Model is trying to remove. In a linear approach when factor A is examined in relation to factor B, then the researcher wants to isolate these factors as much as possible from the surrounding environmental 'noise'. Removing the environmental noise is the stated aim of 'model reduction' in the Generalised Linear Model. Another word for 'context' then could be 'extraneous variables', and the researcher wants it removed from their final equation. Over time, research that lacks context is built on other research lacking context. Earlier research may have associations which are relatively weak in their strength of association, that is, whilst one factor is associated with another factor, and this relationship is 'statistically' significant, there may only be a very small effect size. But the mathematical assertion of statistical significance will probably be interpreted in a conclusion or later summary as being gospel truth that factor A 'causes' factor B. Yes, psychologists are aware of this problem, but it is a systemic problem and simple awareness does not stop the inevitable reductionism of empirical research that has a mechanistic drumbeat marching on to greater and greater levels of 'less'. And yes, by way of admission, in this section there are some references to material that will suffer from this very same dilemma.

The problem for contemporary empiricism using the linear approach is that the supposed 'randomness' is removed from the conclusions, and all the formulations (regressions) are essentially single slices in time or a difference equation between a few slices in time. But the randomness is also an intrinsic INPUT into the conditions that set up the observation being made. The feedback loops, which appear to be quite common in engineering and biology (Dawkins 1986; Gleick 1986), physics, mathematics, climatology and economics (Gleick, 1986), and in the psychology of relationships (Berscheid, 1983; Fitness, 2006). However, particularly in the later, it would appear that the essentially non-linearity of individuals is not being understood. This may happen because the feedback systems that expose randomness as an underlying 'chaos' or non-linear system, are actively removed by the linear approach through model reduction. Popper's empiricism, and the linear approach it favours, has a fundamental problem with non-linearity (Gleick, 1986). Born out of a Newtonian ideal, the mechanistic universe of cause and effect is nice and clean long as there is no turbulence. When you attempt to study feedback systems, such as fluid dynamics, then randomness is just assumed, the noise is everything, and turbulence is king. When the output of a system is also the input to the same system then a non-linear system is actually operating and the Newtonian world view, and with it the reductionist empirical approach, is fundamentally compromised.

Take for instance a very simple formula, one used by ecologists to establish what will happen to a population over time. Will it reduce toward extinction? Will it outgrow it's medium, that is food, shelter or any other dimension of an n-dimensional hyper-volume (niche). Or will it stabilise at a certain level. Sorry, but there has to be at least one formula here, but the good news is that it's extremely simple, which is a crucial point of non-linear systems as you will see. The formula is:

Xn+1 = RXn(1-Xn)

This formula repeats on itself. X is population size, where the subscript "n" denotes the "current population" and subscript "n+1" is the "next" population size in this repeating formula. What this formula is saying is the next population size is determined by the existing population size and some value called 'R'. That's it. That's all there is, so don't panic! For the most part ecologists work with this formula all the time. It's a logarithmic formula where the output from one iteration is used as the input for the next. With R=2.7 for instance, then a steady state is reached at 0.6292 which exists between 0 (extinction) and 1 (theoretical saturation) (Gleick, 1988).

Gleick (1988) reports that Robert May however was a biologist with an extensive background in mathematics. He took this formula and asked what will happen for different values of R. So May repeated it many times for different values of R. For very small values of R (low power) extinction is the end result. As R is slowly increased, it will result in a steady state. The end value of this steady state will climb as you increase the R value. This can be graphed on what is called a "logistic map" which is a form of bifurcation diagram.

The graph depicted, has the x-axis showing the R value and the y-axis showing the end point population size. As R is ramped up even more, something bizarre happens to this steady state where 2 possible end populations sizes are possible and very small increments in R will oscillated between these two curves. This is indicated with a dashed line and is called the 'period 2' set of outcomes with the endpoints appearing to cycle like lines on an oscilloscope, i.e.: it has a regular looking pattern to it. Increase the value of R even more and the end population starts to appear to swing wildly, leading James Yorke to write a paper called "Period Three Implies Chaos", which is where the use of the much popularised term 'chaos' has it's origins.

Reading Gleick's book "Chaos" it's is clear that a pattern exists in the researchers like Lorenz, May, Yorke, Mandelbrot, Feigenbaum etc., who have taken to the field of non-linear systems. Wether from mathematics, theoretical physics, experimental physics, climatology, ecology, fluid dynamics, engineering, or economics, the scientist typically starts out by not being happy with the current state of research (note this affective/emotional basis to their reasoning). They, often independently, start modelling very basic non-linear problems more as a exercise, not really sure what they are doing but completely absorbed by the data they are finding, either as weird printer output or diagrams that plot the data. They often keep it hidden from their colleagues, it's just too bizarre to accept yet, and when they do start producing scientific papers, these papers are ignored at first, but inevitably become cornerstones of a new way of thinking.

Lorenz made a point of masking the pure mathematics he was doing with a few superficial references to the climate in his preface (Gleick, 1988). But then a critical mass will grow in the field, first in the researcher, and then slowly in their colleagues. Eventually a conversion experience happens and colleagues start to see the fundamentally new way of viewing the world, where the very stuff that is removed and dismissed as noise or turbulence or 'randomness' becomes the central feature to their new world view. In response to Lorenz's work, Gleick (1988) reports how Willem Malkus a mathematics professor at M.I.T. initially says "Ed, we know - we know very well - that fluid convection doesn't do that at all" indicating that regular steady motion would hide the complexity. But a decade later Malkus says "Of course, we completely missed the point." (Gleick, 1988:31)

Out of the chaos there is is apparent order, and with order there is an underlying ferment of chaos. It is so tempting to invoke Kant here, but I'll resist. You'll either have the experience yourself, or you won't, but if you want an easily digestible overview of the history of non-linear system, then Gleick's "Chaos" (1988) is required reading.

- - - - - - - - - -  

Non-Linear Dynamical Systems - A Brief History of Understanding The Noise

Where linear models remove randomness, non-linear systems invoke this turbulence as critical. A crucial point about non-linear systems is that they work on feedback, and feedback is not necessarily linear.

Feedback loops are a way that science understands a lot of things in the natural world and the physical universe. Dawkins (1986:195-220), for example, shows how feedback loops can be used to understand speciation, populations, and predator / prey relationship in biology. With the industrial revolution we have harnessed the potential of feedback systems and put them to good use. From the steam engine, to amplifiers and the integrated circuit, in engineering, there are feedback systems everywhere.

A feedback loop occurs when the "output" of the system will have an effect on the "input" to the same system. There are two fundamental types of feedback loops, which in engineering terms are called "negative" and "positive". However because of the obvious additional meaning that these words carry, they are referred to here in the following way. The negative feedback loop is one which is essentially "balancing" in nature where the system attempts to balance and correct itself, here the output from a system will cause the system to move in an opposite direction so that it will return to a set level. The balancing feedback loop attempts to maintain a fixed or homeostatic state. In gross terms, thermostat of a refrigerator is an example of this. When the temperature rises beyond a set point, thermostat will turn on the compressor, and when it drops below a set point it will turn off. The end result is a relatively balanced temperature.

