Action Research

O R I G I N A L P A P E R

Action Research: Its History and Relationship to Scientific Methodology

John Stephens Æ John Barton Æ Tim Haslett

Published online: 13 August 2009 � Springer Science+Business Media, LLC 2009

Abstract Scientific Methodology (SM) has long suited those who favour analytical and quantitative research in management. Thus the dilemma between the rigour and relevance

of contemporary management research methods is fuelled by action researchers who keep

wanting to contrast Action Research (AR) with SM. This paper presents a Western

philosophical view on the development of belief systems and theory-based methods over

time. It thus links the progressive and cumulative development of SM with the contem-

porary AR methodology. In doing this it presents a different point of view—that the

traditions of SM and AR have much closer relationships than people often give them

credit for.

Keywords Scientific methodology � Action research � Induction � Deduction � Abduction � Philosophy

Introduction

Scientific Methodology (SM) has long suited those who favour analytical and quantitative

research in management. Thus since the notion of a fundamental epistemological paradigm

shift surfaced (Kuhn 1962; Feyeraband 1970) the dilemma between rigour and relevance of

research has been fuelled by action researchers who keep wanting to contrast AR with SM

(Reason and Heron 1995; Garrick 2000). In this regard, Reason and Heron (1995, p. 122)

forcibly make the point that ‘‘since scientific research is such a powerful force in our lives

J. Stephens (&) Greyhound Racing Victoria, Melbourne, VIC, Australia e-mail: jstephens@grv.org.au

J. Barton John Barton Associates, Melbourne, VIC, Australia e-mail: bartcons@bigpond.net.au

T. Haslett Department of Management, Monash University, Melbourne, VIC, Australia e-mail: thaslett@bigpond.net.au

123

Syst Pract Action Res (2009) 22:463–474 DOI 10.1007/s11213-009-9147-7

it is shocking that its techniques largely ignore the epistemological and political signifi-

cance of participation’’. Hence in acknowledging that there is no simple answer to the

question ‘‘What is AR?’’ (Checkland and Holwell 1998; Reason and Bradbury 2001) this

paper presents a truncated Western philosophical view on the development of belief sys-

tems and theory-based methods over time. In doing this we link the progressive and

cumulative refinement of the SM with aspects that are common in contemporary AR. Thus

we argue a different point of view—that the traditions of SM and AR have much ‘‘closer

relationships’’ than people often give them credit for.

A Philosophical Basis for Closer Relationships in SM and AR: The Greeks

As research methodologies, both contemporary SM and AR may be traced back to the

ancient Greek ‘‘rise of mental coherence’’ (Singer 1959a, b). Then the Greeks built on pure

observation, to integrate inquiry, debate and discovery into concepts of reality. Hence

while predecessors had asked ‘‘what is it made of?’’ Pythagoras (569–c.500 BC) asked

‘‘what is its pattern?’’ thus introducing mathematics and geometry to thinking and dis-

course. Or, from a broadly inductive perspective, Greek concepts of reality which had sprung from observation (in noting basic enumerative issues), ended up becoming theories.

From this time, some of the first theories about the conceptualisation of reality followed

(Singer 1959a, b; Spangenburg and Moser 2004). It thus transpired that forevermore,

religious philosophies became challenged by the fundamental concept of a ‘‘patterned’’ or

mathematical Universe (Checkland 1993) and of consequence, increasingly rigorous

research processes. Hence, in momentarily jumping to the present, we think that the sorts

of pattern founded in the Greek rise of mental coherence, which are intrinsic to both the

SM and AR contemporary research methodologies, provide support for our argument.

However it is acknowledged that the steadfastness of [Pythagorean] pattern was hotly

debated by the Greeks. On one hand Heraclites (c.500 BC) suggested that conceptual

thinking involved a ‘‘becoming’’ or emergent form of reality. Thus all things involved

unending struggles and strife in a continual state of flux. The only thing of permanence was

the principle of change itself (Singer 1959a, b). On the other hand, Parmenides (c.515–450

BC) presented a ‘‘being’’ or static concept of reality interpreting it not experientially, but

by using an early form of dialectic inquiry. Attacking observational science, Parmenides

asserted the primacy of logic claiming that the senses were deceptive and that observation

was inferior to logical view. Also preferring the dialectic, Socrates (469–399 BC) broadly

favoured deductive inference where concepts of reality started with theories, and from which, sets of action principles were synthesized. Importantly, Socratic ‘‘truth by discus-

sion’’ perhaps first provided for a mixture of inductive and deductive inferences. Thus from this question and answer technique early abductive1 inferences allowed for an embryonic SM, via the formation of hypotheses and promotion of contextual action. We stress the

importance of this hypotheses formation within contextual action in SM research, because our experience (Barton 2007; Barton and Haslett 2007) is that similar processes are

intrinsic to the AR methodology. This commonality also contributes to our argument, but

more of that later.

