Complexity and Systems: Concept 2 - Feedback processes both inhibit and promote change within systems

The Concept: Feedback can be understood as the influence or message that conveys information about the outcome of a process back to its source (Capra 1996). At its most basic feedback can, on the one hand, be positive/amplifying which causes pressure for escalating change or deviation in the system. On the other hand, feedback can be negative/dampening which resists change in the system.

Although feedback plays out very differently in simple and complex systems, the basic principles of feedback is similar. The difference is that in simple systems, feedback is understood to be linear, predictable and consistent. In complex systems, feedback "is about the consequences of non linear, random changes over time" (Byrone 1998 cites in Ramaligam 2008).

Munro in her Review of Child Protection uses this concept to explain how a system learns and how it is monitored. In her first report she introduced the concepts of single and double loop learning which distinguishes atomistic (isolated) from holistic (reflective) approaches to understanding child protection work.

Her diagram of single loop learning illustrates how a system is regulated for compliance to performance targets in child protection. Although in linear terms compliance to these targets is desirable, inevitably these targets and the approach used ripples through the system producing effects that were not intended.

"Ripple effects such as this are not ‘side effects’, in that they are no less ‘effects’ than those originally intended. What they are is effects that the changes were not meant to produce but which do result from the complex connections in the system. They are unintended consequences" (Munro 2010).

Munro then goes on to produce a series of complicated diagrams of feedback loops which illustrates organisational addiction to prescription. Perceived failures in linear approaches are hidden and critical reflective learning (and the need for variety in approaches) does not take place.

Complexity & Systems: Concept 1 - Interconnected and interdependent elements and dimensions

The Concept: A complex system is easily understood as one made up of multiple elements (which may be elements or processes) which are connected to and interdependent on each other and their environment (Nicolis and Prigogine, 1989). Complex systems are also characterised by multiple variables, or dimensions, which are interconnected and interdependent. Complex systems frequently have multiple levels of organisation. The degree of connectivity between these elements, dimensions and levels has a profound influence on how change happens within the broader system. (Ramaligam 2008).

The idea that complex systems comprise of interdependent and interconnected elements and processes is very familiar. The dimensions of the child, parents and their environment has become a well known and standardised framework in the UK for social workers assessing children in need of services and protection.

However, what seems to be further developed in complexity theory is the nature of connection networks between elements, which can vary in different parts of the system and over time. The degrees of interconnection (whether systems are tightly or loosely coupled) is crucial in understanding how the system behaves and how change occurs.

A tightly coupled system is seen as comprising a high degree of interconnection within the system with increased opportunities for the exchange of information and the options of rapid change within the system. However, this is not always an advantage as the more connections there are the more unstable the system will be. This is when relatively trivial changes in one element or dimension can spread rapidly and unpredictably through the system and have dramatic and unpredictable effects.

Those systems where elements are not tightly linked or interdependent with many other components are called loosely coupled systems. In these systems, elements influence each other over longer timeframes, and in more diffuse and subtle ways. It is also considered that although in loosely coupled systems, for example, may contain certain problems because of their lack of connectivity, they also do have benefits in supporting more independency, stability and resilience.

Contribution to Munro's Review

Although the Munro Review had sought quite limited and specific questions from practitioners, I decided to contribute the following thoughts:

How can we create a system for learning from practice which counteracts blame and allows for critical professional reflection? How can the government most effectively support this?

Learning from ordinary practice.

· Serious case reviews can be seen to be limited in their learning experiences as they focus on cases where things go wrong, and mostly, horribly wrong. It doesn’t seem possible in those cases to consider how learning can be conducted without blame. Of course there are cases where things work well and clearly there is more motivation and desire to learn from best practice than from serious case reviews.

· However, very few cases either go horribly wrong or fantastically well. Most cases appear to be rather messy with a combination of good, average and even poor practice elements evident. In my view, there are considerable amount of learning experiences that can and need to happen within the majority of the cases, those which are most familiar and most common to practitioners. Learning from practice is more likely to occur if professional weaknesses and strengths are acknowledged and understood as being part and parcel of ordinary practice rather than associated with the exceptional or extraordinary, the worst or best practice examples on either side of the spectrum.

