Understanding Systems Science: a Visual and Integrative Approach

Home , Systems science

Systems Research and Behavioral Science Syst. Res. 30, 580–595 (2013) Published online 18 October 2013 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/sres.2215 ■ Research Paper

Understanding Systems Science: A Visual and Integrative Approach

Andreas Hieronymi* University of St. Gallen, Switzerland

Systems thinking is considered a much-needed competence to deal better with an increasingly interlinked and complex world. The many streams within systems science have diversified perspectives, theories and methods, but have also complicated the field as a whole. This makes it difficult to understand and master the field. Short introductions to fundamental questions of systems science are rare. This paper is divided into three parts and aims to do the following: (1) to provide a broad overview of the structure and purpose of systems science; (2) to present a set of key systems principles and relate them to theoretical streams; and (3) to describe aspects of systems-oriented methodologies within a general process cycle. Integrative visualizations have been included to highlight the relationships between concepts, perspectives and systems thinkers. Several new attempts have been made to define and organize system concepts and streams in order to provide greater overall coherence and easier understanding. © 2013 The Author. Systems Research and Behavioral Science published by John Wiley & Sons, Ltd.

Keywords systems science; systems theory; complexity; philosophy of science; systems thinking

INTRODUCTION: BACKGROUND AND multidisciplinary. A comparative study (Wiek PROBLEM et al., 2011) came to the conclusion that one of ﬁve key competencies for a sustainable future is What skills are needed for the 21st century that ‘systems thinking competence’. Peter Senge, have been neglected in the past? It has become one of the key promoters of organizational increasingly clear that the problems and challenges learning and systems thinking in management we face are highly interlinked, complex and (1990), argues that three core capabilities are necessary: We need to increase ‘collaboration across ’ ‘ ’ * Correspondence to: Andreas Hieronymi, University of St. Gallen, boundaries , see systems as a part of larger sys- Switzerland. tems and learn to ‘create a desired future’ (Senge E-mail: [email protected] et al., 2010, p. 44). These three challenges are

This is an open access article under the terms of the Creative closely related to the three foundational aspects Commons Attribution-NonCommercial-NoDerivs License, which per- of systems science explored in this paper. mits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modiﬁcations or The goal of this paper is to make key perspec- adaptations are made. tives and concepts of systems thinking and systems

science more understandable to researchers and to persons involved practically in fields such as education, consulting or management. To achieve this goal in the limited space of this paper, emphasis will be on visual maps that help us to integrate systemic knowledge from diverse streams and to highlight relations. More detailed descriptions of the concepts mentioned can be obtained in the cited references. Troncale (1985, p. 30) states ‘There is a need to make general systems theory more user- friendly’. Among other solutions, he recommends overcoming obstacles by use of graphic techniques. Figure 1 Three faces of systems science and respective parts There exist several definitions of systems science. of this paper Systems science is the ordered arrangement of knowledge acquired from the study of systems one), followed by theories of systems (part two) in the observable world, together with the and systems approaches (part three). application of this knowledge to the design of The history of systems science has its begin- man-made systems. (M’Pherson, 1974, p. 229) nings in the years around 1950 (Hammond, 2003) with the work of founding fathers such as Bertalanffy, Wiener, Rapoport, Boulding and [Systems science] does not aim to find the one Miller. The emergence of systems thinking is true representation for a given type of systems closely linked with the endeavour to overcome (e.g. physical, chemical or biological systems), previous boundaries within academia and practice but to formulate general principles about how (interdisciplinarity and transdisciplinarity). Re- different representations of different systems searchers in fields such as biology, psychology, so- can be constructed so as to be effective in ciology and technology collaborated on urgent problem-solving. (Heylighen, 1990, p. 423) real-life problems and on investigating general principles and theories on how systems function in general (theories of systems). Terms were newly Systems science is a science whose domain of created or specified, such as feedback, autopoiesis, inquiry consists of those properties of systems chaos and complexity. These concepts received in- and associated problems that emanate from the creasing interest in applied fields so that many general notion of systemhood. (Klir, 2001, p. 5) methods and methodologies emerged that incorpo- rated aspects of systems theory in order to improve practice (systems approaches). The goal is to better Systems science can be defined as the scientific understand how different sorts of systems work exploration and theory of systems in the various and how to deal with complex situations and re- sciences, such as biology, sociology, economics, duce unwanted side effects. etc., while general system theory concerns the Over 40 years ago, Ackoff (1971, p. 661) was principles that apply to all. (Strijbos, 2010, p. 454) insisting on more coherence on the theoretical side of the systems field: Klir (2001, p. 5) states that systems science, like any other science, needs to distinguish among Despite the importance of systems concepts three components: ‘A domain of inquiry’, ‘a body and the attention that they have received and of knowledge regarding the domain’ and ‘a are receiving, we do not yet have a unified or methodology’. These three components will also integrated set (i.e., a system) of such concepts. be addressed in the three parts of this paper. […] This state is aggravated by the fact that Figure 1 gives an overview on how this paper is the literature of systems research is widely structured, starting with transdisciplinarity (part dispersed and is therefore difficult to track.

© 2013 The Author. Systems Research and Behavioral Science Syst. Res. 30, 580–595 (2013) published by John Wiley & Sons, Ltd. DOI: 10.1002/sres Understanding Systems Science 581 RESEARCH PAPER Syst. Res.