Hence to close this very brief Ancient Greek introduction to belief systems and theory-

based research methods, we acknowledge that Aristotle (384–322 BC) further added to an

emergent SM basing his knowing on more rigorous testing, inquiry and experience. Thus

1 Abductive reasoning—reasoning in which explanatory hypotheses are formed and evaluated.

464 Syst Pract Action Res (2009) 22:463–474

123

generically, the Greeks had ‘‘argued for the sole purpose of arriving at the truth and, with

argument as their chief weapon; argument used deliberately, consciously, and carefully,

developed into a technical method’’ (Checkland 1993). Specifically, from rudimentary

observational science (given the then lack of instruments and measuring devices), and

different forms of inferential logic, distinct patterns began to emerge in regards the rigour

and relevance of research according to SM. In that regard, as the dynamics of pattern in

both the SM and AR methodologies are important, at this point we think it sufficient to say

that rigour and relevance create similar issues for AR. Thus, having first employed the

Ancient Greek era to establish some primary background linkages for our close rela- tionships argument, we now turn to the Scientific Age to further expand on some of the traditions of the SM and AR methodologies.

A Philosophical Basis for Closer Relationships in SM and AR: The Scientific Age

Leading into the scientific age, Ackoff (1993) paints a bleak picture of the intermediary

Middle Ages, the time after Ancient Greeks. Singer (1959a, b) agrees, referring to this time

as ‘‘the failure of knowledge’’. However, during this time, the Arabs translation and

preservation of Greek texts ensured that the systemic development of knowledge was based

on research that was progressive and cumulative. Later we stress the cumulative impor-

tance of reflective thinking in contemporary AR, but as the scientific era dawned, deductive

logic increased in its legitimacy as a method for securing truth. Thus following the Middle

Ages, the ‘‘revival of learning’’ of the Renaissance (Singer 1959a, b) saw what Walsh

(1992) encapsulates as man’s focus shifting from a fulfillment beyond this world, to a state of perfection within it. Observation and measurement (given new instruments) became rigorous co-contributors to the ‘‘neo-sciences’’, the intermediary phases that preceded what

were later to become methods of SM. In this regard, Checkland (2000, p. 36) provides a

succinct account of method and methodology and we take his view that a methodology is at

a meta-level with respect to a method. Methodology is therefore a body of methods where

the associated theory supports why those methods are appropriate.

With the application of observation and logic to the neo-sciences came great discoveries

and an even greater attention to the methods of their creation. Soon ‘‘incredible’’ dis-

coveries in the new field of science began to threaten the conformist theological doctrine. Francis Bacon (1561–1626) emphasised the importance of data, but Harvey (1578–1657)

went further showing how to record accurate data from observations. His seminal theory of

blood circulation demonstrates the importance of three aspects of SM: speculative reason

in the formation of hypotheses, accurate quantitative analysis and a blend of inductive and

deductive inference. Thus while accurate quantitative analysis is a cornerstone of SM, we later show how hypotheses formation and a blend of inductive and deductive (and

abductive) inference are paramount to contemporary AR methods.

Thus a rudimentary holistic scientific technique had emerged from the progressive and

cumulative Greek sequence of observation, mathematics, dialectic and logic, and as a

consequence, the iterative interplay of inductive and deductive inference brought maturity

to an emergent SM. Hence, in setting prediction as a standard for SM, man moved obli-

quely away from divine determinism. In doing this, classical Socratic thought processes

had come to evaluate both the scientific and theological doctrines or, as Einstein (1879–

1955) put it ‘‘science without religion is lame, religion without science is blind’’. Impor-

tantly, for our argument, two points now fundamental to AR stand out from this

emergence. Point one recognizes that reflective assessment and iteration had became

Syst Pract Action Res (2009) 22:463–474 465

123

established tools of refinement for SM. Point two acknowledges the immense power of the

dialectic in establishing concepts of reality through discussion.

