**1. Introduction**

Checkland [1] explained soft systems methodology (SSM) as an approach to understanding complex problem situations one wishes to learn more about and are poorly understood. SSM is a methodology that has shown increasing use over time to address multifaceted business, learning, and engineering problems involving complicated factors and structures about which stakeholders may not agree. The difficulty of translating outcomes from analysis of "messy" or ill-structured problems into usable solutions resulted in some mild criticism of the speed and viability of the SSM process for moving the analytic outcomes into strategic or product improvement recommendations [2]. There is also limited research into the effectiveness of soft systems approaches over time for generating solution, in part because some analysts find the method's requirement of stringent research that requires interactions with stakeholders and complex data sets to be onerous. However, this may be a misguided view because the goal of the approach is often to learn more about a situation, not generate clearly testable solutions. In past attempts, some researchers applied only a portion of the SSM stems in their research, which may limit the method's effectiveness [3]. Other criticisms of the method include its slow speed, difficulty in using it with stakeholders during analysis or to implement suggested solutions; further, the complexity of the problems under study may limit its acceptance as a valuable research methodology by some academics or managers [4].

Checkland [5] noted that there are different ways of thinking about systems. Systems are a group of interacting elements or subsystems with a unified goal and defined by its boundary as well as the nature of the internal structure linking its elements (e.g., physical, logical, functional). "Hard systems thinking assumes that the world is a set of systems (i.e. is systemic) and that these can be systematically engineered to achieve objectives. In the soft tradition, the world as it naturally exists is assumed to be problematic; but, it is also assumed that the process of inquiry into the problematic situations that make up the world can be organized as a system" (p. S49–50). "Hard" systems may be considered as organized systems with features and goals upon which stakeholders agree. "Soft" systems often contain identified elements such as outcomes, processes, strategies, or other features about which there is imperfect stakeholder agreement. Further, the system elements may change dynamically in response to local needs as system participants learn new information about their own or external, related systems. However, it is difficult to retain agreemen<sup>t</sup> about the best way to organize complex systems permanently, requiring further refinement of agreed upon understanding among participants over time that is generated by continued discourses among participants.

Since early work on SSM by Mingers [6] in the late 1970s and early '80s along with Checkland's groundbreaking 1981 *Systems Thinking*, *Systems Practice* book [1] that formalized SSM as a research approach, it has become a means of understanding problems with no single answer for those conducting research in assorted social science fields. SSM is valuable in complex situations that involve multiple stakeholders and systems, especially where learning or making sense of the problem situation is the goal of the study. Other, modified versions of the methodology by Boardman [7] and other authors are used to support visualization of SSM data for the purpose of improving problem situations.

While many students and professors recognize the term "soft systems methodology", SSM's impact on in engineering, business, and other social sciences in the form of vetted, academic research outputs (e.g., peer reviewed articles, committee reviewed dissertations, etc.) is less understood. This leaves additional questions such about the measurable reach of SSM in these academic research areas. Who are the authors using the method to produce research? How extensively is the approach used for research? It is also important to understand where the outcomes of SSM research are published and, using commonly accepted impact scores for associated journals, as a general measure of the degree to which they are perceived to be impactful on academic research and thinking distributed in broadly available publications. Past research by van de Water, Schinkel, and Rozier [8] explored the publications where SSM was published up to 2007, along with an examination of the countries from which the authors and journals were located. Since that time, a significant number of publications have entered public databases, which requires examination of SSM on academic research and dissemination in the ensuing period. A constraint on this research is that many practical applications of SSM used in corporate and other organizational studies that employ SSM are not published. Therefore, their findings are not available for review and remain outside the scope of this study. Such studies are reserved for future research with other, more appropriate methods to that task.

This piece examined the impact of SSM on academic works through a bibliometric meta-analysis of pieces that discussed as an approach or employed SSM for research. We begin with an examination of how systems thinking and the related SSM approach are defined. An exploration of the use of SSM in the fields of business, engineering, and other social sciences then illuminates this research. A bibliometric analysis follows, focused on publications found to discuss or employ SSM to depict the impact of the approach over the last 35 or more years.

