**1. Introduction**

There are many aspects of systems science that occupy the minds of its practitioners. Increasing the widespread use of high-quality systems thinking is among them. Enthusiasm for this goal is based on the conviction that thinking systemically and applying systems knowledge to critical issues are of evolutionary significance to our world. Fueled by this conviction, the systems body of knowledge is growing, being codified, and evaluated for its progress in occupying a rightful place among other scientific disciplines.

Recent work on systemology (e.g., [1,2]) is working to formalize systems knowledge so that educational efforts can give people "clear concepts and a common language that gives them the capability to articulate and reflect on this (systemic) sensibility, and act upon it in a considered way" [3] (p. 3). Further, formalization and other efforts to increase the quality of systems knowledge can enable systems thinkers to "gain influence in supporting organizations, and through that influence to better enable systems thinking and acting of individuals and groups, which may (in turn) lead to more quality in how people deal with complex challenges" [3] (p. 5). At a macro level, the systemology initiative is working to organize our understanding of the body of knowledge about systems relative to other scientific disciplines and to identify key gaps in that body of knowledge.

At a micro level, individual people who consciously work with systems inhabit varying roles—systems engineers, systems scientists, systems dynamicists, and systems practitioners among them. People in each of these roles deal with varying issues and situations and therefore face varying technical demands; a grea<sup>t</sup> range of systems competencies is required for the different roles that systems workers play. Future research would be useful to identify the range of roles people do play with respect to systems to clarify the characteristics of the varied personae they require to do systems work. The present paper addresses a different matter. Alongside the ongoing macro-level work to identify the maturity of the systems body of knowledge, there is a need to address, at a more micro level, the levels of competence required for systems thinkers to be effective.

From a "resource-based view of the firm" [4], the unique competencies of people are a resource as crucial as the combination of capital, physical, and other tangible resources that make up a human system. Measuring competence levels, it is reasoned, should be as important to an organization as measuring operational efficiency or financial performance [5]. In business, things ge<sup>t</sup> measured so that action can be taken to manage and improve them. Increasingly, this premise is being applied to the degree of maturity that an organization's people exhibit, i.e., the degree to which they are applying the knowledge and behaviours necessary to achieve excellence. There is increasing interest in identifying reliable ways of measuring such competence. Without the ability to reliably measure competence, it is difficult to strategically focus organizational attention to things such as "staff development, recruitment and selection, professional registration, training needs, analysis and planning, job descriptions, assessment and appraisal" [6] (p. 16). Without this measurement ability, pragmatic action to increase competence maturity is difficult to take.

From the academic standpoint, Rašula, Vukši´c, and Štemberger have noted that without reliable ways to measure competence maturity, "a comprehensive theory of knowledge or knowledge assets is very difficult to develop. Consequently, there is no visible progress in the effort to treat knowledge as a variable to be researched, or asset to be managed" [5] (p. 48). The argumen<sup>t</sup> here is that maturity assessment is important in developing a common understanding of what comprises fundamental concepts such as *knowledge* and *development*. Creating a maturity assessment demands "refinement of a general, unified representation" of such concepts across specialties [7] (p. 83), which can enable consistency in communicating and operationalizing these ideas. These considerations, arising from organizational theory, are relevant to the issues identified by the editors of this special issue on systems thinking. Without reliable ways to measure the maturity of anyone's competence in systems thinking, it is difficult to develop a theory of systems knowledge. It is difficult to argue for the treatment of systems thinking as an asset for an individual, an organization, a profession, or an academic discipline. It is difficult to agree on how to define exactly what is systems thinking and difficult to know how to develop it.