Hence during the Scientific Revolution we had a time labeled as the most extraordi- narily productive period in the entire history of pure science (Debus 1992). Here we had a

fundamental, but still emergent SM, where new tools of measurement represented a tri-

umph of observation (broadly inductive empiricism) over logic (broadly deductive infer-

ence). Thus also important to our argument, an ongoing dynamic or ping-pong interplay

between inductive-deductive inferences came to the fore. We associate this dynamic with

the sort of reflective iteration produced by double loop learning processes now charac-

teristic in contemporary methods of AR (Bateson 1964; Argyris 1982; Nonaka 1994;

Nonaka and Takeuchi 1995; Checkland 2000). But perhaps more important to our argu-

ment, we then further suggest that reflective iteration underpins the refinement of most AR

methods. This view is evident in Checkland’s (2000, p. 19) description of his Soft Systems

Method[ology] (SSM);

…the [method] is presented as a sequence of stages with iteration back to previous stages, the sequences being: analysis; root definition of relevant systems; concep-

tualization; comparison and definition of changes; selection of change to implement;

design of change and implementation; appraisal

We thus contend that the dialectic between the broad connectivity of analysis/

induction, and synthesis/deduction, added to the creation of cyclic processes within a

universal SM paradigm. The fundamental steps of that paradigm: observation, hypoth-

esis, experimentation and generalization thus became indoctrinated into all particular

methods of science. Even Newton’s thinking often began with intuition, the great

abductive process, later supported by experiment, deductive inference and mathematical

confirmation. In support of this view Stuewer (1970) says that ‘‘Newton invents theories,

he proposes a radically empiricist methodology, and he claims that he has obtained the

former with the help of the latter’’. Philosophically, the SM paradigm thus suggested that

belief secured from the pillars of Physics, Chemistry and Biology could be explained

by relatively simple sets of Mathematical laws. However, problems in Physiology,

Psychology and some of the then newer sciences, on the whole, struggled with attempts

to integrate Newtonian mechanics into their methodologies. Hence, the appropriateness

of subjecting the social sciences to strains of quantitative measurement, appropriate to

the physical sciences, came to the fore.

The subsequent development of ‘‘softer’’ perspectives on reality may be linked to the

times of the German Idealist Immanuel Kant (1724–1804). From Kant’s reflection on

failed attempts to explain various aspects of human nature came his concept of a priori

knowledge where, though common principles have to be accepted as existing indepen-

dently from our sense perceptions, they cannot be proved. Kant declared that the methods

of the physical sciences were founded on a priori concepts ‘‘as well as’’ sensory experi-

ences. Hence he suggested that we can only have access to phenomena (human knowledge)

not reality itself. This is critical to the importance we place on the development of

reflective inquiry in both SM and AR over time. Kant went on to say that the securing

of knowledge in the physical world really occurred by using subjective perspectives of

rational systems such as Mathematics. He therefore contested the appropriateness of SM as

a means for dealing with reality in the metaphysical and social science worlds.

In framing our argument, we have found Kant’s thinking to be an important precursor

from Piaget, Peirce and Pavlov. In this regard, we highlight the significance of Peirce’s

logic, semiotics and four main methods of fixing belief (tenacity, authority, a priori and

466 Syst Pract Action Res (2009) 22:463–474

123

science) in linkages we make in a paper (Barton et al. 2007) through James, Singer, Simon,

Churchman, Forrester and Ackoff to culminate in Beer’s (1972, 1979, 1985) Viable

Systems Diagnosis (VSD). Further, Applebaum (1992, p. 169) says that Peirce (1839–

1914) rectified an empirical imbalance by suggesting not only the justification of analysis

or ‘‘scientific’’ knowledge, but by positing the issue of knowledge growth through the

circular reflection and a blend of empiricist views and a priori concepts of logic to improve

knowledge. This ‘‘pragmatic’’ view of epistemology thence gained popularity in areas

other than the physical sciences, as an acceptable way of securing reality. Thus by the turn of the nineteenth century, empiricism and induction contributed to

rigor in the observational and formation of hypothesis steps on one hand, while on the

other, the deductive or rational approaches advanced rigor in the experimentation and

generalization steps of SM. The importance of this is shown by Kemeny (1959) quoted in

Quade and Miser (1985) who describe the SM by way of Einstein (1879–1955):

As Einstein has repeatedly emphasized…. First of all the scientist is an observer. Next he tries to describe in complete generality what he saw, and what he expects to

see in the future. Next he makes predictions on the basis of his theories, which he

checks against the facts again. The most characteristic feature of the method is its

cyclic nature. It starts with facts, ends in facts, and the facts ending one cycle are the

beginning of the next cycle. A scientist holds his theories tentatively, always pre-

pared to abandon them if the facts do not bear out his predictions. If a series of

observations, designed to verify certain predictions, force us to abandon our theory,

then we look for a new and improved theory.

