Chapter 8: Existing and Emerging Methods for Integrating Theories Within and Between Disciplines

Wallis, S. E. (2014). Existing and emerging methods for integrating theories within and between disciplines. Organizational Transformation and Social Change, 11(1), 3-24.

This paper is based on a presentation at the 56th annual meeting of the International Society for Systems Sciences (ISSS). July 15-22, 2012, at San Jose State University, California.

Abstract

Our conceptual systems (including theories, models, policies, schema, etc.) all help us to understand the world around us. For highly complex situations such as those found in natural systems and service systems, it is important to understand them from an interdisciplinary perspective because these real-world systems do not respect the boundaries of any single discipline. While many conceptual systems exist, they have not proven highly effective for understanding issues that are the focus of their disciplines. Still fewer conceptual systems have been developed that cut across disciplinary boundaries—and they have not been shown to be any more effective than their mono-disciplinary companions. This paper investigates emerging and existing methods for creating and integrating theories within and between disciplines. This includes "soft" methods (ad hoc, cherry picking, and intuitive) as well as "rigorous" (Formal Grounded Theory (FGT), Reflexive Dimensional Analysis (RDA), and Integrative Propositional Analysis (IPA)). The present paper demonstrates that soft methods are relatively easy to use, but they do not produce conceptual systems of great or lasting value. In contrast, it is proposed that the rigorous methods are more likely to yield conceptual systems that are measurably more systemic, more useful and more effective for understanding and engaging the highly complex systems of our world.

Introduction

If we are to enact change in society and organizations, we must have better conceptual tools. That is to say, we need better conceptual systems for better understanding organizations, society, and change. In this paper, I investigate how we might accelerate our ability to create more effective conceptual systems by integrating conceptual systems across disciplines.

For natural systems or service systems most research may be categorized as inductive or deductive. While these are good for “normal” science, more interesting revolutions in science may occur when a deep thinker considers two conceptual systems and seeks to compare, contrast, and combine them. Galileo and Einstein began with this kind of approach. Because of the paradigmatic revolutions they triggered, we all lead much richer lives. Were they unique in their ability to seek and find new insights from existing conceptual systems? Or, is this an approach that we all may use? In this paper, we will investigate multiple methods for integrating conceptual systems to determine which methods might be more useful. The results suggest that more rigorous methods provide a more useful and more systemic approach to integrating conceptual systems.

Although natural systems and social systems are often considered to be desperate disciplines, we may legitimately investigate conceptual systems of the natural and social spheres using the same tools because both sets of conceptual systems rely on causal propositions (Wallis, 2010a). “A proposition is a declarative sentence expressing a relationship among some terms.” (Van de Ven, 2007: 117). When that proposition expresses a causal relationship, “two or more aspects [concepts] are related so that a change in one causes a change in one or more others. A causal relationship is often expressed as a proposition, hypothesis, or a diagram” (Wallis, 2011: 100). For example, “More A causes more C.”

In short, although conceptual systems of service systems and theories of natural systems are from different disciplines, they are made of the same propositional “building blocks.” Therefore, it is possible to integrate them on the level of conceptual systems.

As scholars, we like to think that we are good at evaluating conceptual systems and deciding the extent to which those conceptual systems might be useful or effective for research and/or practice. Unfortunately, while we are able to evaluate the work of others with some level of objectivity, we have less objectivity when evaluating ourselves (Dunning, Heath & Suls, 2004). Worse, those who are low-performers do the worst job of self-evaluation (Ehrlinger, Johnson, Banner, Dunning & Kruger, 2008). In the social sciences, in general, we (as a community) are low performers. By that I mean no disrespect for any individual scholar. Instead, I mean that the social sciences, in general, appear to be “low performers.”

For example, public policies frequently fail (Wallis, 2011), organizational theory seems to have failed (Burrell, 1997), psychology is not held in high regard by other professionals (e.g., Kovera & McAuliff, 2000), and social change theory does not seem to be effective (Appelbaum, 1970; Boudon, 1986). Indeed, the promise of the social sciences is “largely unfulfilled” (Spicer, 1998). This creates an additional level of challenge because we are trying to integrate the low-performing theories of the social sciences with the more reliable theories of the natural sciences—although they too are sometimes held in low favor (Smolin, 2006).

The systems community strives to take a new view—but there is no guarantee that our views will be any more effective than previous ones. Mainly, our community looks at natural systems and/or service systems using a systemic view to better understand the world around us. Relatively few scholars study conceptual systems (Wallis, 2013). This is an important concern because our conceptual systems determine how we understand and engage the world around us. Those conceptual systems are our theories, models, mental models, schema, policies, and so on. In the present paper, I will use the term “conceptual system” to refer to theories, although I may occasionally use other terms to refer to more specific forms of conceptual systems.

Like previous schools of thought, most systems research may be broadly categorized as either inductive (begin with data and move toward creating a conceptual system) or deductive (begin with a conceptual system and then conduct experiments to test that conceptual system) (Hitt & Smith, 2005). These are well and good for “normal” science because they lead to the creation of conceptual systems, data, and insights.

In contrast to those approaches of normal science, revolutionary advances in science are sometimes caused when deep thinkers consider two (or more) conceptual systems and seek to compare, contrast, and combine them.

