I. General Systems TheoryAnatol Rapoport
III. Political SystemsWilliam C. Mitchell
IV. International SystemsMorton A. Kaplan
V. Psychological SystemsDavid S. Gochman
General systems theory is best described not as a theory in the sense that this word is used in science but, rather, as a program or a direction in the contemporary philosophy of science. The outlook represented by this direction stems from various sources, and its adherents emphasize different aspects of the program. However, all the variants and interpretations have a common aim: the integration of diverse content areas by means of a unified methodology of conceptualization or of research.
The scientific background . The traditional outlooks of the physical and biological sciences may be taken as examples of divergent methodologies or conceptualizations. In the eighteenth century, theoretical physics, at least the branch known as mechanics, already appeared in full mathematical garb. So firmly established were the mathematical principles of mechanics that this discipline seemed to be a realization of the rationalists’ program— the derivation of knowledge from first principles by deduction alone. Indeed, the theorems of mechanics were in no way less rigorously derived nor less certain of experimental corroboration than were the theorems of geometry. Because mechanics was the branch of physics which matured first, the notion was not uncommon in the beginning of the nineteenth century that all the laws of being and becoming were manifestations of mechanical laws —in other words, that the universe was a strictly determined clockwork, whose operation would be completely understandable to an intelligence sufficiently vast to grasp the totality of its components and the relations among them.
Biology, in contrast, was at that time an almost wholly descriptive—at most an inductive, hardly ever a deductive—science. Life was tacitly assumed to be a phenomenon sui generis, apart from events governed by mechanical laws. At least, no serious attempts were made to derive the former from the latter. Therefore, a gap existed between the physical and the biological sciences. The basic terms of the latter—for example, organism, survival, reproduction, development, behavior, senescence, death —had no counterpart in the physical sciences.
Reductionism and vitalism . With the fundamental discoveries in the middle of the nineteenth century—the laws of thermodynamics—and with the maturation of chemistry, the relationship between the physical and the biological sciences began to change. Physiologists began to look upon basic life processes as they looked on physicochemical events, and, as such, these processes seemed in no way to differ from similar events occurring in nonliving environments. In particular, the laws of conservation of matter and energy were shown to be valid in living organisms, and the living organism began to appear to the physiologists as a machine. Thus the point of view known as reductionism emerged. Reductionism is essentially a program which seeks to derive events occurring at one level of organization from those occurring at another, presumably both a simpler and a more fundamental level. The reduction of chemistry to physics has been largely successful. The reduction of physiology to chemistry and physics was viewed by the reductionists as their most significant task.
Opposed to the reductionists’ program were the vitalists, who maintained that life is a phenomenon sui generis and that therefore the reductionists’ program was futile.
It should be clear that such a controversy can never be settled to the satisfaction of either party. As long as the reduction of all life phenomena to physics and chemistry has not been carried out, the vitalists can keep insisting that it never can be carried out. On the other hand, there is no ground for supposing that something cannot be done just because it has not been done.
Bertalanffy’s theory of systems. On occasions, the vitalists have attempted to support their position by specific evidence: for example, the apparent ideological nature of some life processes (the so-called principle of entelechy or “equifinality” emphasized by Hans Driesch 1908) and the apparent violations by living organisms of the second law of thermodynamics. Although these arguments have since been shown to be irrelevant to the issue, they stimulated lively discussions which led to one of the early formulations of general systems theory by Ludwig von Bertalanffy (1956; 1962).
Bertalanffy pointed out that apparent goal seeking was not an exclusive characteristic of living systems. He called attention to an essential difference between an isolated system of chemical reactions and an open one, in which sources and sinks were present. In an isolated system, after equilibrium has been attained, the relative concentration of substances depends, of course, on the initial concentrations of the reactants (because of the conservation of mass); thus, the final state of the system depends on the initial conditions. In an open system, however, a steady state may be attained in which the final concentrations are virtually independent of the initial conditions. Moreover, if the steady state is disturbed, as by adding or removing quantities of reacting substances, it will re-establish itself, being determined by the characteristics of the entire system rather than by any specific state of the system. Thus, an open system will appear to exhibit “equifinality” to a naive observer. It will appear to have a “will of its own” or a “purpose”: to maintain the steady state— which, incidentally, is just what living systems are to a large extent engaged in doing by means of their well-known homeostatic (steady-state restoring) mechanisms.
It is noteworthy that the systems cited by Berta-lanffy as examples of entities exhibiting characteristics of equifinality were open systems, i.e., those to which the operation of the classical version of the second law of thermodynamics does not apply. Thus, by calling attention to the fundamental feature of Jiving organisms as open systems, Berta-lanffy refuted both of the specific arguments put forward by the vitalists.
What is a system?
The classification of systems by the nature of their relation to their environments and the search for laws governing the behavior of each class can be said to be problems posed by a general systems theory. Once one has raised the general questions about possible laws governing the behavior of systems, the problem of rigorous definition of a “system” comes to the fore. In common usage, the word refers to widely separated concepts. Engineers are concerned with systems as functionally related aggregates of technological devices. Physiologists single out functionally related portions of living organisms (circulatory, digestive, nervous systems). Social scientists speak of economic and political systems; philosophers, about systems of thought.
It is, of course, by no means necessary to derive from what may be an accident of usage the idea that all the “systems” which have been so named have something important in common. On the other hand, one need not dismiss such an idea out of hand. Thus the question looms as to what to include in and what to exclude from the definition of “systems,” in order to stretch the concept to the limit of generality without at the same time destroying its usefulness.
I subscribe to the view that a definition of “system” must be such that other than physical entities (perhaps language) are included. At the same time, the definition must exclude entities whose principles of organization cannot be at least partially specified. Therefore, I accept the definition of a system as (1) something consisting of a set (finite or infinite) of entities (2) among which a set of relations is specified, so that (3) deductions are possible from some relations to others or from the relations among the entities to the behavior or the history of the system.
According to this definition, both the solar system and a language qualify as systems. In the former, the entities are the sun and the planets; the relations among them are specifiable as position and velocity vectors and forces of gravitational attraction. Other relations (e.g., Kepler’s laws of planetary motion) and the history of the system, past and future, are derivable from the given relations. In a language, there are also identifiable entities—phonemes, morphemes, sentences, and the like—and relations between these are given in terms of syntactic rules. In a larger sense, a language system may also include the referential world and even the speakers. In this sense, semantic and pragmatic relations are added to the syntactic ones. “Social system” is a term so widely used that its meaning is assumed to be obvious. However, in the context of a systems theory, “social system” would have to be defined de novo every time some class of entities (individuals, families, institutions) and relations among them (communication channels, influence, obligations) are singled out for attention.
The organismic approach
As has been said, early explicit programmatic formulations of general systems theory were outlined by Bertalanffy. Another biologist, Ralph W. Gerard (1958), has offered a formulation which carries an even stronger biological flavor. According to Gerard, a “system” is primarily a living system, and the process which defines it is the maintenance of an organization which we know as life. There is a hierarchy of systems, in which the larger ones frequently include the smaller ones as components or subsystems. Thus, cells form organized aggregates known as tissues or organs; these, in turn, are components of a biological individual. Individuals stand in relation to each other as families or tribes (social arrangements) or as species (interbreeding aggregates). Along the scale of social organization, we have the aggregates characteristic of human beings—institutions, political units, and societies. Along the scale of biological organization, organisms and populations stand in symbiotic, predatory, or parasitic relation to each other, forming ecological systems (ecosystems). To view an ecosystem as an “epiorganism” is not merely to indulge in metaphorical analogies. Metabolic chains and cycles are traceable through a biological community quite as precisely as through the various specialized cells of a single organism. Herbivores eat plants; carnivores eat herbivores and smaller carnivores. Under proper conditions, the ecosystem may reach an equilibrium quite analogous to the homeostatically maintained metabolic equilibrium of the individual organism.
Thus, in Gerard’s scheme, the hierarchy of living systems from cell to society, or the entire biota, constitutes one dimension. The levels of organization are the horizontal rows of a matrix, of which the vertical columns are three aspects of living systems: (1) structural, (2) behavioral, and (3) evolving. Structure, in Gerard’s view, is a description of the interrelations among the components of a system: the arrangement of its parts and the potential influence which they may have upon each other. For example, the topology of neural tracts, together with the catalogue of their potential action (excitatory or inhibitory), reveals the structure of a nervous system; an organizational chart reveals the structure of an institution.
According to Gerard, behavior refers to the short-term reversible changes of state of a living system, its immediate responses to environmental stimuli, the functions performed by its homeostatic devices in maintaining certain steady states, etc. Nervous activity and metabolic processes belong under this rubric, as well as the behavior patterns of higher animals and the short-term actions of organized social bodies. Finally, the third rubric deals with the long-term, typically irreversible changes—the development of the embryo, the growth of an individual, the history of a society, the evolution of a species.
The three aspects just described could be called “being” (structure), “acting” (behavior), and “becoming” (history). The intersections of their respective columns with the rows of the matrix (the levels of organization) determine particular fields of inquiry. For example, anatomy is the study of structure at the level of the individual; history is the study of development at the level of a society; embryology is the discipline in the same column as history but at the level of the individual (early stage); and histology deals with structure at the level of the cell.