This definitive expression of SM, which may also be depicted as a four stage, contin- uous and iterative learning cycle, thus brings some common assumptions to contemporary AR learning frameworks, further assisting our closer relationship argument. For example Dewey (1943), Deming (1982), and Flood (1999) base their action-learning frameworks on

the ‘‘cyclic’’ method as described by Einstein above. Such learning frameworks have their

roots in Lewin’s (1948, 1951) seminal AR work, which was founded as an applied science. Much of Lewin’s early work relied on the observation and experiment of real social

contexts to promote learning. Thus his divergence from Psychology to create action

research was absolutely foundational for the AR community. However as Walsh (1992,

p. 192) explains, the study of man and society is different from the study of planets and

electrons because human reality becomes largely known through participation in it. Thus while promoting our closer relationship argument; we clearly acknowledge disparities between SM and AR practices, and the necessity of taking into account multiple stand-

points on perceptions of reality. Thus in closing this philosophical basis for some closer

relationships in SM and AR from the Scientific Age, we now move away from specific

epochs to give our perspective on worldviews over all time. In doing this, we will link

various world hypotheses with an underlying logic to further progress our argument.

A Philosophical Basis for SM and AR: Worldviews

The expression worldview may be traced back to Immanuel Kant. Kant ([1790 (1987)])

translated the German weltanschauung to mean—look onto the world. Worldviews are thus ways in which people make subjective representations of external reality. Lackoff and

Johnson (1999, p. 511) describe worldviews as:

Syst Pract Action Res (2009) 22:463–474 467

123

a consistent constellation of concepts, especially metaphorical concepts over one or

more conceptual domains. Thus one can have philosophical, moral and political

worldviews. Worldviews govern how one understands the world and therefore

deeply influence how one acts.

Hence, given the subjective inadequacies of all world hypotheses, we reiterate the view

(Barton 2007; Barton and Haslett 2007, p. 149/50) that Pepper’s four ‘‘world hypotheses’’

(described in Table 1) may be used to provide a useful foundation for reviewing epistemic

development over all time. In that regard, Pepper’s world hypotheses are associated with

‘‘root metaphors’’ that sit between the extremities of dogmatism and utter scepticism.

Further, and consistent with the dynamic or ping-pong interplay of inductive-deductive

inferences we have described as emergent from the Scientific Revolution, we concur with

the importance that Pepper places on the inter-relationships of these classifications (Fig. 1)

thus linking each world hypothesis with an underlying logic. Hence, in reading down the

columns of Table 1, the first two logics increase in analytic power and the second two

increase in synthetic power. However in this regard, we side with Russell’s (1946, p. 1)

view that ‘‘almost all the questions of most interest to [modern] speculative minds are such

as science cannot answer, and the confident answers of theologians no longer seem so

convincing as they did in former centuries’’.

Now rather than attempt to answer questions such as: ‘‘Which hypothesis may or should

come first?’’ or ‘‘In what order and how many revisits to each hypothesis are required?’’ we

take a broad view that there are acceptable roles from contemporary management system

constructs which may be associated with each metaphor. Thus for example we say (Barton

and Haslett 2007) that management classification systems such as financial accountability,

market segmentation and fundamental organisational structures best link with Formism, Mechanism may associate with strategic management and dynamic strategy models (Warren 2002; Sterman 2000) and highly non-linear (Organicism) models can ally with

Table 1 Pepper’s root metaphors (Pepper 1942)—adapted from Barton and Haslett (2007)

World hypothesis Root metaphor School of philosophy

Formism Similarity Realism or platonic idealism

Mechanism Machine Naturalism or materialism

Organicism Organism Absolute idealism

Contextualism Historical event American pragmatism

Analytic Theories Synthetic Theories

Formism Mechanism Contextualism Organicism

Dispersive Theories

Integrative Theories

(Inadequacy of precision) (Inadequacy of scope)