In this paper we will identify and discuss a range of existing and emerging methods for integrating conceptual systems within and between academic disciplines. These include soft approaches such as “ad-hoc,” “cherry picking,” and “intuitive” methods. Soft methods have been used throughout the history of the social sciences without impressive results (as noted above). Next, we will give more attention to relatively rigorous methods including Formal Grounded Theory (FGT), Reflexive Dimensional Analysis (RDA), and Integrative Propositional Analysis (IPA).

To demonstrate these methods, I will integrate two conceptual systems using each of the soft and rigorous methods. One conceptual system is from the study of service systems while the other conceptual system is from the study of natural systems. Finally, I will analyze and compare the resulting integrated conceptual system to determine the extent to which it might be considered an advancement of the field.

The goal of this paper is to provide a better understanding of how to create more effective conceptual systems by using rigorous methods. Additionally, this paper is expected to engender new learning and new insights into bridging the “theory-gap” between disparate disciplines (such as the separate studies of natural and service systems) through a deeper understanding of conceptual systems.

Data Set: Two Conceptual Systems (Theories)

We are assuming, for the purpose of this conversation, that these conceptual systems are based in some sort of rigorous empirical analysis. In this section, I simply present two sample conceptual systems. All of the subsequent analyses will refer to these two conceptual systems as the data source for their analyses. Each conceptual system contains a set of propositions as shown here:

A sample Service Systems Theory

A Complex Adaptive Systems Model of Organization Change from: (Dooley, 1997: 82; drawing on Thietart and Forgues, 1995)

A sample Natural Systems Theory

From: (Allison & Hobbs, 2004, drawing on Guderson et al., 2002)

These two conceptual systems (one of natural systems, the other of service systems) are both based in systems/complexity field. For scholars interested in creating more effective conceptual systems, the following section presents multiple methods for integrating these two conceptual systems.

Intuitive Methods of Integration

In this section I will present soft methods of integration including ad-hoc, cherry-picking, and intuitive methods. Then, in the following section, I will present the more rigorous methods of integration including Formal Grounded Theory (FGT), Reflexive Dimensional Analysis (RDA), and Integrative Propositional Analysis (IPA).

This is intended to provide a set of examples for these methods of integration. The goal is to highlight some strengths and weaknesses of each method. Because this paper is focused on the level of metatheory, we will look at the concepts within the conceptual systems, rather than on the application of the conceptual systems or the research from which the concepts were derived.

There are many soft methods available. For example, Mintzberg (2005: 361-371) suggests that personal characteristics are key to developing good conceptual systems. He suggests the benefit of creativity, intuition, and bravery. Ritzer (2009) suggests personal reflection while Hall (1999) suggests that social construction is a good path. It is difficult or impossible to test the extent to which these methods will yield useful conceptual systems. Primarily because there is no good way to test the bravery of a theorist or to say how much reflection is applied (or needed) to create or integrate conceptual systems.

Because things like bravery and reflection are difficult or impossible to measure, this paper will focus on the concepts and causal relationships within each conceptual system to see how they change as different methods of integration are applied.

Cherry Picking

Cherry picking is quite simply the process of choosing specific elements from two or more conceptual systems and combining them to create a new conceptual system. While it is expected that the scholar will use some form of reasoning to support the choice of concepts, it is understood that alternative reasoning would lead to alternative choices. For example, drawing on the above two conceptual systems, we might combine: “Resilience derives from functional reinforcement across scales and functional overlap within scales” with “Organizations are potentially chaotic.” The derived conceptual system might be stated as: For chaotic organizations, resilience derives from functional reinforcement across scales and functional overlap within scales.

This may have created a conceptual system that is (to some extent) true and/or useful; however, it is important to note here that the derived conceptual system is smaller, less complete, than either of the two source conceptual systems. In short, cherry picking is reductionist. Additionally, in creating a cherry picked conceptual system, there has been some fragmentation of the field as a new theoretical focus has been created. Finally, there is the question of how we decide which part of a conceptual system to separate from the remainder of the conceptual system. Here, there are no rules in the academic world except that the choice should be supported by some rational argument. That, however, is a weak standard as anything may be rationalized. In short, it is problematic and inappropriate to use partial conceptual systems (Ritzer, 1990; Wallis, 2012a).

Ad-Hoc

Creativity and innovation are hallmarks of the ad-hoc process, which involves the combination of concepts from multiple conceptual systems. Because of the creative effect, additional concepts may be added that are not necessarily part of the original set of conceptual systems under investigation. Starting with the two conceptual systems presented above, an ad-hoc method might be narrated something like this:

From the CAS on organizational change, particularly proposition three, it is clear that there is very little opportunity for predicting the future of the organization. It can also be recognized that the lack of predictability also applies to natural systems. For example, where “complex systems can exhibit alternative stable organizations.” Therefore, it appears that the important similarity between natural and service systems is a duality between predictability and chaos. Moving into other sources, an individual who is managing an organization (or, presumably, a natural system) should “expect to be wrong" (Richardson, 2009, 2009: 49). An idea that makes perfect sense if we have partial theories and chaotic situations. From this, a manager might conclude that there is no reason to study complexity, systems, or anything else for that matter—because that manager will always be in chaos and always be wrong.

To some extent, this is a straw man argument—it is easy to argue against it and break it down. That is exactly the point. Indeed, most of our conceptual systems have a certain amount of ad-hoc logic within them. Thus, none stand for long. Each is replaced rapidly with some combination of other conceptual systems, concepts, and notions from a variety of sources until a “Frankentheory” is created. That conceptual system rampages across the pages of our publications until it is dismembered and recombined to create some new golem.