The other type of feedback loop (positive), is where the output of the system will cause the input to the system to increase in one direction. This type of system has an "explosive" quality to it and is best understood through the analogy of the amplifier. When you put a microphone in front of a speaker you can hear the effect of this explosive feedback loop as the relatively low background noise produced by the speaker is amplified after entering the microphone, then played back through to the speaker and then back into the microphone and so on. The quality of this explosive feedback loop is that it is often quick and can lead to rapid changes in state.

Another example of the explosive feedback loop is the mirror that faces another mirror, to the observer between these two mirrors, it appears to reflect into infinity, but it can also appear to have a fixed point that it recourses into. This fixed point is an illusion created by our brain where the infinitively regressed point is generalised and any finer detail is ignored or lost. Magnification of the end point will only show more detail up to the magnification power provided, then the infinite point is again assumed by our brain. Of course, should either mirror or observer move, then the generalised end point will shift.

Linear systems can interpret simple feedback loops in a very general way, but as much as the engineer would like it to be so, nothing is ever that simple. With the operation of the refrigerator above, where a graph is created of to map the on/off state of thermostat over a period of time, it appears that the temperature of the air in the refrigerator cavity is the most important factor. To an engineer this air temperature is the only important "factor", as this is what determines the operation of the system. But extending beyond this, and acknowledging that the refrigerator is to be used in the real world, then other factors are present. What is the barometric air pressure? What is the humidity in the house? How many people use the refrigerator? What happens if someone opens the door and ruminates on what to eat as they stare into it? Is the refrigerator near a water heater? What about the latitude or altitude of the house in which the refrigerator exists? Is there any air conditioning in the house? Does the owner have an eating disorder? What kind?

These additional factors are what engineers call "externalities" and they cause a problem to linear systems. Whilst they can have an effect on the operation of a system, they are very hard to pin down and measure. An engineer would argue that the externalities are not important, because it's still the temperature of the air that determines the operation of the compressor on the refrigerator. The problem is that refrigerators are meant to be used by humans, and so externalities matter even though the may seem inconsequential. Solar flares, for instance, might only have a vanishingly small effect on the operation of a refrigerator on Earth, but if you consider them in your equation then it means you need to change the context of how you understand the linear system. And ultimately it is this changing context that confounds simple linear systems. Reducing the factors to a manageable subset of statistically significant factors, a process called 'model reduction', also fails to capture the overall picture of what is happening to the refrigerator. From a cross-sectional point of view, that is when looking at a single point in time, this reduced model may be statistically valid, but it has also not explained reality very well.

Up until the mid 20th century, these smaller, less important factors were simply pushed aside and ignored as not contributing to the main effect being observed. The problem is that these small perturbations to so many systems just keep getting in the way. In the very way that science students have been schooled, they have been taught to ignore the minor factors as inconsequential. Now most fields of research is being sequestered into smaller and smaller areas of study. Stated this way "Implicitly the job of many twentieth-century scientists - biologists, neurologists, economists - has been to break their universes down into the simplest atoms that will obey scientific rules. In all these sciences a kind of Newtonian determinism has been brought to bear" (Gleick, 1999).

Furthermore, all of the external factors could be party to other feedback systems. Staying with the refrigerator for a moment, a persons hunger level throughout the day is a separate feedback system which could be understood in terms of carbohydrate load and body mass. With multiple factors and their interacting feedback loops we end up with a seemingly "chaotic" operation in the refrigerator. This is the key to understanding chaotic systems, whenever there are multiple and overlapping balancing and explosive feedback loops in operation you will end up with non-linear, dynamical or "chaotic" behaviour. The generalised linear model tells us that the fridge is averaging 5 degrees Celsius with maybe 1 degree of deviation, but the non-linear modelling shows us that a whole lot more may be occurring, and there may be an "order masquerading as randomness" (Gleick, 1998)

Of course the analogy of a refrigerator can only be carried so far. So long as the mains power is connected and the door is kept shut, the food in the fridge will stay fresh longer, and the overarching engineering behind it will make this happen. But in humans, and the myriad of complex social interactions and environmental pressures, the behaviour they express can not be distilled down to a few linear factors measured at a single point in time. This is why a non-linear investigation is required.

Another way to understand non-linear systems is to think of the motion of a pendulum as it swings around a central point. Assume that no friction is in operation on the system, when there is one central point this is a linear system and the movement of the pendulum is highly predictable. This central point is termed an "attractor" and this is how the refrigerator would seem (to an engineer at least), i.e. with only one significant attractor, that of the temperature in the refrigerator cavity. However friction on the pendulum axle and from the atmosphere is going to influence the motion of the pendulum arc. To a theoretical physicist, this friction can be mostly ignored as negligible and an 'ideal' state can be worked from instead. Problem is that for an experimental physicist, working on fluid dynamics, turbulence is everything.

There is an image produced here which shows what appears to be a bimodal pendulum in operation. Actually it is a set of points in three dimensional space which is governed by three simple formulas that cannot be 'solved' in the traditional sense because they are seen as non-linear. It was presented by Edward Lorenz in a 1963 paper called "Deterministic Nonperiodic Flow", and has been cited an exeedingly large number of times since (Gleick, 1988).

Working principally on climate modelling, Lorenz became fascinated with mathematical equations that could not be 'solved' in the traditional sense because they were essentially non-linear in nature. As often happens in science, he stumbled across the problem by accident. Gleick (1988) describes this in his book, where Lorenz had built a very simple modelling program, using one of them new fangled vacuum tube calculating machines, which simply iterated through calculations that reflected real world climate conditions. It was a set of twelve numerical rules, calculated over and over again, where the output from the last set of calculations became the input for the next set (feedback loop). Each row of numbers would represent one day in the model, and in a way he generated a simple graphical representation of the output as it printed out. One day he wanted to take a shortcut and redo a portion of the printout, so he started from a place somewhere in the middle of the output and entered the starting conditions. Because the printout only had three decimal places, but the computer stored six, he entered in what was a very slight difference in initial conditions ...and got a completely different looking output. Unlike a linear (deterministic) system, Lorenz was able to show how very small perturbations in the initial conditions can lead to what appears to be a radically different output.

In 1972, after a decade of playing with his toy weather machines, Lorenz gave a talk entitled "Predictability: Does the flap of a butterfly's wings in Brazil set of a tornado in Texas?", which gave rise to the term "butterfly effect" (Brooks, 2005). Interestingly this image looks a bit like a butterfly (Gleick, 1988), and the name given to the talk was well phrased. It is catchy and has popular appeal. If he had used a less poetic title, then it may have had less effect. Instead, the appeal of the phrase has served theory well. Ironically this is a good example of the "butterfly effect" in operation, and the phrase itself has entered into pop-culture. Notice that there is tendency to understand the "butterfly effect" as having an explosive quality (i.e. positive feedback loop) however again, this is the trap of mechanistic thinking, where an inference is made between cause and effect. Always keep in mind that most butterflies don't cause tornadoes and that there are probably overlapping and interacting balancing and explosive feedback loops in operation.