Fig. 1 World hypotheses (Pepper 1942, p. 146)—from Barton and Haslett (2007)

468 Syst Pract Action Res (2009) 22:463–474

123

chaos theory (Guastello 1995). Further, we suggest that Contextualist models may offer explanatory powers when the co-evolution of businesses and their environments produce

innovative strategies. In this regard, we view Contextualism as the only hypothesis that admits of human, purposeful behaviour, as Formism and Mechanism are the results of such behaviour. For all these reasons, we believe the Contextualist hypothesis may reasonably be viewed as a background for the sort of theory that underpins AR. Thus from an epoch

perspective, we associate the time of the Ancient Greeks with Formism and Newtonian Science with Mechanism. Further, we now associate Pepper’s third root metaphor Organicism with the Modern Era worldview to look at the formalization of Systems Thinking and the development of the Organisational Behaviour (OB) and AR intellectual

streams.

However we are not saying that Formism and Mechanism have disappeared during the Modern Era. Whereas the mechanistic view had claimed that analysis and reductionism

could explain all objects and events, Vitalists were among those biologically minded who

queried the capacity of Mechanism and/or Formism to sufficiently explain behaviour in complex entities where intrinsic properties could not be derived merely from an under-

standing of their parts. In describing equifinality as a mechanism whereby a system has a specified goal or final state, which it may reach in different ways from different working

conditions—the goal is equifinal, we agree with Beer’s (1959, p. 168/9) inference that von Bertalanffy’s (1968) proof that closed systems cannot behave equifinally, effectively

rebuked Vitalism.

In the Modern Era, von Bertalanffy’s General Systems Theory (GST) and the biological

metaphor underpinned methods that explored complex systems including then embryonic

organizations and laid the claims of the Vitalists to rest. Our views on organisations and

thinking in the Modern Era are however strongly influenced by Systems Thinking with its

foundations in the both biological and mechanistic metaphors. The concept of inherent

feedback systems was fundamental to knowledge about the functionality of organisational

systems. Systems Thinking is now strongly endorsed (Simon 1979; Senge and Sterman

1992; Park 1999) as a valid approach for securing of knowledge about functionality in

organisational systems. Our belief is that the enhanced systemicity of the 1990s introduced

new ways of thinking to people concerned about understanding through Kant’s a priori

knowledge.

Our line of thinking suggests that some discontent with mechanistic thinking led to the

formalisation of Systems Thinking around the middle of the twentieth century. From that

time, OB has progressed to become a justifiable way of securing knowledge about man-

agement practices in complex organisational systems. However disappointment with some

of the OB methods, basically generated through the analysis-synthesis dialectic, then

germinated a different but not unrelated interest in knowledge created in and of organi- sations, rather than that created by the application of static prescriptions onto them. Hence AR, a different stream of thinking, was founded.

We repeat that there is no simple answer to the question ‘‘What is AR?’’ (Checkland and

Holwell 1998; Reason and Bradbury 2001). But foundationally, we think of Lewin’s

seminal AR work in Sarton’s (1952) terms, as a particular ‘‘cut’’ in social history where the

subsequent interconnection of Systems Thinking, OB and AR gave substance to a para-

digm shift about how research might be conducted in and about organisations. As Midgley (2003a, p. xviii) puts it ‘‘by drawing upon the full variety of systems ideas, we should be

able to produce a more rounded understanding of people, organisations, societies and the

world we live in, than could emerge from any of the traditional [deterministic] scientific disciplines’’ [our emphasis]. Here, Midgley (2003b, 2007) means that the traditional

Syst Pract Action Res (2009) 22:463–474 469

123

scientific approaches are useful, but limited, unless they are integrated into a more systemic

perspective. Scientific methods can therefore become part of a systems thinker’s toolkit,

but whether those purposes are appropriate needs to be assessed systemically. This view

emphasises a potential shift in thinking as research is conducted in and by rather than on organisations.