It has already been said that general systems theory is not, strictly speaking, a scientific theory but, rather, an outlook. Gerard’s scheme represents this outlook in its most purely programmatic garb, inasmuch as the matrix of levels and of their three aspects does not imply any theoretical assertion. However, the scheme does represent a possibly fruitful way of viewing the world of living systems, in the sense that it is suggestive of dependencies and analogies.
James G. Miller (1955) has proposed a program for listing hypotheses (which, when verified, could become general propositions) concerning similarities or differences between analogous events taking place on different levels of system organization. These “levels” in Miller’s conceptual scheme are identical with those in Gerard’s. For example, living systems grow; is there one law of growth on the cell level, and another law—perhaps similar, perhaps very different—which applies to the growth of the individual, the group, the society, etc.? Living systems process information and utilize it to maintain their viability. Are there propositions concerning information processing which can be maintained (with possible modifications) about all levels of organizations?
The mathematical approach
In my view, the most fundamental feature which distinguishes a system from other aggregates or from an arbitrarily circumscribed portion of the world is the possibility of describing it in purely structural terms. Here the word “structural” refers not necessarily to specific components or physical features but, rather, to relations (which may be relations among parameters as well as relations among parts). A system, roughly speaking, is a bundle of relations. For this reason a general systems theory, in my opinion, ought to single out purely relational isomorphisms that are abstracted from content.
As an example, consider a mathematical formulation of the growth of some system. Specifically, let a physical system be a solid body with a boundary, and let growth be the result of ingestión of substances from the outside through the boundary at a constant rate per unit of surface. Moreover, let the substance which makes up the system break down inside the system at a constant rate per unit mass and be excreted. Then, since the surface is proportional to the two-thirds power of the volume, while mass is proportional to the volume, we have the equation
dm/dt = am* — bm,
where m is the mass and a and b are constants (Bertalanffy 1957). Such will be the ’law of growth” of all physical systems of this sort, regardless of size or internal organization. On the other hand, if the system is essentially one-dimensional (i.e., grows only at the ends at a constant rate, while breaking down at a constant rate per unit mass), its law of growth will be
dm/dt = a — bm.
Evidently not the “level” of the system but, rather, its geometry is likely to be the determinant of its law of growth. If so, then attempts to specify particular laws of growth for “cells,” “populations,” “corporations,” etc., will prove futile.
Isomorphisms . The objections to the so-called organismic general systems theory, centered on levels of organization, have stimulated an altogether different approach to the subject: one founded on mathematical homologies rather than organismic homologies. The strictest mathematical homology is called an isomorphism. Two mathematical objects are isomorphic if there exists a one-to-one correspondence between the elements of one and those of the other and if the relations among the elements are preserved by the same correspondence. If two physical systems obey the same mathematical law, they are also isomorphic to each other. A famous example of such isomorphism is that between a mechanical harmonic oscillator and an electrical circuit with an inductance, a resistance, and a capacitance. As is well known, the differential equation of the former is
where x is the displacement of the mass m; r is a coefficient of friction; k is the elasticity modulus, associated with the restoring force; and f(t) is an impressed force, which may be a function of time. The differential equation of the electrical system is given by
where q is charge, L the inductance, R the resistance, C the capacitance, and E(t) an impressed electromotive force.
The isomorphism is apparent from the identical forms of the equations. Any law of behavior derived from the equation with respect to one system has an exact analogue with respect to the other.
Moreover, a fundamental set of “homologies” is established between mass and inductance, between electrical resistance and friction, between capacitance and elasticity—homologies which possibly would not have occurred to one preoccupied with the specific content of the events rather than with their mathematical structure. Yet the homologies are quite “real.” For example, the heat produced by “overcoming” electrical resistance is the same sort of heat as that produced by overcoming friction.
Classification of systems . Here, then, is a unifying principle, which truly abstracts from the content of phenomena and concentrates on the structural and dynamic relations, in terms of which the phenomena are described. If we follow the definition of “system” given above to its logical conclusion—namely, as a specified set of entities and a set of relations among them—then it would seem that the method of mathematical homology is the most natural foundation of a general systems theory. For an exact specification of relations is practically synonymous with a mathematical specification. The system is specified as a particular mathematical model and is seen at once to be isomorphic to all systems specified in terms of models of the same type.
In this light, the classification of systems derives from a classification of mathematical models. For example, all systems involving monomolecular chemical reactions are representable by systems of the first-order, first-degree linear differential equations. Moreover, closed systems are isomorphic to homogeneous systems of equations (without constant terms), while open systems are isomorphic to nonhomogeneous systems of equations (which include constant terms). The absence or the presence of the “equifinality” that was attributed by the vitalists to specific life forces is directly derivable from the nature of such equations, depending on whether they are or are not homogeneous. Bimolec-ular reactions are represented by systems of the second degree. These systems are much more complex than linear ones and may have special features like thresholds, which divide the phase space into “watersheds,” so that the steady state which finally obtains may be determined by the direction of a chance fluctuation about an unstable equilibrium.
Integration of knowledge
Realizations of such systems are not confined to chemistry. Ecological systems can also, in principle, be represented by systems of differential equations, from which their characteristic features, such as presence or absence of stable equilibria, oscillations, etc., can be derived. To the extent that systems of equations of this sort can be supposed to underlie any phenomenon whatsoever—chemical, biological, or social (e.g., mass behavior)—these phenomena must exhibit homological laws, so that concepts from one field of investigation are com-pellingly translated into those of another, in the same way that capacitance is translated into elasticity, whatever our preconceived notions may be about the nature of either.
This interchangeability of concepts can already be seen in the fusion of biological and social theories: for example, in the way Malthus influenced Darwin, who, in turn influenced Herbert Spencer and Karl Marx. The essential orientation of all these writers is characterized by an emphasis on the “massive,” deterministic aspects of both biological and social phenomena. The general systems orientation led to a much more precise, mathematical statement of similar ideas. Some examples of mathematico-sociological and mathematico-historical theories were elaborated by Lewis F. Richardson (1960) and Nicolas Rashevsky (1953). Likewise, John W. Thompson (1961) has outlined a physicalist approach which links the concepts of meteorology and sociology.
The mathematical homology method solves the problem of “integrating” knowledge stemming from disparate disciplines via the translation rules that are rigorously derived from mathematical models. The method provides a basis for resolving the interminable controversies regarding terminology in the behavioral sciences—whether, for example, “power,” as it is understood in political science has any relation to “power” as it is understood in sociology, or whether either has any relation to “influence” or “status” as these are understood in social psychology, or whether “energy” as this term is used in psychodynamics has any bearing on physical energy. If a term enters as a homologous variable or parameter in two or more isomorphic models, then the term plays the same part in the respective theories; if it does not, then the opposite is true.
Precision and specification . The mathematical model approach to general systems theory has one serious, at times crippling, drawback. To define a system, a much more precise specification of entities and relations is required than our knowledge usually warrants. Here a word must be said to forestall a possible misunderstanding of what is meant by “precise” in a rigorous theory. Precision is often understood in terms of accuracy of measurement or in terms of the degree of determinism of deduced conclusions. Thus, celestial mechanics is precise (it is sometimes referred to as an exact science) in the sense that positions of heavenly bodies are calculated with great accuracy and the predictions of the theory are extremely reliable. In this sense, meteorology is not “precise,” because the determination of variables relevant to the prediction of weather is much more difficult than the determination of variables relevant to celestial mechanics.
Nevertheless, the systems studied in meteorology are as precisely defined as those studied in celestial mechanics. One knows what one means by the state of a meteorological system—namely, the distribution of temperatures, pressures, wind velocities, etc.—and one is sure that the state of the system determines a certain succession of states in time. Only the difficulty of determining the precise state at a given time and the enormity of the calculations make it impossible for us to predict the weather as precisely as we predict eclipses. In short, we must distinguish between precise results and precise specifications.
To be precisely specified, a system need not even be deterministic. The variables of interest may be the probabilities of states; by means of stochastic models, we can calculate the distributions of these probabilities at future times, given some initial distribution. This sort of theory is no less precise than a deterministic one. In short, a system is precisely defined if the states in which it can be are precisely specified (not necessarily actually determined) and the laws of progression of these states (which may be probabilities) are precisely postulated (not necessarily actually verified).
The organismic approach reconsidered . When we say that many events do not lend themselves to the above sort of description, we mean that it is difficult to specify the states and to postulate the dynamic laws which determine their progression. At times we can say a great deal more about systems if we forgo attempts at such precise specifications. For example, a great deal can be said about living systems without any rigorous specification of “states” and of dynamic laws. We know that all living systems come into being; maintain themselves in more or less steady states in the midst of a changing environment; enter into symbiotic or predatory relations to each other; mature; reproduce themselves, if they are individual organisms—and also, frequently, if they are aggregates (e.g., bee hives); grow old; and cease to exist as organized systems. Furthermore, detailed analysis establishes more specific laws of existence for living systems: for example, the need for outside sources of energy. Similar analysis establishes general conditions of existence of organized social aggregates: for example, the existence of channels of at least internal communication of internalized codes of behavior.
There is no doubt that analysis of this sort yields knowledge and that much of this knowledge can be organized into systematic descriptions and predictions even without the aid of rigorous systems analysis in the mathematical sense of the word “rigorous.” To the extent that the organismic approach to a general systems theory can bypass the obstacles to mathematical analysis mentioned above, the organismic approach has its special heuristic advantages. It can therefore be viewed as complementary to the mathematical approach to a general systems theory.