The talk given by Lorenz was superficially about climatology and how weather systems are fundamentally balancing in nature. Really what needs to be understood is the idea of feedback systems within feedback systems. In one context there is a harmless, soft and gentle butterfly flapping its wings. In another context there is a devastating tornado. In yet another global context, the weather system is just balancing out the temperature difference between the poles and the equator. This is why "fractal" theory is so relevant to non-linear dynamical systems theory, because the same fractal can look similar, yet also different, when viewed from alternate perspectives. The reference point of the observer is being considered, as well as the context in which something, the 'it' is being observed.

In the bifurcation diagram of the logistics map presented earlier there is a very simple expression of the fractal effect. If an area within the chaos is sufficiently magnified (i.e. the change in bifurcation parameter is sufficiently small) then within this seeming turbulent data there appears islands of order. There is a point where the initial period one / period two split appears to be replicated at a later point. This repetition is apparent to a human eye, but not necessarily a computer because human perception may be a chaotic system, and the best way to interpret one chaotic system is with another ({ref} on communications theory).

To get an idea of how a fractal can look the same and yet different, depending upon your perspective, take a look at the classic fractal called the Mandelbrot Set []. There are some tools on the web that allow you to drill down onto the 'edge' of the fractal, where upon increasing the magnification a new fractal appears from this edge that looks similar to initial position. In fractals, this is called "self-similarity" and as the magnification increases images appear that seem surprisingly similar. This idea of self similarity is key to theoretical proposition laid out in this work, where the formation of an individual "self" is "self-similar" to the formation of a dyadic structure and the formation of a group political system, that is, all three are parts of the same fractal, and that this can be measured and is falsifiable if the strange attractors of the dynamical system are presented.

With empirical observations in psychology, the difference in initial starting conditions (context), or the fact that any study or the natural world is most often a slice in time, or even a few slices that show a difference equation, it is suggested that this mechanistic simplification of the world is no longer viable. Even very small differences in starting conditions can be reflected in a non-linear system. More importantly for psychology, the position of the observer is absolutely crucial in understanding that which is being observed. The psychologist and the sociologist like to put on white coats and pretend that they are independent of the observation(s) being made. They will construct rules around the ethics of not being involved, and establish statistical methods for coping with observer bias, but how ever hard they try, it can never be ignored. In fairness the best method presented by empirical work is the 'double-blind' study where the subject is entered into either a control condition or a test condition and the process of this assignment does not suggest to the subject what condition they are in. The subject is blind, and the immediate interface (person or process) with the research is also blind. In drug trials this is the justifiable gold standard of methodology, however it is mostly not possible in psychological research because humans are highly evolved at reading implicit information, processing it unconsciously and intuiting the conditions (Cozolino, 1996), and even these very small perturbations to the initial conditions can have a chaotic (non-linear) effect.

Non-linear systems theory however proposes that a 'whole' system is made up of many very small perturbations. These variables are usually inaccessible and result in the system as a whole being unpredictable, or to be more precise "seemingly unpredictable". Lorenz was able to show how small changes in the initial conditions can produce large changes in the long-term outcome. He also argued that because these variables are not immediately recognisable, the entire system will appear as unpredictable. However this does not mean that it is "random". Rather it is the product of many small perturbations to the system (Brooks, 2005; Gleick, 1988).

However the advantage of using a non-linear approach is that when a chaotic system is in operation there is often an occurrence where a sort of periodicity or regularity appears out of the chaos. In the language of non-linear systems there is said to be 'self-emergent' properties of the chaotic system. It needs to be explicitly stated that the 'self-emergent' properties of a chaotic system may not necessarily be the same thing as the 'self' as understood in psychology, however more that the commonality of language suggests, it is being argued that they are related to some extent. In the next section the 'self', as defined in a psychological context is examined more closely.

- - - - - - - - - -  

The Self

The 'self' of a person is typically understood as a 'thing', as some sore of 'core' personality structure that exists somewhere 'within'. Another way to view the 'self' is that it is a 'process', that is, a process of understanding everything out there that is experienced as well as everything that occurs within.

Consider, as Freudians do, that the self (called the ego) is at the boundary between unconscious and conscious experience. Like a bouncer at the door to the nightclub, where everything that is outside is unconscious and everything that is inside is conscious. It's the bouncer's job to decide who gets in and who doesn't, or what thoughts are conscious and what shall remain unconscious. The question then is wether the bouncer is conscious or not. That is: is the self inside or outside the nightclub? If the bouncer however is both inside AND outside, then the ego, or the self is essentially an abstract macro scale example of the quantum superposition problem that Schrodiger wrestled with. From another angle, when a person is talking about their 'self', then it is the 'self' itself that is doing the talking. In the language of subjective and objective, the subjective talker is also the content of the object being discussed. It really is all a bit confusing, because these are examples of recursion, and it is encountered frequently with talking about the self. It's also where most people give up and leave it to the philosophers.

But if even the discussion of the 'self' is prone to recursion, then what is recursion. As an aside, if you look up recursion in the dictionary, if the author is being half-way humorous the definition should read 'see recursion'. More seriously though, recursion is a 'process' or 'function' that has to refer to itself. It has the property of 'self-similarity', as with fractals, and could be viewed as and endless loop in computing if there is no logical escape sequence. In essence, recursion is a self-referencing feedback loop. Just like the simple population formula (logistic map) previously discussed.

The self is a recursive feedback loop. In fact it may be the only one.

When Descartes wrote that thought itself must exist, he was having a bit of trouble dealing with this recursive problem of knowledge, such as how does a person know anything exists. He suggested that because one thinks about the existence of thought, then thought itself must exist. Sure, it's a step on the path to enlightenment, it's also just a bit unsettling (note: a feeling occurs). Instead he famously said 'cogito ergo sum' - I think therefore I am. And that was it. Game over. Thinking is the only pure form of experience. Perception, if it could never be trusted, would be discarded in place of thought which was separated from the body, and which is partly where dualism was born.

Returning to the feedback system that occurs when a person stands between two roughly parallel mirror surfaces, they will see their image reflected off into infinity. They can mark out on the mirror the rough (average) end point of the reflection as they get smaller and smaller. However this suggested end point is an illusion defined by the level of magnification in the system, and it's initial starting position. Move slightly and the end point will shift. This means the end point is relative to the position of the observer. Just like the mirrors (a metaphor) the self is dependent on the position of the observer, however the observer is also the thing that is being observed.

There is nothing really complex here, but it can mess somewhat with your head. It is simply an extrapolation of some very basic ideas, but a person doesn't spend their whole day wondering about their subjective position, they just assume this starting position / end point / reference point, make the decision that needs to be made and then move on. An observer standing between two mirrors may be somewhat unsettled, but they can easily infer an end point and thus cope with staring into infinity. A child may have fun with this in an amusement park, but to a person from a lost Amazon tribe who hasn't even seen a mirror, it might cause severe trauma. The point is that the emotional organisation experience is present, even when talking about something as simple as standing between two mirrors, and, in general, we are ok with assuming an endpoint into infinity.