We can consider another difficulty. The presence of maturity models can assist the development of a comprehensive theory about systems thinking. At the same time, a comprehensive theoretical foundation would be invaluable in developing a maturity model. Which should come first: models of systems thinking maturity, or theories about systems thinking maturity on which models could be based? Perspectives on how such efforts should be sequenced differ according to stakeholder group. Businesses want to focus on what is known about success, what works, and what can be improved. They want models that enable employees to take pragmatic action on present-day demands. They do not have the luxury to await fully developed theories and formalized knowledge before action can be taken. Yet, given the pressing nature of many of the problems to which systems thinking is being applied, undertheorized systems thinking also seems to be a risk. Here, the academic community can help to mitigate the risk, clarifying knowledge and other relevant constructs, thus helping industry to have confidence in the validity of what gets measured and is believed to create success. Scholarly efforts can improve the advancement of science (while also complementing business' concerns) by focusing on the development of understanding while being less constrained by demands for short-term focus and commercial profitability [8]. Thus, the question of how to prioritize systems thinking maturity models versus theory is a question of objectives: is the goal to make contributions to systems thinking knowledge, or contributions to successful systems projects? Academics and industrialists would answer this differently.

Despite assumptions that people who run organizational systems and people who study them might have similar interests [9], disjoints between academics and practitioners have long existed.<sup>1</sup>

<sup>1</sup> For instance, most organizations do not employ theoretically sound managemen<sup>t</sup> practices even though mounting evidence shows that certain empirically tested business practices are related to greater employee performance, productivity, and overall financial performance for the entire company [9,10].

"The gap is not restricted to the organizational sciences, but rather it is found in nearly all fields in which there are both researchers and practitioners" [8] (p. 340). There are reasons for this. Among them, systems scholar Ian Mitroff has observed that academics and practitioners differ in the types of information they believe constitute valid bases for action [11]. Such difference notwithstanding, the efforts of those interested in research and applied practice have been shown to benefit both [9,10]. Studies have shown that both quality and rate of knowledge creation are enhanced when tensions are present between differing stakeholders, disciplines, hypotheses, theories, prototypes, and implementation strategies [12–16]. This would sugges<sup>t</sup> that two parallel streams of focus—pragmatic maturity model development, and scholarly development of theory about systems thinking maturity—would both be valuable; each would be incomplete without the other, and each would be strengthened by continued improvements in the other.

Thus, given the prospect of gains for both theory and practice, maturity models have emerged in many disciplines: psychology, business processes, project management, and knowledge managemen<sup>t</sup> among them. In each case, the models have enabled comparisons among people and have had the effect of normalizing specific skills and behaviours deemed important to effectiveness [17]. As such, maturity models are having increased significance in aiding those interested in pursuing performance excellence in many realms of human endeavour [18]. Were systems communities to embark on the task of developing a means of assessing systems thinking maturity, we would be closer to clarifying what comprises systems competence across the widely varying subspecialties of systems theory and practice. Identifying unified, disciplined ways of representing systems thinking competence would go far to promote understanding and cohesion among these varied schools of systems thought. Likewise, it could contribute to the communities' ability to promote the value of systems knowledge in society and improve the effectiveness of cross-disciplinary dialogue that systems communities could have with other scientific disciplines.

The purpose of this paper is not an exhaustive survey of existing maturity models. Rather, it will discuss construct definitions, types of maturity models, critiques, and design considerations with an eye to examining how maturity models could be useful evaluation tools for systems thinking proponents to consider.

#### **2. Maturity Models: Fundamentals**

Maturity models are premised on the idea that successful performance is the result of effectively used knowledge. As such, models are tools for practitioners striving for excellence, groups interested in promoting collective expertise, and for scholars working to build developmental theories [19,20]. However, when applied to systems thinking, an immediate problem presents itself: what *is* systems thinking actually? A ubiquitous term [21], in the past 50 years many scholars have posited many definitions, e.g., [22–25]. Additionally, a sizable popular literature has emerged to teach people what it is and how to do it [26]. A central feature of systems thinking articles and books, in most cases, is the claim that becoming a skilled systems thinker rests on understanding what a system is and why systems tend to operate as they do.

