Ashby s law of requi- System Dynamics in the Evolution of ...

System Dynamics in theEvolution of the SystemsApproach

Markus SchwaningerInstitute of Management and Strategy, Universityof St. Gallen, St. Gallen, Switzerland

Article Outline

GlossaryDefinition of the SubjectIntroductionEmergence of the Systems ApproachCommon Grounds and DifferencesThe Variety of Systems MethodologiesSystem Dynamics: Its Features, Strengths, and

LimitationsStrengths of SDLimitations of SDActual and Potential RelationshipsOutlookAppendix – Milestones in the Evolution of the

Systems Approach in General and SystemDynamics in Particular

Bibliography

Glossary

Cybernetics The science of communication andcontrol in complex, dynamic systems. The coreobjects of study are information, communica-tion, feedback, adaptation, and control or gov-ernance. In the newer versions of cybernetics,the emphasis is on observation, self-organization, self-reference, and learning.

Dynamical system The dynamical system con-cept is a mathematical formalization of time-dependent processes. Examples include themathematical models that describe the swing-ing of a clock pendulum, the flow of water in ariver, or the evolution of a population of fish ina lake.

Law of requisite variety Ashby’s law of requi-site variety says “Only variety can destroyvariety.” It implies that the varieties of twointeracting systems must be in balance, if sta-bility is to be achieved.

Organizational cybernetics The science whichapplies cybernetic principles to organization.Synonyms are management cybernetics andmanagerial cybernetics.

System dynamics (SD) A methodology and dis-cipline for the modeling, simulation, and con-trol of dynamic systems. The main emphasisfalls on the role of structure and its relationshipwith the dynamic behavior of systems, whichare modeled as networks of informationallyclosed feedback loops comprising stock andflow variables.

System There are many definitions of system.Two examples: A portion of the world suffi-ciently well defined to be the subject of study;something characterized by a structure, forexample, a social system (Anatol Rapoport).A system is a family of relationships betweenits members acting as a whole (InternationalSociety for the Systems Sciences).

Systems approach A perspective of inquiry,education, and management, which is basedon systems theory and cybernetics.

Systems theory A formal science of the struc-ture, behavior, and development of systems. Infact there are different systems theories. Gen-eral systems theory is a transdisciplinaryframework for the description and analysis ofany kind of system. Systems theories havebeen developed in many domains, e.g., mathe-matics, computer science, engineering, sociol-ogy, psychotherapy, biology, and ecology.

Variety A technical term for complexity whichdenotes the amount of (potential or actual)states or modes of behavior of a system. Rep-ertoire of behavior is a synonym.

© Springer Science+Business Media LLC 2019R. A. Meyers (ed.), Encyclopedia of Complexity and Systems Science,https://doi.org/10.1007/978-3-642-27737-5_537-5

1

http://crossmark.crossref.org/dialog/?doi=10.1007/978-3-642-27737-5_537-5&domain=pdf

https://doi.org/10.1007/978-3-642-27737-5_537-5

Definition of the Subject

The purpose of this chapter is to give an overviewof the role of system dynamics (SD) in the contextof the evolution of the systems movement. This isnecessary because SD is often erroneously takenas the systems approach as such, not as part of it. Itis also requisite to show that the processes of theevolution of both the systems movement as awhole and SD in particular are intimately linkedand intertwined. Finally, in view of the purpose ofthe chapter, the actual and potential relationshipsbetween system dynamics and the other strands ofthe systems movement are evaluated. This way,complementarities and potential synergies areidentified.

Introduction

The purpose of this contribution is to give anoverview of the role of system dynamics in thecontext of the evolution of the systems movement.“Systems movement” – often referred to briefly as“systemics” also cyber systemics – is a broadterm, which takes into account the fact that thereis no single systems approach but a range ofdifferent ones. The diagram in Fig. 1 is not acomplete representation but the result of anattempt to map the major threads of the systemsmovement and some of their interrelations.Hence, the schema does not cover all schools orprotagonists of the movement. Why does the dia-gram show a dynamic and evolutionary systemsthread and an additional cybernetics thread, ifcybernetics is about dynamic systems? The latterembraces all the approaches that are explicitlygrounded in cybernetics. The former relates toall other approaches concerned with dynamic orevolutionary systems. The streamlining made itnecessary to somewhat curtail logical perfectionfor the sake of conveying a synoptic view of thedifferent systems approaches, in a language thatuses the categories common in current scientificand professional discourse. Overlaps exist, e.g.,between dynamic systems and chaos theory, cel-lular automata, and agent-based modeling, etc.

The common denominator of the different sys-tems approaches in our day is that they share a

worldview focused on complex dynamic systemsand an interest in describing, explaining, anddesigning or at least influencing them. Therefore,most of the systems approaches offer not only atheory but also a way of thinking (“systems think-ing” or “systemic thinking”) and a methodologyfor dealing with systemic issues or problems.

System dynamics (SD) is a discipline and amethodology for the modeling, simulation, andcontrol of complex, dynamic systems. SD wasdeveloped by MIT professor Jay W. Forrester(e.g., Forrester 1961, 1968) and has been propa-gated by his students and associates. The terms“methodology” and “method” should not be usedinterchangeably (Jackson 2003). Our workingdefinition for “methodology” embraces twokinds of use: (a) a set of principles that frame theway in which research (theoretical, empirical,practical) is to be undertaken and (b) a coherentgroup of complementary methods used in a givenmethodology. The term “method” derives fromAncient Greek: ὁdó§ (hodós) for path and m«tᾰ́(metá) for in pursuit of. The working definition weare using for “method” refers to a technique orway toward a purpose or serving a specific para-digm. “Paradigm” is a mode of thinking or a set ofassumptions shared by a (scientific) community.SD has grown to a school of numerous academicsand practitioners all over the world. The particularapproach of SD lies in representing the issues orsystems in focus as meshes of closed feedbackloops made up of stocks and flows, in continuoustime, and subject to delays. Given these features,SD modeling is widely applicable to all kinds ofsystems – economic, social, ecological, etc. – andthe construction of generic models is facilitated.

The SD view is focused on continuous asopposed to discrete processes. This makes itextraordinarily powerful in dealing with manage-rial issues. Other modeling and simulationapproaches exist. One is the discrete eventapproach, which is focused on the modeling ofsystems conceived as queuing networks(Brailsford 2014). It is a widely used techniqueof operations research, particularly for fine-grained modeling, simulation, and control, in thedomains of logistics, production, distribution sys-tems, and transportation networks. SD is normallyused for modeling and simulation at relatively

2 System Dynamics in the Evolution of the Systems Approach

System

Dyn

amicsin

theEv

olutionof

theSy

stem

sApproach,F

ig.1

Systemsapproaches:A

nov

erview

System Dynamics in the Evolution of the Systems Approach 3

higher levels of aggregation. Its perspective is top-down, which makes broad and long-term issuestangible and manageable. As a complement, amethodology has emerged which follows abottom-up logic: the agent-based object approach(Holland 1995; Miller and Page 2007). Agent-based models are in discrete time. They representsets of individual agents, whose interactions aresimulated. Hence, the emergent behavior atmacrolevel is generated from the bottom, by theindividual behaviors at the microlevel. The threemethodologies are complementary. Hybrid formsof modeling have been developed, which buildbridges and enable synergies between the respec-tive strengths of these different methodologies(Brailsford et al. 2014).

The development of the system dynamics meth-odology and the worldwide community that appliesSD to modeling and simulation in highly variedcontexts suggest that it is a “systems approach” onits own. Nevertheless, taking “system dynamics” asthe (one and only) synonym for “systemic thinking”would be going too far, given the other approachesto systemic thinking as well as a variety of systemstheories and methodologies, many of which arecomplementary to SD. In any case, however, theSD community has become the strongest “school”of the systems approach, if one takes the numbers ofmembers in organizations representing the differentschools as a measure (by early 2019, the SystemDynamics Society has 1150 members).

The rationale and structure of this contributionare as follows. Starting with the emergence of thesystems approach, the multiple roots and theoreticalstreams of systemics are outlined. Next, the com-mon grounds and differences among differentstrands of the systems approach are highlighted,and the various systems methodologies areexplored. Then the distinctive features of SD areanalyzed. Finally comes a reflection on the relation-ships of SD with the rest of the systems movementand on potential complementarities and synergies.

