Communications of the ACM
Home/Magazine Archive/July 2009 (Vol. 52, No. 7)/Churchman's Inquirers as Design Templates For Knowledge.../Full Text
Virtual extension
Churchman's Inquirers as Design Templates For Knowledge Management Systems
A number of years ago, Peter Keen6 implored information systems researchers to develop theory they could call their own, which would build a cumulative tradition in the area, while maintaining close ties to practice; as it is central to the field, not peripheral. Even before that, Herbert Simon12 called for the development of a rigorous science of design that would also be relevant to the practice of computer science, information systems, business and engineering.
Relating to design, C. West Churchman's work on the design of inquiring systems1 has had a profound influence on research in management information systems over the years. Mason and Mitroff8 introduced this work into the MIS literature early on effectively making it endogenous to the field. Working both together and independently they went on to publish a host of articles based on inquiring systems and other books and papers by Churchman. They have inspired other IS researchers to follow in their footsteps. Ulrich has a website dedicated to Churchman and his work (http://www.geocities.com/csh_home/cwc_appreciation.html). His inquiring systems have been used as the basis for learning organizations,2,3,4 for knowledge management systems11 and proposed as kernel theories for knowledge management systems in general.7
Despite all the work following on Churchman's original inquiring systems, Mason and Mitroff9 wrote recently: "To say that Singerian and Churchmanian systems are underrepresented is putting it kindly. They are virtually nonexistent." We believe that there are several reasons for this. First, the book is out of print and inaccessible. In addition, it is admittedly a difficult read, especially the passages on the Singerian inquirer. Even in lieu of the work that has been done on inquiring systems, perhaps many IS researchers believe the concepts are passé. After all, it's over 35 years old now! On the contrary, we believe that continued development of inquiring systems theory can provide continuity in IS research, serve as a theory that IS researchers can call their own, and contribute to the practice of information systems design.
The objective of this article is, first and foremost, to gain additional recognition for inquiring systems theories themselves, and to suggest ways in which they can provide the basis for contemporary design research. Essentially we argue that the inquirers can serve as logical design templates for various forms of information systems. We use an object oriented approach to describe the inquirers themselves, and then indicate how the inquirers can be viewed as logical design templates for instantiations or physical designs.
Churchman's Inquirers
In his work, The Design of Inquiring Systems, C. West Churchman1 translates the philosophies of Liebniz, Locke, Kant, Hegel, and Singer into the language of design to form the basis for his Inquiring Systems. The purpose of each of the inquiring systems is the creation of knowledge; however each philosopher's view on how to gather knowledge and build models to represent their view of the world differs.
All of the inquirers, being that they are systems, have inputs, outputs and processes.3 These all differ from system to system, dependant on the philosophy that they follow. The one constant that the systems share is that they have a sort of self-validation mechanism known as the "guarantor" that ensures that only "true" knowledge is created by the system.
As we discuss the inquirers from the perspective of the object oriented methodology, the different inquiring systems will be viewed as subclasses of the inquiring systems class. Their processes will become the methods of the subclass, and their inputs and outputs the attributes. All of them will have a method called "validation" that will be the object oriented incarnation of the guarantor. Despite the fact that the guarantor acts differently in each of the systems, we can refer to the validation method in each inquirer class by the same name through the principle of polymorphism.
As the philosophies of the inquiring systems move through the philosophers chronologically, their complexity also increases and some of the later inquirers incorporate the qualities of earlier ones. This is accounted for in the object oriented methodology by applying the principle of inheritance. The object orientated diagram that depicts the inquiring system class and the relations between the subclasses is shown in Figure 1.
By performing the translation from Churchman's language of design and inquiry to the language of the design practitioners of today (object orientated language), it is our hope that the logical characteristics of the inquiring systems will be more understandable to practitioners of design. To this end, we have listed the types of systems design projects to which the inquiring systems design templates would be applicable.
The Leibnizian Inquirer
The Liebnizian inquiring system builds "fact nets" out of an elementary set of axioms and formal logic. The system is closed in that, once built, it does not accept input from outside the system. The process for creating the fact net consists of creating an hypothesis and then testing it against the basic axioms for consistency. Once the hypothesis is verified, then it joins the fact net.
In the language of object oriented methodology, the Liebnizian class uses the validate hypothesis and identify hypothesis methods. Its attribute is the validation status of the tested hypothesis. This class of inquiring system can serve as the logical design template for systems that take a very formal and analytical approach to make inferences about cause and effect relationships such as mathematical modeling systems, accounting information systems, and systems that describe policies and procedures through document management technologies.10
An example of a Leibnizian system in practice is a system that redacts confidential data from public documents. This system uses a well defined rule base and optical character recognition technology to identify patterns consistent with confidential information stored in the systems internal documents. When a possible pattern of confidential information is recognized (hypothesis identification) the pattern is tested against the rule set of the system (hypothesis testing). A choice is then made on whether to redact the information or not based on the pattern's consistency with the rule set.