In this ad-hoc process, it may be seen that concepts were chosen through a reasonable process. Thus, there appears to be some validity. However, the same level of reasoning might have been used to choose different concepts. Thus, the ad-hoc process cannot be relied upon to integrate multiple conceptual systems with any useful level of rigor or repeatability.

The ad-hoc process does not require the scholar to engage the entire conceptual system—or any conceptual system. In short, like cherry picking, ad-hoc integration allows the scholar to look at conceptual systems in a non-systemic fashion.

Intuitive Integration

The intuitive process is, almost by definition, a process that cannot be governed by a rigorous process and explicit set of rules or guides. By way of brief explication, however, I will note that I have just read the conceptual systems closely… then put them aside and let my intuition emerge as I type…

Service and natural systems are different because natural systems become more brittle with decreasing complexity. Service systems in chaos, on the other hand exhibit similarity across levels of scale. It seems to me (intuitively) that such similarity would be quite the opposite of the complexity needed to maintain the natural systems. Thus, it seems that the two cannot be completely integrated. Another part of this (which may or may not have been mentioned in one of the theories) is that the social system is geared toward purposefully creating value for the other participants. Members of a natural system, on the other hand, seem geared toward the creation of value for themselves. Service systems create value for themselves by harvesting natural resources. Often without concern for the long-term sustainability of those resources. However, as the service systems evolve, they may learn to manage natural systems for long-term success.

Returning now to look at the conceptual systems, and reflecting on my intuitive effort, I note how I neglected to include a number of concepts. I may have misrepresented some concepts and created some conceptual linkages out of thin air. This example shows how intuition, like cherry picking and ad-hoc, is a soft approach to conceptual system integration.

Discussion on soft approaches to conceptual systems integration

My experience in investigating conceptual systems and their sources suggests that the most commonly applied methods are the soft ones. This is problematic for the field because intuition is not reliable (Meehl, 1992: 370). We are like gamblers—using intuition to make our wagers—only to go home empty handed.

An important key to the scientific validity of any study is the ability to replicate that study. If another scholar cannot replicate the study and arrive at similar conclusions, the validity of the study is thrown into doubt. The same criteria should apply to metatheoretical studies (Wallis, 2010b). If, for example, we ask ten graduate students to analyze the same ten conceptual systems using the same rigorous methodology they should arrive at similar conclusions. If not, the metatheoretical methodology must be called into question.

Because replication is a mainstay of science, it must be concluded that soft approaches cannot be considered useful from a science of metatheory perspective.

To summarize this subsection on soft methods of the integration of conceptual systems, scholars who use these methods are forced to rely on intuition and reduction. While it may be easy to use, intuition is clearly not to be trusted. And, similarly, reductionism leads us to address partial models that can sometimes lead to fragmented understanding and false assumptions.

Of course, there are benefits to a reductive approach. That direction may be useful and lead to more nuanced knowledge. However, it is also necessary that the deconstruction is followed by reconstruction for creating conceptual systems (Ritzer, 1990: 11); particularly if we want that knowledge to be systemic.

The lack of scientific replication of soft methods drives the final nail. The limited usefulness of soft methods is evidenced by the poor progress of the social sciences. Those methods, which have been used for so many decades, have resulted in low performing conceptual systems—clearly we need something better. In the next section, I will present and discuss more rigorous methods.

Rigorous Methods

In this section, I begin my briefly presenting three older methods of theory evaluation/integration. Finding that these provide limited levels of rigor, I go on to present three rigorous methods for integrating multiple conceptual systems. Formal Grounded Theory (FGT), Reflexive Dimensional Analysis (RDA), and Integrative Propositional Analysis (IPA). These methods follow a prescribed path, and are thought to provide a more repeatable approach to the scientific analysis and integration of conceptual systems. Next, each of those methods is used to analyze the subject conceptual systems.

Previous approaches have been suggested for the evaluation and analysis of theory. For example, Dubin (1978) suggests a measure of “efficiency” (pp. 109-111). These relate to:

  1. The presence (or absence) of some unit of measure within a conceptual system
  2. The casual directionality between two units of measure
  3. The covariation of those units of measure.
  4. The rate of change between them.

These are certainly good and useful things to have within any conceptual systems. And, all scholars should strive to include them. One limitation of this approach is that there is no concrete method to measure what a conceptual system contains within (or between) those levels. Dubin only provides a general guide.

Ritzer (2001: 53-55) suggests “architectonics” as an approach to compare and integrate conceptual systems of sociology. However, that approach is geared toward identifying fundamental similarities in human actions, rather than engaging in a highly rigorous study of the conceptual system, itself.

More recently, Shoemaker, Tankard Jr., & Lasora (2004: 170-178) suggest key steps to building a theory including:

  1. Problem recognition;
  2. Identification of key concepts;
  3. Observation and creativity to identify and suggest the causes and effects of those key concepts;
  4. Specify theoretical and operational definitions for all concepts;
  5. Link some concepts to form hypotheses;
  6. Specify rationale for hypotheses;
  7. Attempt to think in terms of multiple hypotheses, and;
  8. Attempt to place the hypotheses in an organized system.

They suggest that those steps will allow conceptual systems to be tested as to their testability, falsifiability, parsimony, explanatory power, predictive power, scope, and cumulative nature of the field, degree of formal development, heuristic value, and aesthetics. While their suggestions provide a good starting point, their steps are open to ad-hoc reasoning, cherry picking, intuition, and poorly defined measures.