In Festinger's ({ref} 1956) theory of Cognitive Dissonance the accumulation of new data is driven by a need to reduce discomforting emotions caused by holding conflicting cognitions. Often however the new data is constructed or viewed in a different way in order to relieve the pressure of the cognitions that don't make sense any more. One example often used is from Aesop where the fox want to reach some grapes that are too high in the tree, when he fails to obtain them then the fox 'reasons' that the grapes would have been sour anyway and he did not really want them, from which we get the phrase 'sour grapes'. However in the social psychology literature of Festinger, a 'cognition' can be both conceptual and/or emotional, again supporting the suggestion illusion of a single self orientation. Of course Festinger was not the first or the last to come at this same issue. Freud would have argued that psychological defence such as projection and repression are used to help an individual cope with reality. Similarly Carl Rogers instructs a therapist to be on the look out for incongruence within the client, where what is being said does not have internal consistency with other information or behaviour. And Kohut's self-psychology and the intersubjectivists are aware that the ultimate goal in therapy is to help reintegrate previously disavowed affective states by helping the client to making sense of their phenomenological world.

When the world makes sense, decisions are easier because the affective component of the decision is in line with the conceptual information about the decision. Damasio (1994) calls this alignment process a 'somatic marker', where, in simple terms, a decision will 'feel' right. In a challenge Descartes assumption of a separate mind and body, Damasio pointed out that a person can only make good clear and simple choices because their body will feel right about the decision. Neurologically the 'it feels right' is occurring at an unconscious (or implicit) level, in the amygdala, and it is only with the associated input from the Orbito-Medial Pre-Frontal Cortex (OMPFC) and somatosensory cortex that the whole system works. Damage to the OMPFC or somatosensory cortex will lead to poor decision making because the somatic markers were either lost, in the case of brain damage, or were never there, in the case of autism like disorders (Cozolino, 2006:73). Without the PFC, the affective organisation of experience has no counterbalancing (antagonising) system that is based on more abstract reasoning and context.

The somatic layering of experience is stored as a subjective emotional experience in the brain, and it is most likely the amygdala that works to encode and decode this information (Cozolino, 2006; Shore, 2002). Over time, and through experience the amygdala builds up a set of schemas of understanding. Think of schemas as a sort of map for navigating through the phenomenological world that is both the physical environment and the social environment with other humans. But the amygdala is not working alone, these same experiences are also being encoded by the later evolved cortex, specifically the OMPFC. This is the duel-route model of encoding information (Cozolino, 2006) and it may be the reason that Descartes came to the conclusions he did.

The limbic brain (amygdala) evolved fairly early on, but it is the core feature of the mammalian brain. Unlike the older brain stem which is shared with a reptilian ancestry, the mammalian brain evolved within a warm-blooded body that was capable of so much more by way of interaction with the environment. This early mammalian brain could see threat, recognise it and evoke massive resources of energy by way of the Hypothalamic Pituitary Axis (HPA) that sits atop the limbic brain. The sudden signal change to danger could mobilise blood to be diverted from the internal organs to the external muscle and perception system. It would increase metabolic activity through the use of adrenalin as a signalling hormone and cortisol which, like a central bank in a time of war, would be an open check book saying "what ever it costs".

The problem with the limbic brain and the sympathetic arousal of the fight and flight response is that it creates very simplistic models of the world and extrapolates them as much as it can. The schemas are affective in nature and somatically based, with the memory involved being 'implicit' in form. These schemas quick to lay down and extremely difficult to forget (low neuroplasticity), are laterally right-hemisphere dominant, which has it's first growth spurt before the left hemisphere starts to develop (Cozolino, 2006).

But the later mammalian system developed in the cortex proper. It is far more malleable, that is, it has a higher level of neuroplasticity, but it is able to abstract schema's of understanding and reapply them to later occurring contexts that may be slightly altered, in addition, pattern recognition becomes possible, and through the use of 'explicit' memory systems in the hippocampus the potential to use reason and create abstract logical thought (Cozolino, 2006). However it is also extremely expensive from an energy perspective, because the cortex in humans is 'huge'. Far bigger than other mammals relative to expected body mass (Encephalization Quotient), with the closest equivalent being that of bottlenose dolphins having an Encephalization Quotient of around 4 compared to humans at 7.5 (Curtis, 1986). The human brain is a very expensive gamble that evolution has taken on.

The refinements offered by the parasympathetic system also meant that humans did not have to live in perpetual state of being in survival mode (sympathetic fight / flight). The OMPFC took on the role of moderating the other arm of the Autonomic Nervous System, and with the development of language in the left-hemisphere, it would be able to trigger the vagal response {ref} and calm down. Reinvesting resources into the internal processes like digestion, rebuilding the immunological functions that had been mobilised by cortisol (Curtis, 1986). This parasympathetic response meant, perhaps most of all, being able to make sense of the experience that just happened. Through this making sense of the experience a better road map could be built that would make better sense of the world out there.

With a better map available, a more successful future experience was possible. From this process, and starting with simple, clear logical (unemotional) observation of the world out there, humans could develop more sophisticated ways of dealing with threat. The long term payoff was massive. Threat would always be there, as proof by the highly conserved nature of the limbic structures, however if response time was not as critical then the better navigation and executive planning could be invoked in order to be better adapted to the world.

When Descartes was stating cogito ergo sum, he was really heavily invested in this logical/conceptual way of viewing the world. But the body also thinks, a point the Freud was trying to make about the unconscious {ref}. Which is where Damasio picks up the story and isolates case study data of brain damage to supply proof for this proposition, that it is the conceptual / explicit / logical / objective / parasympathetic organisation of experience working in a fundamentally interconnected way with the affective / implicit / emotional / subjective / sympathetic organisation of experience.

The take home point is that there are two modes for encoding experience as described by the duel route hypothesis, and that it is the interaction of these two systems that create the 'self' in a way suggested by Marks-Tarlow (1999), a process that is suspected to occur within the cingulate and insulate cortices, which sit on the junction between the limbic system and the cortex (Cozolino, 2006).

The affective / implicit / emotional / subjective / sympathetic system is essentially one devoted to motivation and rapid action. Action without really thinking, but rather just 'knowing'. It has the hallmarks of a feedback loop that has an explosive quality to it. Whilst the conceptual / explicit / logical / objective / parasympathetic organisation of experience is about balancing back down to a resting state in order to make sense of experience, diametrically prone to thinking without acting. So here there are two overlapping and interacting positive (explosive) and negative (balancing) feedback loops, fundamentally interconnected and where the emergent 'self-function' or 'self-organised' identity is thus a chaotic, non-linear dynamical system.

Marks-Tarlow (1999) in an essay to a little known journal which focused on non-linear systems in psychology, spoke about the self in these terms, so in a way it's not a new idea. The closely parallel language with dynamical systems theory is suggestive of the proposition being put forward that humans are self-organising analysers (Dawkins, 1986), or the emergent function of explicit and implicit memory (Greenwald & Banaji, 1989 as cited in Cozolino, 2006:338).

From the same journal where Marks-Tarlow publish his essay, "Non-linear Dynamics, Psychology, and Life Sciences", which admittedly was only short lived, there are a few other research efforts that attempt to encapsulate psychological concepts with non-linear systems. They are presented here because the support the notion that chaotic or non-linear interactive process are being observed in the emotional, perceptual, behavioural, neurological and physiological process "within" an individual.