In the Appendix, a timeline overview of somemilestones in the evolution of the systemsapproach in general and system dynamics in par-ticular is given. Elaborating on each of the sourcesquoted therein would reach beyond the purpose ofthis chapter. However, a synoptic view of the

systems approaches has been provided in the dia-gram in Fig. 1.

Emergence of the Systems Approach

The systems movement has many roots and facets,with some of its concepts going back as far asancient Greece. What we name as “the systemsapproach” today materialized in the first half ofthe twentieth century. At least two important com-ponents should be mentioned: those proposed byvon Bertalanffy and by Wiener.

Ludwig von Bertalanffy, an American biolo-gist of Austrian origin, developed the idea thatorganized wholes of any kind should be describ-able and, to a certain extent, explainable, bymeans of the same categories and ultimately byone and the same formal apparatus. His GeneralSystems Theory (Von Bertalanffy 1950, 1968)triggered a whole movement which has tried toidentify invariant structures and “mechanisms”across different kinds of organized wholes (e.g.,hierarchy, teleology, purposefulness, differentia-tion, morphogenesis, stability, ultrastability, emer-gence, and evolution).

In 1948 Norbert Wiener, an American mathema-tician at the Massachusetts Institute of Technology,published his seminal book onCybernetics, buildingupon interdisciplinary work carried out in coopera-tion with Bigelow, an IBM engineer, andRosenblueth, a physiologist. Wiener’s opus becamethe transdisciplinary foundation for a new science ofcapturing as well as designing control and commu-nication mechanisms in all kinds of dynamic sys-tems (Wiener 1948). Cyberneticists have beeninterested in concepts such as information, commu-nication, complexity, autonomy, interdependence,cooperation and conflict, self-production (“auto-poiesis”), self-organization, (self-) control, self-reference, and (self-) transformation of complexdynamic systems.

Along the genetic line of the tradition whichled to the evolution of general systems theory (vonBertalanffy, Boulding, Gerard, Miller, Rapoport)and cybernetics (Wiener, McCulloch, Ashby,Powers, Pask, Beer), a number of roots can beidentified, in particular:


• Mathematics (e.g., Newton, Poincaré, Lyapunov,Lotka, Volterra, Rashevsky).

• Logic (e.g., Epimenides, Leibniz, Boole, Rus-sell and Whitehead, Goedel, Spencer-Brown).

• Biology, including general physiology andneurophysiology (e.g., Hippocrates, Cannon,Rosenblueth, McCulloch, Rosen).

• Engineering and computer science, includingthe respective physical and mathematical foun-dations (e.g., Heron, Kepler, Watt, Euler, Fou-rier, Maxwell, Hertz, Turing, Shannon andWeaver, von Neumann, Walsh).

• Social and human sciences, including econom-ics (e.g., Hume, Adam Smith, Adam Ferguson,John Stuart Mill, Dewey, Bateson, Merton,Simon, Piaget).

In this last-mentioned strand of the systemsmovement, one focus of inquiry is on the role offeedback in communication and control in (andbetween) organizations and society, as well as intechnical systems. The other focus of interest is onthe multidimensional nature and the multilevelstructures of complex systems. Specific theorybuilding, methodological developments, and per-tinent applications have occurred at the followinglevels:

• Individual and family levels (e.g., systemicpsychotherapy, family therapy, holistic medi-cine, cognitive therapy, reality therapy)

• Organizational and societal levels (e.g., mana-gerial cybernetics, organizational cybernetics,sociocybernetics, social systems design, socialecology, learning organizations)

• The level of complex (socio-)technical systems(systems engineering).

The notion of “socio-technical systems” hasbecome widely used in the context of the designof organized wholes involving interactions ofpeople and technology (for instance, Linstone’smultiperspectives framework, known by way ofthe mnemonic TOP (technical, organizational,personal/individual)).

As can be noted from these preliminaries, dif-ferent kinds of systems theory and methodology

have evolved over time. One of these is a theory ofdynamic systems by Jay W. Forrester, whichserves as a basis for the methodology of systemdynamics. Two eminent titles are Forrester (1961)and (1968). In SD, the main emphasis falls on therole of structure and its relationship with thedynamic behavior of systems, modeled as net-works of informationally closed feedback loopsbetween stock and flow variables. Several othermathematical systems theories have been elabo-rated, for example, mathematical general systemstheory (Klir, Pestel, Mesarovic, and Takahara), aswell as a whole stream of theoretical develop-ments which can be subsumed under the terms“dynamic systems theory” or “theories of non-linear dynamics” (e.g., catastrophe theory, chaostheory, and complexity theory). Under the latter,branches such as the theory of fractals(Mandelbrot), geometry of behavior (Abraham),self-organized criticality (Bak), and network the-ory (Barabasi, Watts) are subsumed. In this con-text, the term “sciences of complexity” is used.

In addition, a number of mathematical theories,which can be called “systems theories,” haveemerged in different application contexts, exam-ples of which are discernible in the followingfields:

• Engineering, namely, information and commu-nication theory (Shannon and Weaver), tech-nology and computer-aided systems theory(e.g., control theory, automata, cellular autom-ata, agent-based modeling, artificial intelli-gence, cybernetic machines, neural nets,machine learning)

• Operations research (e.g., modeling theory andsimulation methodologies, Markov chains,genetic algorithms, fuzzy control, orthogonalsets, rough sets)

• Social sciences, economics in particular (e.g.,game theory, decision theory)

• Biology (e.g., Sabelli’s Bios theory of creation)• Ecology (e.g., E. and H. Odum’s systems

ecology)

Most of these theories are transdisciplinary innature, i.e., they can be applied across disciplines.


The Bios theory, for example, is applicable toclinical, social, ecological, and personal settings(Sabelli 2005). Examples of essentially non-mathematical systems theories can be found inmany different areas of study:

• Economics, namely, its institutional/evolution-ist strand (Veblen, Myrdal, Boulding, Dopfer)

• Sociology (e.g., Parsons’ and Luhmann’ssocial systems theories, Hall’s cultural systemstheory)

• Political sciences (e.g., Easton, Deutsch,Wallerstein)

• Anthropology (e.g., Levi Strauss’s structuralist-functionalist anthropology, Margaret Mead)

• Semiotics (e.g., general semantics (Korzybski,Hayakawa, Rapoport), cybersemiotics (Brier))

• Psychology and psychotherapy (e.g., systemicintervention (Bateson, Watzlawick, F. Simon)and fractal affect logic (Ciompi))

• Ethics and epistemology (e.g., Vickers,Churchman, von Foerster, van Gigch).

Several system-theoretic contributions havemerged the quantitative and the qualitative in newways. This is the case, for example, in Rapoport’sworks in game theory as well as general systemstheory, Pask’s conversation theory, von Foerster’sCybernetics of Cybernetics (second-order cyber-netics), and Stafford Beer’s opus in managerialcybernetics. In all four cases, mathematical expres-sion is virtuously connected to ethical, philosoph-ical, and epistemological reflection. Furtherexamples are Prigogine’s theory of dissipativestructures, Mandelbrot’s theory of fractals, com-plex adaptive systems (Holland et al.), Kauffman’scomplexity theory, and Haken’s synergetics, all ofwhich combine mathematical analysis and a strongcomponent of qualitative interpretation.

A large number of systems methodologies,with the pertinent threads of systems practice,have emanated from these theoretical develop-ments. Many of them are expounded in detail inspecialized publications (e.g., François 2004;Jackson 2019) and, under a specific theme,named Systems Science and Cybernetics ofthe Encyclopedia of Life Support Systems

(Encyclopedia of Life Support Systems 2002). Inthis chapter, only some of these will be addressedexplicitly, in order to shed light on the role of SDas part of the systems movement.