We have now presented a Western philosophical view on the development of belief

systems and theory-based methods in the light of recognised worldviews. That view is

grounded on our core belief that knowledge may be developed through an analysis-syn-

thesis dialectic. In promoting Systems Thinking, OB and AR as meaningful ‘‘new’’ ways of

producing knowledge, our view is that Pepper’s Organicism root metaphor has been influential in the development of knowledge. It is also our belief that Pepper’s fourth root

metaphor Contextualism grows in importance, particularly for emerging worldviews such

as Post-modernism. In this regard, Midgley (2007) says that it is not just Post-modernism

in which Contextualism is important. It is as important to acknowledge that all of the soft systems approaches and the most recent of epistemic theories embrace Contextualism.

Close Relationships in the SM and AR

Barton and Haslett (2007) argue that both SM and the development of knowledge can be

described in terms of the analysis-synthesis dialectic and that this process can be used to

describe both the ‘‘grand’’ developments from science and the ‘‘micro’’ developments at

organisational and personal levels. When supported by the system of inferential logic

described by the nineteenth century pragmatist philosopher, Charles Sanders Peirce (Peirce

1931/1958) this process also provides a rigorous framework for defining both AR and

Positivist Research and demonstrates that they are complementary processes. The process

As a result of the dialectic process, new “categories” emerge with

increasing power of explanation of data

Observable data ( “Surprising facts” ) Initial hypothesis

Synthesis using Abductive Reasoning Generating new hypotheses

Action & Analysis using Deduction and Induction generating new data

New hypothesis/ category

New hypothesis/ category

Emphasis on Wholes

E m

p h

a s

is o

n P

a rt

s

New data

New data

Time

Fig. 2 The scientific method as analysis and synthesis dialectic

470 Syst Pract Action Res (2009) 22:463–474

123

is described in Fig. 2. It starts with the observation of a number of events for which there is

no obvious and immediate explanation and for which there is a desire to gain a coherent

explanation. In Ackoff’s terms these situations are described as ‘‘messes’’. Peirce describes

them as ‘‘surprising facts’’.

A new synthesis of these ‘‘surprising facts’’ leads to a primary ‘‘thesis’’. Then, analysis,

experimentation and/or action taken on the basis of this thesis, leads to differences between

the explanatory powers of the thesis and observation. This leads to the development of an

‘‘antithesis’’ and so the dialectic is established. The resolution of this dialectic results in a

new synthesis or hypothesis, which becomes the thesis for the next cycle of dialectic. The

concept of the cycle of dialectic is central to SM method, in the same way, that the process

of cyclic reflection is important to AR. The essential distinction is however between an

open and closed system of enquiry. Traditionally scientific enquiry takes place in a closed

system where the influence of the environment is minimised whereas in AR the influence

of the environment is part of the experiment.

To illustrate this, Senge (1990) identified ‘‘five disciplines’’ that define a set of capa-

bilities for any organisation aspiring to become a ‘‘learning organisation’’. Importantly, in

each case these factors form part of a coherent, systemic framework, which can be used to

explain what initially appears as an ill-defined and largely intuitive set of observations and

ideas. In each case, a number of ‘‘surprising facts’’ are synthesised into a systemic

framework in the sense that they describe an integrative framework based on an organising

principle—the systems principle (Emery 2000). Over time, these frameworks will be tested

and compared with other possible frameworks leading to the emergence of new frame-

works and theories. For example, Senge’s learning framework can be compared and tested

against Nonaka’s (1994) knowledge creation construct based on a tacit knowledge-explicit

knowledge dialectic (Nonaka and Takeuchi 1995). What is significant is that each of these

frames is ‘‘systemic’’. Of significance to our argument is Einstein’s identification of a

similar process in physics (Fig. 3) as the jump from the observed facts to the set of

‘‘Axioms of Fundamental Principle’’:

Speculative leap based on hunch, conjecture, inspiration, and guesswork…. We are dealing, after all, with the private process of theory construction or innovation,

the phase not open to inspection by others and indeed perhaps little understood by the

originator himself. But the leap to the top of the schema symbolizes precisely the

precious moment of great energy, the response to the motivation of ‘‘wonder’’ and

the ‘‘passion of comprehension’’. (Holton 1998, p. 31)

Fig. 3 Einstein’s model for constructing a scientific theory (Holton 1998)