If non-linear processes appear as oscillations or a flowing between alternate states on any given continuum, then the bipolar condition would be relatively low hanging fruit for this approach. Thomasson et al. (2002) were able to perform a non-linear analysis on a bipolar patient using EEG measurements. As part of theoretical premise for their work, they point to the many overlapping feedback systems involved in depression including the "circadian rhythm.. thyroidic secretion.. , and sleep and temperature rhythms." (Thomasson et al., 2002:260)

In a similar vein, only working to a shorter time scale, Schuldberg & Gottlieb (2002) looked at how Affect has non-linear attributes "Thus, when an individual reports on current “mood,” this may represent the output of multiple, coupled, modular, and semi-autonomous affect subsystems." (Schuldberg & Gottlieb, 2002:341) In another study Guastello et al. (2002) found that neural activity associated with memory and vision as measured with a FMRI could be explained in a non-linear approach. More broadly, Kahn et al. (2002:311) argue that dreaming has "chaos-like stochastic properties that are highly sensitive to internal influences."

Whilst devices like EEG's and fMRI's can be useful for objectifying these non-linear psychological constructs, they do not provide an underlying theoretical structure. As packets of mechanistic information, the projected implications for what happens in therapy room are only suggestive at best. Even non-linear theory is falling into the trap of ignoring the objective / subjective nature of experimental science.

What is lacking is a more fundamental connection to the broader theoretical work, particularly in the psychoanalytic tradition, object relations theory, cognitive behaviour theory and evolutionary psychology. The empirical researchers cited immediately above can "see" that there is a non-linear nature to human psychology, but they have come to the research still holding the vestiges of reductionist / linear world view and there is very little theoretical basis in the work they have done. Or maybe, what they were attempting to do was so beyond the scope of a researcher schooled in linear techniques that they did not have the backing to pursue it more fully, or their respective institutions and peers, like those of Lorenz, May, Mandelbrot etc, were completely unable to support them in this new way of thinking. Some will even admit this, such as Ward and West (1998:280) who detail a study which looks at interpreting human behaviour as non-linear process, but can only conclude that they do not have these "psychologically meaningful and empirically manipulable... parameters" which underlie the "mechanisms" of human behaviour (Ward & West, 1998:280).

However, Marks-Tarlow (1999) describes the self as being "dynamically fluid and ever changing, mediated by complex, recursive, feedback loops existing simultaneously at physical, social, cultural and historical levels" (Marks-Tarlow, 1999:311). In self psychology and intersubjectivity we see:

"If the self, the organisation of experience, is viewed as being in transition, it has the characteristics of an open system. A system as von Bertalanffy (1968) defines it, is an entity identified by its function, rather than its physical attributes; it is a stable, information-processing collective made up of a hierarchy of interacting feedback cycles."
(Lee & Martin., 1991:188)

What is new is the placing of emphasis on the self-emergent properties of the 'self' as being the result of overlapping and interacting non-linear dynamical feedback systems, which are based conceptually on the sympathetic and parasympathetic arms of the ANS and their associated neurological structures.

- - - - - - - - - -  

What we can learn from Attachment and Trauma

These two pathways of encoding experience are important because they are directly connected to the study of trauma (Levine, 1997; Soloman & Seigal, eds., 2003; Schore, 2001; Rothschild, 2000; Cozolino, 2002; Cozolino, 2006; Shane et al., 1997). Francis (2004) points out that it is the activation of both systems simultaneously that may indicate that a traumatic event is being "re-experienced" somatically by a client. Adding to this Shane et al., (1997) maintain that "... most relevant to our work, we conceptualise all disorders of the self as based in trauma, and we contend that these disorders are connected in a bi-directional fashion to insecure (traumatic) attachment pattern." (Shane et al., 1997:97)

But because we develop as part of a social structure, the role played by others, especially caregivers, when it goes wrong this relationship can severly upset the process of self formation. As Cozolino (2006) states: "The depth of the harm caused by neglect, abuse, and inadequate nurturance rests on the fact that the human brain is a social organ... when stress is early and prolonged, neurons, neurological structures such as the hippocampus, processes of immunological functioning, and the construction of the self can all be compromised" (Cozolino, 2006:240)

Trauma is identified as resulting from an extreme event of hyper-arousal. A person who is "traumatised" expresses symptoms that seem to be connected to the traumatic event. This connection can be seen, for instance, through recurrent episodes of hyper-arousal, flashback and night terrors (Rothschild, 2000). To summarise this process, the adrenal fight or flight, in its initial phase, will be brought about without a conscious understanding (executive reasoning) of what is occurring. When threatened by an external event, the affect of fear or terror will begin this process of quickly becoming ready for danger. Blood will leave the internal organs associated with digestion and move to the periphery of the body, arms and legs for movement, lungs for breath, heart for blood circulation and head for perception. The bowels and bladder may empty in order to reduce weight, perception will be sharpened and the sense of time even slowed as more information is being processes in much the same way as a slow moving picture actually contains more frames per second (Curtis, 1986:766; Rothschild, 2000:8-10).

Here there is a contention with Rothschild on a key point. The two branches of the Autonomic Nervous System (ANS), whilst they often have a balancing affect on each other are actually separate systems. This is where the linear interpretation diverts to the non-linear interpretation. The sympathetic and parasympathetic arms of the ANS are overlapping and interacting positive and negative feedback loops. But they are also fundamentally independent. It is this independent interaction that has a clue to the non-linear nature of the self.

This difference between linear and non-linear can be seen with the interpretations of what happens when the ANS is compromised and a "freezing response" occurs. Referring to when a cat catches a mouse, Rothschild, (2000:49) acknowledges that something goes "wrong" during the mammalian hyper-arousal of the Sympathetic Nervous System. She states, "In this extreme circumstance the SNS will remain activated, while the Parasympathetic Nervous System simultaneously becomes highly activated, masking the sympathetic activity, causing the mouse to 'go dead' ". This same argument is also used by Levine (1997:95-97), however whilst stating that something has gone wrong, only a very poor explanation of what has happened is supplied.

A non-linear interpretation can easily understand this freezing response. Using the idea of the Triune Brain (MacLean as cited in Cozolino, 2006:29) where the brain is said to be made up of components of a reptilian brainstem and cerebrum, an early mammalian system in the limbic brain (that governs the ANS) and a late mammalian system in the cortex. The freezing response is a conserved reptilian process, that during 'normal' conditions is masked by the later evolved mammalian system. However, the later evolved mammalian system is also less robust because it is attempting to do a lot more by way of the organisation of experience, with the later evolved cortex attempting to do even more with a higher degree of neuroplasticity. Extreme terror can fundamentally compromise the ANS (like a microphone put in front of a speaker, you could blow the transistor in the amplifier). When this happens, the reptilian system is still there and it will cause a freezing response.

Reptiles vision is based on movement, so in a evolutionary sense, freezing means you become invisible. It's also likely that this reptilian system has been conserved in mammals, because the mammalian system of coping with threat (ANS) can be subjected to hyper-arousal and be compromised. It may also be conserved because by being absolutely still, the prey is able to hear the other noise, see other movement, and have more of a chance to get ready for fight/flight (Cozolino, 2006). In a computer analogy, the later evolved mammalian system could be the operating system of a computer, when it crashes, the bios clock on the motherboard (or reptilian system) will still be working.