Common Grounds and Differences

Even though the spectrum of systems theories andmethodologies outlined in the preceding sectionmay seem multifarious, all of them have a strongcommon denominator: They build on the idea ofsystems as organized wholes. An objectivistworking definition of a system is that of a whole,the organization of which is made up by interre-lationships. A subjectivist definition is that of a setof interdependent variables in the mind of anobserver, or, a mental construct of a whole, anaspect that has been emphasized by the positionof constructivism. Constructivism is a synonymfor second-order cybernetics. While first-ordercybernetics concentrates on regulation, informa-tion, and feedback, second-order cyberneticsfocuses on observation, self-organization, andself-reference. Heinz von Foerster establishedthe distinction between “observed systems” forthe former and “observing systems” for the latter(Von Foerster 1984).

From the standpoint of operational philosophy,a system is, as Rapoport says, “a part of the world,which is sufficiently well defined to be the objectof an inquiry or also something, which is charac-terized by a structure, for example, a productionsystem” (Rapoport 1953).

In recent systems theory, the aspect of relation-ships has been emphasized as the main buildingblock of a system, as one can see from a definitionpublished by the International Society for theSystems Sciences (ISSS), “A system is a familyof relationships between its members acting as awhole” (Shapiro et al. 1996). Also, purpose andinteraction have played an important part in reflec-tions on systems. Systems are conceived, in thewords of Forrester (1968), as “wholes of ele-ments, which cooperate towards a commongoal.” Purposeful behavior is driven by internalgoals, while purposive behavior rests on a func-tion assigned from the outside. Finally, the aspects


of open and closed functioning have been empha-sized. Open systems are characterized by theimport and export of matter, energy, and informa-tion. Avariant of particular relevance in the case ofsocial systems is the operationally closed system,that is, a system which is self-referential in thesense that its self-production (autopoiesis) is a func-tion of production rules and processes by whichorder and identity are maintained, which cannot bemodified directly fromoutside.Aswe shall see, thisconcept of operational closure is very much in linewith the concept of circularity used in SD.

Another generic concept is that of complexadaptive system (CAS). The CAS is defined as asystem that has a large number of components,often called agents, which interact and adapt orlearn (John Holland 2006). It has features such asself-similarity (adaptation takes place at bothlevels – component and whole), complexity,emergence, and self-organization. The behaviorof the whole system cannot be simply derivedfrom the actions of its components; it emergesfrom their interaction.

At this point, it is worth elaborating on thespecific differences between two major threadsof the systems movement, which are of specialinterest because they are grounded in “feedbackthought” (Richardson 1999): the cyberneticthread, from which organizational cyberneticshas emanated, and the servomechanic thread inwhich SD is grounded. As Richardson’s detailedstudy shows, the strongest influence on cybernet-ics came from biologists and physiologists, whilethe thinking of economists and engineers essen-tially shaped the servomechanic thread.

Consequently, the concepts of the former aremore focused on the adaptation and control ofcomplex systems for the purpose of maintainingstability under exogenous disturbances.Servomechanics, on the other hand, and SD, inparticular, take an endogenous view, being mainlyinterested in understanding circular causality asthe principal source of a system’s behavior.Cybernetics is more connected with communica-tion theory, the general concern of which can besummarized as how to deal with randomly vary-ing input. SD, on the other hand, shows a strongerlink with engineering control theory, which is

primarily concerned with behavior generated bythe control system itself and by the role of non-linearities. Managerial cybernetics and SD bothshare the concern of contributing to managementscience but with different emphases and withinstruments that are distinct but in principle com-plementary. Finally, the mathematical foundationsare generally more evident in the basic literatureon SD than in the writings on organizationalcybernetics, in which the formal apparatus under-lying model formulation is confined to a smallnumber of publications (e.g., Beer 1981, 1994),which are less known than the qualitative trea-tises. The terms management cybernetics andmanagerial cybernetics are used as synonymsfor organizational cybernetics.

The Variety of Systems Methodologies

The methodologies that have evolved as part ofthe systems movement cannot be expounded indetail here. The two epistemological strands inwhich they are grounded, however, can beidentified – the positivist tradition and theinterpretivist tradition.

Positivist tradition denotes those methodologi-cal approaches that focus on the generation of“positive knowledge,” that is, a knowledge basedon “positively” ascertained facts. Interpretivist tra-dition denotes those methodological approachesthat emphasize the importance of subjective inter-pretations of phenomena. This stream goes back toGreek art and science of the interpretation andunderstanding of texts.

Some systems methodologies have beenrooted in the positivist tradition and others in theinterpretivist tradition. As proposed in works onintegrative systems methodology (see below), thedifferences between the two can be describedalong the following set of polarities:

• An objectivist versus a subjectivist position• A conceptual-instrumental versus a communi-

cational/cultural/political rationality• An inclination to quantitative versus qualita-

tive modeling• A structuralist versus a discursive orientation


A positivistic methodological position tendstoward the objectivistic, conceptual-instrumental,quantitative, and structuralist-functionalist in itsapproach. An interpretive position, on the otherhand, tends to emphasize the subjectivist, com-municational, cultural, political, ethical, andesthetic, that is, the qualitative and discursiveaspects. It would be too simplistic to classify aspecific methodology in itself as being “positivis-tic” or “interpretative.” Despite the traditions theyhave grown out of, several methodologies haveevolved and been reinterpreted or opened to newaspects (see below).

In the following, a sample of systems method-ologies will be characterized and positioned inrelation to these two traditions, beginning withthose in the positivistic strand:

• Hard OR methods. Operations research(OR) uses a wide variety of mathematical andstatistical methods and techniques, for exam-ple, of optimization, queuing, dynamic pro-gramming, graph theory, and time seriesanalysis, to provide solutions for organiza-tional and managerial problems, in the opera-tional domains of production, logistics, andfinance, among others.

• Living systems theory. In his LST, James GrierMiller (1978) identifies a set of 20 necessarycomponents that can be discerned in livingsystems of any kind. These structural featuresare specified on the basis of a huge empiricalstudy and proposed as the “critical subsys-tems” that “make up a living system.” LSThas been used as a device for diagnosis anddesign in the domains of engineering and thesocial sciences.

• Viable system model. To date, Stafford Beer’sVSM is probably the best known product oforganizational cybernetics. It specifies a set ofmanagement functions and their interrelation-ships as the sufficient preconditions for theviability of any human or social system (seeBeer 1981). These are applicable in a recursivemode, for example, to the different levels of anorganization. The VSM has been widelyapplied in the diagnostic mode but also tosupport the design of all kinds of social

systems (Espejo and Reyes 2011; Pérez Ríos2012; Malik 2016). Specific methodologies forthese purposes have been developed, forinstance, for use in consultancy. The term via-ble system diagnosis (VSD) is also used.

The methodologies and models addressed upto this point have by and large been created in thepositivistic tradition of science. Other strands inthis tradition do exist, e.g., systems analysis andsystems engineering, which together with ORhave been called “hard systems thinking”(Jackson 2000, p. 127). Also, more recent devel-opments such as mathematical complexity andnetwork theories, agent-based modeling, gametheory, and machine learning can be classified ashard systems approaches.

The respective approaches have not altogetherbeen excluded from fertile contacts with theinterpretivist strand of inquiry. In principle, all ofthem can be considered as instruments forsupporting discourses about different interpreta-tions of an organizational reality or alternativefutures studied in concrete cases. In our time,most applications of the VSM, for example, areconstructivist in nature. To put it in a nutshell,these applications are (usually collective) con-structions of a (new) reality, in which observationand interpretation play a crucial part. In this pro-cess, the actors involved make sense of the systemunder study, i.e., the organization in focus, bymapping it on the VSM. At the same time, theybring forth “multiple realities rather than strivingfor a fit with one reality” (Harnden 1989, p. 299).

The second group of methodologies is part ofthe interpretive strand:

• Interactive planning. IP is a methodology,designed by Russell Ackoff (1981) and devel-oped further by Jamshid Gharajedaghi (1999),for the purpose of dealing with “messes” andenabling actors to design their desired futures,as well as to bring them about. It is grounded intheoretical work on purposeful systems;reverts to the principles of continuous, partici-pative, and holistic planning; and centers onthe idea of an “idealized design.”