We have traced the current interest in general systems theory to two classes of investigators, the biologists and the mathematicians. The former have had a long-felt need to spell out some features of the organismic view (always predominant in biology) which can contribute to the pressing problem of integrating knowledge. The latter have contributed the rigor of formal analysis to the formulation of a systems theory whose integrative power derives from the process of abstraction characteristic of mathematical thought. There is still a third source of ideas which feeds general systems theory: the modern conception of a technological system.
The structure of a technological system (i.e., an aggregate of interrelated technological devices) is completely known, since such systems are designed by men. Consequently, the problem is not to discover what the important elements are and how they are related but, rather, to determine the overall behavior of a system whose structure is specified. Once methods of solving these problems are developed, the systems engineer can pose the problem of optimal system synthesis: what elements to use and how to relate them to each other in order to achieve a performance that is optimal in some given respect.
General systems theory contributes to the solution of such problems by placing them in a general structural context abstracted from specific content. For example, the emergence of cybernetics can be viewed as a development in the spirit of general systems theory. Cybernetics is a science which deals with the information-processing aspects (as distinguished, say, from the energy-transforming aspects) of all systems, regardless of their physical nature. This point of view has greatly facilitated the development of automatic control, telecommunications, and computing technologies. The influence of cybernetics, however, has not been confined to technology. The ideas of information theory (which underlie cybernetics) have helped to unify thinking in such apparently very disparate fields as systems engineering, economics, and neuro-physiology by singling out the concepts underlying all of them, such as homeostasis (maintenance of dynamic equilibria) and transmission of information.
The future of general systems theory
In short, the task of general systems theory is to find the most general conceptual framework in which a scientific theory or a technological problem can be placed without losing the essential features of the theory or the problem. The proponents of general systems theory see in it the focal point of resynthesis of knowledge. There was a time when the man of knowledge was a generalist rather than a specialist, that is, he embodied the knowledge of principles rather than skills. He was the philosopher and the sage, and his epistemological creed was most clearly stated by Plato, who believed that all real knowledge comes from within rather than from without, that is, from the contemplation of what must be rather than what seems tobe.
The rise of science and of experimental method puts this extreme rationalist view under a shadow. The legitimate sources of scientific knowledge came to be restricted to data derived from direct contact with the observable world. However, there was a price to be paid for this: the fragmentation of knowledge into specialties. Hand in hand with the fragmentation, however, new syntheses have appeared. Mathematical physics is the best-known example, and the evolutionary principle as the key theme of biology is another. “Systematization” of the social sciences has also been attempted, the works of Marx and Toynbee being among the most ambitious of such attempts.
The main theme of general systems theory is, I believe, the explicit fusion of the mathematical approach with the organismic. The key task of general systems theory is to show how the organismic aspect of a system emerges from the mathematical structure. At times, classical mathematical methods suffice to bring this out; for example, organismic aspects emerge from the properties of systems of differential equations, including trends toward equilibrium states that are independent of initial conditions, stability properties, etc.
However, classical mathematics is not able to handle complex structural features. Organization is best depicted as a network, and the mathematical theory of networks derives largely from certain branches of topology and abstract algebra rather than from analysis, which underlies classical mathematics. Thus the salient feature of a nervous system, of an institution, or of international systems may well reside in the vastly complex network of relations which constitute them: for example, functional neural pathways, lines of communication and authority, links of alliances or rivalries in international trade. If the “nature” of the system is indeed embodied in the quality and interrelations of these connections, then there is hope that knowledge of wholes will emerge from knowledge of the parts.
Moreover, in the system theoretic view, the whole can be viewed as a unit no less than a part can. Hence, the fate of the components of a system may be viewed as determined by the fate of the whole system as legitimately as the other way around. For example, the organism goes through certain stages of maturation because its cells differentiate, but the process of differentiation is also the result of the maturation. The organism dies because the cells die, but the converse is also true. National policies are set by leaders, but the selection of leaders depends at least in part on the inertia of ongoing policies.
These observations are rather commonplace and do not in themselves constitute theories. However, a rigorous deduction of these principles as system properties may well involve profound theoretical discoveries. Therein lies the promise of general systems theory.
[Directly related are the entries Cyberneticsand Information Theory]
Ashby, W. Ross 1958 General Systems Theory as a New Discipline. General Systems 3:1-6.
Bertalanffy, Ludwig Von 1956 General System Theory. General Systems 1:1-10.
Bertalanffy, Ludwig Von 1957 Quantitative Laws in Metabolism and Growth. Quarterly Review of Biology 32:217-231.
Bertalanffy, Ludwig Von 1962 General System Theory: A Critical Review. General Systems 7:1-20.
Boulding, Kenneth E. 1956 General System Theory: The Skeleton of Science. General Systems 1:11-17.
Driesch, Hans (1908) 1929 The Science and Philosophy of the Organism. 2d ed. Vol. 2. London: Mac-millan. → See especially Part 3.
Foster, Caxton C.; Rapoport, Anatol; and Trucco, Ernesto 1957 Some Unsolved Problems in the Theory of Non-isolated Systems. General Systems 2: 9-29.
Gerard, Ralph W. 1958 Units and Concepts of Biology. Behavioral Science 3:197-206.
Miller, James G. 1955 Toward a General Theory for the Behavioral Sciences. American Psychologist 10: 513-531.
Miller, James G. 1965a Living Systems: Basic Concepts. Behavioral Science 10:193-237.
Miller, James G. 1965b Living Systems: Structure and Process; Cross-level Hypotheses. Behavioral Science 10:337-411.
Rapopoht, Anatol 1967 Mathematical Aspects of General Systems Analysis. Unpublished manuscript.
Rashevsky, Nicolas 1953 Outline of a Mathematical Approach to History. Bulletin of Mathematical Biophysics 15:197-234.
Richardson, Lewis F. 1960 Arms and Insecurity. Edited by Nicolas Rashevsky and Ernesto Trueco. Pittsburgh: Boxwood; Chicago: Quadrangle Books.
Thompson, John W. 1961 Meteorological Models in Social Dynamics. Human Relations 14:43-62.
“System” is the concept that refers both to a complex of interdependencies between parts, components, and processes that involves discernible regularities of relationship, and to a similar type of interdependency between such a complex and its surrounding environment. System, in this sense, is therefore the concept around which all sophisticated theory in the conceptually generalizing disciplines is and must be organized. This is because any regularity of relationship can be more adequately understood if the whole complex of multiple interdependencies of which it forms part is taken into account.
Social systems and the action system
Methodologically, one must distinguish a theoretical system, which is a complex of assumptions, concepts, and propositions having both logical integration and empirical reference, from an empirical system, which is a set of phenomena in the observable world that can be described and analyzed by means of a theoretical system. An empirical system (e.g., the solar system as relevant to analytical mechanics) is never a totally concrete entity but, rather, a selective organization of those properties of the concrete entity defined as relevant to the theoretical system in question. Thus, for Newtonian solar system mechanics, the earth is “only” a particle with a given mass, location in space, velocity, and direction of motion; the Newtonian scheme is not concerned with the earth’s geological or human social and cultural characteristics. In this sense, any theoretical system is abstract.
As a theoretical system, the social system is speciflcally adapted to describing and analyzing social interaction considered as a class of empirical systems. These systems are concerned with the behavior, as distinguished from the metabolic physiology, of living organisms. Among the categories of organisms, our interest in this article centers on human social interaction, which is organized on the symbolic levels we call “cultural.” However, one should remember that such interaction is a late evolutionary product and is continuous with a very broad range of interaction phenomena among other organisms. All bisexual reproduction, for example, requires highly structured interactive relations between the organisms of the two sexes. Various kinds of interspecies ecological relations constitute another example, one to which human relations with domesticated animals are relevant.
The aspects of behavior which directly concern “cultural-level” systems I call action. Action in this technical sense includes four generic types of subsystems, the differentiation among which has gained fairly clear definition during modern intellectual history.
The first is simply the organism, which, though quite properly treated as a concrete entity in one set of terms, becomes, on a more generalized level, a set of abstract components (i.e., a subsystem) in the culturally organized system of action.
A second subsystem is the social system, which is generated by the process of interaction among individual units. Its distinctive properties are consequences and conditions of the specific modes of interrelationship obtaining among the living organisms which constitute its units.
Third is the cultural system, which is the aspect of action organized about the specific characteristics of symbols and the exigencies of forming stable systems of them. It is structured in terms of patternings of meaning which, when stable, imply in turn generalized complexes of constitutive symbolisms that give the action system its primary “sense of direction,” and which must be treated as independent of any particular system of social interaction. Thus, although there are many ramifications into such areas as language and communication, the prototypical cultural systems are those of beliefs and ideas. The possibilities of their preservation over time, and of their diffusion from one personality and/or social system into another, are perhaps the most important hallmarks of the independent structure of cultural systems.
Fourth, the analytical distinction between social and cultural systems has a correlative relation to the distinction between the organism and those other aspects of the individual actor which we generally call the personality. With the achievement of cultural levels of the control of behavior, the primary subsystems of action can no longer be organized—or structured primarily—about the organic base, which, in the first instance, is anatomical or “physical.” Personality, then, is the aspect of the living individual, as “actor,” which must be understood in terms of the cultural and social content of the learned patternings that make up his behavioral system. Here, “learned” refers not only to the problem of the origin of the patterns in the heredity-environment sense, but also to the problem of the kind and level of their content. The connection between these two problems partly reflects the fact that we have no evidence that cultural content is, at what we call here the level of pattern, determined through the genes. Thus, there is no evidence of a hereditary “propensity” to speak one language rather than another, although the genetically determined capacities to learn and use language are generally fundamental.