· Although there are concerted efforts to improve learning from serious case reviews and even from best practice examples there are currently limited opportunities to learn from the majority, run of the mill cases. For example, the way the system is structured into specialist teams on the one hand can assist in accessing specialist expertise and services but on the other hand, it fragments practice and makes it difficult to encourage learning and see the bigger picture. Overall casework outcomes are often not evident to those involved as families and children pass through various workers, teams and services. Although clients do contribute to reviews of child protection or care plans their focus is on reviewing specific safeguarding plans or statutory care responsibilities at the time of review. Sometimes client feedback is sought for inspection or advertising purposes but rarely is feedback sought with the purpose of learning and development and letting those practitioners who were involved be part of this learning experience. Therefore learning from practice is very difficult if social workers are frankly not aware of the outcomes and impacts of what social work is done and the way it is done in cases they were involved with.

Critically applying research and theory development to practice

· Learning from practice also raises the need for undertaking more and better research and developing theory. However, outside of academic studies, exploring applicability of theory to social work practice is not usually encouraged. Theory is seen as too esoteric and abstract to inform and develop practice and too divorced from the real world. The problem is not only that thinking about theory is discouraged but also that there are limited intellectual and theoretical social work frameworks available for practitioners.

· In addition, undertaking basic research in social work is not included in current social work qualifications. Research and theory testing or development is therefore viewed by most frontline practitioners and managers as something relevant only in the world of academia with little relevance to practice as it is happening on the ground. Not surprisingly, what little social work research is conducted and theory tested and developed has no or little impact for most qualified practitioners as many do not have the skills to engage in theoretical and research reflection beyond the superficial.

· I anticipate the College of Social Work will promote research and theory development in practice that is professionally led rather than politically led. However, I believe that conducting a research project as part of the basic degree for social work (as it is the case in South African qualifications in social work for example) will promote a bottom up interest and input for better quality, more ground-based learning and research and more deeper understanding of research that is conducted.

Promoting independent professional supervision

· Critical reflection and learning from practice involves not only developing better intellectual skills and abilities but also a critical awareness of emotions and values. Social workers in child protection are confronted with some of the darkest, most difficult, value ridden and emotionally charged human situations any profession has to deal with. These situations can challenge fundamental values and evoke intrapersonal responses in involved practitioners and their managers. These human responses impact powerfully, either negatively or positively, on professional relationships and on individual and collective practice. Value dilemmas and emotive factors are also many times not obvious or explicit, occurring within practitioner and organisational blind spots, which requires considerable and specialist skills and abilities in identifying and exploring them in supervision.

· Much has been written about the need for effective supervision to support reflective practice and to address emotional and value issues that may be involved. However, the problem with how supervision is currently implemented and promoted is as an employer and managerial led process, primarily for monitoring of performance and conduct. I also believe the employment dynamics inherent in the relationship between a social worker and their line manger are not conducive to the required trust that is often necessary to promote critical professional reflection and learning from practice. For example, social workers will be not inclined to openly share their own difficulties that a client’s situation may evoke with their line managers (let’s say in dealing with experiences of discrimination, abuse, violence, addictions or mental health issues) out of fear that raising these will adversely affect their employment opportunities. As a result, these issues are often not effectively explored or addressed within current supervision practices.

· Further, line managers may be both unwilling and unable to identify or challenge wider working practices within the team that impacts negatively on casework, as managers are part of those practices themselves, and may indeed be the driving force behind them. Frontline managers themselves also often experience poor quality supervision from their line managers and they themselves have nowhere to go to explore in any detail the value dilemmas or emotional stresses that supervisees may evoke in them. In addition, line managers are also under obligations for implementing corporate priorities and policies and are under pressure from senior managers who have sometimes vested interests in implementing dubious processes and procedures (for example, in the top down implementation of the ICS based on my experiences).