#### **2. State of Knowledge and Practice**

Current knowledge regarding SSM originated in theory and practice work around General Systems Theory dating back to at least the 1950s and 1960s in a period following World War II, as organizations sought to build complex physical and human systems. Since that time, some authors have built new analytic tools for producing more rapid depictions of SSM analyzed complex, including the conceptagon and Systemigrams [9] in the case of Boardman's [7] version of SSM. The goal of these improvements was to use the better visualizations of the complexity as a means of developing improved systems, whether they are often well-defined manufacturing and software products, or instead, more poorly structured organizational systems upon which agreemen<sup>t</sup> about their shape remains elusive or in flux, leading to new problem situations. The following sections review the relevant history of soft systems methodology and its development.

#### *2.1. General Systems Theory*

Systems thinking, as a term and set of processes, was first introduced and formalized in the 1950s. Originally labeled General Systems Theory (GST), it was developed as both conceptual framework and mathematically expressed theory, most notably by Ludwig Von Bertalanffy [10]. His original conception was that problems identified symptomatically in complex systems across different disciplines affected one another, but they had to be first described independently and then in terms of their interrelations to help researchers clearly understand how they affected one another. While this was only a starting point, GST allowed systems thinking to flourish across disciplines such as ecology [11], engineering [12,13], business and academia [2,12], and education [14]. Furthermore, significant original academic work was done over the last thirty years in multiple disciplines to meaningfully grow the value and use of systems thinking [3,9].

Systems thinking describes the act of examining and seeking to understand a system, an interlinked set of objects, people, actions, and subsystems in a cooperating mechanism or set of activities, as a complex, Gestalt whole [9]. Rather than requiring a person to attempt perception of a multifaceted system one small piece at a time without its inter-relationships; systems thinking seeks to present the entire picture as a means of identifying where different components meet, perform well, or require change. This approach still requires that an analyst shift their gaze from the whole system to the parts in a back-and-forth effort. It is this process that allows comprehension of how components fit together, interact, and depend on one another for the entire system to operate and achieve its overall function. Such holistic thinking permits the mind to discover patterns among each element that may not be immediately evident or emerge over time. This approach requires acknowledging situations surrounding the whole as they change and interact with other, interrelated wholes to create the system as stakeholders are likely to perceive it.

For some authors, systems thinking is synonymous with holistic judgement regarding a coherent phenomenon or conceptual framework. At its core, this systems thinking approach examines the interconnectedness of each part of the system, uncovering patterns regarding how different components work together to produce certain systemic outcomes, as well as what may hinder desired results. The Society for General Systems Research founders stated that systems theory "provide(d) a meta-level language and theory in which the problems of many different disciplines could be expressed and solved" [5]. Unfortunately, GST has not resulted in the substantive investment by scholars to produce a generalized, holistic view across disciplines. Instead, systems thinking expanded slowly over the last 65 or more years in fields such as education, biology, engineering, and increasingly focusing on components of supply, demand, and logistics in the fields of business and related engineering disciplines.

#### *2.2. Soft Systems Methodology*

To foster systems thinking about complex organizations and processes, Checkland [15] formalized separate definitions of hard and soft systems, calling his own research approach and related conceptual framework "Soft Systems Methodology" (SSM) [1]. This means of analyzing ill-structured, soft systems emerged from General Systems Theory and includes a seven-stage process for applying SSM. This approach requires the analyst to think both about the real world and the conceptual model of the system under study. The stages include: (1) Entering the unstructured problem situation; (2) expressing the problem situation; (3) formulating root definitions of relevant human activity systems; (4) building conceptual models from the root definitions; (5) comparing models with the real world; (6) defining desirable and feasible changes; and (7) taking action in the problem situation. SSM is a process of generating understanding that may be used by analysts to generate systemic improvement recommendations for a system under study and while it also provides stakeholders with a rich picture of an identified problem situation and related systems one wishes to learn more about. Depending on a researcher's view, the research outcome could be a clearer view of shared systems of knowledge organized by stakeholders, organizational structures, activities, processes, physical objects working in concert, or others, depending on how they are defined at the outset of a study [9].