Thus, we treat the social system, when evolved to the action level, as one of four primary subsystems of action, all of which articulate with the organic bases of life and with organic adaptation to the environment in the broadest biological sense.
There is a sense in which the social system is the core of human action systems, being the primary link between the culture and the individual both as personality and as organism—a fact for which “culture and personality” theorists have often not adequately accounted. As the principal source of the independence of cultural systems from restrictive organic and environmental conditions, it has been the primary locus of the “operation bootstrap” of human evolution. The secret of this evolutionary capacity evidently lies in the possibility for “reverberation” among the intercommunicating members of a social system, each of whom is both an actor orienting himself to his situation in terms of complex, cultural-level, intended meanings and an object of orientation meaningful to orienting actors. Furthermore, each person is both actor and object to himself as well as to others. Interaction at the symbolic level thus becomes a system analytically and, very appreciably, empirically independent of its presymbolic bases (though still grounded in them), and is capable of development on its own.
Insight into this basic complex of facts constitutes a principal foundation of modern social science theory. It has been attained by convergence from at least four sources: Freud’s psychology, starting from a medical-biological base; Weber’s sociology, which worked to transcend the problems of the German intellectual tradition concerning idealism-materialism; Durkheim’s analysis of the individual actor’s relations to the “social facts” of his situation; and the social psychology of the American “symbolic interactionists” Cooley and Mead, who built upon the philosophy of pragmatism. [See Interaction, article onsocial Interaction.]
In dealing with social systems, one must distinguish terminologically between an actor as a unit in a social system and the system as such. The actor may be either an individual or some kind of collective unit. In both cases, the actor within a system of reference will be spoken of as acting in a situation consisting of other actor-units within the same system of reference who are considered as objects. The system as a whole, however, functions (but does not “act” in a technical sense) in relation to its environment. Of course, the system references are inherently relative to particular scientific problems. When a collective (i.e., social) system is said to act, as in the case of a government conducting foreign relations, this will mean that it and the objects of its action constitute the social system of reference and that these objects are situation, not environment, to the acting collectivity.
The social system and its environments
A social system, like all living systems, is inherently an open system engaged in processes of interchange (or “input-output relations”) with its environment, as well as consisting of interchanges among its internal units. Regarding it as an open system is, from some viewpoints, regarding it as a part of—i.e., a subsystem of—one or more super-ordinate systems. In this sense, it is interdependent with the other parts of the more comprehensive system or systems and, hence, partly dependent on them for essential inputs. Here the dependence of the organism on its physical environment for nutrition and respiration is prototypical. This is the essential basis of the famous concept of function as it applies to social systems, as to all other living systems.
For any system of reference, functional problems are those concerning the conditions of the maintenance and/or development of the interchanges with environing systems, both inputs from them and outputs to them. Functional significance may be determined by the simple criterion of the dysfunctional consequences of failure, deficit, or excess of an input to a receiving system, as asphyxiation is the consequence of failure in oxygen input, and so oxygen input is judged to be functionally significant for the organism. Function is the only basis on which a theoretically systematic ordering of the structure of living systems is possible. In this context functional references certainly need beg no question about how structural arrangements have come about, since the biological concepts of variation, selection, and adaptation have long since provided a framework for analyzing the widest variety of change processes.
Goal-attaining processes explicitly intended to fulfill functional requirements constitute a limiting, but very important, case. Outputs in this sense have primary functional significance only for the system which receives them and which is situa-tional or environmental to the system of reference, although they have secondary functional significance to the latter. For example, although economic output (”produced” goods) goes to “consumers,” the maintenance of certain levels of salable output clearly has great significance to producing organizations. It is its inputs that have primary functional significance for any given system of reference. The “factors of production” of economic theory are classic examples, being the critical inputs of the economy.
In a crucial sense, the relation between any action system—including the social—and any of its environments is dual. On the one hand, the particular environment constitutes a set of objects which are “exterior” to the system in the Cartesian-Durkheimian sense. On the other hand, through interpenetration, the environmental system is partially and selectively included in the action system of reference. Internalization of cultural and social objects in the personality of the individual is certainly the prototypical case of interpenetration, but the principle it involves should be generalized to all the relations between action systems and their environments.
Thus, neither the individual personality nor the social system has any direct relation to the physical environment; their relations with the latter are mediated entirely through the organism, which is action’s primary link with the physical world. This, after all, is now a commonplace of modern perceptual and epistemological theory (Ayer 1956, pp. 130-133). In essentially the same sense, neither personalities nor social systems have direct contact with the ultimate objects of reference, with the “ultimate reality” which poses “problems of meaning” in the sense sociologists associate above all with the work of Max Weber. The objects that personalities and social systems know and otherwise directly experience are in our terminology cultural objects, which are human artifacts in much the same sense as are the objects of empirical cognition. Hence, the relations of personalities and social systems with ultimate “nonempirical reality” are in a basic sense mediated through the cultural system.
Emphasis on their lack of direct contact with what is “out there” concerns in both cases certain qualities of the environing systems as objects. There is, however, important contact with the physical and supernatural environments through the interpenetration of the latter into action systems. Hence, such concepts as knowledge are not naive illusions but modes of the organization of the relations between the various action systems and their environments (Whitehead  and Mead  based their analysis of action on philosophical positions similar to that assumed here). We must regard the relations between the subsystems of action, and between the action system and the systems of nonaction, as pluralistic. That is, there will be no one-to-one correspondence between any two interdependent and interpenetrating systems, but there will be a complex relation which can perhaps be understood by theoretical analysis. This is true of “heredity and environment,” “culture and personality,” and the “ideal” and “real” factors in social systems.
It is necessary to consider the various environments of a living system, because each such environment is engaged in one of the interchange relations with the system, and the specialized natures of these relations serve as the primary bases of the internal differentiation of the system. For instance, the nutrition and elimination systems, the respiratory system, and the locomotor system of an organism are differentiated from each other on this basis. This, as noted, is the essential meaning of the controversial (in social, not biological, science) concept of function. The basis of differentiation is functional, since it consists in the differing input-output relations of the system with its various environments and, following from that, the internal relations between the differing parts of the system itself.
Society and societal community
On the understanding that all social systems are systems of interaction, the best reference point among their many types, for general theoretical purposes, is the society. The definition of this concept presents considerable difficulties, the history of which cannot occupy us here. For present purposes, I shall define society as the category of social system embodying, at the requisite levels of evolutionary development and of control over the conditions of environmental relations, the greatest self-sufficiency of any type of social system. [See Society.]
By self-sufficiency (a criterion which has figured prominently in Western thought on the subject since Aristotle at least), I mean the capacity of the system, gained through both its internal organization and resources and its access to inputs from its environments, to function autonomously in implementing its normative culture, particularly its values but also its norms and collective goals. Self-sufficiency is clearly a degree of generalized adaptive capacity in the sense of biological theory.
The term “environment” is pluralized here to emphasize the fact that the relevant environment is not just physical, as in most formulations of general biological theory, but also includes the three basic subsystems of action other than the social, which have been outlined above.
The core structure of a society I will call the societal community. More specifically, at different levels of evolution, it is called tribe, or “the people,” or, for classical Greece, polis, or, for the modern world, nation. It is the collective structure in which members are united or, in some sense, associated. Its most important property is the kind and level of solidarity—in Durkheim’s sense—which characterizes the relations between its members.
The solidarity of a community is essentially the degree to which (and the ways in which) its collective interest can be expected to prevail over the unit interests of its members wherever the two conflict. It may involve mutual respect among the units for the rights of membership status, conformity with the value and norms institutionalized in the collectivity, or positive contribution to the attainment of collective goals. The character of solidarity varies with the level of differentiation in the society, differentiation which is evident in the structures of the roles in which a given individual is involved, of the system’s subcollectivities, and of its norms and specified value orientations. The best-known basis for classifying the types of solidarity is Durkheim’s two categories, mechanical and organic (see Parsons 1960a).
Both types of solidarity are characterized by common values and institutionalized norms. In the case of mechanical solidarity, however, the patterns of action expected from units are also uniform for all units in the system: relative to one another, the units are segments, since they are not functionally differentiated. Durkheim analyzed crime as the prototypical violation of the obligations of mechanical solidarity. For full members of the community, no matter how highly differentiated the society, the treatment of the criminal should ideally be always the same, regardless of who commits the crime, even though this ideal is frequently and seriously deviated from. At the societal community level in differentiated societies, the core of the system of mechanical solidarity lies in the patterns of citizenship, in T. H. Marshall’s sense (1949). These patterns can be conveniently subdivided into the components of civil-legal citizenship, political citizenship, and social citizenship. In modern American society, the bill of rights and associated constitutional structures, such as the fourteenth amendment, comprise the most directly relevant institutions in this field.
Organic solidarity concerns those aspects of the societal system in which roles, subcollectivities, and norms are differentiated on a functional basis. Here, though common value patterns remain of the first importance to the various subsystems at the relevant levels of specification, expectations of behavior differ according to role and subcollectivity. Solidarity, then, involves the integration of these differing expectations with respect to the various bases of compatible functioning, from mutual noninterference to positive mutual reinforcement. [See Integration.]