• Soft systems methodology. SSM is a heuristicdesigned by Peter Checkland (1981; Checklandand Poulter 2006) for dealing with complexsituations. Checkland suggests a process ofinquiry constituted by two aspects: a conceptualone, which is logic based, and a sociopoliticalone, which is concerned with the cultural feasi-bility, desirability, and implementation ofchange.

• Critical systems heuristics. CSH is a method-ology, which Werner Ulrich (1983, 1996) pro-posed for the purpose of scientificallyinforming planning and design in order tolead to an improvement in the human condi-tion. The process aims at uncovering the inter-ests that the system under study serves. Thelegitimacy and expertise of actors and particu-larly the impacts of decisions and behaviors ofthe system on others – the “affected” – areelicited by means of a set of boundary ques-tions. CSH can be seen as part of a widermovement known as the “emancipatory sys-tems approaches” which embraces, e.g.,Freire’s critical pedagogy, interpretivesystemology, and community OR (see Jackson2000, p. 291 ff).

All three of these methodologies (IP, SSM, andCSH) are positioned in the interpretive tradition.Other methodologies and concepts which can besubsumed under the interpretive systems approachare, e.g., Warfield’s science of generic design,Churchman’s social system design, Senge’s softsystems thinking, Mason and Mitroff’s strategicassumptions surfacing and testing (SAST), Edenand Ackermann’s strategic options developmentand analysis (SODA), and other methodologies ofsoft operational research (for details, see Jackson2000, p. 211 ff). The interpretive methodologieswere designed to deal with qualitative aspects inthe analysis and design of complex systems,emphasizing the communicational, social, politi-cal, and ethical dimensions of problem solving.Several authors mention explicitly that they donot preclude the use of quantitative techniques, orthey include such techniques in their repertoire(e.g., the biocyberneticist Frederic Vester).

In an advanced understanding of systemdynamics, both of these traditions positivist andinterpretivist are synthesized. The adherents ofSD have conceived of model building and valida-tion (the pursuit of model quality) as a semiformal,relativistic, and holistic social process. Validity isunderstood as usefulness or fitness in relation to thepurpose of the model and validation as an elaborateset of procedures – including logico-structural,heuristic, algorithmic, statistical, and also discur-sive components – by which the quality of and theconfidence in a model are gradually improved (seeBarlas 1996; Barlas and Carpenter 1990;Schwaninger and Groesser 2009).

System Dynamics: Its Features,Strengths, and Limitations

The features, strengths, and limitations of the SDmethodology are a consequence of its specificcharacteristics. In the context of the multiple the-ories and methodologies of the systems move-ment, some of the distinctive features of SD are(for an overview, see Richardson 1999; Jackson2000, p. 142 ff):

• Feedback as conceptual basis. SD model sys-tems are higher-order, multiple-loop networks ofclosed loops of information. Concomitantly, aninterest in nonlinearities, long-term patterns, andinternal structure rather than external distur-bances is characteristic of SD (Meadows 1980,p. 31). However, SD models are not “closedsystems,” as sometimes is claimed, in the sensethat (a) flows can originate from outside thesystem’s boundaries, (b) representations of exog-enous factors or systems can be incorporated intoany model as parameters or special modules, and(c) new information can be accommodated viachanges to a model. In other words, the SD viewhinges on a view of systemswhich are closed in acausal sense but not materially (Richardson1999, p. 297).

• Focus on internally generated dynamics. SDmodels are conceived as causally closed sys-tems. The interest of users is in the dynamicsgenerated inside those systems (Richardson


2011). Given the nature of closed feedbackloops and the fact that delays occur withinthem, the dynamic behavior of these systemsis essentially nonlinear.

• Emphasis on understanding. For systemdynamicists, the understanding of the dynam-ics of a system is the first goal to be achievedby means of modeling and simulation. Con-ceptually, they try to understand events asembedded in patterns of behavior, which inturn are generated by underlying structures.Such understanding is enabled by SD as it“shows how present policies lead to futureconsequences” (Forrester 1971, Sect. VIII).Thereby, the feedback loops are “a majorsource of puzzling behavior and policy diffi-culties” (Richardson 1999, p. 300). SD modelspurport to test mental models, hone intuition,and improve learning (see Sterman 1994).

• High degree of operationality. SD relies onformal modeling. This fosters disciplined think-ing; assumptions, underlying equations, andquantifications must be clarified. Feedbackloops and delays are visualized and formalized;therewith the causal logic inherent in a model ismade more transparent and discussable than inmost other methodologies (Richmond 1997).Also, a high level of realism in the models canbe achieved. SD is therefore apt to supportdecision-making processes effectively.

• Far-reaching requirements (and possibilities)for the combination of qualitative and quanti-tative aspects of modeling and simulation. Thisis a consequence of the emphasis on under-standing. Qualitative modeling and visualiza-tion are carried out by means of causal loopdiagrams and stock and flow diagrams, whichare then underlaid quantitatively with equa-tions. The focus is not on point-precise predic-tion but on the generation of insights into thepatterns generated by the systems under studyand the underlying structures which producethose patterns.

• High level of generality and scale robustness.The representation of dynamic issues in termsof stocks and flows is a generic form, which isadequate for a wide spectrum of potentialapplications. This spectrum is both broad as

to the potential subjects under study and deepas to the possible degrees of resolution anddetail (La Roche and Simon 2000). In addition,the SD methodology enables one to deal withlarge numbers of variables within multipleinteracting feedback loops (Forrester 1969,p. 9). SD has been applied to the most diversesubject areas, e.g., global modeling, environ-mental issues, social and economic policy, cor-porate and public management, regionalplanning, medicine, psychology, and educationin mathematics, physics, and biology.

The features of SD just sketched out result inboth strengths and limitations. We start with thestrengths.

Strengths of SD

1. Its specific modeling approach makes SD par-ticularly helpful in gaining insights into thepatterns exhibited by dynamic systems, aswell as the structures underlying them. Closed-loop modeling has been found most useful infostering understanding of the dynamic func-tioning of complex systems. Such understand-ing is especially facilitated by the principle ofmodeling the systems or issues under study in acontinuous mode and at rather high aggrega-tion levels (Forrester 1961; La Roche andSimon 2000). With the help of relativelysmall but insightful models, and by means ofsensitivity analyses as well as optimizationheuristics incorporated in the application soft-ware packages, decision-spaces can be thor-oughly explored. Vulnerabilities and theconsequences of different system designs canbe examined with relative ease.

2. The generality of the methodology and itspower to crystallize operational thinking inrealistic models have triggered applications inthe most varied contexts. Easy-to-use softwareand the features of modeling via graphic userinterfaces provide a strong lever for collabora-tive model building in teams (Andersen andRichardson 1997; Vennix 1996).


3. Its specific features make SD an exceptionallyeffective tool for conveying systemic thinkingto anybody. Therefore, it also has an outstand-ing track record of classroom applications forwhich “learner-directed learning” (Forrester1993a) or “learner-centered learning” is advo-cated (Forrester 1993b, 1997). Pertinent audi-ences range from schoolchildren at the levelsof secondary and primary schools to managersand scientists.

4. The momentum of the SD movement is remark-able. Due to the strengths commented above thispoint, the community of users has grownsteadily, being probably the largest communitywithin the systems movement. Lane (2006,p. 484) has termed SD “one of the most widelyused systems approaches in the world.”Hence, acritical mass exists to spark future developments.

Given these strengths, the community of users hasnot only grown significantly but has also trans-cended disciplinary boundaries, ranging from theformal and natural sciences to the humanities andcovering multiple uses from theory building andeducation to the tackling of real-world problemsat almost any conceivable level. Applications toorganizational, societal, and ecological issueshave seen a particularly strong increase. Thisfeeds back on the availability and growth of theknowledge upon which the individual modelercan draw.

The flip side of most of the strengths outlinedhere embodies the limitations of SD; we concen-trate on those which can be relevant to a possiblecomplementarity of SD with other systemsmethodologies.

Limitations of SD

1. The main point here is that SD does not pro-vide a framework or methodology for the diag-nosis and design of organizational structuresin the sense of interrelationships among orga-nizational actors. This makes SD susceptible tocompletion from without – a completion whichorganizational cybernetics (OC), and the VSMin particular but also living systems theory

(LST), especially can provide. The choicefalls on these two approaches because of theirstrong heuristic power and their complemen-tary strengths in relation to SD (Schwaninger2006; Schwaninger and Pérez Ríos 2008).