Syst Pract Action Res (2009) 22:463–474 471

123

In summary, each cycle of dialectic described in Fig. 2 can be described as a learning

process starting with ‘‘surprising facts’’ which lead to the formation of a systemic framing

of the ideas; that is, an explanatory hypothesis. Barton and Haslett (2007) argue that this is

describes the role of the systems concept in science—the framing of explanatory

hypotheses. There is a close similarity between this description of SM and that identified

by Peirce as being analogous to Cuvier’s evolutionary process. This process emphasizes

that science advances in leaps, compared to the more stable evolutionary processes

identified with Darwin and Lamarck (Sharpe 1970; Tuomi 1992). But it must be noted that

as modern theories of biological evolution also involve leaps, this comparison looks to be

problematic. To relate our dialectic version of SM to AR, it is necessary to show its linkage

with experiential learning theory. In particular, the process described in Fig. 2 can be

related to Dewey’s ‘‘Spiral of Learning’’ shown as Fig. 4;

But to make this connection explicit, it is necessary to articulate the role of inferential

logic in this process, and abductive inference in particular. Dewey’s experiential logic is

drawn from Peirce’s notion of the continuity of inquiry supported by his description of

inferential logic (Dewey 1938). However that matter will be the subject of a later paper

which will further support our argument that SM (which relates to closed systems thinking)

and AR (which relates to open systems thinking) are both essential to any complete scientific

approach. Finally, we think we can define AR clearly, if we use the open-closed systems

framing as distinct from the soft-hard systems framing, which amongst other things, ends up

confusing ontology and epistemology. As such, that the traditions of SM and AR have much

closer relationships than people often give them credit for, will be further pursued.

Conclusions

We have presented a Western philosophical view on the development of belief systems and

theory-based methods over time and linked the progressive and cumulative development of

Fig. 4 Dewey’s experiential learning cycle (Kolb 1984)

472 Syst Pract Action Res (2009) 22:463–474

123

SM with the contemporary AR methodology. From this, we conclude that for both SM and

AR the development of knowledge comes through an analysis-synthesis dialectic. And

while SM takes place in a closed system where the influence of the environment is

minimised, AR recognizes and integrates the influence of the environment into the enquiry

process. This places AR in an open, rather than a closed, systems context. As such, we have

contributed to the view that the traditions of the SM and AR have much closer relationships

than people often give them credit for.

Acknowledgement We acknowledge and thank Gerald Midgley for his ideas, suggested reference materials and constructive critique which have greatly assisted in the preparation of this paper.

References

Ackoff RL (1993) From mechanistic to social systemic thinking. Pegasus Communications, Cambridge, MA Applebaum W (1992) Epistemological and political implications of the scientific revolution. In: McKnight

SA (ed) Science, pseudo-science, and utopianism in early modern thought. University of Missouri Press, London, UK

Argyris C (1982) The executive mind and double loop learning. Organ Dyn 11(2):5–22 Barton J (2007) Systems thinking: philosophy, methodology and applications to knowledge management.

Business and economics (management). Unpublished PhD dissertation, Monash University, Melbourne, Australia, p 297

Barton J, Haslett T (2007) Analysis, synthesis, systems thinking and the scientific method: rediscovering the importance of open systems. Syst Res Behav Sci 24:143–155

Barton J, Haslett T et al (2007) Action research: its logic and relationship to the scientific method. ANZSYS, Auckland, NZ

Bateson G (1964) Steps to an ecology of mind. Ballantine Books, New York Beer S (1959) Cybernetics and management. English Universities Press, London Beer S (1972) Brain of the firm. The Penguin Press, Allen Lane Beer S (1979) The heart of the enterprise. Wiley, London Beer S (1985) Diagnosing the system for organizations. Wiley, London Checkland PB (1993) Systems thinking, systems practice. Wiley, Chichester, England Checkland PB (2000) Soft systems methodology: a thirty year retrospective. Syst Res Behav Sci 17:S11–58 Checkland PB, Holwell S (1998) Action research: its nature and validity. Syst Pract Action Res 11(1):9–21 Debus AG (1992) Science and history. The birth of a new field. In: McKnight SA (ed) Science, pseudo-

science, and utopianism in early modern thought. University of Missouri Press, London, UK Deming WE (1982) Out of crisis. Mass. Inst. of Technology, Cambridge, MA Dewey J (1938) Logic: the theory of inquiry. Holt, Rinehart and Winston, New York Dewey J (1943) The child of the curriculum and the school and society. University of Chicago Press,

Chicago Emery M (2000) The current version of Emery’s open systems theory. Syst Pract Action Res 13(5):623–644 Feyeraband PK (1970) Against method. University of Minnesota Press, Minneapolis, MN Flood RL (1999) Rethinking the fifth discipline: learning within the unknowable. Routledge, London Garrick J (2000) The construction of ‘working knowledge’ and (mis) interpretive research. In: Garrick J,