According to social psychology, a fundamental first step in any epistemic is a knowledge domain [27]. In many fields, such a requirement is institutionally recognized: to be an accountant, one must have accounting knowledge; to be an historian, one must study history; to be an engineer, one must know one of several core disciplines; etc. [28]. It follows that the same logic would apply to an epistemic of systems thinking. However, regarding systems thinking, Checkland has been dismissive about anything that might be considered a knowledge domain:

This is about the limit of what we can say about every example of systems thinking ... there will be an observer who gives an account of the world, or part of it, in systems terms; his purpose in so doing; his definition of his system or systems; the principle which makes them coherent entities; the means and mechanism by which they tend to maintain their integrity; their boundaries, inputs, outputs, and components; their structure ... [24] (p. 102).

Despite this pessimistic tone, a case can be made that a knowledge domain about systems themselves does exist.

For example, the literature pertaining to sociotechnical systems (e.g., [29] and elsewhere) is consistent in describing fundamental tenets such as these:

People exist in relationship [21] and "systems ... are intrinsically concerned with relationships" [24] (p. A24). Anyone attempting a systemic view must attend to "detail complexity" (i.e., the number and individual characteristics of a system's members [30]). Also important are the number and qualities of relationships that a system contains.

Members of systems require one another to achieve their goals. They are connected in relationships of interdependence [31,32]. The strength of interdependence (or "interconnection" [33]) among members varies, affecting the degree to which a system can be viewed as cohesive.

Human systems are purposive [24]. Such purposes arise from the ways members organize meanings [25]. A system's behaviour expresses explicitly espoused purposes [34] and also tacit purposes which may be unrecognized by members themselves [35,36].

The manner in which systems are organized (intentionally or otherwise) arises from interactions among the system's members [23,37]. While such interactions tend to be patterned, they are also dynamic [21,38]. Hence, emergence is a feature of human systems [39].

Sociotechnical systems contain dichotomies and tensions [26]. Working effectively with these requires, for example, that one consider both a system's parts [25,38] and its nature as a coherent whole [24,36,39], akin to working with both "figure and ground", as described by Gestalt principles of perception [40].

Though not exhaustive of the principles elucidated in systems books and articles, these tenets illustrate central ideas in the systems knowledge domain. Still, how can we know when a person is relating to such knowledge maturely or not?

Designers of maturity model tools imply that the way one uses knowledge reflects one's location on a scale of immaturity-to-maturity.<sup>2</sup> Popular encyclopedias such as Wikipedia convey that maturity involves:


People exhibit these abilities to varying degrees (that is, at various levels of maturity). The means by which maturity is demonstrated is *competence*, a concept of increasing interest since a seminal publication in the educational testing literature in the 1970s [43]. Since then, it has become a construct of increasing interest in theories of organizational behaviour, popularized by Boyatzis [44] and other

<sup>2</sup> Despite connotations that immaturity involves undesirable deficits (e.g., [7]), it is worth noting that psychologists have proposed that the stage of immaturity is an important time of experimentation [41,42] that is valuable to the development of individuals (and the evolution of a field's theoretical understanding of a developmental phenomenon [41]).

scholars. A comprehensive review of competence is beyond the scope of this paper, but for our purposes, in the realm of maturity models, competence at systems thinking would encompass a collection of traits, motives, self-image, and perceptions of social norms and behaviour enabling a person or group to direct systems knowledge in such a way that desired results are consistently achieved [6,45].

"Desirability" of the results achieved is readily measured in organizational settings. For example, in education, a curriculum gets delivered that is shown to enable students to meet intended learning objectives; in systems engineering, a complex structure is delivered that is designed to function effectively across its life cycle; in artificial intelligence, machines are created to accurately mimic human cognition. In any of these cases, success is measured in how well a team identifies and solves necessary problems in such a way that they develop and execute projects that meet sponsors' intended goals and success criteria [29,38,46].