· Similar to research, I believe there is scope for including supervision theory and practice (again as is the case in South African social work training for example) as part of current training for social workers. This would provide newly qualified social workers with basic supervision skills that will promote the value of supervision (including peer supervision) and develop better quality practice if they do wish to specialise in professional supervision as they progress. In addition, I would think that there is further scope of government piloting the provision of professional supervision outside of local authority direction and control. In Sweden, for example, external professional supervisors promote learning and reflection in combination to line managers who retain responsibility for case direction and decisions. This seems like a good idea as independent supervisors are more likely to allow practitioners the opportunities for a much more critical and deeper awareness of both individual factors and wider working and organisational practices that many times cloud professional judgement and decision making.

· In conclusion, from my perspective as a practitioner, with more focus on learning from the average, run of the mill cases, more ground-based social work research and theory development and better quality professional supervision independent of line management, is fundamental in supporting critical professional reflection and practice learning.

Application of Chaos Theory to Social Work

Hudson in 1999 noted that unlike systems theory or general systems theory which appeared in 120 social work articles, chaos theory only appeared in two social work abstracts. Subsequently, over the last decade, chaos or complexity theory appears to continue on the fringe of thinking about social work, in only a handful of literature, at least that is accessible. Interestingly, the little that has been written about complexity theory in the UK is all related to exploring its implications and applications to child protection social work.

Professor Christopher Hudson (1999):

Hudson, an academic in the United States, has recently written a book on Human Behaviour and Complex Systems, which I have not yet read but is based on complexity theory. In 1999 he explored chaos theory and its implications to social work.

Hudson (1999) identified a major problem of chaos theory is that it presents an inability for social work to predict the outcomes of interventions. This is understood as uncertainty and unpredictability is a central characteristic of complex, non-linear and chaotic systems. Unpredictability of interventions has also lead to a similar dilemma in related fields. For example, Hendrick (2009) explained uncertainty "has always been a part of interventions in complex systems it is just that it was very difficult to admit this for personal or professional reasons - an unwillingness to undermine one's own sense of usefulness or control, the pressure of results-dependent funding from donors, fear of it being viewed as an excuse for incompetence or failure - and this has led to unrealistic expectations, less than illuminating evaluations and, in some cases, fantasy strategising."

As previously mentioned, uncertainty emerges as a significant theme in Munro's Review, where she identifies that "child protection work involves working with uncertainty: we cannot know for sure what is going on in families; we cannot be sure that improvements in family circumstances will last. Many of the problems in current practice seem to arise from the defensive ways in which professionals are expected to manage that uncertainty".

Although predicting outcomes of interventions is impossible, Hudson explains that chaos theory can assist in predicting overall patterns, trends and processes in client's lives. The use of the concepts of self-organisation, attractors and feedback processes, for example, is helpful in doing this. Predicting overall patterns and processes can aid in decision making of social work intervention.

Emergence, bifurcation and the edge of chaos is also used as central concepts to understand how change and intervention in social work can be approached from a chaos perspective. The edge of chaos, for example, is regarded as "one of the most important conditions for creativity and effective problem solving" (Richards 1996). This is reflected in Hendrick who also found that the edge of chaos has been incorporated in some problem-solving and mediation approaches in the past. She goes further to cite Wils et al (2006) arguments that it may be necessary to push the system to the edge of chaos in order to create the conditions necessary for change and possibly to induce system transformation.

Hudson concludes in his review that it is "premature to conclude that chaos theory or complex systems represent a new paradigm for social work, if for no other reason than the sparsity of its use in the social work profession to date". However, he stresses that both concepts and methodologies of chaos theory have the potential to extend social work's ability to understand increasingly complex systems, moving beyond the limitations of general systems theory. This includes an understanding of natural processes of change, growth, and the self-organising and emergent properties of complex systems which are found at the edge of chaos. This is something the dated general systems and ecosystems theory is unable to explain.

Dr Irene Stevens and Pat Cox (2008):

Stevens and Cox (2008) from the Scottish Institute for Residential Child Care explore the potential contribution of complexity theory and concepts that have relevance to child protection, in both field and residential practice. They explore the relevance of concepts from chaos theory to understanding child abuse and child protection. These concepts will be revisited in future posts when I will explore and critically reflect on their applications and implications to practice on the ground.