2. Another limitation of SD is related to theabsorption of variety (complexity) by an orga-nization. Variety is a technical term for com-plexity, which denotes a (high) number ofpotential states or behaviors of a system(based on Ashby 1956; Beer 1966). SD offersan approach to the handling of variety whichallows modeling at different scales of a prob-lem or system (Odum and Odum 2000). Itfocuses on the identification, at a certain reso-lution level or possibly several resolutionlevels, of the main stock variables which willbe affected by the respective flows. These, inturn, will be influenced by parameters and aux-iliary variables. This approach, even though itenables thinking and modeling at differentscales, does not provide a formal procedurefor an organization to cope with the externalcomplexity it faces, namely, for designing astructure which can absorb that complexity.In contrast, OC and LSToffer elaborate modelsto enable the absorption of variety, in the caseof the VSM based explicitly on Ashby’s law ofrequisite variety. It says “Only variety candestroy variety,” which implies that the varie-ties of two interacting systems must be in bal-ance, if stability is to be achieved (Ashby1956). The VSM has two salient features inthis respect. Firstly, it helps design an organi-zational unit for viability, by enabling it toattenuate the complexity of its environmentand also to enhance its eigen-variety, so thatthe two are in balance. The term variety engi-neering has been used in this context (Beer1979). Secondly, the recursive structure of theVSM ensures that an organization with severallevels will develop sufficient eigen-varietyalong the fronts on which the complexity itfaces unfolds. Similarly, LST offers the condi-tions for social systems to survive, bymaintaining thermodynamically highlyimprobable energy states via continuous inter-action with their environments. The difference


between the two approaches is that the VSMfunctions more in the strategic and informa-tional domains, while the LST model essen-tially focuses on the operational domain. Insum, both can make a strong contributionrelated to coping with the complexity facedby organizations and therefore can deliver astrong complement to SD.

3. The design of modeling processes confrontsSD with specific challenges. The original SDmethodology of modeling and simulation wasto a large extent functionally and technicallyoriented. This made it strong in the domain oflogical analysis, while the sociocultural andpolitical dimensions of the modeling processwere, if not completely out of consideration, atleast not a significant concern in methodolog-ical developments. The SD community – alsounder the influence of the soft systemsapproaches – has become aware of this limita-tion and has worked on incorporating featuresof the social sciences into its repertoire. Thefollowing examples, which document thiseffort to close the gap, stand for many. Exten-sive work on group model building has beenachieved, which explores the potential of col-laborative model building (Vennix 1996).A new schema for the modeling process hasbeen proposed, which complements logic-based analysis by cultural analysis (Lane andOliva 1998). The social dimension of systemdynamics-based modeling has become a sub-ject to intensive discussion (Vriens andAchtenbergh 2006) and other contributions tothe special issue of Systems Research andBehavioral Science, Vol. 51, No. 49. Finally,in relation to consultancy methodology,modeling has been framed as a learning pro-cess (Lane 1994a; Kineman 2017) and assecond-order intervention (Schwaningeret al. 2006).

4. System dynamics modeling adopts a top-downlogic. It operates with high-level aggregatesand is not convenient for models that representthe individual behaviors of high quantities ofagents. In cases where microagents producebehavior at macrolevel, agent-based models(ABM) are the prevalent methodological

solution. A combination of SD and ABM canleverage the complementarities of the twoapproaches (Duggan 2007; Martin andSchlüter 2015; Nava Guerrero et al. 2016).This way a great contribution should be possi-ble, for example, to the modeling of processesof emergence and the understanding of emer-gent phenomena, an area where more work isneeded (Kineman 2017).

As has been shown, there is a need to complementclassical SD with other methodologies, whenissues are at stake which it cannot handle by itself.VSM and LST are pertinent choices when issuesof organizational diagnosis or design are to betackled.

The limitations addressed here call attention toother methodologies which exhibit certain fea-tures that traditionally were not incorporated, orat least not explicit, in SD methodology. Oneaspect concerns the features that explicitly addressthe subjectivity of purposes and meaningsascribed to systems. In this context, support forproblem formulation, model construction, andstrategy design by individuals on the one handand groups on the other are relevant issues. Also,techniques for an enhancement of creativity (e.g.,the generation and the reframing of options) inboth individuals and groups are a matter of con-cern. Two further aspects relate to methodologicalarrangements for coping with the specific issuesof negotiation and alignment in pluralist and coer-cive settings.

As far as the modeling and simulation pro-cesses are concerned, group model building hasbecome first a complement of SD and then acrucial feature of most modeling activities(Kineman 2017). In addition, there are othermethodologies which should be considered aspotentially apt to enrich SD analysis, namely, thesoft systems approaches commented upon earlier,e.g., interactive planning, soft systems methodol-ogy, and critical systems heuristics.

On the other hand, SD can be a powerful com-plement to other methodologies which are moreabstract or more static in nature. This potentialrefers essentially to all systems approacheswhich stand in the interpretive (“soft”) tradition


but also to approaches which stand in the positiv-ist traditions, such as the VSM and LST. Theseshould revert to the support of SD in the event thattrade-offs between different goals must be han-dled, or if implications of long-term decisions onshort-term outcomes (and vice versa) have to beascertained, and whenever contingencies or vul-nerabilities must be assessed.

Actual and Potential Relationships

It should be clear by now that the systems move-ment has bred a number of theories and method-ologies, none of which can be considered all-embracing or complete. All of them have theirstrengths and weaknesses and their specific poten-tials and limitations.

Since Burrell and Morgan (1979) adverted toincommensurability between different paradigms ofsocial theory, several authors have acknowledged oreven advocated methodological complementarism.They argue that there is a potential complementaritybetween different methodologies and, one may add,models andmethods, even if they come from distinctparadigms. Among these authors are, e.g.,Brocklesby (1993), Jackson (1991), Midgley(2000), Mingers (1997), Schwaninger (1997), Yolles(1998), and García-Díaz and Olaya (2018). One ofthese protagonists proposes “that well-designed andwell-executed multi-method research has inferentialadvantages over research relying on a singlemethod” (Seawright 2016, p. 1). Seawright advo-cates a “well-constructed integrative multi-methoddesign” (Ibidem: 9). In other words, a new perspec-tive, in comparison with the non-complementaristicstate of the art, has been opened.

In the past, the different methodologies have ledto the formation of their own traditions and“schools,” with boundaries across which notmuch dialogue has evolved. The methodologieshave kept their protagonists busy testing them anddeveloping them further. Also, the differencesbetween different language games and epistemo-logical traditions have often suggested incommen-surability and therewith have impairedcommunication. Prejudices and a lack of knowl-edge of the respective other side have accentuated

this problem: Typically, “hard” systems scientistshave been suspicious of “soft” systems scientists.For example, many members of the OR commu-nity, not unlike orthodox quantitatively orientedeconomists, adhere to the opinion that “SD is toosoft.” On the other hand, the protagonists of “soft”systems approaches, even though many of themhave adopted feedback diagrams (causal loop dia-grams) for the sake of visualization, are all toooften convinced that “SD is too hard.” Both ofthese judgments indicate a lack of knowledge, inparticular of the SD validation and testing methodsavailable, on the one hand, and the technicaladvancements achieved in modeling and simula-tion, on the other (see Barlas and Carpenter 1990;Schwaninger and Groesser 2009-2019; Sterman2000).

In principle, both approaches are complemen-tary. The qualitative view can enrich quantitativemodels, and it is connected to their philosophical,ethical, and esthetical foundations. However,qualitative reasoning tends to be misleading ifapplied to causal network structures withoutbeing complemented by formalization and quan-tification of relationships and variables. Further-more, the quantitative simulation fosters insightsinto qualitative patterns and principles. It is thus amost valuable device for validating and honingthe intuition of decision-makers, via corrobora-tion and falsification.

Proposals that advocate mutual learningbetween the different “schools” have been formu-lated inside the SD community (e.g., Lane 1994b).The International System Dynamics Conference of1994 in Stirling, held under the banner of “Trans-cending the Boundaries,” was dedicated to thedialogue between different streams of the systemsmovement.