The Lockean Inquirer
As opposed to the Liebnizian inquirer, the Lockean inquirer does not rely solely on internal mechanisms, rather it gets its inputs from the outside world in the form of observations. Each of these inputs is then assigned labels. The validity of the labels is determined by a consensus of the community of inquirers. It is also capable of observing its own processes through a reflective process. In this process, the assigned labels are traced back to the most elementary labels in the system.
Represented in the context of object oriented methodologies the Lockean class possesses the methods of observation, assigning labels (interpretation), reflection, and consensus seeking (communication). The attributes are the common language and the complex and elementary labels assigned by the system.
The systems types for which this logical template is applicable are ones that help bring about consensus, such as group decision support systems and those for computer supported cooperative work.10
In practice, graphical information systems (GIS) are good examples of the Lockean logical design template. These systems take observations in the form of satellite imagery and assign meaningful labels (assigning labels) to them. The labels must then be agreed upon by all other members of the community through a communicative process (communication). It is also possible to trace the labels backwards from more complex labels (street address) to more elementary ones (state or nation).
The Kantian Inquirer
The Kantian inquirer contains components of both the Liebnizian and the Lockean inquirers. It is Lockean in the sense that it accepts observations from the outside world and it has the Liebnizian property of hypothesis testing. The system possesses a kinematic component that records the time and space of the inputs. It has multiple models and an executive routine that can turn models on or off depending on how well they are performing. The system guarantor is the extent to which the model fits the data.
With the Kantian inquring object, we see our first notion of inheritance between the subclasses as the Kantian inquirer is essentially a combination of the Liebnizian and Lockean inquirers. Therefore, the methods of the aforementioned classes are shared by the Kantian class. Thus, in addition to the shared methods, the Kantian class has the fit model method. It possesses the attribute of the goodness of model's fit. The goodness of fit of the model is also the determinant of the validate method that serves the guarantor function.
The Kantian class can be used as a logical design template for any system that evaluates multiple interpretations of a situation and has to choose the best interpretation. Forecasting applications, business/competitive intelligence applications, and decision support applications that use multiple models are all examples of this type of system.
The practical instantiation of this type of inquirer is evidenced by any software that takes external data from the environment and fits it to models. The hurricane modeling applications in use by the National Hurricane Center immediately come to mind as a specific example.
The Hegelian Inquirer
The Hegelian Inquirer is constructed on the premise of conflict. In this system, two diametrically opposed views of the world are constructed, a thesis and its arch-nemesis an antithesis, the two of which are then entered into a dialectical process. The views are evaluated in this process and eventually synthesized into a new view that subsumes both of the original views. The guarantor of the process is an "over-seer" that is viewed as a "bigger mind" and is neutral in the debate.
The Hegelian class inherits all of the methods of the Kantian class and in addition to those methods it adds the methods of the dialectic process and synthesis. The attributes of the system are the worldviews that are put in conflict and the newly synthesized worldview that results from the synthesis. The arbitration process provided by the over-seer constitutes the validate function that acts as the guarantor method.
The types of systems best suited for this logical design template are exemplified by groupware designed to support arbitration functions or any systems designed to provide conflict resolution.
In practice, these systems would take the form of legal simulators that take the arguments from the opposing sides of the isle, allows them to debate, and then finds a compromise position. It is important to note that the synthesis is not a consensus as found in the Lockean inquirer; rather the new position in the sense of the Hegelian inquirer is formed when the over-seer takes the most plausible elements of both the thesis and anti-thesis and synthesizes these elements into a revised position that subsumes them both.
The Singerian Inquirer
The strategy of disagreement and a system of measures are the two distinguishing principles that guide the Singerian inquirer. The principle of measurement states that disagreements that occur between community members can be resolved through a system of measures. If the system does not resolve the disagreement, then the disagreement is taken to a more refined measure. The second principle states that disagreement may occur for various reasons, including but not limited to different observer perspectives and inadequate explanatory models. When models fail to explain a phenomenon, new variables and laws are "swept in" to provide guidance and overcome inconsistencies. Despite these efforts to resolve disagreement, it is actually encouraged in the Singerian inquirer. It is through disagreement that world views come to be improved. Complacency is avoided by continuously challenging system knowledge.