Mammals evolved a far more useful way to avoid being food ...when it works. Through the sympathetic branch of ANS it is possible to move into extreme states of arousal, more blood to peripheral muscle systems, hormones that stimulate the release of stored energy, better perception from increased blood flow etc, such that greater resources in the body are available for the fleeing or fighting that needs to occur. With the fight/flight response action is quick and is based on the best map of experience available at the time. There is no time to sit down and ruminate on the possible courses of action. But the amygdala which governs this process is quick to learn and never really forgets such that the 'self emergent' part of this process is based on fears that tend to stay with us throughout life and are not easy to overcome, where ... "the role of the amygdala is to remember a threat, generalize it to other possible threats, and carry it into the future." (Cozolino, 2006:319).

However mammals also need to rest back down from this explosive feedback system, and in humans, work out what happened and how to avoid it in the future. In a normal situation, after the threat has passed, the body will rebalance itself through the parasympathetic nervous response. Here the body will calm down, reinvesting blood into the other parts of the body (such as the alimentary canal) which were abandoned during the fight or flight response, repairing damaged systems and boosting immunological functions (Cozolino, 2002:70) {ref on vagal response} . This parasympathetic response is predominantly the work of the cortex (orbital-medial pre-frontal cortex - OMPFC), and a more neuroplastic structure that in humans can invoke abstract reasoning to help both understand what has happened and build better schemas that can be applied to future scenarios, i.e.: conscious learning and executive decision making.

But this is when all goes well. When it doesn't then a breakdown in rebalancing occurs. The traumatised person becomes unable to rest back down properly (Rothschild, 2000). The process of making sense of the experience is hindered. In a severe expression of this breakdown a person can be easily provoked into a new bout of stress, anxiety or traumatic flashback. What intensifies this problem is that is it the "self" which needs to organise an experience where the ANS broke down.

In non-linear terms, trauma occurs when the very process of organising experience, the self, needs to makes sense of an event where it was severely compromised.

But the 'self' can't do that because it was the thing that was severely compromised. Imagine, for instance, that you did not know how to use a newly purchased video recorder, yet the manufacturer gave you a instruction manual, on video, to teach you. You don't know how to play the video, so how can you learn how to use it. In a similar way, when the ANS is compromised it is not possible to make sense of this compromised state because the process by which understanding happens (the self) was the very thing that was severely compromised. Again, when looking a trauma, there is a recursive nature of the self which is indicative of a non-linear system.

Levine (1997:112) documents the case of Marius, who as a boy was attacked by a polar bear. For several hours he had to cling to a telegraph pole, utterly helpless as he waited for the bear to leave. When Marius arrived home he was expecting warmth and comfort following his ordeal. But his father was 'out of his mind' with worry, fragmented by concern about his son being so late. In order to re-cohere his fragmented self, the father did what most parents do when they are fragmented with worry, he got angry. When a parent looses a child in a shopping centre this same process happens as they chide "How many times have I told you not to run off on your own!".

Marius was expecting comfort and got an incongruent emotional response from the phenomenological world. When terror is mixed with a sense of helplessness and a break down in relationship, you usually, but not always, get trauma. Here the balancing effect provided by caregivers was absent and the process of making sense of the terrifying ordeal was replaced with one of shame and retribution, and as a consequence of not feeling safe. This shame became an integral part of the trauma that re-occurred through Marius' early life until receiving successful therapy (Levine, 1997:113). Cozolino (2002:194) suggests that shame can be seen as a "rapid transition from a positive to negative affective state and from sympathetic to parasympathetic dominance" all of which correlates with the Levine's case study of Marius.

How a person is, and they organise their experience, is fundamentally connected to the world they inhabit. This phenomenological point of view suggests that what is occurring to someone and how they make sense of the experience are intrinsically connected. In PTSD this breakdown may be reflected in a relationship with another person. The internal system of the Autonomic Nervous System is fundamentally connected to the external environment, which is often expressed as a relationship to another person (Cozolino, 2006; Rothschild, 2000; Schore, 2003; Stolorow et al., 1995). Suggested here is that a larger feedback system between two people is in operation, possibly one that is non-linear in nature and also one that is has 'self-similarity' to the feedback system within an individual.

This is a "subjectivity" focused view of what is occurring, and the context of the events are extremely important because they enable us to see beyond the diminutive and restrictive approach of a reductionist scientific theory. This incongruence between the internal feeling state and the external environment has profound implications for all of psychoanalytic thought, because it cuts to the core of what is considered to be the "self", because as the next section attempts to explain, the self can only be formed (properly) when there is a suitable relationship.

The non-linear system (set of feedback loops) of the sympathetic and parasympathetic ANS are in operation together as part of the 'self' process of understanding the world and behaving in it. One is being depicted as an internal organization of neurology and biochemistry as present in the sympathetic feedback loop and parasympathetic feedback loop. The other is what happens when TWO of these dynamical systems interact - i.e.: in the relationship between two individuals. This dyad might be between mother and child, therapist and client, boss and employee, abuser and victim, lovers, friends or enemies, but however it plays out, the relationship is reflective of a higher order, and self-similar, fractal of the same system defined by the Autonomic Nervious System (or the Affective/Conceptual organisation of experience). So the relationship may be the very same dynamical system as the sympathetic and parasympathetic arms of the autonomic nervous system, only at different level of magnification.

- - - - - - - - - -  

The Dyad

Psychology has been struggling in an pre-paradigmatic position in that it appears to lack consensus (Kuhn, 1962). Berscheid (1995) argues that a 'grand theory' of relationship is required, and the only one that appears to be coming close is Attachment Theory. In psychology, nothing in the research really fits everything, there are problems everywhere, which is why there are so many theories and so many different takes on the research data being gathered. Not just problems like the nature/nurture debate, or determinism / non-determinism, but problems in either theories presented, or in the empirical assumptions that are made. If you want to challenge a classical psychoanalyst, all you need to do is say "counter transference" as Ferenczi, Reich, Jung and Klein did (Mitchell & Black, 1995). If you want to challenge behaviourism all you need to do is say "language" as Chomsky did ({ref} 1959), which led to the 'cognitive revolution' of behaviourism. If you want to challenge a humanist (Rogers), ask them if the empathic method will work on a psychopath (Hare, 1999:204) or machiavellian (McIlwain, 2003). If you want to challenge the intersubjectivists, turn them around, take their wallet, and then lie to them about your subjective point of view, followed by a question of how this sort of thing would play out in therapy (Mills, 2005).

If you want to challenge most contemporary fields of research in psychology, just say "context" and their house of empirical cards falls apart, backing down to comments like "but we can't really define motivation / resilience / distal traits (or whatever)". If you want to challenge a clinical approach like Cognitive Behavioural Therapy (CBT), ask the clinician about what happens when a client does not undertake therapy 'properly'. They will respond with something like, "well you can't treat a client that is 'resisting' therapy or 'resistant to' the CBT method". At which point you say "so when YOUR methodology fails it's actually THEIR fault!"

There are hundreds of different modalities for treating someone. There are mainstream approaches like CBT, but also psychodynamic methods like classical Freudian Psychoanalysis and it's many offshoots. There is Rogerian Humanism, Gestalt, Self-Psychology, Intersubjectivity. But you can also go off and study somatic psychotherapy, music therapy, dance therapy ... the list goes on for quite a bit and you will eventually get to charkas, incense and bizarre forms of meditation involving groups of people sitting around in pyramid structures. But ask this one simple question. Given that all these modalities exist in a very real marketplace of ideas, an assumption could be posited any one of them may show some success to some people some of the time, and that given this proposition, what is the one common factor they all have? Even excluding the placebo-likely modalities at the extreme, the question still holds.