Rhodes C (eds) Research and knowledge at work: perspectives, case studies and innovative strategies. Routledge, London, UK

Guastello SJ (1995) Chaos, catastrophe and human affairs – application of non-linear dynamics to work, organisations and social evolution. Lawerence Erlbaum Associates, Mahwah, NJ

Holton G (1998) The advancement of science and its burdens. Cambridge University Press, Cambridge, UK Kant I ([1790 (1987)]) Critique of judgment. USA Hackett, Indianapolis Kemeny JG (1959) A philosopher looks at science. Von Nostrand Reinhold, New York Kolb DA (1984) Experiential learning. Prentice-Hall, Englewood Cliffs, NJ Kuhn ST (1962) The structure of scientific revolutions. University of Chicago Press, Chicago Lackoff G, Johnson M (1999) Philosophy in the flesh: the embodied mind and its challenge to western

thought. Basic Books, New York Lewin K (1948) Resolving social conflicts. Souveneir Press, London Lewin K (1951) Field theory in social science: selected theoretical papers. McGraw-Hill, New York

Syst Pract Action Res (2009) 22:463–474 473

123

Midgley G (2003a) Systems thinking, vol I–IV. Sage, London, UK Midgley G (2003b) Science as systemic intervention: some implications of systems thinking and complexity

for the philosophy of science. Syst Pract Action Res 16:77–97 Midgley G (2007) Traditional scientific disciplines and systems thinking. Melbourne, Australia: e mail

communication Nonaka I (1994) A dynamic theory of organisational knowledge creation. Organ Sci 5(1):14–37 Nonaka I, Takeuchi H (1995) The knowledge creating company: how the Japanese companies create the

dynamics of innovation. Oxford University Press, New York Park P (1999) People, knowledge, and change in participatory research. Manage Learn 30(2):141–157 Peirce CS (1931/1958) Collected papers, vol I–VIII. CD-ROM from Past Masters Pepper SC (1942) World hypotheses. University of California Press, Berkeley, CA Quade ES, Miser HJ (1985) Handbook of systems analysis: overview of uses, procedures, applications and

practice. In: Midgley G (ed) Systems thinking, vol 2. Sage, London, UK, pp 241–270 Reason P, Bradbury H (2001) Handbook of action research. Participative inquiry and practice. Sage,

London, UK Reason P, Heron J (1995) Co-operative inquiry. In: Smith JA, Harré R, Van Langenhove L (eds) Rethinking

methods in psychology. Sage, London, UK Russell B (1946/2004) History of western philosophy. Routledge Classics, London, UK Sarton G (1952) A guide to the history of science. The Ronald Press Company, New York Senge P (1990) The fifth discipline. The art and practice of the learning organisation. Random House,

Melbourne, Australia Senge PM, Sterman JD (1992) Systems thinking and organisational learning: acting locally and thinking

globally in the organisations of the future. Eur J Oper Res 59(1):137–150 Sharpe R (1970) Induction. Abduction, and the evolution of science. Trans Charles S Peirce Soc 6/1:17–33 Simon HA (1979) Rational decision making in business organisations. Am Econ Rev 69:493–513 Singer CJ (1959a) Short history of scientific ideas to 1900. Oxford University Press, London, UK Singer EA (1959b) Experience and reflection. University of Philadelphia Press, Philadelphia Spangenburg R, Moser DK (2004) The birth of science: ancient times to 1699. Facts on File, New York Sterman JD (2000) Business dynamics: systems thinking and modeling for a complex world. Irwin/

McGraw-Hill, Boston, MA Stuewer RH (1970) Historical and philosophical perspectives of science. University of Minnesota, Gordon

Breach, New York Tuomi J (1992) Evolutionary synthesis: a search for the strategy. Philos Sci 59:429–438 von Bertalanffy L (1968) General systems theory: foundations, developments, applications. G.Braziller,

New York Walsh D (1992) A mythology of reason. In: McKnight SA (ed) Science, pseudo-science, and utopianism in

early modern thought. University of Missouri Press, London, UK Warren K (2002) Competitive strategy dynamics. John Wiley and Sons, Chichester

474 Syst Pract Action Res (2009) 22:463–474

123

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.