Also, from the 1990s onward, there were vig-orous efforts to deal with methodological chal-lenges, which traditionally had not been animportant matter of scientific interest within theSD community. Some of the progress made inthese areas is documented in a special edition ofSystems Research and Behavioral Science (Vol.21, No. 4, July–August 2004). The main point isthat much of the available potential is based on the


complementarity, not the mutual exclusiveness, ofthe different systems approaches.

In the future, much can be gained from leverag-ing these complementarities. Here are two exam-ples of methodological developments in thisdirection, which appear to be achievable andpotentially fertile: first, the enhancement of qual-itative components in “soft” systems methodolo-gies in the process of knowledge elicitation andmodel building (Vennix 1996) and second, thecombination of cybernetics-based organizationaldesign with SD-based modeling and simulation(Schwaninger and Pérez Ríos 2008). One mustadd that potential complementarities exist notonly across the quality-quantity boundary butalso within each one of the domains. For example,with the help of advanced software, SD modeling(“top-down”) and agent-based modeling(“bottom-up”) can be used in combination.

From a meta-methodological stance, generalistframeworks have been elaborated which containblueprints for combining different methodologieswhere this is indicated. The following are twoearly and two more recent examples:

• Total systems intervention (TSI) is a frameworkproposed by Flood and Jackson (1991), whichfurnishes a number of heuristic schemes andprinciples for the purpose of selecting and com-bining systems methods/methodologies in acustomized way, according to the issue to betackled. SD is among the recommended “tools.”

• Integrative systems methodology (ISM) is aheuristic for providing actors in organizationswith requisite variety (Schwaninger 1997,2004). It advocates (a) dealing with bothcontent- and context-related issues during theprocess and (b) placing a stronger emphasis onthe validation of qualitative and quantitativemodels as well as strategies, in both dimen-sions of the content of the issue under studyand the organizational context into which thatissue is embedded. For this purpose, the toolsof SD (to model content) and organizationalcybernetics – the VSM (to model context) – arecogently integrated.

• Evolutionary learning laboratory (ELLab)denotes an approach to the management ofcomplex issues developed by Bosch and

coauthors (Bosch et al. 2013). The frameworkincludes a process design for collective learn-ing, involving participatory systems analysis,the interpretation of system structures to iden-tify levers for systemic interventions, strategydesign, and implementation. The methodologyis based on several methods including qualita-tive system dynamics.

• Social systems engineering (SSE) is anapplication-oriented framework for copingwith complexity, based on a nonmechanisticengineering viewpoint. It has a traditiongoing back to the 1970s and is now beingreactivated (Garcia-Díaz and Olaya 2018).The approach considers technical, human,and social perspectives, which are crucial tosolving complex problems. It focuses on sys-temic interventions aimed at the design, rede-sign, and transformation of social systems.Different approaches to multimethodology areproposed (Ibidem), e.g., actor-network theoryas a heuristic for designing and building agent-based models (Méndez-Fajardo 2018) or aholistic approach to the design of healthcaresystems, involving SD and the VSM(Schwaninger and Klocker 2018).

Recently, multimethodological frameworkshave been developed, which are focused on inte-gration (a systemic principle), but not explicitlytied to systemic methodologies (e.g., Morgan2014; Seawright 2016). Given their meta-methodological position, these frameworks arein principle open to including systems methodsor methodologies into their repertoire.

In principle, SD could make an important con-tribution in the context of any multimethodologicalframework, far beyond the extent to which this hasbeen the case. Systems methodologists and practi-tioners can potentially benefit enormously fromincluding SD methodology in their repertoires.

Outlook

There have been recurrent calls for an eclectic“mixing and matching” of methodologies. Inlight of the epistemological tendencies of our


time toward radical relativism, it is necessary towarn against taking a course in which “anythinggoes.” It is most important to emphasize that thedesirable methodological progress can only beachieved on the grounds of scientific rigor. Thispostulate of “rigor” is not to be confused with anencouragement of “rigidity.” The necessary meth-odological principles advocated here are disci-plined thinking, a permanent quest for models ofhighest quality (i.e., thorough model validation),and full transparency in the formalizations as wellas of the underlying assumptions and sourcesused. Scientific rigor, in this context, also impliesthat combinations of methodologies, methods, ormodels reach beyond merely eclectic add-ons sothat genuine integration toward better adequacy tothe issues at hand is achieved.

The contribution of system dynamics can comein the realms of the following:

• Fostering disciplined thinking• Understanding dynamic behaviors of systems

and the structures that generate them• Exploring paths into the future and the con-

crete implications of decisions• Assessing strategies as to their robustness, vul-

nerabilities, and impact, in ways precluded byother, more philosophical, and generally “soft”systems approaches

These latter streams can contribute to reflectingand tackling the meaning- and value-laden dimen-sions of complex human, social, and ecologicalsystems. Some of their features should and can becombined synergistically with system dynamics,particularly by being incorporated into the reper-toires of system dynamicists. From the reverseperspective, incorporating system dynamics as astandard tool will be of great benefit for theadopters of broad methodological frameworks.Model formalization and dynamic simulationmay even be considered necessary componentsfor the study of the concrete dynamics of complexsystems.

There are also many developments in the“hard,” i.e., mathematics-, statistics-, logic-, andinformatics-based methods and technologies,

which are apt to enrich the system dynamics meth-odology, namely, in terms of modeling and deci-sion support. For example, the constantlyevolving techniques of time series analysis, filter-ing, neural networks, control theory, machinelearning, etc. can improve the design of system-dynamics-based systems of (self-)control. Also, abridge across the divide between the top-downmodeling approach of SD and the bottom-upapproach of agent-based modeling has been builtwith the help of modern software.

Furthermore, a promising perspective for thedesign of genuinely “intelligent organizations”emerges if one combines dynamic modeling andcybernetic modeling, namely, system dynamicsand the viable system model. Also, combiningthe “soft” systems methodologies and the “hard”methods from operations research could enablebetter, model-based management.

The approaches of integrating complementarymethodologies, as outlined in this contribution,definitely mark a new phase in the history of thesystems movement.

Appendix – Milestones in the Evolutionof the Systems Approach in General andSystem Dynamics in Particular

The following table gives an overview of thesystems movement’s evolution, as shown in itsmain literature. This overview is not exhaustive.

Foundations of general systems theory

Von Bertalanffy Zu einerallgemeinenSystemlehre

1945

An Outline ofGeneral SystemTheory

1950

General SystemTheory

1968

Bertalanffy,Boulding,Gerard, Rapoport

Foundation of theSociety for GeneralSystems Research

1953

Klir An Approach toGeneral SystemsTheory

1968

Simon 1969

(continued)


The Sciences of theArtificial

Pichler MathematischeSystemtheorie

1975

Miller Living Systems 1978

Mesarovic andTakahara

Abstract SystemsTheory

1985

Rapoport General SystemTheory

1986

Foundations of cybernetics

Macyconferences(Josiah Macy,Jr. foundation)

Cybernetics,Circular Causal,and FeedbackMechanisms inBiological andSocial Systems

1946–1951

Wiener Cybernetics orControl andCommunication inthe Animal and inthe Machine

1948

Ashby An Introduction toCybernetics

1956

Pask An Approach toCybernetics

1961

Von Foerster,Zopf

Principles of Self-Organization

1962

McCulloch Embodiments ofMind

1965

Foundations of organizational cybernetics

Beer Cybernetics andManagement

1959

Towards theCybernetic Factory

1962

Decision andControl

1966

Brain of the Firm 1972

Von Foerster Cybernetics ofCybernetics

1974

Foundations of system dynamics

Forrester Industrial Dynamics 1961

Principles ofSystems

1968

Urban Dynamics 1969

World Dynamics 1971

Meadows et al. Limits to Growth 1972

Richardson Feedback Thoughtin Social Scienceand Systems Theory

1991

Systems methodology

Churchman Challenge to Reason 1968

1968

(continued)