While older kinds of methods offer some improvement over soft methods, the level of rigor does not seem sufficient to develop more effective conceptual systems because they still employ or allow for the application of soft methods. Therefore, for this section, I will focus on the more rigorous approaches that follow a specific methodology.

Formal Grounded Theory

Grounded Theory was developed by Glaser & Straus (1967) as a transparent process to create theory that is grounded in real word contexts (Glaser, 2002). In brief, experiences and insights are coded, categorized and related in a specific methodology to create a theory with a specific focus. Since then, others have used a Formal Grounded Theory (FGT) approach that uses extant theory as the data to create a new theory. For example, Apprey (2006) who suggests, FGT can be used to combine multiple theories and so gain more meaning and insight in an area of study. According to Charmaz (2006). The process includes:

  1. Gathering data;
  2. Coding the data and categorizing theoretical concepts;
  3. Constant comparison between concepts;
  4. Memo writing, and;
  5. Creation of a theoretical construct.

The following subsections detail my efforts and results from following this approach.

Gathering data

This part is easily accomplished by choosing (as examples) the two conceptual systems presented above.

Coding the data

Initial Codes are:

  1. Vulnerability;
  2. Bifurcation process;
  3. Sources of novelty;
  4. Scale;
  5. Forecasting possibility;
  6. Functional diversity;
  7. Actions lead to the same (or different) result;
  8. Time;
  9. Complex systems that have more states can exhibit more alternate stable organizations;
  10. Emergence of organized regional resource systems;
  11. Functional overlap within scales;
  12. Functional reinforcement across scales;
  13. Interactions of few variables;
  14. Organization is attracted to identifiable configurations;
  15. Resilience;
  16. Similarity across levels of scale;
  17. Organization is in chaos;
  18. One dynamic state of organizations;
  19. Organizations are potentially chaotic;
  20. Second dynamic state of organization, and;
  21. Cross functional replication.

Categorizing theoretical concepts

Constant comparison between concepts

I accomplished this part of the process by keeping the conceptual systems on a single page and referring back to them frequently.

Memo writing

Some memos include:

Create theoretical construct

For FGT, the theoretical construct is based on a central question or focus. This introduces another point of ambiguity to the process. After all, if two scholars approach the conceptual data with different questions, they will likely create different constructs. Here, in addition to the subjectivity of creating themes, is another source of subjectivity for the FGT process. The combined theoretical construct I developed came out as:

Within and across levels of scale, there are overlaps and reinforcements. Organizations in chaos tend toward recognizable configurations and cross-functional replication. The larger and more complex system emerges from interactions of smaller systems. And, conversely, the larger system exhibits more stable systems and alternative stable states.

Conclusion

In this process, some complex concepts were fragmented into multiple simpler concepts before being combined into categories. It is not clear if that extra step supports the creation of improved conceptual systems. Also, the categories are not rigorous—another scholar might legitimately undertake the same analysis and develop different categories. Thus, the process is not necessarily repeatable. Because this process is focused on concepts, rather than their relationships, it is too easy to find one’s self with a conceptual system that is a collection of ideas, rather than a set of interrelated propositions. Thus, one may end up with a construct that is hardly a theory (or a system) at all. So, the usefulness of the resulting construct seems questionable.

Multiple concepts were categorized into fewer—suggesting that reductionism may be taking place. This might be countered in future versions by creating a new method of FGT that requires that each category represent an abstraction of the concepts. This opens some interesting possibilities. For a rather abstract example, if a conceptual system contained concepts of “square” and “rectangle” the abstract categories that are suggested might include “width” and height.” The idea is not to force many ideas into fewer ideas. Rather, the goal is to seek highly abstract categories that can fully represent the concepts within the conceptual systems. As such, it is entirely possible that identifying all the abstractions might result in a conceptual system that is much larger and more complex than the subject conceptual system upon which it is based.

Reflexive Dimensional Analysis

Reflexive Dimensional Analysis (RDA) was derived in part from insights found in GT and has been used to integrate conceptual systems of Complex Adaptive Systems (Wallis, 2006) and Complexity Theory (Wallis, 2009). RDA differs from FGT because RDA specifically calls for the scholar to identify causal relationships at the sub-category level—and apply them to the category level. This provides an additional level of rigor above the FGT approach. RDA has six steps (Wallis, 2006: 7) :

  1. Define a body of theory (conceptual systems).

  2. Investigate the literature to identify the concepts that define it.

  3. Code the concepts to identify relevant components.

  4. Clump the components into mutually exclusive categories.

  5. Define each category as a dimension.

  6. Investigate those dimensions through the literature, looking for robust relationships.

    Define a body of conceptual systems (theory).

The scope of the conceptual systems includes service systems and natural systems.

Investigate the literature to identify the concepts that define it.

This step has been accomplished by choosing the two conceptual systems presented above

Code the concepts to identify relevant components.

This step has already been accomplished in the FGT process above.

Clump the components into mutually exclusive categories.

This approach to categorization is more rigorous than other approaches to categorization. By calling for categories that are “mutually exclusive” there is more work to be done—and we end up with more categories. For ease of comparison, I will simply break out the category of “miscellaneous” into new categories of “time” and “forecasting possibility.” I will also re-focus the category of “conditions of the system” to focus on vulnerability and resilience as that seems to be a state of the system that is of particular importance. Next, I will break out “scales” into overlap and reinforcement because they may be understood as exclusive.