The object oriented instantiation of the inquirer is the Singerian object. The object is the most complex object in the Inquiring Systems class. It inherits the properties of the Hegelian object. In addition to these properties, it also has the methods of measurement, sweeping-in, and challenging knowledge. The attributes of the class are the variables that are swept in and the level of refinement of the measure. The guarantor represented by the validation method is similar to the one found in the Hegelian object.
This class can form the foundation for the logical design of complex competitive intelligence applications. Additionally, systems that assist in the solution of social issues that require information from various sources to overcome inconsistencies in the models due to the complex nature of the issues are good candidates for the Singerian design template.
A very relevant practical example of this logical design in practice is executive information systems. These systems sweep-in environmental variables and combine them with internal data and measurement standards to provide real-time assistance to high level managers in organizations.
Conclusion
Over 35 years have passed since C. West Churchman translated the philosophies of Liebniz, Locke, Kant, Hegel, and Singer into the language of design and inquiry to describe his inquiring systems. These inquirers all had the same core purpose: the creation of knowledge. Despite the fact that much time has passed, Churchman's inquiring systems are relatively unknown in the realm of systems design. Part of this obscurity can be attributed to the obtuseness of the work from which they were taken.
We argue that the inquirers have great potential as logical templates for the design of information systems and especially in the design of knowledge management systems. This paper attempts to remedy this situation by re-introducting inquiring systems to the academic and practitioner communities. To make them a bit more palatable to these communities, we have recast Churchman's language of design and inquiry in the terms of object oriented methodology. This methodology was chosen as a vehicle to represent the inquirers for two reasons: to leverage the strength of the methodology, which is reducing complexity by breaking down large concepts into smaller, more manageable components; and to present the inquiring systems in a commonly used language among today's system designers.
Future areas of research in this area could be the application of these design templates in conjunction with the principles of design science to create artifactual instantiations of the inquiring systems. Through the creation of artifacts based on the logical design templates, the templates would be tested as to their validity as true design templates. Other areas of research could include the application of these design templates in the realm of decision support.2
In our opinion, recognition of the potential of inquiring systems has been overlooked by many IS researchers. It is our sincere hope that this paper helps to change that trend and elevate them towards a higher profile in the design of information systems. In so doing, we believe that this Churchman's inquiring systems can contribute to the evolution of Keen's goals of continuity in IS research that maintains a close link with practice.
References
1. Churchman, C.W. The Design of Inquiring Systems: Basic Concepts of Systems and Organizations, Basic Books, N.Y., 1971.
2. Courtney, J.F. Decision making and knowledge management in inquiring organizations: Toward a new decision-making paradigm for DSS. Decision Support Systems 31, 1, (2001), 1738.
3. Courtney, J.F., Croasdell, D.T., and Paradice, D. B. Inquiring organizations. Australian Journal of Information Systems 6, 1, (Sept. 1998), 315.
4. Hall, D.J. and Paradice, D.B. Philosophical foundations for a learning-oriented knowledge management system for decision support. Decision Support Systems 39, 3, (May 2005), 445461.
5. Hall, D.J., D.B. Paradice, and Courtney, J. F. Building a theoretical foundation for a learning-oriented knowledge management system. Journal of Information Technology Theory and Application.
6. Keen, P.G.W. MIS research: Reference disciplines and a cumulative tradition. Proceedings of the First International Conference on Information Systems, Phildelphia, 1980, 918.
7. Linden, L.P., Kuhn, J.R., Parrish, J.L., Richardson, S.M., Adams, L.A., and Courtney, J.F. Churchman's inquiring systems: Kernel theories for knowledge management. Comm. of the Association for Information Systems 20, 52 (2007). Available at: http://aisel.aisnet.org/cais/vol20/iss1/52.
8. Mason, R., and Mitroff, I. A program for research on management information systems. Management Science 19, 5, (Jan. 1973).
9. Mason, R.O. and Mitroff, I.I. Foreward: A dedication to C. West Churchman. Inquiring Organizations: Moving from Knowledge Management to Wisdom, Courtney, J.F., Haynes, J.D., and Paradice, D.B. (Eds). Idea Group Publishing, Hershey, PA, 2005.
10. Parrish, J.L. and Courtney, J.F. IT Support for Inquiring Organizations. Handbook on Decision Support Systems. Burstein, F. and Holsapple, C.W. (Eds.) Springer-Verlag, forthcoming in 2007.
11. Richardson, S.M., Courtney, J.F. and Haynes, J.D. Theoretical principles for knowledge management system design: Application to pediatric bipolar disorder. Decision Support Systems 42, 3, (Dec. 2006), 13211337.
12. Simon, H.A., The Sciences of the Artificial. MIT Press, Cambridge, MA, 1969.
©2009 ACM 0001-0782/09/0700 $10.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2009 ACM, Inc.
No entries found