However, binary thinking (e.g., desirable results vs. undesirable results, maturity vs. immaturity) could readily lead us to view systems thinking as something one can do or cannot. Writers have argued to the contrary, that systems thinking can be developed [21,29]. The maturity model literature takes pains to state that the use of specific knowledge is deemed important both in the capacity to both operate in a competent manner [6] and to improve that capacity: "Improvement ... require[s] some guidance on what to improve, and in identifying improvement efforts that will provide the most value ... Conducting assessments provides guidance in terms of current capabilities and identification of performance gaps, helping to identify where improvement is possible, necessary, or desirable" [47]. As we can speak of improving the capacity to deliver desirable results, so too does the maturity model literature speak of improving maturity. It does this by conceptualizing maturity in terms of gradations.

A central assumption of maturity models is that development of mature performance occurs in predictable patterns. This assumption is evident throughout scholarly papers in many disciplines that address maturity, for example:


Proponents of maturity models believe that maturity is comprised of "tightly defined, repeatable, and predictable processes [that] directly contribute" to capable behaviour [51] (p. 147). The sense-making patterns underlying capable behaviour with respect to systems is largely unknown at this time. If they were understood, then models could be built to diagnose the maturity of a person's (or organization's) current systems thinking practices (i.e., the stage at which practice has stabilized [52]), in order to understand that person's (or organization's) current standing relative to others who are competent in systems thinking [5,53].

Maturity models are guided by the assumption that specifically interlinked collections of competently used knowledge and skills [52] comprise coherent *levels* or *stages* of maturity. Levels are ordered, thereby creating a hierarchical concept system that enables comparative ranking of different persons (or groups) and models a process of evolution by which a person (or group) can move toward increasingly sophisticated and reliable performance [17,54]. Different maturity stages are understood to be appropriate for achieving tasks of varying levels of complexity, giving rise to the notion of competence *fit*, which is another facet of maturity models. Rather than assuming that the highest possible level of maturity is inherently necessary, adherents of maturity models generally agree that maturity level should be matched to the difficulty of the task at hand, problems to be solved, and environmental context in which one is operating. At present, there are no consensually agreed-upon levels of sense-making based on systems thinking skills that could inform the assignment of systems thinkers to systemically complex projects.

If it is true that developmental behaviours are predictable, then maturity models are diagnostic, anticipating the likelihood of success a systems thinker would have when faced with a systems

problem and given that thinker's current maturity level. Maturity models also claim to be prognostic, taking the idea of developmental predictability to mean that likelihood of success can be reliably increased by identifying areas of improvement that will progress one toward excellence. They do this by identifying areas of consensually defined weakness (or "fragility" [20]) that hinder optimal functioning [52]. As such, maturity models could facilitate planning, guidance, and control over future systems thinking performance by outlining what sense-making approaches in systems thinkers should be reinforced and prescribing the sense-making approaches of more mature stages as areas for prioritized development. "Organizations regularly invest in capability development; the capability maturity model aims to provide valuable guidance" in targeting training investment [20] (p. 146), in such a way as to strategically exploit existing capabilities and to strengthen potential ones [55]. Description and prescription lie at the heart of the maturity model's purpose and promise.

As indicated above, Checkland's claim was that every example of systems thinking is merely a matter of an observer with a purpose who defines a system and identifies the mechanisms that make its structure coherent [24] (p. 102). This view obscures the fact that observers differ in their capacities to define a system, identify its mechanisms, and understand its structural cohesion. Systems thinkers vary widely in their knowledge and skill, as do professionals of every sort. Accordingly, diverse parties have embarked on maturity model initiatives.

#### **3. Varieties of Maturity Model**

A variety of groups have made claims to understanding systems thinking maturity and codifying it. For example:


In each instance, maturity was defined in terms of a particular application or domain of human activity, ranging from innovation-generating situations like project management, the collaborative development of new computer software, and the ability to leverage Big Data [51,52,61], to routine organizational operations where effectiveness-enhancing actions such as practices of reflection are said to signify and enhance an organization's functioning [52]. Oriented toward different fields of human endeavour, what all of these maturity models have in common is the intent of formalizing and institutionalizing particular knowledges, skills, behaviours, values, and practices that are considered necessary for effective (i.e., mature) modes of operating in a particular context. The knowledge, skills, behaviours, values, and practices demanded in different fields of systems theory and practice are vast. However, the different contexts of human activity of interest to maturity modelers give rise to a variety of common characteristics in their models.