[…] Contributions are not as interactive and disciplines in the same way that universities additive as they might be. are’ (Ackoff, 1960, p. 6). A similar critique is made by Warfield (2003, The so called systems approach is often p. 507). He considers systems science to be ‘very portrayed as a counter-current to the increas- broad in its scope and far reaching in its implica- ing fractionation of science into highly tions for practitioners’, but in his view, systems spezialized branches resulting in a breakdown science is still in ‘a formative stage’. of communication between the specialists. (Rapoport, 1986, preface) There is not a corresponding understanding concerning the content of systems science. Today, systems science and the field of interdis- On the contrary, there are dozens of small ciplinarity and transdisciplinarity still share many systems societies that speak to widely differ- similar goals and partly overlap. One way to think ing points of view as to what constitutes about the structure and boundaries of disciplines systems science for their members. (Warfield, is by means of visual science maps, including the 2003, p. 508) simple ones in this paper, or more sophisticated ones such as those Börner (2010) is elaborating by Warfield insists that every science—including sys- useofimmenseamountsofdata. tems science—needs a ‘central purpose’, ‘a corpus […] that consists of foundations and theory’ and a Through maps of science, we can begin to see ‘methodology as a defined process’ (2003, p. 508). all that we know as landscape – viewed as if Troncale (2006) makes a comparable point: ‘The from above or from a great distance. Science overall field of systems science is still in formation’ maps provide guidance for navigating, (p. 554), and ‘there is still insufficient integration of understanding and communicating the dyna- the many different strains of systems theory and mic and changing structure of science and systems tools’ (p. 560). technology. (Börner, 2010, p. ix) The hypothesis of this paper is that steps Other maps describing the landscape of the sci- towards a better integration of the systems field ences and the emerging field of systems science can be made by exploring and visualizing are provided in Müller (2011), for example. systems science as a system itself. What function does systems science have within its larger system? What are the functions of the subsystems Systems Science within the Science System of systems science? And how do the parts of systems science relate to each other and create a What are the role, purpose and place of systems bigger whole? science in the landscape of the sciences? How would a librarian classify the books in this field? Is systems science closer to computer science, or PART ONE: SYSTEMS SCIENCE AND management, or theoretical biology, or is it more TRANSDISCIPLINARITY similar to mathematics? The following map (Figure 2) provides an overview regarding the Many theoretical and practical problems cannot position that systems science could be assigned be understood and solved by a single discipline within science as a whole. Five horizontal system alone. Interdisciplinarity involves work between categories are combined with five vertical knowl- fields and learning about each other, whereas edge dimensions. The result is a visual map of transdisciplinarity involves work in which new the sciences with two axes. M’Pherson (1974, p. shared concepts are needed, and work that 223, p. 229) and Max-Neef (2005) use similar clas- bridges theoretical and practical issues. Systems sifications to distinguish systems and disciplines. science grew out of the need to communicate Figure 2 presents a map with five major science across disciplinary boundaries. ‘We must stop fields on the horizontal line, which revolve around acting as though nature were organized into the following concepts: physical system, living

Figure 2 Map of science—with a special focus on systems science and systems design

system, cognitive system, social system and technolog- Seen in this way, systems science and systems ical system. The map is extended by further knowl- design provide a bridge between natural science edge fields on a vertical line: ‘Logic and and the humanities, as well as between descrip- mathematics’ are mentioned in the direction of tive research and normative practice, thus ‘formal’ sciences because they are the most ab- making a contribution in terms of inter- and stract; ‘values and aesthetics’ are mentioned in transdisciplinarity. the ‘normative’ direction because they involve reflection and judgments of action. This diagram in- cludes two further fields: systems science and Mapping Systems Thinkers systems design. Systems science is located on the map as a formal science between mathematics In the following, the structure of Figure 2 is and the reality-based sciences. Boulding (1956, p. used to provide a short overview of a variety 197) states that ‘General Systems Theory […]lies of systems thinkers (Figure 3). A special feature somewhere between the highly generalized con- of the systems movement is the fact that its structions of pure mathematics and the specific important exponents come from very different theories of the specialized disciplines’. Systems de- science fields. Many of the systems thinkers sign refers to the heterogeneous field of general presented here are discussed in more detail in methods and practices relevant for all applied Ramage and Shipp (2009). Other visual repre- fields. Banathy (1996, p. 34) states that ‘In design, sentations of the systems field are given by Ison we focus on finding solutions and creating things et al. (1997), Castellani and Hafferty (2009) and and systems of value that do not yet exist’. Sayama (2012).

Figure 3 Overview of systems thinkers and their position in the ﬁeld of the sciences

The graphical representation in Figure 3 in- PART TWO: THEORIES OF SYSTEMS cludes over 100 people considered to be systems thinkers. They are organized according to There is no generally agreed on ‘systems the- key contributions to the field, although most ory’. The focus in this section is on different no- of these systems thinkers could be assigned to tions of defining systems and a variety of multiple places. The list is by no means systems principles. exhaustive. The names of those pictured are given in bold. The purpose of the map is to The task of a general theory of systems would stimulate interest in and discussion on systems include that of defining a system, of formulat- thinkers. This map is based on the author’s fa- ing a taxonomy of systems, of singling out miliarity with the systems thinkers presented properties that various systems have in in it, as well as additional feedback obtained common, and of explaining how this approach at two systems-oriented conferences. Most sys- can help us to a better understanding of our tems thinkers have developed their own unique world. (Rapoport, 1986, p. 1) understanding of systems definitions and concepts. Often, terms and concepts of systems ‘A scientific field can arise only on the base of a theories are optimized for a certain field, such as system of concepts. Systems science is not an for managers, social scientists or engineers. Part exception’ (Ackoff, 1971, p. 671). Researchers in two, on the other hand, attempts to create a several fast-growing scientific fields realize that multidisciplinary view of systems. the theoretical foundations of their own fields

are strongly related to or even directly based on between the objects and between their attri- systems science – such as in sustainable develop- butes’ Hall and Fagen (1956, p. 18). Others use ment (Weinstein and Turner, 2012), public health the word ‘system’ to connote a relatively com- (Luke and Stamatakis, 2012), service systems plex adaptive system, which has many interre- (Maglio, Kieliszewski and Spohrer, 2010) and lated subsystems and is once again a part of systems engineering (Pyster and Olwell, 2013). alargersystem:‘CAS [Complex adaptive sys- This situation demands from the subdisciplines tems] are systems that have a large number of within systems science increased collaboration to components, often called agents, that interact and provide a coherent body of knowledge, including adapt or learn’ Holland (2006, p. 1). deﬁnitions, concepts, classiﬁcations and related Figure 4 indicates that, depending on the general methods to deal with complex situations. perspective, different properties are connoted Additional educational strategies and tools are with the word ‘system’. The next section takes a needed to facilitate teaching and learning systems closer look at this problem and proposes a hierar- science and to creatively apply its principles in a chical approach to differentiating between several wide variety of settings. systems principles and types.