Organic solidarity seems to be particularly important in three primary structural contexts. Most familiar is the one Durkheim himself particularly stressed, the economic division of labor, where the most important institutional patterns are contract and property. Second is what we ordinarily call the area of political differentiation, that of both the organization of authority and leadership and the various modes of participation in collective decision making, which involve the interplay of information and influence bearing on collective action. The third is the area of the society’s relations with its cultural involvements. This particularly concerns the society’s articulation with the religious system, but also (and the more so, the more differentiated both the society and the culture) with the arts, the system of intellectual disciplines, and the relationship between the patterns of moral obligation and those of law.
Organic solidarity and pluralism . In all three contexts, organic solidarity is associated with the phenomenon generally called pluralism. In none of these cases is the structure of a subsystem articulating with the societal community ascribed to the structure of the latter. On the contrary, as a function of the level of differentiation among the articulating subsystems, there is an increasing flexibility that facilitates the concrete relations coming to be established by relatively specific processes.
Thus, there is, first, a pluralism of economic interests which, if uncontrolled, would tend to destroy the solidarity of the societal community—indeed, it may be suggested that an exaggerated anxiety about this underlies much of the modern socialist dogma that only the central societal collectivity, the state, can be trusted with any interest which seems important to the public welfare. However, there is a second pluralism of “interest groups” in the political context which, though of course linked with the economic pluralism, is by no means the same. The political process, as that leading to collective decision making, is in part a “political struggle” among such interest groups. Thus, it has great potential for disrupting societal solidarity. However, the latter can also not merely contain the struggle but, even more positively, further integrate the disparate groups by virtue of various mechanisms of integrative control. Finally, the more differentiated societal community tends also to be culturally pluralistic. This is particularly conspicuous in the few Western societies which have attained a certain level of religious pluralism. Thus, at the very least, contemporary American society is a multidenominational, Judaeo-Christian society which also includes secular humanists who prefer not to affiliate with any explicitly religious association. In one sense, it has “transcended” the historic bases of religious conflict which prevailed in the Western world for centuries. The basis of this is genuine denominational pluralism, not only before the law but also in terms of acceptance in the community.
Very closely associated with this is the pluralism among the intellectual disciplines which has gained institutionalization in modern society, especially in the university system (Parsons 1965). The rise of the sciences was, in the first instance, a profound symptom of this pluralization. But it has now become a major factor in the future development of modern society in a variety of ways. The problem of “ethical” pluralism is analytically more difficult and complex. The trend seems to be away from the special kind of moral uniformity which characterizes societies in which mechanical solidarity predominates. The essential point concerns the level of generality at which common moral standards are defined: if a pluralistic society is to integrate its many various kinds of units into a solidary societal community, what counts as moral obligation cannot be defined in terms specific to each kind of unit but must be sufficiently general to apply to the considerable range of differentiated classes of units. MoraHsm ties morality to the specifics of a subgroup or a particular stage of social development and must be distinguished from concern with maintaining control of action in accord with more generalized moral standards.
Cultural system and political system
The societal community in the present sense is articulated most directly with the cultural and political subsystems of the society. Furthermore, it is in these two relationships that the main connections between organic and mechanical solidarity are lodged.
The cultural (or pattern-maintenance) system centers on the institutionalization of cultural value patterns, which, at the general cultural level, may be regarded as moral. Institutionalized societal values, and their specifications to societal subsystems, comprise only part of the relevance of moral values to action; moral values are also involved, through internalization, in structures of the personality and behavioral organism; and, more generally, they articulate with religion, science, and the arts within the cultural system.
Community in the present sense is never a simple matter of the “acting out” of value commitments. It also involves differentiated acceptance, in valuational terms, of the conditions necessary for the functioning of societies and their subsystems. Essentially this latter element draws the line between utopianism—making an imperative of “pure” value actualization—and realistic social idealism. Avoiding the Utopian dilemma involves organizing the value system so as to include the positive valuation of social relationships for their own sake, not only as being rigidly instrumental to specific value patterns.
But this is not the whole story. In addition to a general “set” establishing a presumption of legitimacy for the social system as such, there must also be a more flexible set of mechanisms providing for adaptation between the cultural subsystem of the society and the societal community itself. These mechanisms concern the capacity for handling the changing needs and exigencies of various associational relationships in the light of both their developing interrelations and their relations with the value system; the more particularized commitments must be a function of changing conceptions of the imperatives of relationship, as defining the nature of “valued association.” The commitment to the societal community is, so far as this interchange develops flexibility, no longer ascrip-tive but dependent on the need for such commitment and on an evaluation of its compatibility with deeper moral commitments at the cultural level. One aspect of this flexibility is the individual’s enhanced moral independence from imperatives of unquestioning obligations to conformity. But the obverse aspect is the “right” of the community to expect appropriate flexibility in the adaptation of moral demands to exigencies of realistic implementation.
The minimum imperatives of specified common value commitment define one pole of the structures of the societal system organized with mechanical solidarity. There is a place for organic solidarity in this context so far as such commitments are so firm as not to be “negotiable” and so general as to permit the kind of flexibility in adapting to particular “exigencies” which has just been discussed. What I above called moralism is the limiting case where lack of generality (and perhaps firmness of commitment) forecloses such flexibility. The basic rights of members in the societal community constitute, in negative definition, the limits of application of these value commitments. Members’ complementary obligations to the societal community constitute the obverse expectations of contribution to the functioning of a social system to which they are committed.
In a sense, the “payoff” on such obligations comes in the relation between the societal community and the political subsystem, since the latter is concerned with collective goal attainment as a function of the total society and, pari passu, of each subsystem grounded in communal solidarity. This relation between the societal community and the political subsystem concerns a further step toward mastering exigencies in the interest of the implementation of values. It is a matter not just of establishing particular relationships of solidarity as the “setting” for value implementation but, further, of committing the interests of that community to particular collective goals—which involves dealing with the exigencies of particular environmental conditions. For the individual, then, this concerns not merely his personal commitment to the goal but his obligations as a member of the community. Committing the community implies a solution to the problem of integrating the community with reference to the “policy” in question, whether this involves developing a broad consensus or ruthlessly suppressing minority, or even majority, views. As a somewhat extreme case, entering a war commits the national community, whatever various membership elements think about it, short of their mustering a resistance which would favor the enemy cause.
Here, as in the relation of the national community to the “cultural” subsystem, two importantly different levels are involved. One concerns the general “authority” of differentiated elements in the society to commit or bind the collectivity as a whole in the pursuit of particular goals in particular situations. One extreme in this context would be an absolutist or despotic “government” which presumed to act as it pleased, regardless of consent or opposition in the broader societal community. An opposite extreme would be a community which made any collective action dependent on virtually unanimous and explicit consent.
By differentiating the two levels, modern governmental systems avoid being caught in the above dilemma. They set up procedural rules denning the level of support needed to authorize collective action binding the collectivity as a whole, including minorities that dissent in various contexts. For this to work, the minorities must be committed to the legitimacy of the governmental system, even though they refuse to support particular policy decisions of the moment.
For the individual (or political minority groups), however, such situations may present a moral dilemma. In his role as a responsible member of the societal community, which includes an obligation to support its government (not particular decisions or parties) the member of a minority subgroup is, up to a point, obligated not only to accept but often also to cooperate actively in implementing a policy of which he disapproves. There may, however, be a point beyond which his conscience will not allow this. He will then be driven into various levels of resistance, ranging from withdrawal of active participation, through public protest, conspicuous noncooperation, and militant attempts to prevent or sabotage its implementation, to revolution.
The development of political differentiation and pluralism, including the generalization of the crucial levels of political obligation, tends to broaden the range of individual freedom for dissent and also to draw the lines between politically institutionalized—as distinguished from moral—rights of dissent and opposition and those institutionally defined as illegitimate. The basic independence of the cultural-moral and the socially institutionalized systems, however, precludes any social community from being completely immune to the kind of political opposition which can lead to the disruption of its basic solidarity.
The element of mechanical solidarity here concerns the legitimation of collective decision-making authorities. Such legitimation must derive from common value commitments to the societal community and, hence, to the kinds of collective action considered legitimate, including the identification of the agencies entitled to take such action. Obversely, this also concerns the rights of membership elements to give or withhold support for particular policies and, more generally, particular claims to leadership status. The appeals for such support, however, must be on grounds of organic rather than mechanical solidarity. The procedural rules become the focus of common commitments, while particular outcomes become matters for legitimate contest.
Solidarity and the economy
At this point, we may recall that Durkheim introduced the concept of organic solidarity in analyzing the division of labor in the economic sense. This was quite logical in the light both of the utilitarian theories to which he was critically orienting himself and of the economy’s relative remoteness from the setting of the system of mechanical solidarity as that which was just discussed. Focus on the economic system was the most convenient way to set up a clear conceptual dichotomy.
Nevertheless, it now seems better to approach the problem of the economic system indirectly, through its relations to the other aspects of a social system. We conceive of the economy as the functional subsystem of a society differentiated about producing and allocating fluidly disposable resources within the society. As put in a quite familiar paradigm, it operates through combining the factors of production—e.g., land, labor, capital, and organization—to produce the two primary categories of output: commodities and services. The economic categories are not the physical objects or the physical behavior involved as such, but certain ways of controlling them: in the case of commodities, essentially property rights; in the case of services, the kind of authority or power over the performer we associate with the status of employer.