This gives us the following categories:

  1. Time;
  2. Forecasting possibility;
  3. Emergence;
  4. Vulnerability;
  5. Conditions of the system;
  6. Overlap within scales;
  7. Reinforcement across scales;
  8. Similarity across scale, and;
  9. Actions/Attractions.

Define each category as a dimension.

Here, we simply define each category as a scalar dimension. This gives it the ability to represent a wider variety of states. For example, “time” may be seen as “more time” (or, conversely, less time).

  1. More Time;
  2. More Forecasting possibility;
  3. More Emergence;
  4. More Vulnerability;
  5. More (differing) Conditions of the system;
  6. More Overlap within scales;
  7. More Reinforcement across scales;
  8. More Similarity across scale, and;
  9. More Actions/Attractions.

Investigate those dimensions through the literature, looking for robust relationships.

Here is a difference between FGT and RDA. FGT simply asks the scholar to identify relationships between the categories. Thus, the scholar may intuitively assign relationships. This kind of approach is not so rigorous as it might be. RDA, in contrast, calls for those relationships to be defined by the data itself. Therefore, at this stage, we must go back to the propositions within each category to see if they contain linkages to other categories.

From the service systems model, More Time causes Less Forecasting ability—thus casually linking those two categories. Also from the service systems model, the states as one Condition of the system will lead to bifurcation found in the Emergence, which leads back to create more states of the organization in Conditions of the systems. Therefore, there are some linkages between those categories. I continued the process in this way—for each category, investigating the causal propositions of the concepts within that category. Those concept-level connections were then used to justify category-level connections. The result is a RDA model integrating the two conceptual systems that looks like this:

Figure 1

Figure 1 RDA integrated model of Service Systems and Natural Systems

To conclude this subsection, RDA is a more rigorous way to integrate conceptual systems from the service systems and natural systems. However, room remains for interpretation and intuition. This may be beneficial if one values creativity (which, I hope, we all do). However, that openness and flexibility becomes problematic when we are trying to make a more rigorous science.

To apply one test—a thought experiment—we might consider giving these two conceptual systems to ten scholars and ask that they all use the RDA method to create an integrated conceptual system. My hunch is that they would end up with ten different integrated conceptual systems. In short, where advances have been made in the area of rigor, there are accompanying difficulties in the convolutions that may work to reduce the effectiveness of the results. A more straightforward approach may be found in Integrative Propositional Analysis.

Integrative Propositional Analysis

Propositional analysis (PA) is used to determine the complexity and interrelatedness of a conceptual system or body of conceptual systems. This is a structural approach which extends and deepens the work of well-known authors who suggest a correlation between the structure of a conceptual systems and the effectiveness of that theory (Dubin, 1978; Kaplan, 1964; Stinchcombe, 1987) by providing reliable quantification. Where PA is generally understood to analyze the set of propositions within a single conceptual system, Integrative Propositional Analysis (IPA) explicitly accepts propositions from multiple conceptual systems as an input to the process and integrates them into a single conceptual system as an output to its process.

The process of PA includes the following six steps (Wallis, 2008):

  1. Identify propositions within the conceptual system.
  2. Compare with one another to identify overlaps, and drop redundant aspects.
  3. Investigate propositions for conceptual relatedness.
  4. Link causal aspects with resultant aspects.
  5. Identify “Concatenated” aspects (those aspects that are explained by, or resultant from, two or more other concepts).
  6. Divide the number of Concatenated aspects by the total number of aspects in the conceptual system (to provide a number between zero and one).

In short, PA starts by creating a diagram of the causal relationships found in the propositions of a conceptual system or body of conceptual systems. Below, I have diagramed each of the subject theories for clarity. Each concept is placed within a box (and numbered); an arrow represents each causal relationship.

Figure 2

Figure 2 Sample Service Systems conceptual system

By counting the aspects within the conceptual system it is clear that the complexity of the conceptual system is C = 11. There is only one aspect that is the resultant of two or more causal concepts (see box #6). Therefore, the Robustness or interrelatedness of the system of conceptual system is R = 0.09 (the result of one divided by eleven). Performing the same analysis on the Natural systems conceptual system, we have:

Figure 3

Figure 3 Sample Natural Systems conceptual system Diagramed

Here it can be seen that there are nine different aspects, so the complexity of the conceptual system is C = 9. There are two aspects that are concatenated (#6 & #9) because they are the resultant of two or more other aspects. Therefore, the Robustness of this conceptual system is R = 0.22 (the result of two divided by nine). There is some small possibility for alternative interpretations. For example, it may be that resilience is the inverse of vulnerability. However, here we will stay with a direct representation of the author’s text in order to maintain rigor.

Reflecting briefly on the two studies from a PA perspective, it seems that the Robustness for both conceptual systems is rather low. This is not unusual for conceptual systems of the social sciences. That low level essentially reflects how the aspects of each conceptual system are interconnected. That level of interconnection may be understood as systemicity (systemicity may be defined as the extent to which a set of things may be considered a system—in this case, a conceptual system).

The presented conceptual systems are not highly systemic. Thus, neither is likely to be highly useful in practical application. Each conceptual system may be improved through research that identifies causal linkages between the aspects within the conceptual system.