Checkland [5] explained hard systems research as focused on analyzing well-defined systems with a goal of describing and understanding problem situations in which stakeholder agreemen<sup>t</sup> is lacking. Such systems interact with one another and, through examination of the points at which they touch, can be depicted to identify how they may be engineered to perform better. Thus, hard systems analysis is commonly used to seek solutions to a well-defined, agreed upon problem. By comparison, a soft systems research approach is often used to learn more about a poorly defined situation that one seeks to better understand, without necessarily generating testable solutions, though researchers sometimes do so. These soft systems are complex and, when viewed from outside, may be deemed mysterious. While not always the case, when humans are part of the system under study, involve complex cultural mores and multiple systems that interact in uncertain ways with unpredictable outcomes. This challenge stems from the often poorly defined boundaries and conceptual definitions of a system's component parts, subsystems, or complex relations between similarly sized systems. This problem is often because the system emerged and evolved organically in response to its environment and needs, so its form may appear chaotic at the outset of analysis. Since system changes may have been done quickly where acute problems exist, without consideration of the consequence of a decision, there may be many ways for the system to be improved to perform more effectively. Engineers and analysts are meant to inquire into whether the soft system can be organized into what Checkland [15] called a learning system. With a hard systems approach, the Observer sees the world as full of systems that they can engineer; that is, they see the world as systemic. By contrast, in soft systems approaches, the Observer sees the world as full of complexity and confusion. However, these features can be organized for exploration as a learning system, using a systems inquiry process. What is presented here is our understandings of Checkland's concepts resulting from a synthesis of readings. However, Holwell [16] offered historic a critique of many authors' views of SSM, leaving our own open to similar criticism. However, the focus of this research is not our depiction of SSM; rather, we sought to examine the academic impact of the methodology on disseminated literature in social sciences fields. The research methods employed to that end follow in the next section.

#### **3. Materials and Methods**

This study depicts the impact of soft systems methodology in the engineering, business, and other social research fields. To do so, a multi-strategy, bibliometric analysis [17] was performed on the term "soft systems methodology". This research approach, commonly used in information science studies, involved multiple data sources and analytic approaches. This methodology was deemed appropriate to gather the complex academic evidence available from different sources to help tell the story of the impact of SSM, as evidenced in published research and theory disseminated publicly. To meet this goal, a positivist research method and conceptual framing was employed [18] as it was appropriate to our questions, despite SSM being a non-positivistic research methodology. Our numeric, though descriptive approach allowed longitudinal description of the academic use of the methodology, focused on qualities of the academic publications and public metrics. Taking this path allowed us to first observe and visualize the state of use over an extended period with the ability to interpret findings from that observation regarding the subjective impact of SSM generally on academic research outputs. Given SSM's as a qualitative methodology, an interpretivist or hermeneutic research approach may have been in better alignment; however, given the size of the data corpus and scope of each publication, analysis and explanation of hundreds of articles was impractical. Further, such an approach likely would have failed to meet our research goal, which was to provide a historical view of the impact of SSM as a method on academic research as evidenced in published pieces.

This methodology required capturing written pieces that we could directly evaluate for evidence that the pieces included direct discussion of SSM as a methodology, or employed it as a research approach with observable outcomes. Bibliometrics, as a data analytics research methodology, comes from the library and information sciences. It is used to capture quantitative outputs of information sources using descriptive and network statistics based on citations, authors, keywords, texts, and dissemination outlets. In this study, the method was used to identify trends, impacts by use, subjects, and fields that have adopted SSM. Some analyzed data and presented here is meant to provide context for readers less familiar with SSM as a research approach, including who the major authors and journals in the area are. By contrast, other outcomes are valuable for visualization to depict the impacts of published research over time. To hold the rigor of the publication outlets steady for this study, we examined only published pieces we could fully read through to determine whether SSM was discussed or employed for research. Therefore, we did not include conference proceedings, unless they were available for review beyond a short-form abstract. This approach left as our data sources peer-reviewed articles, books, book chapters, white papers, and dissertations/theses. To gather these sources, our data collection methods, which sought to be exhaustive, employed data mining and Boolean searches from multiple sources using the following approach.