The actual combinatorial processes, which we call economic production, take place in goal-oriented organizational units that economists call firms. The strictly economic functions concern the management of the boundary relations of these units through what is ordinarily called the market system, and should be distinguished from the technological functions. The economic functions involve procuring control of the factors of production (including determination of requirements for them) and disposing of the outputs of production through marketing. These processes operate by adjusting relations between supply and demand through establishing terms for the transfer of control that equate quantity and price for both parties to the exchange.
Here the primary institutional focus of organic solidarity is the institution of contract, which is essentially the set of procedural rules regulating transfers of both factors of production and economic outputs. This institutional complex not only regulates the actual settlement of contracts but also defines what types of contract may—and may not—be entered into, how agreements may be arrived at, their bearing on the interests of third parties, and the obligations of parties under various special contingencies, such as the development of unforeseen obstacles to the fulfillment of terms.
The institution of property, then, is the normative system regulating acquisition, disposal, control, and use of physical objects in relation to the contractual system, whether the objects be factors of production or commodity outputs. And the institutional complex we call employment regulates the acquisition and utilization of human services, either as factors of production or as ultimate agents of valued consumption.
Generalized media of interchange
In sufficiently developed and differentiated systems, a central role in economic process is played by money, as both a symbolic medium of exchange and a measure and store of value in the economic sense. Money may be defined as the capacity of a societal unit to command economically valuable resources through the exchange process, i.e., through contractual agreements, without giving commodities or services in return. The payment of money constitutes the transfer of such capacity from one unit to another. In most transactions in a developed economy, entities that have “value in use” figure on only one side of an exchange relationship, being balanced by a monetary “consideration” on the other. To “pay” money is to accept certain economic obligations, defined by ’a proportionate diminution in one’s capacity to command economic “values” in other transactions. To accept money in payment, on the other hand, is to gain the right to an expectation that others will make economically valuable goods and services available at the times and places of one’s own choosing, within the limits defined in the market nexus. It has long been a commonplace of economics that only a far-reaching institutionalization of the monetary mechanism can make an extended division of labor possible (see, for example, Adam Smith’s classic statement in 1776, book 1, especially chapter 3), though it is known that politically controlled administrative allocation of resources can substitute for the contractual-monetary mechanisms up to a point, as in the “command” economy of the Soviet Union, which reached its highest development in the late Stalinist period (see Grossman 1962). Nevertheless, the extent of an economy’s “monetization” is undoubtedly the most important single index of the mobilizability of its resources and, hence, the flexibility of their allocation, at all combinatorial stages, from ultimate natural resources and human energies or skills, to finished consumption goods and services.
Money is also important theoretically as the best-understood member of the family of generalized symbolic media of interchange involved in social interaction processes. Political power and influence as used in political leadership processes certainly belong to this family (Parsons 1963a; 1963b).
The economy, as here conceived, articulates with the societal community primarily through the institutional complex of contract, property, and the employment-occupation system. Its solidarity is maintained by keeping its transactions in line with certain integrative imperatives, e.g., by protecting the interests both of parties to contractual relations and of third parties and by providing a basis in solidary relations for effective collective action, especially through making economic resources available to collective units, including particularly the government.
Money, like the other members of the family of media, is a symbolic medium which, without being too farfetched, we may call a specialized language. Like all such media, it expresses and communicates messages having meanings with reference to a code—that is, a set of rules for the use, transformation, and combination of symbols. (The theory of the operation of such types of messages and codes of rules has been developed in the field of linguistics by, particularly, Jakobson & Halle 1956; Chomsky 1957; 1964). In the case of money, as institutionalized, it is highly important to recognize that the relevant code is part of the legal system; this is most clear in societies having a sufficiently high level of differentiation. As we have put it, the institutions of contract, property, and employment, as parts of the legal system, constitute the code in terms of which transformations between money and commodities or services and among different forms of monetary assets operate. Financial transactions, therefore, constitute a certain type of “conversations.”
This paradigm is also applicable to relations between the societal community and the other primary functional subsystems of the society. In the case of the polity, the medium which corresponds to money is power. This I conceive as the generalized medium of mobilizing capacities for effective collective action, utilizable by members of collectivities to contribute toward binding the collectivity to particular courses of action, either determining or contributing to the implementation of specific policy goals. (This usage of the concept of political power is clearly different from those most common in both sociology and political science; for a discussion of the issues involved in the usage of this concept, see Parsons 1963b.) The code within which power as a medium operates centers about the institution of authority, which in turn articulates with the patterns of institutionalized leadership and administrative responsibility for maintaining regulatory norms.
In the sphere of articulation with the cultural system, the operative medium is what I call commitments. This concerns the specification of the general value patterns to the levels necessary for their workable combination with the other factors requisite to their implementation in concrete action. Commitment to valued associations of the societal community type is the prototype here (unfortunately, I have not yet been able to develop for publication an analysis of the commitments medium on the same level as I have done for money, power, and influence). The relevant code is the set of institutions which constitutes the underpinning of society’s mechanical solidarity—in American society those formulated in the bill of rights, etc., as noted. Within this context, the civil component holds precedence, because it formulates the valua-tional basis of community membership.
Finally, the societal community itself is the focus of operation for a fourth generalized medium, which I have called, in a special technical sense, influence (Parsons 1963a). Here the relevant code is comprised of the norms underlying organic solidarity, as they relate to the pluralistic structure of differentiated societies. Since their primary context is that of the solidarity of the society, we may consider their major focus to be justification for the allocation of loyalties. Here justification must be carefully distinguished from legitimation. Justification is less absolute and operates at a lower level in the cybernetic hierarchy. The system may well be legitimated while questions of the justification of certain choices between alternative subsidiary solidarities are still left open where actual or potential dilemmas are posed.
These different code components are more or less adequately integrated in a going societal system, where they constitute its basic normative structure. They should be distinguished from the primary normative components of a pattern-maintenance system, since the latter are made up of value patterns and their specifications, not of differentiated norms. The integratively oriented code of the societal system must be anchored in a value system if it is to have a basis of legitimation. But its structure is determined not only by value specification but also by adjustment to the exigencies of the other functional subsystems. But in this process of adjustment the integratively oriented code still maintains a certain level of integrity with respect to the value commitment and solidarity of the societal community. In highly differentiated societies, this basic code system is the core of the legal system.
Societies and their environments
We may now return to the problem of the relations of a society as a social system with its environments. The basis of the differentiation between the societal community and the other three primary subsystems of the society should be sought in the basis on which they in turn are differentiated from it and from each other. In general, it can be said that the reason for the existence of these patterned differentiations is that they help the social system to cope with the exigencies imposed on it by its environments.
The organic-physical environment . In dealing with this problem, perhaps we had best begin with the economy, partly because the relevant theoretical analysis is most highly developed there. In the terms of our general paradigm, the intrasocial relation between societal community and economy is paralleled at the level of the general action system by the relation between the social system and the behavioral organism.
First, it should be emphasized that all relations between the social system and the physical environment are mediated through the behavioral organism. The perceptual processes of the organism are the source of information about the physical environment, which gains cultural organization from its conceptual and theoretical components. The organism is also the source of the “instinctual” components of the motivation of individuals’ personalities.
The relation between the organism and the society’s economic subsystem, which is of direct concern here, constitutes the technological system. This involves the utilization of empirical knowledge, structured by perceptual feedback through the cultural system, for the design and production of commodities having utility for human social functions. What is to be produced, in what quantities relative to alternative uses of the factors of production (cost factors), is economically determined; how it is to be produced is a technological problem. Technology involves not only the use of ultimate “natural” resources (analytically a “land” factor) and “equipment” (a benefit from previous production) but also labor—a factor that, sociologically speaking, takes the form of service. This is a particularly important category of the inter-penetration of the economy with other parts of the societal system. We conceive of service as an output from the economy which “corresponds” to labor as a factor of production but which should definitely not be identified with labor. Very importantly, however, service is a crucial factor in technological efficiency. This apparently paradoxical conception derives from the fact that technological processes always occur within a framework of social organization, never as “purely physical” phenomena. This means that the physical, behavioral operations of persons in technological settings are a function of their commitments, as members of the societal community and its relevant subsystems, to devote their energy and skill to productive uses in the economic sense. This human component is then combined, at the general action system level, with empirical knowledge of standards of socioeconomic utility to produce facilities which can be relatively freely allocated to the various functional needs of societal units. Analyses in these terms can contribute much toward resolving the old controversy about whether the material basis of a complex societal system is “ultimately” economic or technological, or whether the distinction between these categories should be abandoned.
Physical location is a particularly important involvement of technological systems, deriving from the necessity to bring together physical materials, plant, equipment, and organisms as performers of service. Role differentiation between the occupational and residential units tends to involve physical separation of workplaces from places of residence, although the involvement of the same persons in both units sets certain requirements for the physical interrelations of the units’ locations. In particular, the modern urban community is very largely built about the relationship between these two sets of locations.
Residence, like occupation, also articulates physical location and the organism into the social system. But it operates in the context of the organic rhythm, such as sleep, nutrition, and sexual activity, to which human beings are bound. Another limiting factor is that the household (which, in spite of many exceptions, remains the usual unit of residence), has at its core kinship units centering on one or more nuclear families. The place of residence is the human individual’s residual location, the place where he is likely to be, and is often normatively expected to be, when he is not engaged in such other specific activities as work and special recreation.