Seeking to integrate the two conceptual systems, a strict application of PA requires that we identify aspects within each conceptual system that are identical. Where identical aspects are identified, overlaps exist and the conceptual systems may be connected. While there are a number of similar aspects between the two conceptual systems, there do not seem to be any exact matches.

Allowing for some interpretive license (for the purpose of the present discussion), there are some possibilities for linking the two conceptual systems. First, we might interpret the conceptual systems to suggest that some of the aspects are really the same thing—only with different names. This kind of renaming is not uncommon in the social sciences! Another approach would be to seek a higher level of abstraction—and so link the two conceptual systems under a more abstract concept that adequately accounts for the more concrete phenomena. Third, we might infer casual linkages between aspects of the two models (although, in the name of rigor, this should not be done without empirical analysis).

In these kinds of integration it is much too easy to fall into the trap of intuition or ad-hoc thinking. For example, N4 and N5 include the concept of scales, as does S11. So, one may be tempted to integrate all of those aspects. However, N4 and N5 discuss Reinforcement and Overlap, while S11 is about Similarity. Therefore, it is not clear from the propositions that Similarity would be causal to Overlap and/or Reinforcement. This view suggests a fourth approach.

We are on more solid ground by addressing simpler aspects. For example, the linear logic represented in S1-S2-S3 might be abstracted to a derived proposition, SD1 “More bifurcation process creates More states” Similarly, N3 (which is represented as a single aspect because of the wording provided by the author) might be deconstructed to a derived proposition, ND1 “More states cause more complex systems that cause more alternatives to be exhibited. The derived propositions would be diagramed as in Figure 4.

Figure 4

Figure 4 derived propositions

This opens the door for a match between SD1b and ND1a to create an integrated model as in Figure 5:

Figure 5

Figure 5 Integrated derived propositions

With their derived aspects legitimately integrated, the other aspects of the conceptual systems may be added to the structure in Figure 6:

Figure 6

Figure 6 Integrated conceptual systems

Integrated Propositional Analysis (IPA) is a useful approach to integrating conceptual systems from within and between disciplines. The approach has more methodological rigor than other approaches and requires the use of whole conceptual systems and a systemic perspective. The integrated conceptual system in Figure 6 has 20 aspects, therefore it has a Complexity of C = 20. Three of those aspects are concatenated so the Robustness is R = 0.15 (the result of three divided by 20).

The complexity of the integrated conceptual system is much higher than either of the source conceptual systems. This may be seen as a step forward in the evolution of the conceptual system. However, level of interrelatedness between the aspects of the conceptual systems has not increased. Therefore, while the integrated conceptual system may be more effective than either of the source theories individually, it is not expected to be highly effective in practical application.

To increase the effectiveness of the derived conceptual system, there are two basic approaches. First, a scholar may add additional concepts to increase the Complexity of the conceptual system. Thus may be done by including additional theories and propositions. Recall here that the preference is for using whole theories to maintain the integrity of the conceptual systems. The second approach is to identify causal relationships between the concepts to increase the Robustness. This may be accomplished my additional research and/or by identifying casual relationships from the literature.

One potential weakness of this approach is the creation of highly complex conceptual systems. Some scholars (as well as editors and potential clients for consulting engagements) may look on complex conceptual systems with some suspicion because they are too large to print or too difficult to understand. For editors considering the inclusion of complex conceptual systems into their journals, I suggest creating links between print and online sources. This is one way that the key ideas may be communicated in print without losing the large-field view of the entire conceptual system.

In order to use highly complex conceptual systems to inform organizational interventions in service and/or natural systems, I suggest that practitioners adopt a team-based approach. Each practitioner might focus on a “chunk” of the larger conceptual system. Causal relationships between the chunks should be clearly defined and directly related to the coordination between practitioners.

For scholars, the same kind of approach also seems viable. That is, to involve a team of scholars to research a highly complex theory in a way that has each scholar addressing a specific “chunk” of the broader conceptual system. Importantly, each chunk should have clear causal connections with one or more other chunks.

Another important benefit of the IPA approach is that the integrated conceptual system combines multiple conceptual systems. Thus, instead of fragmenting the field (as occurs with softer methods) this integrative approach serves to unify the field.

It should also be noted that the integration process creates new insights and new challenges for testing each conceptual system. For example if we integrate “more A causes more B,” with “more A causes more C”, we might ask if B & C are the same thing because they have the same causal relationship with A? Or, is there an abstraction that is relevant? Or, is there some dimension of similarity we might find between them? These are challenges that would not arise if we looked at one conceptual system or the other.

Similarly, juxtaposing the conceptual systems creates a challenge and opportunity for research that will clearly advance the coherence of the conceptual system. Using the above set of integrated conceptual systems, it is clear this process is only one step toward a higher level of systemic integration. Each set of disconnected boxes represents an opportunity for research to define the two as a causal relationship. And, defining those relationships will increase the internal coherence and usefulness of the conceptual system.

Conversations and Additional Concepts

In brief, the use of soft methods (cherry picking, ad-hoc, intuitive) for integrating conceptual systems gives the appearance of making sense, but do not seem particularly useful in the creation of more effective conceptual systems. First, they have been used through history, without highly effective results. Second, they are reductionist and non-systemic. Third, they support the fragmentation of the field without providing for re-integration.