Definitions of a System Principles of Systems What is a system? The founders of the systems movement stressed The concept of a system is one of the most the importance of general systems principles widely used concepts in science, particularly and concepts. in recent times. It is encountered in nearly all the fundamental fields of science, such as phys- Today our main problem is that of organized ics, chemistry, mathematics, logic, cybernetics, complexity. Concepts like those of organiza- economics, linguistics, biology, psychology, as tion, wholeness, directiveness, teleology, con- well as in the majority of engineering branches. trol, self-regulation, differentiation and the Klir (1965, p. 29) like are alien to conventional physics. How- ever, they pop up everywhere in the bio- Although the concept of a system is now very logical, behavioral, and social sciences, and widespread, we still have a situation in which are, in fact, indispensable for dealing with multiple definitions co-exist. Rather than an living organisms or social groups. von error, this reflects the multidimensionality of Bertalanffy (1956) the concept, ranging from simple to complex notions (Figure 4). Some researchers understand ‘Proponents of general system theory purport to systems as the simplest form possible: ‘Asystem seek integrating principles sufficiently general to is a set of objects together with relationships apply to many different contexts: physical,

Figure 4 Two extremes of deﬁning a ‘system’. Left: A ‘system’ in the sense of simpel elements with relations. Right: A ‘system’ in the sense of an adaptive agent interacting within a dynamic environment, that itself is a system of systems

biological, psychological, and social’ Rapoport chosen key terms appear in quotation marks. (1986, preface). What general principles/charac- Criteria for their selection were that they are short, teristics/features can be defined that differentiate easily understandable and can be used in several or unify certain instances of systems in a logical areas of application. Other synonyms might be and coherent way? This is one of the most central more appropriate depending on the context and sci- questions in systems science, and there has been entific perspective. More important than the terms no consensus. One of the earliest hierarchies of themselves are the described functions behind the system types was proposed by Boulding (1956). terms. This process of concept building and classifi- Miller (1978) proposed in his ‘living systems cation is not yet finished, it is still at an early stage, theory’ a collection of 20 general system compo- open to discussion and further improvement. nents that can be found on many levels of systems. Other suggested systems principles and types can •‘Boundary’: A system exists as a unity by be found in Ackoff (1971), Mingers (1995), means of the relations between its elements Martinelli (2001), Meadows (2008) and Lin et al. and the boundary that differentiates the system (2012). The following list of principles (Figure 5) from its environment. In that basic sense, the has been newly elaborated by the author and is system can still be a ‘static system’ (connected the result of extensive research on the literature in elements within a boundary). multiple fields of knowledge. •‘Energy’:Ifenergyflows through a system that These principles comprise a core element of the enables movement, it becomes an ‘active sys- author’s intended future research. Certain similar tem’ (energy flow, motion). principles/characteristics of systems can be found •‘Computation’: If a system is capable of in theoretical biology (e.g. Koshland, 2002; Elitzur, processing and computing relevant data and 2005), in developmental psychology (e.g. Piaget, directing its actions based on rules, a system be- 1971; Fischer, 1980; Commons et al., 1998) as well comes a ‘rule-based system’ (data processor, as in artificial intelligence and robotics (Braitenberg, rules, computation). In this general sense, every 1984; Pfeifer and Scheier, 2001; Russell and Norvig, living cell is in the process of computing data— 2003). What kind of features does a system need in not just computers. order that the parts can build an integrated whole, •‘Perception’: If a system is able to detect and the system is able to adapt to changing environ- perceive certain signals of its environment, such ments and the system itself can finally be a func- as through sensors, and uses this information in tional part of a bigger (group) system and the form of feedback, it becomes a ‘cybernetic sys- contribute to it? The proposed principles (Figure 5) tem’ (sensors, perception, monitoring, feedback). are described in the following text in the sense of a •‘Robustness’: If a system can store energy and simple system evolving step by step into a more make use of stored energy, this can enable the complexone.Ineachstep,thefocusischangedto system to reliably maintain critical processes an additional type of functionality or process. The and structures—and endure and survive even

Figure 5 List of proposed general principles of systems. The metaphor of a circular staircase expresses the idea of hierarchical systems principles moving from bottom to top to reach the next systems level