Stevens and Cox also use the perspective of complexity theory to viewing child protection policies as they have developed or emerged over time. They are rightly critical of linear thinking which they argue is fundamental to how serious case reviews have been approached and implemented. Indeed linear approaches is also characteristic of child protection assessment and practice on the ground, "which can lead social service practitioners with a false sense of security".

They argue that complexity theory can help practitioners working to keep children safe, that it offers new ways of working, away from a risk-averse approach. "In conclusion, the authors believe that complexity theory offers tools for understanding and analysing many of the complex adaptive systems within which practitioners operate in protecting children".

Dr Stevens and Peter Hassett (2007)

Stevens and Hassett, in 2007, have explored the application of complexity theory to risk assessment in child protection practice. They focused on using the concepts of non linear dynamics and fractals (a diagramme of risks to illustrate non linear patterns of risk) in order to identify and analyse risks. They also use spatial analysis in approaching risk assessment and explain that the Scottish Institute for Residential Child Care over the last two years have used this and complexity theory to train practitioners in its application.

Stevens and Hassett illustrate the model by applying it to an example in practice which includes the following process:

1. Risk and hazard identification: Approaching risk as multilayered and multifaceted, identifying risks that are present in a situation and constructing a fractal (a diagramme of non-linear dynamics which illustrate how risks are related).
2. Assessing impact and likelihood of harm: For each risk (and sub-branches) that are identified, the potential impact (the severity) and likelihood (of re-occurring) is assessed and quantified (1 - 3).
3. Spatial analysis of risk: Simply put, spatial analysis considers the relevance of space, place and time. Space is the social construct of place (what happens within a place) which aid where and how things happen within specific times. Spatial analysis relates the risks to social circumstances. This assessment gives a picture of the range, degree and circumstances of risks in a situation.
4. Developing a series of quantitative and qualitative statements of risk: This involves analysing the spatial analysis and the likelihood and impact of harm (judgement of seriousness) and deriving a series of statements to inform intervention.
5. Control mechanisms and monitoring: Identifying control mechanisms brings together the highest ranked risks and the spatial anaysis. These identify what is to happen, where and when, who is responsible and how the risk will be monitored and reviewed.

Steven and Hassett conclude that although complexity theory "itself may sound quite esoteric when discussed in isolation" if it is utilised as an analytical tool and as a paradigm, it can provide a much more realistic framework to understand risk and keep children safe.

Thomas Coram Research Unit: University of London (2009)

John Rowlands (2009), a visiting fellow to the Institute of Education, had used complexity theory, in particular the concept of "self organising criticality", that is suggested to underpin the generation of children in need as defined by the Children's Act 1989. He states that complexity theory provides a framework for the phenomena involving the interaction of multiple agents that produce unpredictable outcomes and can not be dealt with by using at linear and reductive analytical methods. He suggests that the population of known children in need could be such a phenomena and provides evidence to support this.

Rowlands (2009) uses secondary analysis of children in need census conducted in England in 2005. The census, well known to social workers at the time, was developed in response to Treasury's concern about the rising costs of children social services and to gain a better understanding of "where and why the money goes". The census required local authorities to survey all the children being worked with during the census week and gather information about the costs and activities related to each child.

Taking the idea of self organisation, Rowlands uses the census data to test the concept statistically. He noted that previously Down (2007) had argued that there appears to be a lack of any published research on the application of complexity theory to social care. Therefore, by using quantitative methods, Rowlands "tries to demonstrate that complexity theory is at work within a dynamic system that gives rise to children in need".

He concludes that the finding are suggestive that indeed complexity theory could be applied and this supports the the desirability of pursuing complexity theory as a means of gaining a better understanding on the dynamics of children in need. He argues against policy development that is based on linear and reductionist understanding and suggests that "research into children’s social care might always provide limited insights because the interactions of agents are too complex

to allow a full description of the efficacy of intervention".