I have asked a lot of colleagues this question, and they usually look a bit dumbfounded when I do. It is such a radical thought that befuddlement is the likely first response. But think about it as a serious hypothetical question. What is the one factor that unites the many different modalities of therapy where one person interacts with another? Ahh, I've given the answer in the question, can you see it now?

The answer is they all have a 'relationship' of some form or other. Now a whole other world view emerges which supports how growth, development and therapy can only happen in relationship with other people.

For over a century, psychology has been teetering on the edge of getting the immovable role that relationship plays in self formation and therapy. Every now and then a theorist or researcher will cross over and then start talking about the 'relationship' from their particular theoretical position. Freud spoke of the instincts (drives) as being shaped through a persons vicissitudes (Freud, 1915/1995). Freud also wrote about the relationship (transference) that needs to develop between the client (analysand) and therapist (analyst) (Freud, 1912/1995).

Klein looked at the parenting role in more detail and along with others spoke of how a 'good enough' parent is a child's first embodiment of object functions which are subsequently internalised by the child (Mitchell & Black, 1995). Cozolino states that, "In one sense, a child 'borrows' the prefrontal cortex of the parent while modelling the development of its own nascent brain on what is borrowed" (Cozolino, 2006:86). This object relations function could even manifest for a small child as something inanimate, like a towel, blanket, doll or spoon, things which Winnicott called a transitional object (Mitchell & Black, 1995). Carl Rogers and the humanist therapeutic approach looked closely at empathy and 'unconditional positive regard' as key parts of the relationship which forms in therapy (Mearns & Thorne, 2002). Rene Spitz radically confronted the childcare/adoption industry by looking at the absence of relationship, specifically maternal and emotional deprivation (Mitchell & Black, 1995).

This relationship between parent and child was also explored observationally by Bowlby, who developed Attachment Theory, and which was later supported empirically through Ainsworth's work with the landmark 'strange situation' observational experiments (Karen, 1998). Even narcissism, which was seen by Freud as "untreatable" (Freud 1912/1915), was later understood, and coped with, by Kohut through the understanding of the maladaptive caregiver interaction which forms in childhood, and how the narcissist stage of development has to be navigated sufficiently either in childhood, or in adulthood through a Self-Psychological therapeutic modality (Lee & Martin, 1991). From the philosophical tradition, existential phenomenology places the relationship as central to therapeutic process (Spinnelli, 2001). And resting heavily on Attachment theory, Self-Psychology and Existential Phenomenology, the Intersubjective approach suggests in more detail how the relationship dyad itself is manifest from the combined (interacting) subjectivities of two people (Stolorow & Attwood, 1992; Stolorow, Brandchaft & Atwood, 1995).

So relationship is somewhat important.

- - - - - - - - - -  

The Social Brain.

Cozolino (2006) in a profoundly important work that gathers data from over 1400 sources, talks about the neurological underpinnings of what he terms the "Social Brain", the basis of how relationships form and their function in everyday life. Throughout this text there is an allusion to both the separateness of function as well as their basic interconnectedness and Cozolino has been cited often. Importantly Cozolino never uses any terms from non-linear dynamical systems theory, but when viewed with a fresh eye, these separate yet connected systems are the same strange attractors that go toward making up the 'self' in an individual. Presented here is a simple table, largely extrapolated from Cozolino, but adapted to include theoretical non-linear functions being proposed in this work.

Table 10: The Duel/Interconnected Organisation of Experience
adapted from Cozolino, 2006 table 9.1, p.129


Neurology Sympathetic ANS Parasympathetic ANS
limbic based (amygdala) cortex based (OMPFC)
right hemisphere dominance left hemisphere dominance
early developing later developing
Schema Makeup affective/emotional conceptual/logical
implicit memory explicit memory
subjective in nature objective in nature
context free (no source attribution) Contextual / known memory source
Specific problems arising brash (act without thinking) ruminate (think without acting)
cognitive dissonance depression
Freud unconscious conscious
From Graves theory of consciousness A|N (survival) B|O (animistic)
C|P (ego/adventuring) D|Q (order / Absolutionist)
E|R (Multiplistic-achievist /science / consumerist) F|S (global / environmental / bounded)
Jung (spiralling out to the DYAD) animus (the masculine in the female) anima (the feminine in the male)

--- IN THE DYAD ---
Predisposition in the literature (choose any area you want really) masculine feminine
directed communicative open communicative style
(action) tending to negatively framed comm. (relationship) tending to positively framed comm.
lack of emotional balance, may lead to physical violence lack of physical action, may lead to emotional violence
Spiralling out to the GROUP Individual Group
attachment schema identity, social information
transference narratives
superego autobiographical descriptions
background affect social rules, norms, expectations
individual morality laws
decisions politics

--- IN THE GROUP ---
Politics right wing left wing
conservative liberal
patriarchy matriarchal (later feminist)
From George Lakoff authoritarian father nurturing parent

Taking a brief look from several areas, some of which were covered by Cozolino, this is how the social brain works.

Eckman (2003) lays the groundwork for culturally independent, or biological, ability of humans to interpret facial expressions. Humans, regardless of culture, will know the anger, disgust, joy, surprise, sadness and fear when they see it in another person. If, for example, this system goes wrong, as with Borderline Personality Disorder, Social Phobia, Autism (Cozolino, 2006) or Psychopathy (Cozolino, 2006; Hare, 1999; McIlwain, 2003) then the self of an individual may form differently. With psychopathy for instance, because the psychopath is unable to detect fear and sadness in another person, which are viewed as possible cues of submission, they never really develop empathy (Blair, 2002).

As well as detecting the facial expression in others, there are automatic responses like blushing, that happen in an individual in order to communicate that the truth is being told. Someone who is lying cannot blush involuntarily, but someone who is telling the truth may blush given the context of what is being discussed (Cozolino, 2006).

Humans also have distinctly different eyes when compared to other animals, and even other primates. There is significantly more white sclera as well as horizontal exposure of this sclera relative to the iris. This allows for communication through of direction of gaze and pupil dilation (in conjunction with automatic responses like blushing), all of which give the human face it's characteristic 'tell' (Cozolino, 2006). It's for a very good reason that professional poker players wear hats and dark sunglasses when they compete.

Olfaction plays an important role in the additional modes of communication. Endorphins, oxytocin and vasopressin can interact on the monoamines in another person (dopamine, serotonin, noradrenaline) (Cozolino, 2006). For example, when a mother undergoes childbirth there is often a flood of oxytocin in their system. This ramps up when they first see the child. Interestingly there is also a rush of oxytocin in the father when they see their child for the first time.

All these factors of information are exchanged in a largely unconscious way. The information becomes part of the implicit (amygdala based) organisation of experience. And this information acts as a sort of glue which builds relationships.