without a continuous flow of energy from groups or even be a functional part of a broader outside. It becomes a ‘robust system’ (energy organization. These aspects have parallels in storage, redundancy, robustness, stability). many areas, for instance, about how to define •‘Identity’: If a system can build up memory and characteristics of organizations (Katz and Kahn, refer to some of its own present and past states 1978) and how communities collectively manage and conditions, it can thereby create representa- common goods (Ostrom, 1990). tions and a sort of identity. The system might On the basis of the principles described here, a now exhibit increasingly path-dependent, type of a system can be classified in terms of the nonlinear, unpredictable and chaotic behaviour. most prominent or highest principle reached. A This can be described as a ‘self-referential ‘type’ refers to ‘all the agents in a population that system’ (identity, individuality, memory). have some characteristic in common’ Axelrod and •‘Adaptation’: If a system attempts to maintain Cohen (2000). The aforementioned list makes it or ‘bounce back’ to inner conditions that are possible to distinguish between ‘passive systems’ favoured, it assumes a form of homeostasis (such as a table) and ‘active systems’ (such as a and becomes an ‘adaptive system’ (internal steam engine), for example. A company might homeostasis, adaptation, resilience). behaveonthemarketlikeanadaptivesystem(re- •‘Innovation’: If a system can actively search for act to changes to maintain identity) or an innova- and create new internal connections and con- tive system (intensive search for new solutions). nectinformationinnewways,itcanbecome Similarly, Morgan (1986) used system types to an explorative learning system. In this sense, it classify organizations. The aforementioned list of is enabled for innovation and evolution. This hierarchical systems principles and types needs can be described as an ‘evolutionary system’ further ground work to demonstrate its relevance (creativity, innovation, thriving). in different areas of application. •‘Organization’: If a system is able to control given goals and establish future goals that organize and integrate its behaviour and can include multiple Subdisciplines of Systems Science ‘ priorities in a ranking order, it becomes a goal- fi oriented system’ (goals, organization, priorities). If systems science is not a homogenous eld, what •‘Communication’: If a system can comm- subdisciplines belong to it and how do they relate unicate internal conditions, knowledge and to core functions and principles of systems? It fl goals to other systems and prompt them to should be possible to analyze how in uential vari- take action, it becomes a ‘communicative sys- ous traditions of systems science have been in tem’ (messages, interaction, communication). establishing or clarifying some key systemic princi- •‘Next-level boundary’: On the basis of commu- ples. There have been many historical overviews nication, a system can build longer-lasting on the development of systems science (e.g. connections to other systems so that a new François, 1999; Hammond, 2003; Schwaninger, communication-based unit and boundary 2009; Merali and Allen, 2011). Here, we will focus emerges. This leads to a ‘social system’ (relation- on the following traditions: thermodynamics, open ships, networks, alliances). We have thus systems theory, information theory, cybernetics, reached a higher and emergent boundary level. theory of autopoiesis, chaos theory, complexity theory, (multi-)agent modelling and network science. These traditions can be summarized as follows. This closes the circle of the described set of principles, and the evolving process can be repeated on • Classical thermodynamics treats closed systems in the next level: A group of ‘agents’ can build up an energetic equilibrium. Nevertheless, it is not their shared capability to collectively define designed to treat systems in non-equilibrium boundaries, take action, establish rules, monitor (for an overview, refer to, e.g. Atkins, 2010). inner and outer processes, store resources, develop • Open systems theory describes the necessity of an identity, adapt, innovate, organize itself and, as living systems being energetically open to the a group, communicate and collaborate with other environment (von Bertalanffy, 1950).

• Information theory treats the storage, compres- look at what kind of systems principles/concepts sion and transmission of data (Shannon and they manly focus on and how the whole field devel- Weaver, 1949). oped. The systems principles proposed earlier in • Cybernetics describes feedback processes for this section are used now as a classification scheme regulating systems (Wiener, 1948; Ashby, 1956). to organize subdisciplines of systems science. • The theory of autopoiesis clarifies how living Figure 6 provides a combination of the aforemen- systems reproduce and maintain themselves tioned subdisciplines and the set of proposed continually (Maturana and Varela, 1980). systems principles. Systemsprinciples(vertical • Chaos theory indicates the reasons for instability axis) are presented from bottom to top in a and nonlinear change processes (Mandelbrot, suggested hierarchical order from low to high 1983; Gleick, 1987). degrees of complexity. Subdisciplines of systems • Complexity theory describes processes of self- science (horizontal axis) are presented from left organization, adaptation and innovation to right in a rough chronological order. The time- (Kauffman, 1993; Holland, 1995; Kauffman, 1995). line begins with thermodynamics and then • (Multi-)agent modelling and the concept of autono- encompasses around 60 years of systems science. mous agents make it possible to formulate and The ‘correlations stars’ introduced here are based simulate processes of systems (agents) that act on initial suggestions of the author and discus- in a goal-oriented manner, e.g. humans and sions with system thinkers at two conferences. robots (Axelrod, 1997). Both the principles and the subdisciplines are • Network science, finally, is concerned with broad in their scope, and there exist many differ- the interaction of numerous actors, their ent views on how to define them, which can lead processpatternsanddynamicsocialstruc- to different results. If the concepts and subdisci- tures (Watts and Strogatz, 1998; Barabasi, plines are arranged as in Figure 6 then the distri- 2003; Watts, 2004). bution of the ‘correlation stars’ (*) indicate a visual trend line moving from the lower left to The similarities and differences among these the upper right corner. This leads to the follow- streams can be better understood if we take a closer ing two hypotheses.

Figure 6 Comparison chart displaying systems principles and subdisciplines of systems science. One to three stars (*) indicate a suggested correlation between the theoretical ﬁeld and the proposed systems principle.