Other approaches (FGT and RDA) follow a more rigorous methodology. And, they are more systemic because they seek to identify and connect causal relationships between concepts within conceptual systems. However, both of those methods allow the scholar to address partial conceptual systems. FGT, for example, would allow the theorist to cherry pick portions of the subject conceptual systems that seem most relevant to the analysis. In contrast, IPA requires the rigorous integration of whole theories—each as a “closed system” (Dubin, 1978: 116) unto itself.

Investigating theories as conceptual systems unto themselves appears to be a useful and effective path to improving theories of natural systems and theories of service systems. Research suggests that conceptual systems should evolve towards greater complexity first, and afterward evolve toward greater systemic interrelationship (Wallis, 2010a). Therefore, the IPA method presented above seems to be a good first step toward creating more complex conceptual systems—and pointing the way for additional research that will support the development of more systemic conceptual systems.

However, this approach seems to be counter the prevailing current of the social sciences. In preliminary studies of conceptual systems of psychology and sociology, it seems that conceptual systems of the social sciences have been declining in complexity (Wallis, 2012b). This may be due to scholars following the false call of parsimony that has been decried in theory (Dubin, 1978; Meehl, 2002) and research (Wallis, 2010a, 2011).

In the following section, I will discuss some additional insights and approaches that may be useful in developing theories that are more effective in practical application.

The Case Against Ease of Use

The case in favor of parsimonious conceptual systems is simple. “The simplest theory is the best” (Shoemaker et al., 2004: 172). There is, however, no a priori reason or proof that a conceptual system “should” or “must” be easy to use. Indeed, given the astonishing complexity of our lived world, it is more reasonable to assume that our conceptual systems should be highly complex. Indeed, policies which are more complex tend to be more successful in practical application (Wallis, 2011). Also, students with conceptual systems that are more complex and more interrelated tend to have higher scores on papers (Curseu, Schalk & Schruijer, 2010).

It seems reasonable that we will be more effective as individuals and organizations if we use more complex conceptual systems. Given the limitations of the human mind, however, such an approach calls for greater collaboration between scholars. For example, in an organizational intervention, it may be better to have three consultants instead of one—with each consultant focusing on a different area of a shared conceptual system. This perspective also calls for more resources (including funding and administrative coordination) to support teams of scholars and teams of practitioners.

Conclusion

To briefly summarize and conclude, soft methods of integrating conceptual systems (intuitive, ad-hoc, and cherry picking) have been applied through the history of the social sciences without great benefit. Our present conceptual systems for understanding and engaging natural and service systems tend to be simple and have very low levels of internal coherence (systemicity). Indeed, this is part of the reason why we have the problems we do—because we do not have effective conceptual systems to understand and improve our situations.

Soft methods may support the fragmentation of the social sciences, as they lead toward the creation of conceptual systems that are atomistic and spurious. Further, the social sciences seems to be following the false promise of parsimony—and creating conceptual systems that are simpler instead of better.

What we have here is a kind of three-body problem. First, there is the world (which is irreducibly complex). Second, there is the scholar (who has abductive moments of inspiration). Third, there is the theory (the conceptual systems which have been historically difficult to evaluate). None of these provides a clear “fixed point” frame of reference. Instead, each influences the other in ways that have not been well defined. Ultimately, that pattern may be fractal and perhaps indefinable. It should come as no surprise that there is no easy way to understand the situation. Rigorous methodology such as IPA makes the conceptual systems quantifiable and opens the door for a science of conceptual systems.

A systemic view of conceptual systems suggests a number of alternative approaches to reverse the trend and create conceptual systems that we may use to more effectively address our social-ecological issues. First, using more rigorous methods (FGT, RDA, and particularly IPA) will serve to re-integrate the many fragmented conceptual systems. Second, the use of whole conceptual systems should be preferred to the use of partial conceptual systems.

I expect that this paper will provide new tools that scholars and practitioners might use to more effectively decide which conceptual systems will be more useful for research and practice. It will also provide a better understanding of how to more rigorously create more effective conceptual systems. The integrative effort and resulting conversation of this paper is expected to engender new challenges and new insights into bridging the theory-gap between disciplines including the study of natural and service systems through a deeper understanding of conceptual systems. These, in turn, will allow us to gain better understandings of our society and support more effective change.

References

Allison, H. E. and Hobbs, R. J. (2004). Resilience, adaptive capacity, and the “Lock-in Trap” of the Western Australian Agricultural Region. Ecology and Society, ISSN 9(1): 25. http://www.ecologyandsociety.org/vol9/iss1/art3

Appelbaum, R. P. (1970). Theories of Social Change, ISBN 8410-4019-2.

Apprey, M. (2006). A formal grounded theory on the ethics of transfer in conflict resolution. Mind and Human Interaction, ISSN 0533-3164, 14: 51-74.

Boudon, R. (1986). Theories of Social Change (J. C. Whitehouse, Trans.), ISBN 0-7456-0950-3.

Burrell, G. (1997). Pandemonium: Towards a Retro-Organizational Theory, ISBN 0 8039 7777-8 (pbk).

Charmaz, K. (2006). Constructing Grounded Theory: A Practical Guide through Qualitative Analysis, ISBN 9780761973539.

Curseu, P. L., Schalk, R. and Schruijer, S. (2010). The use of cognitive mapping in eliciting and evaluating group cognitions. Journal of Applied Social Psychology, ISSN 0021-9029, 40(5): 1258-1291.

Dooley, K. J. (1997). A complex adaptive systems model of

organization change. Nonlinear Dynamics, Psychology, and Life Sciences, ISSN 1(1): 69-97, link.