Communication and transportation—of both goods and persons—therefore require physical media and must be involved in the physical world, perhaps especially in its spatial aspects. The actual communication of a message from a sender in one physical location to a receiver in another is always problematic, even if the two are engaged in face-to-face conversation in the same room. The same is true of broadcast communication—newspapers must get from the printing plant to the readers, radio and television broadcasts must be transmitted through the “air”—and of the conveyance of persons and goods from place to place.
In certain senses, though, the most fundamental problem here is that the normative orders constitutive of social systems must “apply” to categories of persons and their acts in ways that include specifications of where the persons or acts are located. Very generally, then, the societal community and various of its subsystems “claim jurisdiction” over persons and their acts with reference to particular territorial areas. A most important reason for the prominence of territoriality is that normative obligations, if taken sufficiently seriously, must on occasion be somehow enforced, and this involves resort at some point to physical negative sanctions, which can only be applied to the noncompliant individual where he is. This, in turn, obviously includes enforcing claims to the jurisdiction over, and the utilization of resources within, an area and, hence, a readiness to enforce respect for such control upon outsiders, i.e., the function of defense (Parsons 1960b).
Thus, spatial location is involved in all the functions of social systems. Its articulation with social processes is what we ordinarily call the ecological aspect of the system—the distribution of its various activities in physical space and their orientation to spatial considerations. In principle, all other analytically distinguishable aspects of physical systems are comparably involved with social interaction, but the foregoing will have to serve for illustration.
The core of the social system, the societal community, relates to the physical environment primarily through two mediating systems: the economy, which is primarily social but which interpenetrates with the technological system, and the technological system, which is primarily organic-physical but which interpenetrates with the economy. Organic-physical factors, then, operate in all the other primary subsystems of the society, each of which has its technological and economic aspects, although they are subordinated to other considerations, such as the political.
The cultural environment . There is parallel complexity at the other end of the cybernetic hierarchy, in which action and, hence, social systems are involved. A society, or any other type of social system, has a pattern-maintenance subsystem, the units of which (once the system is sufficiently differentiated) have cultural primacy. These social system units, then, interpenetrate with both the societal community (and other societal subsystems) and with the cultural system proper. With progressing differentiation, they tend to become distinctively different according to whether their primary concern is cultural or social.
Religion comprises the matrix from which cultural institutions in general have become differentiated and remains the “master system” in the cybernetic sense. But secular intellectual disciplines (science), arts in the expressive-symbolic sense, and normative disciplines (e.g., ethics and law) have gained differentiation from it.
This formulates very briefly the main line of internal differentiation of a cultural system. The pattern-maintenance system, however, is not a cultural system in a strict sense (though for simplification the distinction has not always been made in this article), but the subsystem of the social system articulating most closely with the cultural system. Religion as a cultural phenomenon is not part of the pattern-maintenance system. Rather, the relevant structure is the collective organization of religious orientations, e.g., in churches or in prophetic movements. Science as a body of knowledge is cultural; universities as collectivities organized about the development of science through research and about its communication through teaching are parts of the society. Pattern-maintenance structures in this connection have cultural primacy only in that their societal functions concern interchange with the cultural system and in that they interpenetrate with the latter. Thus, religious orientations or scientific “systems of knowledge” are constitutive parts of churches and universities, not only “environments” to them.
Just as man has no direct contact with the physical world independent of the organisms (which, however, is itself part of that world), so he has no direct contact with the ultimate nonempirical “grounds” of his existence, what Weber called the world of “ultimate realities.” His objects in this realm to which he orients himself are not the ultimate entities as such but his representations of them. They are cultural objects—parts of the cultural system in the action sense—and hence interpenetrate with all the other subsystems of action.
As structures of such interpenetration, “theological schools” or “prophetic movements,” though quite distinct from religion as a component of the culture, are cultural subsystems of the society that have religious primacy but also interpenetrate with churches or other forms of the social institutional-ization of religion. In the same way, law schools, as companies of legal scholars, are cultural subsystems, whereas courts of law are the social-system units in which legal doctrines are applied to social situations. In the more strictly cognitive disciplines, “companies of scholars” constitute cultural subsystems, which often involve “schools” at the level of cultural content, whereas universities and other educational collectivities constitute the articulated social system units.
For certain purposes, we may, as above, legitimately equate the pattern-maintenance subsystem of a society with the cultural system, since its primary function is articulating the social system as such—the system constituted by social interaction —with cultural patterns and norms. This, however, is elliptical. In the first instance, there are the more complex relationships just sketched. But there is also a further complication. Any system of cultural content, particularly a value system, must be specified from the most general relevant levels to levels relevant to the highly particular functions and exigencies of many and various subsystems. For instance, every technological system producing a particular commodity has special exigencies that the general principles of the relevant science cannot handle alone; similarly, every medical case is in some sense unique, and the physician must tailor his general medical knowledge to its specificities.
One set of exigencies of human societies has a special bearing here. It concerns the consequences of the fact that culture is learned by the human being; it is not part of his hereditary equipment. If a given society is defined by its institu-tionalization of certain cultural patterns, then the necessity of internalizing those patterns in the oncoming generation is second in functional importance only to maintenance of the adult levels of that culture. This cultural imperative evidently underlies the functioning of kinship institutions in all known human societies and, at higher levels of differentiation, of many kinds and levels of formal education.
This whole subsystem of institutions, as well as those involved with cultural innovation (e.g., research organizations), should be included in the pattern-maintenance subsystem of a society, characterized by primary interpenetration with the cultural system of action. Kinship, however, having special reference to child care, is the substructure of the pattern-maintenance system that operates at the farthest remove from the considerations of the general culture; at the appropriate level of specification of values, however, it has cultural primacy. Furthermore, it also relates quite specially not only to the society but also to the exigencies of both organism and personality, about which a few words must now be said.
The psychological environment . The personality, as analytically distinguished from the organism, constitutes the third primary environment of a social system. It interpenetrates with the individual organism in the obvious and fundamental sense that the storage facilities of learned content must be organic, as must the physical mechanisms of perception and cognition, of the control of learned behavior, and of the bases of motivation.
At the level of this discussion, however, the personality forms a distinct system articulated with social systems through their political subsystems, not simply in the sense of government but of any collective ordering. This is to say that the primary goal output of social systems is to the personalities of their members. Although they interpenetrate crucially with social systems, the personalities of individuals are not core constituents of social systems (nor vice versa) but precisely environments of them. Freud, especially in his later work, was quite clear about the obverse relationship: namely, that the individual personality’s primary environment consists of the social systems into which it becomes integrated. Freud’s famous “reality principle” is the principle of ego adaptation to the social environment.
I am treating the personality last among the primary environments of the social system because, of the three, it is the least commonly conceptualized as such. This conceptualization directly counters the long tradition that a society is “composed” or “made up” of “individuals.” The latter may be true if the society and the individual are conceived of as concrete entities. Here, however, social system and personality—the concrete term “individual” is avoided in this context—are used as abstractly defined systems which are distinguished analytically, though allowance is made for the crucial relation of interpenetration. The unit of interpene-tration between a personality and a social system is not the individual but a role or complex of roles. The same personality may participate in several social systems in different roles.
From the viewpoint of the psychology of the personality, the positive outputs from the social system are rewards. Indeed, I would even say that, at the level of cultural symbolization, except for intermediate cases specially involved at the crux of differentiation between organism and personality (notably, erotic pleasure), all rewards are social system outputs. Conversely, outputs from the personality to the social system are personal goal achievements which, from the viewpoint of the receiving social system, are contributions to its functioning, insofar as the two systems are integrated with each other.
The focus of such integration is the phenomenon of “identification,” through which the personality acquires a motivationally and cognitively meaningful role set and the social system acquires a member who can make meaningful contributions. Malintegration means that this matching relationship has failed in one way or another—“deviance,” “alienation,” and a variety of other phenomena fit in this category. It is also crucial to allow for personal creativity in relation to the social system. The analytical independence of social system and personality is the basic origin of both the prevalence of deviant behavior and the openings for creativity. The frequent allegation that sociology teaches the necessity of flat “conformity” is a conspicuous case of the fallacy of misplaced concreteness. If our analytical generalizations about social systems “applied” without qualification to all the member personalities, this would be the case. The mutual independence of the two categories of system— though accounting for their interdependence and interpenetration—is the theoretical basis for the fundamental and general phenomenon of the autonomy of the individual, so far as the social system is concerned.
Two important considerations reinforce this assertion of the reality of personal autonomy, the degrees and kinds of which must be seen as varying with different types of social system. First, analytically and apart from its direct relation to the social system, the personality system is the primary meeting ground of the cultural system, the behavioral organism, and, secondarily, the physical world. Although there have been serious theoretical difficulties with the “culture and personality” studies of the last generation in behavioral science, they did focus upon a crucial relationship here, as did the “behavioristic” traditions of psychology in studying the interrelations of personality and organism. Hence, it can be said not only that the personality is autonomous as a distinct subsystem of action, but also that this autonomy is importantly grounded in the personality’s interchanges with the cultural and organic levels of the organization of action. These three sets of considerations (plus the uniqueness of the genetic constitution of practically every human organism) go far in explaining the irreducibility of the distinctiveness of all human personalities, as well as their autonomy.