Hypothesis 1: The development of systems change and intervene in systems? Over the past science, with its various subdisciplines, followed 60 years, many systemic methods have been de- a process in the sense of paradigm shifts and veloped to put systemic concepts and principles scientificrevolutions. into practice. The following section takes a closer look at methods used in areas close to manage- Hypothesis 2:Thedevelopmentofsystemssci- ment, problem-solving and design. Not treated ence has approached and clarified principles and here are other systemic methods such as those concepts with increasing levels of complexity. developed in the fields of medicine, therapy and other applied areas. The prevailing subdiscipline describes a particu- Simple tasks do not require major thinking or lar viewpoint and aspect of reality. With time, how- planning. Classical analytical problem solving ever, it reaches its limits in explaining or predicting works fine for isolated problems with pre-stated the anomalies of behaviour in the system in ques- goals. Systemic methods, on the other hand, are tion. In the sense of Thomas Kuhn’s paradigm shift especially helpful when many different stake- (1970), small scientific revolutions have occurred in holders interact in a dynamic complex setting, this respect: changes in perspectives, changes in the where there is no initial consensus on the prob- mental models and methods of the researchers. Of- lem definition, the expected future, or a shared ten, these shifts are also supported by the use of vision of what to reach. Terms such as ‘wicked more sophisticated and more efficient calculation problem’ or ‘mess’ express how traditional prob- methods and computer resources. The interpreta- lem solving has its limits. tion of the visual results in Figure 6 provides a When one enters the field of systems possible way of better understanding some contro- approaches, it becomes difficult to maintain an versies, debates or separation lines between subdis- overview and understand similarities and differ- ciplines. Systems science needs integrating forces to ences. Reynolds and Holwell (2010, p. 9) see overcome the historical borders in order to prog- strengths and weaknesses in the systems move- ress and fulfil its function within science, education ment: ‘In the systems field there is no shortage and practice. of approaches; it is diverse with many concepts, The aforementioned hierarchy of systems methodologies, methods and techniques. […] principles and types is in close agreement with We may well have inadvertently created a com- the argument of Troncale (2006, p. 317): ‘We plex clutter of systems approaches.’ Detailed concluded then that it was more reasonable to discussions of specificsystemicmethodsare expect a hierarchy of partial theories than to provided by Midgley (2000) and Jackson (2003), expect one overarching general theory.’ A for example. Recent practically oriented introduc- similar way of relating concepts with subdisci- tions can be found in in Reynolds and Holwell plines of systems science is proposed by Dent (2010) and Williams and Hummelbrunner (2011). and Umpleby (1998), who discuss eight ‘under- Ulrich and Probst (1995) as well as Gomez and lying assumptions’ in the light of six ‘systems Probst (1995) offer further insights into integrated science traditions’. and systemic problem-solving methods. Figure 7 shows a new way of visually inter- connecting and explaining systemic methods PART THREE: SYSTEMS APPROACHES TO and aspects of several methodologies. This over- CHANGE view is not exhaustive. The spiral in the middle of Figure 7 indicates that learning develops in cy- From a scientific viewpoint, general systems clic steps. The four stages in the illustration have principles are worthy of investigation, as science points in common with concepts derived from is always attempting to find the simplest and Kurt Lewin’s idea of action research (1946), such most general laws, principles and mechanisms as experiential learning (Kolb, 1983), reflective in explaining reality. But how do insights in sys- practice (Schön, 1983), appreciative inquiry temic knowledge also inform the way we design (Cooperrider and Srivastva, 1987) and process

Figure 7 Proposed process cycle involving four general steps to illustrate systemic aspects and methodologies

cycles used in other areas. The stages (act, ana- levels and timelines. Sustainable solutions for lyze, envision and plan) used for this framework complex problems need to build on the existing are explained in the following. resources and strengths while also building up new capabilities. The participation of involved actors in the planning process makes use of •‘Act/Experience/Intervene’: Complex situa- all existing knowledge, reduces the risk of tions are always unique, and therefore, direct resistance and increases the chance of a suc- experience of actors/observers from several cessful implementation that will be in accor- perspectives and on several levels is needed. dance with the main vision. However, •‘Analyze/Understand’: In order to understand complex situations can change fast so that the causal relationships within complex situa- new direct experiences and adjustments be- tions, it is often necessary to look at dynamic come necessary. behaviour over time, series of events, patterns of behaviour and underlying structures (e.g. When dealing with complexity, such a process feedback loops). normally does not end after one cycle but con- •‘Envision/Design’: Complex situations of- tinues iteratively: The defined boundaries might tencannotbesolvedwithpredefined solu- be reduced or enlarged, and the values and tions from the past. It therefore becomes expectations attached to the result might change necessary to take diverse values and stake- as well. holders into account to detect the opportu- Six specific system methodologies have been nities and threats, define a shared vision, mapped onto the illustration (dotted circles in find ideas and select feasible solutions. Figure 7). Each place represents one major strength •‘Plan/Organize’: To plan successfully, it is nec- of each methodology, although all of them are essary to consider interdependencies on various much broader in their coverage. The precise

placing is debatable. Although a methodology is Another way of mapping systemic methodolo- mapped in one corner of the framework, it gies within a process of several steps is provided can nevertheless be used to go through the by Mingers (1997). The timescale to run through whole process cycle without having to make a full cycle can vary greatly. Three examples will use of other methodologies. The following illustrate this. A recent approach that combines is a short description of these six systems several powerful methodologies within a full methodologies. learning cycle is the systems-based ‘evolutionary learning laboratory’ (ELLab) (Bosch et al., 2013). In an ELLab, ‘adiversegroupofparticipants • Critical systems heuristics (Ulrich, 1983) in- engage in a cyclical process of thinking, plan- volves detailed considerations about how to ning, action and reflection for collective learning draw the boundaries when considering towards a common good’ (Bosch et al., 2013, p. systems. The question about what is relevant 118). It has been successfully applied in fields and what is less important involves values such as sustainable development or the design and facts and can have strong political and and improvement of educational programs. The ethical implications. • timescale of an ELLab might be weeks, months System dynamics (Forrester, 1971; Sterman, 2000) or years. The methodology as described in is an elaborated qualitative and quantitative Figure7buildsonexistingtheoriesandexperi- method for understanding, modelling and ence. It has been applied by the author in the simulating dynamic systems. fields of coaching, team development and facil- • Soft systems methodology (Checkland, 1981) itating change. In this case, the timescale to helps to explore multiple perspectives, reach run with participants through the full cycle accommodations between those perspectives can range from a couple of hours to days or fi and de ne action plans that are systemically weeks, depending on the addressed issues. It desirable and culturally feasible. is not the intention to claim in Figure 7 that • Interactive planning (Ackoff, 1981) is a systemic thinking always needs to explicitly methodology that involves idealized design involve all mentioned aspects or be purely se- to define a far-sighted but still actionable quential; many systemic aspects can be found plan. in parallel processes of thinking in action by • Optimization techniques in the tradition of opera- professionals in complex situations that may tional research (OR; Churchman et al., 1957) are just last minutes or seconds. Dancers coodinate a set of methods for improved decision making their movements with the movement of the and efficiency. whole group in real time; a saxophonist coor- • Reflective practice (Schön, 1983) emphasizes dinates his or her solo with the jazz band and the process of continuous and deep learn- the clapping of the audience; a firefighter coor- ing. Good practice requires reflection, and dinates his or her actions with colleagues good learning requires experience. In com- within a collapsing building. All these profes- plex systems, it is nearly impossible to sionals embrace all four quadrants mentioned achieve perfect prediction and error-free here (act, understand, envision and plan) plans. It thus becomes crucial to learn parallel to each other and in real time: acting, through direct interaction with the respec- reflecting on interactions, seeing emerging tive system. possibilities, seeing the situation through the eyes of others, balancing one’sownvalues Every methodology has its own strengths and and those of others, balancing short-term and weaknesses. In the past 30 years, several authors long-term outcomes, finding feasible solutions have presented integrative, multimethodological and empowering others. Such complex multi- frameworks to compare and combine several faceted behaviour has similarities with what systemic methodologies, such as Hall (1989), Flood Hämäläinen and Saarinen (2004) call ‘systems and Jackson (1991) and Schwaninger (2004). intelligent behavior’.