The SystemsApproach

Vester and vonHesler

Sensitivitätsmodell 1980

Checkland Systems Thinking,Systems Practice

1981

Ackoff Creating theCorporate Future

1981

Ulrich Critical Heuristicsof Social Planning

1983

Warfield A Science of GenericDesign

1994

Schwaninger Integrative SystemsMethodology

1997

Gharajedaghi Systems Thinking 1999

Sabelli Bios: A Study ofCreation

2005

García and Olaya Social SystemsEngineering

2018

Selected recent works in system dynamics

Senge The Fifth Discipline 1990

Barlas andCarpenter

Model Validity 1990

Vennix Group ModelBuilding

1996

Lane and Oliva Synthesis of SystemDynamics and SoftSystemsMethodology

1998

Sterman Business Dynamics 2000

Warren Strategy Dynamics 2002, 2008

Wolstenholme ArchetypalStructures

2003

Morecroft Strategic Modeling 2007

Schwaninger andGrösser

Theory-Buildingwith SystemDynamics andModel Validation

2008, 2009

Rahmandad,Oliva, Osgood

Analytical Methodsfor DynamicModelers

2015

Kunc Strategic Analytics 2019

Bibliography

Primary LiteratureAckoff RL (1981) Creating the corporate future. Wiley,

New YorkAndersen DF, Richardson GP (1997) Scripts for group

model building. Syst Dyn Rev 13(2):107–129Ashby WR (1956) An introduction to cybernetics. Chap-

man & Hall, London


Barlas Y (1996) Formal aspects of model validity and val-idation in system dynamics. Syst Dyn Rev 12(3):183–210

Barlas Y, Carpenter S (1990) Philosophical roots of modelvalidity: two paradigms. Syst Dyn Rev 6(2):148–166

Beer S (1959) Cybernetics and management. English Uni-versities Press, London

Beer S (1966) Decision and control. Wiley, ChichesterBeer S (1979) The heart of enterprise. Wiley, ChichesterBeer S (1981) Brain of the firm, 2nd edn.Wiley, ChichesterBeer S (1994) Towards the cybernetic factory. In:

Harnden R, Leonard A (eds) How many grapes wentinto the wine. Stafford Beer on the art and science ofholistic management. Wiley, Chichester, pp 163–225.(reprint, originally published in 1962)

Bosch OJH, Nguyen NC, Maeno T, Yasui T (2013) Man-aging complex issues through evolutionary learninglaboratories. Syst Res Behav Sci 30(2):116–135

Brailsford S (2014) Theoretical comparison of discrete-event simulation and system dynamics. In:Brailsford S, Churilov L, Dangerfield B (eds) Discrete-event simulation and system dynamics for managementdecision making. Wiley, Chichester, pp 105–124

Brailsford S, Churilov L, Dangerfield B (eds)(2014) Discrete-event simulation and system dynamicsfor management decision making. Wiley, Chichester

Brocklesby J (1993) Methodological complementarism orseparate paradigm development – examining theoptions for enhanced operational research. AustJ Manag 18(2):133–157

Burrell G, Morgan G (1979) Sociological paradigms andorganisational analysis. Gower, Hants

Checkland PB (1981) Systems thinking, systems practice.Wiley, Chichester

Checkland PB, Poulter J (2006) Learning for action: a shortdefinitive account of soft systems methodology, and itsuse practitioners, teachers and students. Wiley,Chichester

Churchman CW (1968a) Challenge to reason. McGraw-Hill, New York

Churchman CW (1968b) The systems approach. DelacortePress, New York

Churchman CW (1979) The systems approach and itsenemies. Basic Books, New York

Duggan J (2007) Modelling agent-based systems usingsystem dynamics. In: Proceedings of the 2007 interna-tional conference of the system dynamics society. TheSystem Dynamics Society, Boston

Encyclopedia of Life Support Systems (2002) publishedunder: http://www.eolss.net/

Espejo R, Reyes A (2011) Organizational systems. Man-aging complexity with the viable system model.Springer, Berlin

Flood RL, Jackson MC (1991) Creative problem solving.Total systems intervention. Wiley, Chichester

Forrester JW (1961) Industrial dynamics. MIT Press,Cambridge

Forrester JW (1968) Principles of systems. MIT Press,Cambridge

Forrester JW (1969) Urban dynamics. MIT Press,Cambridge

Forrester JW (1971) World dynamics. Pegasus Communi-cations, Waltham

Forrester JW (1993a) System dynamics and the lessons of35 years. In: KB DG (ed) Systems-based approach topolicy making. Kluwer, Boston

Forrester JW (1993b) System dynamics as an organizingframework for pre-college education. Syst Dyn Rev9(2):183–194

Forrester JW (1997) System dynamics and K-12 teachers.A lecture at the University of Virginia School of Edu-cation, Massachusetts Institute of Technology. SystemDynamics Group Paper D-4665-4

François C (2004) International encyclopedia of systemsand cybernetics, 2nd edn. Saur, München

Garcia-Díaz C, Olaya C (eds) (2018) Social systems engi-neering: the design of complexity. Wiley, Chichester

Gharajedaghi J (1999) Systems thinking. Managing chaosand complexity. Butterworth-Heinemann, Boston

Harnden RJ (1989) Technology for enabling: the implica-tions for management science of a hermeneutics ofdistinction. The University of Aston, Birmingham

Holland JH (1995) Hidden order: how adaption buildscomplexity. Addison-Wesley, Reading

Holland JH (2006) Studying complex adaptive systems.J Syst Sci Complex 19(1):1–8. https://doi.org/10.1007/s11424-006-0001-z. ISSN 1009-6124

Jackson MC (1991) Systems methodology for the manage-ment sciences. Plenum Press, New York

Jackson MC (2000) Systems approaches to management.Kluwer/Plenum, New York

Jackson MC (2003) The power of multi-methodology:some thoughts for John Mingers. J Oper Res Soc54(12):1300–1301

Jackson MC (2019) Critical systems thinking and the man-agement of complexity. Wiley Hoboken, NJ

Kauffman SA (1993) The origins of order. Self-organization and selection in evolution. Oxford Uni-versity Press, New York

Kineman JJ (2017) Modeling and simulation. In: EdsonMD, Buckle Henning P, Sankaran S (eds) A guide tosystems research: philosophy, processes and practice.Springer, Singapore, pp 81–110

Klir GJ (1969) An approach to general systems theory.Nostrand, New York

Kunc M (2019) Strategic analytics: integrating manage-ment science and strategy. Wiley, Hoboken/Chichester

La Roche U, Simon M (2000) Geschäftsprozessesimulieren: flexibel und zielorientiert führen mitFliessmodellen. Orell Füssli, Zürich

Lane DC (1994a) Modeling as learning: a consultancymethodology for enhancing learning in managementin management teams. In: Morecroft J, Sterman JD(eds) Modeling for learning organizations. ProductivityPress, Portland, pp 205–240

Lane DC (1994b) With a little help from our friends: howsystem dynamics and soft OR can learn from eachother. Syst Dyn Rev 10(2–3):101–134


http://www.eolss.net/

https://doi.org/10.1007/s11424-006-0001-z

https://doi.org/10.1007/s11424-006-0001-z

Lane DC (2006) IFORS’ operational research hall of fame.Jay Wright Forrester. Int Trans Oper Res 13:483–492

Lane DC, Oliva R (1998) The greater whole: towards asynthesis of system dynamics and soft system method-ology. Eur J Oper Res 107(1):214–235

Malik F (2016) Strategy for managing complex systems: acontribution to management cybernetics for evolution-ary systems. Campus Verlag, Frankfurt/New York

Martin R, Schlüter M (2015) Combining system dynamicsand agent-based modeling to analyze social-ecologicalinteractions – an example from modeling restoration ofa shallow lake. Front Environ Sci. https://doi.org/10.3389/fenvs.2015.00066. 29 Aug 2017

McCulloch WS (1965) Embodiments of mind. MIT Press,Cambridge

Meadows DH (1980) The unavoidable a priori. In: RandersJ (ed) Elements of the system dynamics method. MITPress, Cambridge, pp 23–57

Meadows DH, Meadows DL, Randers J, Behrens WW III(1972) Limits to growth. Universe Books, New York