All maturity models aim to provide users with conceptual schemas for understanding how maturity is multifaceted in nature. Language used in models includes:


Models vary in the granularity with which they conceptualize these elements and their varied permutations. The degree to which models claim to be "tools" seems related to whether or not they portray maturity as a state resulting from tangible factors conducive to *quantitative measurement* by Likert scales or intangible factors better suited to *qualitative description* (the latter ranging from models using broad-based qualitative descriptions to those utilizing detailed descriptions that managemen<sup>t</sup> theorists would characterize as "thick" or "rich" [62]. The ambition of some maturity models is to elucidate different modes of behaviour with respect to maturity, each useful in their own right. For example, models like these would describe clustered themes of systemic sense-making and the resulting behaviours. Such models would enable users to conceptualize qualitatively different ways that systems thinkers can function (not better or worse; merely different).

In contrast to maturity models that focus on qualities, the ambition of other models is to rank quantity—to rate different modes of functioning in terms of greater or lesser desirability. Such models seek to identify "poor" systems thinking in contrast to "good" systems thinking. Gradations are a feature common to these maturity models: they identify a range of human capabilities and behaviours and locate each in terms of its proximity to what designers understand as a state of ideal systems thinking maturity, thus creating a representation of distance and nearness to that state. Aspirational models like these vary in complicatedness, typically involving four or more levels (depending on whether or not stage zero is accorded any merit [17]), with each model including distinct components, dimensions, behaviours, or capabilities ranging from a few to upwards of 75 [20]. Such models are said to describe an ordered arrangemen<sup>t</sup> of levels, each understood as prerequisite to the next step along a singular evolutionary pathway oriented toward optimal performance (i.e., matureness) [18]. Users of these models can identify their location along the path via rating keys, or in some cases, exemplar situations given to represent how behaviour at each stage should look.

Developmentally oriented maturity models have an explicitly forward-moving telos, clarifying the meanings of desirable states such as *superiority*, *mastery*, and *excellence*. Where forward-moving progress is the aim, maturity models vary in the degree to which they explicitly assist users in advancing their path. Some models focus on within-stage characteristics, clarifying in sometimes grea<sup>t</sup> detail each level; they facilitate progress only indirectly, by merely naming the subsequent stage to be achieved ("benchmarking models" are examples of this [53]). Developmentally oriented models take a more active and direct role in facilitating users' progress, describing constraining forces that account for limited functioning (i.e., explaining why one is at a current level of maturity and not a higher one), and by naming drivers or enabling factors that would facilitate movement toward each next stage [50].

While all maturity models describe different stages of maturity, considerably fewer make claims to have uncovered the mechanisms of movement necessary to ascend between stages. That is, models differ in the claims they make to be *descriptive* of different states of maturity, *comparative* of modes of maturity exhibited by different people, or *prescriptive* of what actions to take in order to better one's level of maturity [54]. Thus, maturity models present themselves for two distinctly different purposes of use: (1) understanding how one operates, and (2) directing how to change that in favour of different or more useful ways of operating.

A final variance in existing maturity models is worth emphasizing. Few models portray maturity as context-free. Indeed, maturity itself should be defined in terms of one's skillful engagemen<sup>t</sup> with contextual factors. Most maturity models identify contextual factors that are pertinent to users and provide descriptions of how effective engagemen<sup>t</sup> with those factors typically looks at each level of maturity [47]. Maturity models in different disciplines vary in the attention they draw to environmental factors as central to the development and display of maturity. Systems thinkers work in wide-ranging settings. Amidst other factors, systems thinkers deal with myriad customers, audiences, organizational cultures, and leadership dynamics [51]. All are exigencies that demand a systems thinker's engagemen<sup>t</sup> if one were to perform at effective levels of maturity.