© 2013 The Author. Systems Research and Behavioral Science Syst. Res. 30, 580–595 (2013) published by John Wiley & Sons, Ltd. DOI: 10.1002/sres Understanding Systems Science 591 RESEARCH PAPER Syst. Res.

CONCLUSION

When we are confronted with the many perspectives, concepts, principles and methods in the systems field, there is a danger of not seeing the forest for the trees. Visual maps help us to integrate knowledge and establish relations between the concepts. This paper has attempted to bring together a wide variety of perspectives and concepts to underpin three aspects of systems science: supporting interdisciplinarity and transdisciplinarity, exploring and formalizing systems con- Figure 8 Mutual influence of four components in the sys- cepts and developing systemic methods for tems field (adapted from Flood and Carson, 1993, p. 4) learning and change. Figure 8 is related to Figure 1 and shows that a desirable cooperation between different compo- fi nents of the systems eld leads to a mutual forces together. Work towards coherent strengthening. Similar descriptions of compo- theories of systems lead to synergetic effects ’ nents and interactions are given in M Pherson and strengthen the function of systems (1974) as well as Flood and Carson (1993). science within the landscape of the sciences. Isolated development of any one of these compo- 3. Systemic methods show many varieties in nents will probably have limitations. A steady detail and points of focus, they provide rich po- fl information ow between these components is tential for dealing with complexity and change. necessary to strengthen the co-evolution of Stronger connections between different schools theories, methods and applications. Further of thought facilitate the use of methodologies investigations could clarify which of these pro- in educational and applied settings. posed linkages are strong, which are weak and how they could be improved. These statements support the idea of an inte- The figures and ideas presented in this paper grated pluralism that appreciates both diver- are works in progress. Other representations sity and unity. Many real-life situations might be appropriate from other perspectives. require holistic solutions that involve working The paper offers initial arguments to support across disciplines, principles and methods. At the following statements: present, some of the older separation lines are disappearing, and the systems field is gaining 1. Researchers and practitioners in interdisciplin- momentum as a whole. fi ary and transdisciplinary elds can make use Figure 9 is related to a quote by Peter Senge. of systems concepts and methods to facilitate The quote agrees very well with the aforemen- practice. Systems science and systems design tioned statements and describes the importance provide a bridge between science and the of making systems knowledge known to a wider humanities, as well as between descriptive audience: research and normative practice. This can improve mutual understanding and enhance A real change is grounded in new ways of communication. thinking and perceiving. […] With nature and 2. The subdisciplines of systems science can be not machines as their inspiration, today’sinno- viewed as different perspectives on a set of vators are showing how to create a different general systems principles, thus forming a future by learning how to see the larger system unity through their interlinked diversity. of which they are a part and to foster collabora- Systems science seems to be an appropriate tion across every imaginable boundary. These name and framework in bringing these core capabilities—seeing systems, collaborating