Dubin, R. (1978). Theory building (Revised ed.), ISBN 0-02-907620-X.

Dunning, D., Heath, C. and Suls, J. M. (2004). Flawed self-assessment: Implications for health, education, and the workplace. Psychological Science in the Public Interest, ISSN 5(3): 69-106.

Ehrlinger, J., Johnson, K., Banner, M., Dunning, D. and Kruger, J. (2008). Why the unskilled are unaware: Further explorations of (absent) self-insight among the incompetent. Organizational Behavior and Human Decision Processes, ISSN 105(1): 98-121.

Glaser, B. G. (2002). Conceptualization: On theory and theorizing using grounded theory. International Journal of Qualitative Methods, ISSN 1(2): 23-38, link.

Glaser, B. G. and Strauss, A. L. (1967). The Discovery of Grounded Theory: Strategies for Qualitative Research, ISBN 0202302601.

Hall, J. R. (1999). Cultures of inquiry: From epistemology to discourse in sociohistorical research.

Hitt, M. A. and Smith, K. G. (2005). Introduction: The process of developing management theory. In K. G. Smith and M. A. Hitt (Eds.), Great minds in management: The process of theory development: 1-6. ISBN 0-19-927681-1.

Kaplan, A. (1964). The conduct of inquiry: Methodology for behavioral science, ISBN 0765804484.

Kovera, M. B. and McAuliff, B. D. (2000). The effects of peer review and evidence quality on judge evaluations of psychological science: Are judges effective gatekeepers? Journal of Applied Psychology, ISSN 85(4): 547-586.

Meehl, P. E. (1992). Cliometric metatheory: The actuarial approach to empirical, history-based philosophy of science. Psychological Reports, ISSN 0033-2941, 71: 339-467.

Meehl, P. E. (2002). Cliometric metatheory: II. Criteria scientists use in theory appraisal and why it is rational to do so. Psychological Reports, ISSN 0033-2941, 91: 339-404.

Mintzberg, H. (2005). Developing theory about the development of theory. In K. G. Smith and M. A. Hitt (eds.), Great minds in management: The process of theory development: 355-372. ISBN 0-19-927681-1.

Richardson, K. A. (2009). Exploring the Implications of Complexity Thinking for the Management of Complex Organizations. In S. E. Wallis (Ed.), Cybernetics and Systems Theory in Management: Tools, Views, and Advancements.

Ritzer, G. (1990). Metatheorizing in sociology. Sociological Forum, ISSN 0884-8971, 5(1): 3-15.

Ritzer, G. (2001). Explorations in social theory: From metatheorizing to rationalization, ISBN 0-7619-6773-7.

Ritzer, G. (2009). Metatheory. In G. Ritzer (Ed.), Blackwell Encyclopedia of Sociology Online, Vol. 2010. Hoboken, New Jersey: Blackwell. http://www.sociologyencyclopedia.com

Shoemaker, P. J., Tankard Jr., J. W. and Lasorsa, D. L. (2004). How to Build Social Science Theories, ISBN 0-7619-2667-4.

Smolin, L. (2006). The trouble with physics: The rise of string theory, the fall of a science, and what comes next, ISBN 0618551050.

Spicer, M. W. (1998). Public administration, social science, and political association. Administration and Society, ISSN 0095-3997, 30(1): 35-53.

Stinchcombe, A. L. (1987). Constructing social theories, ISBN 0-226-77484-8.

Van de Ven, A. H. (2007). Engaged scholarship: A guide for organizational and social research, ISBN 978-0-19-922630-6.

Wallis, S. E. (2006). A sideways look at systems: Identifying sub-systemic dimensions as a technique for avoiding an hierarchical perspective. Paper presented at the International Society for the Systems Sciences, Rohnert Park, California. 0-9740735-7-1.

Wallis, S. E. (2008). From reductive to robust: Seeking the core of complex adaptive systems theory. In A. Yang and Y. Shan (eds.), Intelligent complex adaptive systems: 1-25. ISBN 978-1-59904-717-1.

Wallis, S. E. (2009). The complexity of complexity theory: An innovative analysis. Emergence: Complexity and Organization, ISSN 1521-3250, 11(4).

Wallis, S. E. (2010a). The structure of theory and the structure of scientific revolutions: What constitutes an advance in theory? In S. E. Wallis (Ed.), Cybernetics and systems theory in management: Views, tools, and advancements: 151-174. ISBN 978-1-61520-668-1.

Wallis, S. E. (2010b). Toward a science of metatheory. Integral Review, ISSN 1553-3069, 6(Special Issue: "Emerging Perspectives of Metatheory and Theory"), link.

Wallis, S. E. (2011). Avoiding policy failure: A workable approach, ISBN 978-0-9842165-0-5.

Wallis, S. E. (2012a). The right tool for the job: Philosophy's evolving role in advancing management theory. Philosophy of Management - Special Issue (Guest Editors: Stephen Sheard, Mark Dibben, ISSN 11(3): In press, publication anticipated 2013.

Wallis, S. E. (2012b). Theories of Psychology: Evolving Towards Greater Effectiveness or Wandering, Lost in the Jungle, Without a Guide? Paper presented at the 30th International Congress of Psychology: Psychology Serving Humanity, Cape Town, South Africa.

Wallis, S. E. (2013). Propositional Analysis for Evaluating Explanations through their Conceptual Structures. Paper presented at the International Society for Emergence and Complexity, Paris.