But perhaps the standout parameter of the social brain theory is in the role played by mirror neurons in the cortex. Mirror neurons were first uncovered in function di-Pellegrino et al. (Cozolino, 2006:187) in 1992, and have a special function in how relationships form. For example, when a monkey reaches out to grab an item, these mirror neurons fire. When the same monkey sees another monkey reach out for the same item, these same mirror neurons fire. The mirror neurons are somatically based and contextually dependent and will fire both when observing and performing a task. Sometimes when you see another person fall, and they land heavily on their knee, you can feel the pain in your own knee. In a very basic sense this is the activation of the mirror neuron system.

Cozolino (2002; 2006) argues that mirror system may have evolved initially as a way to transfer skill and abilities to others. They may also be implicated in the evolution of language, as hand-gestures are so often involved in communication. On top of this:

"Mirror neurons and the neural networks they coordinate work together to allow us to automatically react to, move with, and generate a theory of what is on the mind of others. Thus, mirror neurons not only link networks within us but link us to each other. They appear to be an essential component of the social brain and an important mechanism of communication across the social synapses." (Cozolino, 2006:198)

But the empirical work conducted on relationships can be somewhat daunting. There appears to be a lot of it, and not all in agreement. A few general threads emerge as consistent, such as with Attachment theory, but in general there seems to be as many ideas as there are researchers studying them. What follows are a few examples of work from empirically based researchers and how it can end up like a confusing mess. Keep in mind it seems to be is only the tip of a very big iceberg.

Berscheid (1983) discussed the role that emotion plays in an early work on close relationships from an Attributions perspective (Kelley et al. eds.), where she describes the role that the autonomic nervous system plays in relationships, and how the activation of the fight and flight response can serve as an interruption in the dyad such that a reorganisation occurs. In this depiction about relationships, there is a warning about using affect alone to interpret their strength because cognitive appraisals also play a part. Another reason to avoid using affect alone when looking at relationships is because there will be feedback loops in operation between the two people in relationship, called "'continuous loop' of feedback among elements of emotional experience" (Berscheid, 1983:122). In another work on emotions and cognitions in close relationships, it is suggest that: "Simply put, beliefs elicit emotions which motivate behaviours, that provoke responses that reinforce beliefs" (Fitness, 2006, p. 293).

Attributions theory is a mini-theory (Berscheid, 1995) developed from empirical study. It is almost completely divorced in it's hereditary from the classic theories, from trauma theory, from neuropsychology. Yet from the non-linear perspective argument presented in this work, it is another example of the multiple organisation of experience, affective and cognitive (sympathetic / parasympathetic, or amygdalic / cortical, or implicit / explicit etc) that go toward making up the decision making process, both in an individual and between individuals in a dyad.

The key problem is that the empirical methods used, were not really designed for non-linearity, and yet it is non-linearity which is being suggested, by feedback loops, to anyone who has studied dynamical (chaotic) system. In relationship research, apart from the fact that all the research is predisposed to an underlying set of arguments about gender, this core problem of feedback systems at work keeps appearing and is simply pushed aside as having been dealt with by the statistical approach. Here are a few more examples.

In signal detection theory, the idea of rejection has a particular problem. Signal detection is subject to type 1 errors (unnoticed rejection) and type 2 errors (false alarms), and it is the possibility of false alarms from background behaviour (noise) that can lead to a positive feedback loop that explodes out of all proportion. What's more, rejected people feel excluded thus leading to further exclusion {Lakin & Chartrand, 2005; Leary et al., 2004}. This rejection sensitivity is where the rejected individual could behave in a colder fashion leading to more rejection, {Murry et al., 2003} which has the hallmarks of distal insecure attachment styles, functioning as different starting conditions leading to chaotically different outcomes. Often this feedback loop is dismissed as a simple 'self-fulfilling prophecy', as though the unconscious implicit organisation of experience was simply absent from the process.

In another example of feedback loops being interpreted in a linear way, how partners see each other, that is having positive attributions or negative attributions, can lead to another direction of causality problem. Does positive attributions about a partner lead to happiness, or does happiness with the relationship cause of more positive attributions, with the admission that "empirical support for the underlying assumptions (of causality) remains weak" (Karney & Bradbury, 2000:305-306). This reliance on a linear interpretation of the world again dictates the way the information is both gathered and understood. At any point in time the full nature of this obvious dynamical system (because it can feedback on itself) can not possibly be captured and any conclusions drawn are simplistic at best.

And again, in mate selection, the notion of familiarity {Moreland & Beach, 1992} is also predisposed to a basic feedback loop over time, such that the input of to the system, recognition of someone, can build exponentially every time that same person is encountered in the future, without suggesting that distal personality influences may be the subtly different perturbations in initial conditions that make the effect of familiarity more power that is warranted.

Staying with mate selection, there is also data on proximity or functional distance {Myers, 1999); Complementarity {Soloman, 1986}; Reciprocity of attraction {Berscheid and Walster, 1978} and gaining another approval {Walster, 1965}; Similarity {Newcomb, 1961}; Affective (sympathetic / limbic) arousal {Dutton and Aron; 1974}; Self-fullfilling prophecies {Kenrick et al.}; and of course not to forget the role played by other factors such as beauty or facial/body symetry {Cunningham, et al., 1995}, social conditioning {Howard, Blumstein & Schwartz, 1987}, sexual experience {Clark & Hatfield, 1989}, high status/success {Grammer, 1992}, and resources / power {Ellis, 1992}. However ask either man or woman, from any culture, what they want and the top of the list is kindness, loyalty, emotional stability, with the male specific traits being youth and physical attractiveness, and the female specific characteristics being status and resources {Buss, 1989}. This is a long list of factors that have been independently isolated and studied in a linear fashion, taken from Fitness (2009). The question is how can they ALL possibly exist an not be fundamentally chaotic or non-linear?

All this research could benefit from a non-linear perspective. The dyad, how it forms, how it succeeds or fails, and the dynamic interaction and feedback systems that appear to be in play, all suggest that another fundamental dynamical system is in play. The second main thrust of this work suggests that the non-linear system of the dyad is likely connected to the non-linear system of the individual. This means the two are self-similar, and that the dyad is also an emergent function of the interaction of two people.

When a couple has been together a long time, they are often referred to as a single entity (a dyad), and to an outside observer it can seem like they are. Often within the dyad, each individual will take on different roles, which superficially at least could be termed masculine and feminine and through this, become subject to all manner of self-help books about planetary metaphors. Or in an attachment schema sense, if an avoidant member of the dyad (typically male) were to withdraw from their anxiously attached partner (typically female) then a 'demand-withdrawal' patter could emerge. As the female demands more, the male withdraws more, in a viscous cycle (read component of a non-linear feedback system). Add in other components like an open communication style vs. closed (non-talkative) communication style. Or perhaps a communication style that is positively framed vs. one that is belittling, sarcastic and negatively framed. The point is that any set of conditions can be set up so that the different individuals in the dyad take on specific roles of that dyad. In a stereotypical way, the male might take on an action orientation, where as the female might take on a sense-making orientation.

It is at this superficial level of roles within the relationship that the self-similarity of the same dynamical system that occurred in the individual may be at work. At any rate, it's the idea of a 'dyad' as being self-emergent that seems worth researching as opposed to the chunked down causality seeking linear interpretation of the dominant statistical method being used.

NEXT - Group Decision