JOHN P VAN GIGCH
CALIFORNIA STATE UNIVERSITY, SACRAMENTO
School of Business and Public Administration
Abstract: What’s wrong with OR/MS? John van Gigch believes that these fields would grow faster if their proponents devoted more effort to metamodeling. He maintains that these sciences, by being too closely wedded to a single methodology, are restricting their innovative and creative thrust. He proposes that OR/MS and GST should now move toward each other.
When comparing OR/MS (Operations or Operational
Research/Management Science) and GST (General Systems Theory), it is useful to take an epistemological
perspective of both disciplines. This perspective requires
that the following questions be answered
1. What are the objects of study of each discipline?
2. What are the objectives of each discipline?
3. What methodology does each discipline use to meei
its avowed objectives?
COMPARISON OF THE OBJECTS OF STUDY OF THE DISCIPLINES
In order to provide a comparison of the objects of study of the two disciplines, we postulate the existence of three distinct categories of objects of study:
- Metamodeling, where the object of study is the methods which are used to design and formulate models. Metamodeling provides the epistemological framework for modeling (see below)
- Modeling, where the object of study is the design and formulation of models.
- Implementation of Models, where the object of study consists of the application and direct use of the models, derived earlier, to the real world.
The use of these three objects of study can be associated with the three levels of the metasystem paradigm (see van Gigch, 1984 and 1987a), where Modeling takes place at the object level, Metamodeling at the metalevel, and Implementation at the lower level of the hierarchy of control levels. See Figure 1.
On the basis of a study of the papers published in the professional journals of each discipline, distribution depending on whether they deal with the topic of Metamodeling, Modeling, or Implementation can easily be obtained,
It is hypothesized that the proportions of papers devoted to Metamodeling, Modeling, and to Implementation are, respectively, 10 percent, 50 percent and 40 percent in the OR/MS journals such as Management Science and Operations Research. These proportions are reversed to 40 percent, 50 percent and 10 percent in the GST journals such as Systems Research, Behavioral Science, and General Systems. See Table 1. (Note that no study or compilation of actual figures was carried out for this short essay.)
COMPARISON OF THE OBJECTIVES OF THE DISCIPLINES
Two statements gleaned from current publications exemplify distinct objectives:
- “The Institute of Management Sciences is an international society which seeks to identify, extend and unify scientific knowledge pertaining to management.”
- “The International Society for General Systems Research is an international organization for the development of theoretical systems,”
Operations Research and Management Science came into being as disciplines whose objectives were directly related to the improvement of the decision-making ability of management and managers. On the other hand, General Systems Theory was born with a more grandiose objective: that of producing a paradigmatic revolution. It wanted to turn away from the traditional and classical Scientific Method and create a new science of living and open systems which would be devoted to integrating, synthesizing and finding isomorphism across other disciplines.
We will not discuss whether these twp distinct goals have been met or whether the respective disciplines have been successful in achieving their stated purposes. On the whole, it can be said that the disciplines can still be clearly distinguished by those two statements of intent. Accordingly, and for all intents and purposes, the disciplines stand almost at opposite ends of the Pure Science-Applied Science spectrum.
COMPARISON OF THE METHODOLOGIES USED BY EACH OF THE DISCIPLINES
OR/MS has always insisted on the necessity of following the traditional Scientific Method. It demands from its practitioners that they follow the traditional Science Paradigm which goes from observations to theory and from hypotheses to experimentation and measurement, to finally end in generalizations, laws and predictions.
GST has been less traditional and has not insisted that its disciplines follow any established paradigm or scientific method. As a matter of fact, it repudiated the analytical reductionistic approach and adopted the holistic all embracing Systems Approach.
It is possible to draw a list of the possible methodologies available and hypothesize the proportion used by each discipline. The methods include:
Treatises of a Conceptual Nature such as discussions of hilosophical, epistemological and theoretical questions.
Solutions Through the Analysis of Models.
Experimental Methods (including laboratory and field experiments.
Participant-Observer Methods (including clinical methods, surveys and case studies).
The hypothesized proportion used by each discipline is shown in Table 2. While one can argue about the exact proportions, the intent is to give the reader an idea of rough trends.
On the basis of a perusal of current publications, we can conclude that OR/MS definitely favors the Methodology 2 (Solution Through the Analysis of Models) and uses the other methodologies in only a sprinkling of cases. GST favors Methodology 1 (Treatises of a Conceptual Nature) and uses the Solution Through the Analysis of Models sometimes, but seldom uses the other two methodologies.
Based on the comparison of the objects of study, the objectives and the methodologies employed by both disciplines, the following conclusions can be drawn.
- By not devoting enough intellectual effort to the study of Metamodeling, OR/MS is limiting the scope and extent of its own growth. A discipline withers if it does not pursue the development of its own epistemological foundations. (van Gigch, 1987b)
- Based on the diametrically opposite positions which OR/MS and GST hold on the Pure-Science- Applied Science spectrum, it would seem appropriate to encourage both disciplines to move toward each other i. e., to encourage OR/MS scientists to renew their conceptual and theoretical foundations, while the GST scholars should seek to complement their theoretical cogitations with more applied science and practical work.
- The discipline of OR/MS is much too closely wedded to the use of one single methodology which consists of the Solution Through the Analysis of Models, thus restricting its innovative and creative thrust. It should regain some of the boldness of the early years by encouraging departure from its hallowed and traditional approach. On the other hand, GST accepts more esoteric and untried methodologies but only concentrates upon theoretical and conceptual approaches. GST’s originality demonstrates vigor and vitality, but often runs the risk of reduced acceptance from the conservative scientific community.
- The divergence between OR/MS and GST is not widening by design. It is natural that each discipline attracts scientists whose intellectual bent is closer to the ideal definition of what constitutes the object of study, objectives and methodologies characteristic of each. It would appear that, as was spelled out 10, 20 and even 30 years ago, both disciplines would probably benefit from more cross-fertilization. Neither of them can claim to be closer to the truth and consequently the pursuit of knowledge must be carried out along both paths.
1. van Gigch, John, “The Metasystem Paradigm as a new Hierarchical Theory of Organizations”, Proceedings of the 1984 Annual Meeting of the Society for General Systems Research, New York City, May 1984. Also presented at ORSAITIMS 1984 Annual Meeting, San Francisco, May 1984 under the name of “Strategic Planning as a Form of Metadesign”. Reprinted in General Systems, vol. XXIX, 1985 – 86.
2. van Gigch, John, ~ Decision Making About Decision Making, Metamodels and Metasystems, Tunbridge Wells, Kent, England, Abacus Press, 1987a.
3 van Gigch, John, “The Potential Demise of OR/MS: Consequences of Neglecting Epistemology”, Sacramento, California, California State University, 1987b. Submitted for publication.
FIGURE 1: Metasystem hierarchy applied to the objects of study
Metalevel METAMODELING MCR Object Level MODELING CR Lower Level IMPLEMENTATION CS
TABLE 1: Hypothesized comparison of the proportion of articles published at each level of the metasystem hierarchy by OR/MS and by GST
Proportion of Articles Published at Each Level of the Metasystem Hierarchy OR/MS GST METAMODELING 10% 40% MODELING 50% 50% IMPLEMENTATION 40% 10%
TABLE 2: Hypothesized percentage distribution showing the methodologies used by each discipline
Percentage Utilization by each discipline OR/MS GST 1. Treatises of a Conceptual Nature 10% 70% 2. Solution Through the Analysis
60% 20% 3. Expenmental Methods 10% 5% 4. Participant-Observer Methods 20% 5%