REFERENCES

Ackoff RL. 1960. Systems, organizations and interdisciplinary research. General Systems Yearbook 5:1–8. Ackoff RL. 1971. Towards a system of systems concepts. Management Science 17(11): 661–671. Ackoff RL. 1981. Creating the Corporate Future: Plan or be Planned for. John Wiley: New York. Ashby WR. 1956. An Introduction to Cybernetics.Chapman & Hall: London. – Figure 9 Three capabilities for systemic change. (Adapted Atkins P. 2010. The Laws of Thermodynamics A Very from Senge et al., 2010, p. 45) Short Introduction. Oxford University Press: Oxford. Axelrod R. 1997. The Complexity of Cooperation. Agent- Based Models of Competition and Collaboration.Princeton University Press: Princeton. across boundaries and creating versus problem Axelrod R, Cohen MD. 2000. Harnessing Complexity: solving—form the underpinnings, and ulti- Organizational Implications of a Scientific Frontier. Basic mately the tool and methods, for this shift in Books: New York. – Banathy BH. 1996. Designing Social Systems in a Changing thinking. Senge et al.(2010,p.10 11) World. Plenum Press: New York. Barabasi AL. 2003. Linked. Plume: New York. von Bertalanffy L. 1950. The theory of open systems in physics and biology. Science 111:23–29. von Bertalanffy L. 1956. General system theory. General ACKNOWLEDGEMENTS AND REMARKS Systems 1:1–10. Börner K. 2010. Atlas of Science – Visualizing What We This paper reflects the author’slearningjourney Know. MIT Press: Cambridge, MA. through the diverse and rich systems landscape. Bosch OJH, Nguyen NC, Maeno T, Yasui T. 2013. My sincere thanks go to all the people discussing Managing complex issues through evolutionary learning laboratories. Systems Research and Behavioral these issues in many places. Initial insights were Science 30(2): 116–135. obtained at three interdisciplinary Cortona- Boulding K. 1956. General system theory – the skeleton Conferences conducted by ETH Zurich, followed of science. General Systems 1:11–17. by a research visit at the Santa Fe Institute, and Braitenberg V. 1984. Vehicles: Experiments in Synthetic several conferences and seminars on systems-re- Psychology. MIT Press: Cambridge, MA. Castellani B, Hafferty FW. 2009. Sociology and lated topics. This paper is based on earlier versions Complexity Science: A New Field of Inquiry. that were presented at two conferences: European Springer: Berlin. http://www.art-sciencefactory. Meetings on Cybernetics and Systems Research, com/complexity-map_feb09.html [last accessed EMCSR, Vienna, Austria, April 2012; annual meet- 1/2/2013]. ing of the International Society for the Systems Sci- Checkland P. 1981. Systems Thinking, Systems Practice. John Wiley: Chichester. ences, ISSS, San José, USA, July 2012. The author is Churchman CW, Ackoff RL, Arnoff EL. 1957. Introduc- particularly grateful to Prof. Stuart Kauffman for tion to Operations Research. John Wiley: New York. clarifying aspects of complexity and autonomous Commons ML, Trudeau EJ, Stein SA, Richards FA, agents, to Prof. Markus Schwaninger for clarifying Krause SR. 1998. Hierarchical complexity of tasks aspects of cybernetics and system dynamics and to shows the existence of developmental stages. Devel- opmental Review 18(3): 237–278. Prof. Peter Checkland and Prof. Werner Ulrich for Cooperrider DL, Srivastva S. 1987. Appreciative their introduction to systems design. inquiry in organizational life. Research in Organiza- Given the breadth of the topics addressed, it is tional Change and Development 11:129–169. clear that there is much room for delving more Dent EB, Umpleby SA. 1998. Underlying assumptions of several traditions in systems theory and cybernet- deeply into the subject matter. The website www. ’ understandingsystems.net provides links to some ics. In Cybernetics and Systems 98, Trappl R. (ed.). Austrian Society for Cybernetic Studies: Vienna; of the papers, concepts and systems thinkers men- 513–518. tioned in this paper. Feedback and suggestions for Elitzur A. 2005. When form outlasts its medium: a improvement and applications are most welcome. definition of life integrating platonism and

thermodynamics. In Life as We Know It, Seckbach J. Kuhn T. 1970. The Structure of Scientific Revolutions. (ed.). Kluwer Academic Publishers: Dordrecht; University of Chicago Press: Chicago. 607–620. Lewin K. 1946. Action research and minority problems. Fischer K. 1980. A theory of cognitive development: Journal of Social Issues 24:34–46. The control and construction of hierarchies of skills. Lin Y, Duan X, Zhao C, Xu L. 2012. Systems Science: Psychological Review 87(6): 477–531. Methodological Approaches. CRC Press: New York. Flood RL, Carson ER. 1993. Dealing with Complexity: An Luke DA, Stamatakis KA. 2012. Systems science Introduction to the Theory and Application of Systems methods in public health: dynamics, networks, and Science. Plenum: New York. agents. Annual review of public health 33: 357–376. Flood RL, Jackson M. 1991. Creative Problem Solving: Maglio P, Kieliszewski C, Spohrer J. 2010. Handbook of Total Systems Intervention. John Wiley: Chichester. Service Science. Springer: New York. Forrester JW. 1971. World Dynamics. Wright & Allen: Mandelbrot B. 1983. The Fractal Geometry of Nature. Cambridge MA. Freeman: New York. François C. 1999. Systemics and cybernetics in a histori- Martinelli D. 2001. Systems hierarchies and manage- cal perspective. Systems Research and Behavioral Science ment. Systems Research and Behavioral Science 18:69–82. 16:203–219. Maturana H, Varela F. 1980. Autopoiesis and Cognition: Gomez P, Probst G. 1995. Die Praxis des ganzheitlichen The Realization of the Living. D Reidel: Boston. Problemlösens. Haupt: Bern. Max-Neef MA. 2005. Foundations of transdisciplinarity. Gleick J. 1987. Chaos: Making a New Science.Abacus: Ecological Economics 53(1): 5–16. London. Meadows DH. 2008. Thinking in Systems – A primer. Hall AD. 1989. Metasystems Methodology. A New Synthesis Earthscan: London. and Unification. Pergamon Press: Oxford. Merali Y, Allen P. 2011. Complexity and systems think- Hall AD, Fagen RE. 1956. Definition of system. General ing. In The SAGE Handbook of Complexity and Manage- System 1:18–28. ment, Allen P, McGuire S, McKelvey B. (eds.). Sage: Hämäläinen RP, Saarinen E. 2004. Systems intelligence: London; 31–52. connecting engineer thinking with human sensitivity. Midgley G. 2000. Systemic Intervention: Philosophy, In Systems Intelligence, Hämäläinen RP, Saarinen E. Methodology, and Practice. Kluwer/Plenum: New York. (eds.). Systems Analysis Laboratory, Helsinki Univer- Miller JG. 1978. Living Systems. McGraw-Hill: New York. sity of Technology: Espoo; 9–38. Mingers J. 1995. Self-Producing Systems. Implications and Hammond D. 2003. The Science of Synthesis. Exploring Applications of Autopoiesis. Plenum: New York. the Social Implications of General Systems Theory. Mingers J. 1997. Towards critical pluralism. In Multime- University Press of Colorado: Boulder. thodology: Theory and Practice of Combining Management Heylighen F. 1990. Classical and non-classical repre- Science Methodologies, Mingers J, Gill A. (eds.). sentations in physics I. Cybernetics and Systems 21: Chichester: John Wiley; 407–440. 423–444. Morgan G. 1986. Images of Organization. Sage: Beverly Holland JH. 1995. Hidden Order: How Adaptation Builds Hills, CA. Complexity. Addison-Wesley: Reading MA. M’Pherson PK. 1974. A perspective on systems science Holland JH. 2006. Studying complex adaptive systems. and systems philosophy. Futures 6: 219–239. Journal of Systems Science and Complexity 191:1–8. Müller KH. 2011. The New Science of Cybernetics. The Ison RL, Maiteny PT, Carr S. 1997. Systems methodolo- Evolution of Living Research Designs. Vol. II Theory. gies for sustainable natural resources research and Edition Echoraum: Wien. development. Agricultural Systems 552:257–272. Ostrom E. 1990. Governing the Commons: The Evolution Jackson MC. 2003. Systems Thinking: Creative Holism for of Institutions for Collective Action. Cambridge Managers. John Wiley: Chichester. University Press: New York. Katz D, Kahn R. 1978. The Social Psychology of Organiza- Pfeifer R, Scheier C. 2001. Understanding Intelligence. tions. John Wiley: New York. MIT Press: Cambridge, MA. Kauffman S. 1993. Origins of Order: Self-Organization and Piaget J. 1971. The theory of stages in cognitive devel- Selection in Evolution. Oxford University Press: Oxford. opment. In Measurement and Piaget, Green DR (ed). Kauffman S. 1995. At Home in the Universe. Oxford McGraw-Hill: New York; 1–11. University Press: Oxford. Pyster A, Olwell DH (eds.). 2013. The Guide to the Sys- Klir GJ. 1965. The general system as a methodological tems Engineering Body of Knowledge (SEBoK). The tool. General Systems 10:29–42. Trustees of the Stevens Institute of Technology. Klir GJ. 2001. Facets of Systems Science.Kluwer:NewYork. Hoboken, NJ. www.sebokwiki.org [last accessed 1/ Kolb DA. 1983. Experiential Learning: Experience as the 2/2013]. Source of Learning and Development. Prentice Hall Ramage M, Shipp K. 2009. Systems Thinkers. Springer: International: Englewood Cliffs. Berlin. Koshland DE. 2002. The seven pillars of life. Science Rapoport A. 1986. General System Theory. Abacus: 295: 2215–2216. Tunbridge Wells, Kent.