Munro's Review (2010)

As previously outlined, Munro's analysis is based on a systems approach. She does not mention chaos or complexity theory by name but some of the main concepts and themes she is focusing on appear to be based on complexity theory. She focuses on why previous, albeit well intentioned attempts to improve the system have had unintended outcomes. She also applies this to her own review when she states that "in such a complex system as child protection, it is inevitable that any innovations this review recommends will themselves have unexpected consequences as they are put into operation". A wholistic view of the system is argued for.

Munro uses ripple effects and feedback loops to understand how previous reforms have resulted in unintentional outcomes because perhaps to a narrow approach was taken. She illustrates this in a series of diagrammes to a hypothetical understanding of systemic factors in child protection. Interestingly she stresses that the link between variables are no presumption of a linear relationship.

Through the application of feedback loops, ripple effects and requisite variety, Munro suggests how practice can create a self defense mechanism which hides procedural failures. A system's "ability to show that there was compliance allows it to argue that there was due diligence in terms of the procedures used and that, hence, the errors in actual child protection cannot be perceived to be errors in terms of the approach used". Munro's review will be revisited in future posts as more insights are gained by her analysis and her recommendations in the new year.

In conclusion, the increasing use of chaos theory and concepts in recent literature in social work can be seen to be a turning point for advancing theory and its application to practice on the ground. This is significant in a particular social and professional context of child protection social work which can be rightly argued to be on the edge of chaos.

My posts from now identify each concept as I outlined previously, hopefully focusing less on the academics and more on their critical implications and possible applications to practice.

Application of chaos theory: Other disciplines

The most extensive applications of chaos theory have been in the physical and biological sciences. In medicine several researchers have used chaos theory in the study of the brain, nervous system, and perception. Related areas of research application of chaos theory include sleep neurophysiology, neuroendocrinology and psychophysics (Hudson, 1999).

Hudson (1999) further identifies that although the application in psychology appear to be fairly widespread, most of these involve only selected concepts, rather than the methodologies of chaos theory. For example, the use of "butterfly effect" has suggested that "the sources of specific day to day lasting behaviours, as well as the sources of problems and functioning, are not limited only to major events in people's lives, but also include relatively small events, which, over time, can result in significant impact on behaviour" (Duke, 1994). The use of other chaos theory concepts have been more extensively used however, over the last decade, with the formation of the Society for Chaos Theory in Psychology and the Life Sciences and further developments of the theory in psychology and psychotherapy.

Applications in related fields can be found in decision making and organisational behaviour, group process, as well as in education and sociology (Hudson 1999). Recently, from what I have read, there has been increasing application of the theory to international development and to conflict resolution studies.

An extensive but interesting working paper by Dianne Hendrick in 2009 explores complexity theory in relation to peace research and conflict intervention. She critically explores complexity and chaos concepts, including providing suggestions and insights on how these could inform analysis and intervention. This appears very relevant to social work with children and families.

Violence and conflict (whether they occur within societies or families) have similar characteristics, they are both complex and chaotic. In addition, the challenges facing conflict transformation in a wider context is very similar to those facing social workers when protecting children or vulnerable adults. For these reasons, Hendrick's paper will be revisited in more detail in future posts.

Chaos Theory: Key Concepts

In undertaking some amount of reading on complexity and chaos theory, specifically related to social science and social work, it became apparent very early on that there is no unifying theory and that chaos and complexity studies is largely a developing field.

Nonetheless the conceptual framework provided Ramaligam et al (2008) in their working paper exploring the implications of complexity science to humanitarian and development efforts appears to include almost all of the concepts I have read elsewhere and make the most sense. They identify three main sets of complexity science concepts. However, I slightly adapt the framework by including a key concept of self similarity, relevant to social work, within third set that was identified by Hudson (1999).

The key combined key concepts slightly adapted from Ramaligam (2008) therefore are as follows:

Complexity and Systems: The concepts identified relate to the features of systems which can be described as complex.

1. A key starting point is that complex systems are characterised by interconnected and interdependent elements and dimensions.
2. Feedback processes shape how change happens within a complex system, where positive feedback never settles the system into completely predictable patterns.
3. Emergence describes how a complex system emerges, often unpredictably and based on simple rules, from the interactions between the parts to form a whole that is different to the sum of the parts.