Méndez-Fajardo S (2018) Using actor-network theory inagent-based modelling. In: Garcia-Díaz C, OlayaC (eds) Social systems engineering: the design of com-plexity. Wiley, Chichester, pp 157–196

Mesarovic MD, Takahara Y (1985) Abstract systems the-ory. Springer, Berlin

Midgley G (2000) Systemic intervention. Philosophy,methodology, and practice. Kluwer, New York

Miller JG (1978) Living systems. McGraw-Hill, New YorkMiller JH, Page SE (2007) Complex adaptive systems: an

introduction to computational models of social life.Princeton University Press, Princeton

Mingers J (1997) Multi-paradigm multimethodology. In:Mingers J, Gill A (eds) Multimethodology. Wiley,Chichester

Morecroft J (2007) Strategic modelling and businessdynamics: a feedback systems approach. Wiley,Chichester

Morgan DL (2014) Integrating qualitative and quantitativemethods: a pragmatic approach. Sage, Los Angeles

Nava Guerrero G, Schwarz P, Slinger JH (2016) A recentoverview of the integration of system dynamics andagent-based modelling and simulation. In: Proceedingsof the 34th international conference of the systemdynamics society, July 17–21, 2016 Delft, Netherlands,p 1153

Odum HT, Odum EC (2000) Modeling for all scales: anintroduction to system simulation. Academic, SanDiego

Pask G (1961) An approach to cybernetics. Hutchinson,London

Pérez Ríos J (2012) Design and diagnosis for sustainableorganizations: the viable system method. Springer,Berlin

Pichler F (1975) Mathematische Systemtheorie. deGruyter, Berlin

Rapoport A (1953) Operational philosophy: integratingknowledge and action. Harper, New York

Rapoport A (1986) General system theory. Essential con-cepts and applications. Abacus Press, TurnbridgeWells

Richardson GP (1999) Feedback thought in social scienceand systems theory. Pegasus communications, Wal-tham. (Originally published in 1991)

Richmond B (1997) The “thinking” in systems thinking:how can we make it easier to master? Syst Think8(2):1–5

Sabelli H (2005) Bios: a study of creation. World Scien-tific, Hackensack

Schwaninger M (1997) Integrative systems methodology:heuristic for requisite variety. Int Trans Oper Res4(4):109–123

Schwaninger M (2004) Methodologies in conflict: achiev-ing synergies between system dynamics and organiza-tional cybernetics. Syst Res Behav Sci 21(4):1–21

Schwaninger M (2006) Theories of viability.A comparison. Syst Res Behav Sci 23:337–347

Schwaninger M, Groesser S (2008) System dynamics asmodel-based theory building. Syst Res Behav Sci25:1–19

Schwaninger M, Groesser S (2009, new edition 2019)Model validation: the quest for quality in systemdynamics modeling. Encyclopaedia of complexity andsystems science. Springer, New York, pp 9000–9014.Reprinted in: Complex System in Finance and econo-metrics. Springer New York, 2011: 767–781

Schwaninger M, Klocker J (2018) Holistic system design:the oncology Carinthia study. In: Garcia-Díaz C, OlayaC (eds) Social systems engineering: the design of com-plexity. Wiley, Chichester, pp 235–266

Schwaninger M, Pérez Ríos J (2008) System dynamics andcybernetics: a synergetic pair. Syst Dyn Rev24(2):145–174

Schwaninger M, Janovjak M, Ambroz K (2006) Second-order intervention: enhancing organizational compe-tence and performance. Syst Res Behav Sci23:529–545

Seawright J (2016) Multi-method social science: combin-ing qualitative and quantitative tools. Cambridge Uni-versity Press, Cambridge, UK

Senge PM (1990) The fifth discipline. The art and practiceof the learning organization. Doubleday, New York

Shapiro M, Mandel T, Schwaninger M et al (1996) Theprimer toolbox. International Society for the SystemsSciences. http://www.isss.org/primer/toolbox.htm

Simon HA (1969) The sciences of the artificial. MIT Press,Cambridge

Sterman JD (1994) Learning in and about complex sys-tems. Syst Dyn Rev 10(2–3):291–330

Sterman JD (2000) Business dynamics. Systems thinkingand modeling for a complex world. Irwin/McGraw-Hill, Boston

Ulrich W (1983) Critical heuristics of social planning.Haupt, Bern

Ulrich W (1996) A primer to critical systems heuristics foraction researchers. The Centre of systems studies. Uni-versity of Hull, Hull


https://doi.org/10.3389/fenvs.2015.00066

https://doi.org/10.3389/fenvs.2015.00066

http://www.isss.org/primer/toolbox.htm

Vennix JAM (1996) Group model building. Facilitatingteam learning using system dynamics. Wiley,Chichester

Vester F, Von Hesler A (1980) Sensitivitätsmodell.Regionale Planungsgemeinschaft Untermain, Frank-furt am Main

Von Bertalanffy L (1949) Zu einer allgemeinenSystemlehre. Bl Dtsch Philos 18(3/4) (Excerpts: inBiol Gen 19(1):114–129 and in General System The-ory, 1968, Chapter III)

Von Bertalanffy L (1950) An outline of general systemtheory. Br J Philos Sci 1:139–164

Von Bertalanffy L (1968) General system theory. Braziller,New York

Von Foerster H (1984) Observing systems, 2nd edn. Inter-systems Publications, Seaside

Von Foerster H (ed) (1995) Cybernetics of cybernetics,2nd edn. Future Systems, Minneapolis. Originallypublished in 1974

Von Foerster H, Zopf GW (eds) (1962) Principles of self-organization. Pergamon Press, Oxford

Vriens D, Achtenbergh J (2006) The social dimension ofsystem dynamics-based modelling. Syst Res Behav Sci23(4):553–563

Warfield JN (1994) A science of generic design: managingcomplexity through systems design, 2nd edn. IowaState University Press, Ames

Warren K (2002) Competitive strategy dynamics. Wiley,Chichester

Warren K (2008) Strategic management dynamics. Wiley,Chichester

Wiener N (1948) Cybernetics: control and communicationin the animal and in the machine. MIT Press,Cambridge

Wolstenholme E (2003) Towards the definition and use of acore set of archetypal structures in system dynamics.Syst Dyn Rev 19(1):7–26

Yolles MA (1998) Cybernetic exploration of methodolog-ical complement. Kybern 27(5):527–542

Books and ReviewsDangerfield B (2014) Systems thinking and system dynam-

ics: a primer. In: Brailsford S, Churilov L, DangerfieldB (eds) Discrete – event simulation and system dynam-ics for management decision-making. Wiley, Chiches-ter, pp 26–51

JacksonMC (2003b) Systems thinking: creative holism formanagers. Wiley, Chichester

Jackson MC (2019) Critical systems thinking and the man-agement of complexity. Wiley, Chichester

Klir GJ (2001) Facets of systems science, 2nd edn. Kluwer/Plenum, New York

Kunc M (ed) (2018) System dynamics: soft and hard oper-ational research. Palgrave Macmillan, London

Midgley G (ed) (2003) Systems thinking, vol vol 4. Sage,London

Ragsdell G, Wilby J (eds) (2001) Understanding complex-ity. Kluwer/Plenum, New York

Rahmandad H, Oliva R, Osgood ND (eds) (2015) Analyt-ical methods for dynamic modelers. MIT Press, Cam-bridge, MA

Ramage M, Shipp K (2009) Systems thinkers. Springer,Dordrecht

Richardson GP (ed) (1996) Modelling for management.Simulation in support of systems thinking, vol 2 vols.Dartmouth, Aldershot

Richardson GP (2011) Reflections on the foundations ofsystem dynamics. Syst Dyn Rev 27(3):219–243

Schwaninger M (2009) Intelligent organizations. Powerfulmodel for systemic management, 2nd edn. Springer,Berlin

Seawright J (2016b) Multi-method social science. Cam-bridge University Press, Cambridge

Van Gigch JP (2003) Metadecisions. Rehabilitating episte-mology. Kluwer/Plenum, New York


Ashby s law of requi- System Dynamics in the Evolution of ...

Documents

Transcript of Ashby s law of requi- System Dynamics in the Evolution of ...