Reynolds M, Holwell S (eds.). 2010. Systems Approaches to Troncale L. 1985. The future of general systems Managing Change: A Practical Guide. Springer: London. research: obstacles, potentials, case studies. Systems Russell SJ, Norvig, P. 2003. Artificial Intelligence: A Research 2:43–84. Modern Approach. Prentice Hall: New Jersey. Troncale L. 2006. Towards a science of systems. Sayama H. 2012. Complex systems organizational map. Systems Research and Behavioral Science 233: http://www.sandia.gov/CasosEngineering/images/ 301–321. Sayama_Complex_systems_organizational_map.png Ulrich H, Probst G. 1995. Anleitung zum ganzheitlichen [last accessed 1/2/2013]. Denken und Handeln. Ein Brevier für Führungskräfte. Schön DA. 1983. The Reflective Practitioner: How Profes- Haupt: Bern. sionals Think in Action.TempleSmith:London. Ulrich W. 1983. Critical Heuristics of Social Schwaninger M. 2004. Methodologies in conflict: achiev- Planning: A new Approach to Practical Philosophy. ing synergies between system dynamics and organi- Haupt: Bern. Reprint edition 1995. John Wiley, zational cybernetics. Systems Research and Behavioral Chichester. Science 21:1–21. Warfield J. 2003. A proposal for systems science. Schwaninger M. 2009. System dynamics in the evolution Systems Research and Behavioral Science 2: 507–520. of the systems approach. In Encyclopedia for Complexity Watts DJ. 2004. Six Degrees. Vintage: London. and Systems Science, 9, Meyers RA. (ed.). Springer: Watts DJ, Strogatz SH. 1998. Collective dynamics of New York; 8974–8987. ‘small world’ networks. Nature 393: 440–442. Senge P. 1990. The Fifth Discipline. The Art and Practice Weinstein MP, Turner ER (eds). 2012. Sustainability Sci- of the Learning Organization. Doubleday/Currency: ence: The Emerging Paradigm and the Urban Environ- New York. ment. Springer: New York. Senge P, Smith B, Kruschwitz N, Laur J, Schley S. 2010. Wiek A, Withycombe L, Redman CL. 2011. Key compe- The Necessary Revolution: Working Together to Create a tencies in sustainability – a reference framework for Sustainable World.BroadwayBooks:NewYork. academic program development. Sustainability Science Shannon CE, Weaver W. 1949. The Mathematical Theory of 62:203–218. Communication. University of Illinois Press: Urbana, IL. Wiener N. 1948. Cybernetics or Control and Communica- Sterman J. 2000. Business Dynamics. Irwin-McGraw- tion in the Animal and the Machine. MIT Press: Hill: New York. Cambridge, MA. Strijbos S. 2010. Systems thinking. In The Oxford Handbook Williams B, Hummelbrunner R. 2011. Systems Concepts of Interdisciplinarity,FrodemanR,ThompsonKJ, in Action: A Practitioners Toolkit. Stanford University Mitcham C. (eds.). Oxford University: Oxford; 453–470. Press: Stanford, CA.