Complexity and Change: Concepts related to the phenomena in which complexity manifests itself within systems.

1. Within complex systems, relationships between dimensions are frequently nonlinear, i.e. when change happens, it is frequently disproportionate and unpredictable.
2. Sensitivity to initial conditions highlights how small differences in the initial state of a system can lead to massive differences later; butterfly effects and bifurcations are two ways in which complex systems can change drastically over time.
3. Phase space helps to build a picture of the dimensions of a system, and how they change over time. This enables understanding of how systems move and evolve over time.
4. The edge of chaos describe the transition point between order and chaos within a complex system and one of the most important conditions for creativity and effective problem solving.

Complexity and Agency: Concepts related to the notion of adaptive agents and how behavours are manifested in complex systems.

1. Adaptive agents react to the system and to each other, leading to a number of phenomena.
2. Self similarity involves structures within systems that repeat basic features on several different levels of observations.
3. Self-organisation characterises a particular form of emergent property that can occur in complex adaptive systems.
4. Co-evolution describes how, within a system of adaptive agents, co-evolution occurs, such that the overall system and the agents within it evolve together, or co-evolve, over time.

Before I explore the implications of these concepts in more detail to the work I do in the next few posts I will describe how these concepts have been explored in various other fields, including social work.

Chaos Theory: An Introduction

The idea of chaos can be traced back to ancient Greek, Egyptian and Chinese philosophy. Chaos in ancient thinking was regarded as the birthplace of the cosmos. In the 20th Century chaos returned as a critical concept with the decline of metaphysics and the fundamental critique of science (Schirmacher, 1989).

The most important work foreshadowing the field of chaos studies was done by a late 19th Century French mathematician, Poincare who also first introduced the principle of relativity. Other predecessors included Lewin (considered as the founder of social psychology), Thom (the founder of catastrophe theory) and most significantly Lorenz, who in 1962 discovered the existence of chaotic structures, referred to as "strange attractors", in weather patterns (Hudson, 1999).

However the field of chaos theory developed its identity in the 1970s when the mathematician and biologist, Robert May suggested that seemingly simple equations may represent very complicated dynamics (Hudson, 1999). This concept was first introduced by Alan Turing, when in 1954 he proposed mathematical equations to illustrate the formation of patterns and morphogenesis.

Chaos Theory in its inception was developed to deal specifically with systems characterised by the mathematical notion of "chaos". Chaos, in this sense, refers to "systems which can be found at an intermediate point in the continuum which ranges from the completely periodic and predictable to the totally random, and in which there is a type of order which never exactly replicates itself" (Hudson, 1999).

As chaos theory developed, the approach has been applied to a range of complex, dynamic, and non-linear systems which do not technically qualify as representing the narrow mathematical notion of chaos. This broader field has been variously referred to as non-equilibrium theory, self-organisation theory, nonlinear dynamics or complex adaptive systems, each which "have typically attempted to integrate what is known of the three major classes of processes: deterministic, chaotic and random" (Hudson, 1999).

The most extensive applications of chaos theory have been in the physical and biological sciences, less so in medicine, and in economics, psychology, sociology and conflict resolution studies. While chaotic properties are believed to take part in all major categories of systems (conservative, dissipative and quantum) most work has focused on chaos in dissipative systems, "of which biological and social systems are prime examples" (Hudson 1999). Chaos theory "represent a recognition of the limitations of Newtonian, linear scientific paradigm when applied to complex systems" (Hendrick, 2009).

Hudson (1999) stated that although "enthusiasts for chaos theory have gone as far as to argue that the 20th Century has seen three major scientific revolutions -- relativity, quantum, and most recently, chaos theory -- ... much of social work research continues to draw heavily from GST and even older paradigms". He remains critical that even though chaos theory has received attention in related disciplines, discussion of chaos theory in social work literature is almost non-existent.

I will consider in more detail how chaos theory and its concepts have developed and been applied in social work and related disciplines in further posts. However, in order to have a firmer foundation of what chaos theory is, in my next post I will outline key concepts.