The second consideration derives from an internal feature of social systems that is generally called “role pluralism.” That is, not only do individuals have plural role involvements, but also different individuals’ combinations of role participations vary widely. Such variance includes complexes of differing roles which are often categorized together for limited purposes. Thus, one “middle-class suburban mother” may have one child, another three, and another five, and the assortments of the children by age and sex may vary, so that even “being a mother” is not an identical thing for each member of that category, even sociologically. To this we can add differences in occupation of husbands, religion, ethnicity, participation in community affairs, etc.
When so many mutually independent—though also interdependent—factors are operating, anyone familiar with the logic of combinatorial variability should find it difficult to maintain that a modern, highly differentiated society is incompatible with individuality. Of course, there are also matters of the specific kinds of autonomy and individuality which are at stake. However, the arguments alleging that modern societies are repressive of all autonomy and stifling of all individuality are frequently so overgeneralized that they appear to deny altogether the combinatorial argument just outlined. Furthermore, a strong case can be made that the trend of modern society, because it has become so highly differentiated and pluralistic, is positively to favor individuality rather than to suppress it in favor of conformism.
We have confined our attention here to human-level social systems and have emphasized the importance of the symbolic systems, which we call cultural, that become constitutive of them through being involved in action and interpenetrating with social systems. Perhaps the most general matrix of these symbolic systems is language. On various levels, there is great familiarity with the concept of symbolic systems, e.g., of “ideas” having a predominantly cognitive focus and of “expressive symbols” in the arts and in ritual.
The media of interchange revisited
In conclusion, we may carry a little further the discussion, introduced above, of another category of symbolic systems that emerges into great prominence in highly differentiated social systems: the media of interchange. Attention was called above to money as the medium of exchange in economic transactions. Though the science of economics has gone far in understanding the vastly complex phenomena of monetary systems, they have generally been considered as unique. I have suggested that money is not unique in either of two senses.
First, it can be considered a special case of a very general phenomenon: language. It is in fact a very highly specialized language. Crucial here is the recognition that it operates at the symbolic level and that its primary function is communication, though of a special, normative sort. The “monetary system” is a code, in the grammatical-syntactical sense. The circulation of money is the “sending” of messages which give the recipient capacity to command goods and services through market channels. The recipient gains the expectation that he can “request,” by virtue of his holding money, access to goods and services of a given value. There is an institutionalized obligation on those receiving such requests—if they are “in business”—to comply. But the process of money circulation involves literally nothing except communicated messages. A check is only a filled-in form letter to the bank on which it is drawn.
Second, money is not the only specialized language of this sort operating in social systems. Political power is certainly another. It centers on the use of discretionary authority in collective organizations to make decisions which, as binding on the collectivity, require performances of those who are obligated to further their implementation. Not only executive decisions constitute uses of power in this sense, but also the exercise of franchises in many connections, from voting in governmental elections to voting as a member of a small committee.
A third generalized symbolic medium is influence. By this I mean, quite technically, the capacity to achieve “consensus” with other members of an associated group through persuasion, without having to give fully adequate reasons (an adequate reason, in this sense, would be one that gave the recipient sufficient information for making a rational decision himself, or one that was at least fully understandable to him). Thus, a physician, as a technical specialist, may persuade a patient to follow his advice even when it is out of the question that the patient is competent to understand its technical grounds. The patient must, as members of the profession often put it, have “confidence” in his physician.
Fourth is the medium of generalized commitments to the implementation of cultural values, at the level of the social system as such. It is the most difficult to conceptualize, and the least can be said about it.
The need for generalized media of interchange is a function of the differentiatedness of social structures; in this sense they are all partly integra-tive mechanisms. The relations between markets and money and the division of labor are well known, but similar considerations apply in the other cases.
In the political case, the necessity for the mechanism of power stems from the social “status distance” between the loci of decision making and the loci of the performances necessary for the implementation of the decisions. In complex organizations, it is not realistically possible for decision makers to consult in detail with every person upon whose compliance effective implementation of their decisions depends. This may involve reasons of time and urgency, technical considerations, access to special information, or various exigencies of coordination. Thus, elections must lead to a concentration of power in the hands of the candidates elected. There cannot, however, be a simple consensus between all the members of the electorate and the preferred candidate—this would be incompatible with the voter’s freedom of choice. Hence, the individual voter must agree to make a binding decision that he prefers candidate X over Y. If enough voters do likewise, X will be elected. The electoral authorities are obligated to comply with the aggregate of decisions of the voters.
In the case of influence, the functional need involves bridging certain gaps between the bases of accepting “advice” (in the sense of attempts to persuade without either situational inducements or threats of coercive sanctions) and the intrinsically cogent “reasons” for such acceptance. Complex communities cannot wait for fully rational demonstrations of the advisability of all commitments. Therefore, they must rely on influence or, as we sometimes say, prestige, as utilized by persons in responsible roles. The user of influence creates a presumption for the reasonableness of his case, so that the object of his attempts at persuasion feels, in the integrated case, reasonably sure in trusting him.
Similarly, commitments are given to others when an individual enters into a situation (i.e., makes or, more appropriately, gives a commitment) without in fact being fully able to ensure that the process of action implementation will be carried out in a manner conducive to preserving or enhancing the integrity of his values. Thus, in a sense different from that of the influence context, he has either to trust others or to sacrifice the prospect of successful implementation. In turn, others must trust him to gain fulfillment of their commitments. It is in this sense that commitments may be considered a “circulating” medium.
These media appear in generalized and differentiated form only when relatively high levels of differentiation in the relevant spheres have been attained. Primitive societies never have money and market systems, and many archaic societies have them only rudimentarily, if at all. What Weber called “patriarchal” political structures do not have power as a generalized medium, and “patrimonial” regimes show only its first emergence.
Other generalized media seem to operate in the zones of interpenetration between the social system and the other primary subsystems of action. As already noted, what Freud called erotic pleasure is at the same time both organic (i.e., a component of the personality) and, because of its involvement with interpersonal relations, a component of certain elementary social systems. What psychoanalytic and other social psychologists have called affect is probably another such mechanism, operating among persons in the interchange between the personality and social systems rather than in direct relation to the organism. The two famous “wishes” for recognition and for response discussed by W. I. Thomas perhaps designate still another medium which, however, may be a subdivision of the more general mechanism of affect. In the organic-physical set of relations, technological “know-how” and skill are probably well regarded in this way.
Another set of media operate in the zone of interpenetration between the social and cultural systems. Ideology is a conspicuous example. The concept conscience, as used in Puritan traditions especially, seems to belong in this category. Reputation, as that term is used in discussing the social structure of scientific communities, is probably another case. The concept faith, as used in Christian tradition, especially Protestantism, probably refers to a generalized mechanism peculiar to the cultural level of action organization.
The relative salience of the various generalized media of interchange (and of particular cases within them) for specific structures is a useful guide to the structural arrangements among and within the subsytems of more generalized social systems, notably societies. We have also claimed that the core of a society is the societal community, which, functionally regarded, is the integrative subsystem. It interpenetrates and interchanges directly with each of the other primary subsystems: the pattern-maintenance or cultural-primary subsystems; the goal-attainment subsystem, or polity; and the adaptive subsystem, or economy. The medium focal to the societal community is influence, which is interchangeable for power, money, and value commitments.
Each of the other three subsystems constitutes a zone of primary interpenetration and interchange between the social system and one of its intra-action environments. The economy interchanges with the organic-physical environment; and money, in a sufficiently differentiated economy, can be used in exchange for the factors of production, which are then also technologically combined. Though a modern economy is structured primarily about financial institutions and market systems, these latter interpenetrate, in turn, with the technological organization of production.
The polity interpenetrates, in the first instance, with the personality. Power, as the medium having political primacy, can be used to acquire both human services and the demands for collective action which justify leadership initiative. Underlying these two forms of “mobile” human resources are the processes that generate and stabilize them. Here the interpenetration between social system and personality leads toward both the psychological “depths” of the personality and the relational contexts articulating the basic integration of social systems. Above all, family and kinship, as well as neighborhood and education, fit this context but so do complexes such as recreation. These operate, however, at a level quite different from the direct interchanges between personality and polity. For macrosocial purposes, therefore, they should be treated as pattern-maintenance processes.
Finally, the interpenetration between social and cultural systems concerns, most saliently, the place of religion in relation to social structure. Indeed, the primary structures of the most primitive societies fall almost entirely into the two basic categories of kinship and religion. With further differentiation, however, religion becomes more and more clearly distinguished from political organization. It also tends to become distinguished from economic structuring, while the latter remains ascribed to both kinship and, above all, to the polity (in the broad, analytical sense).
In relatively advanced societies, the cultural system itself begins to differentiate, particularly through the appearance of secular cultural disciplines. Thus, law, in close relation to ethical philosophy; the arts, as something other than direct handmaidens of religion; and, generally last, science have become independent cultural realms— though they are always also interdependent and interpenetrating with each other and with the social system. Value commitments constitute the principal societal medium operating in this realm, though various others are involved secondarily. A modern society, then, contains a considerable number of structural units having cultural primacy. Religious collectivities need hardly be mentioned, so conspicuous are they from any comparative point of view. Increasingly, modern societies have universities, which institutionalize the intellectual disciplines that are in some sense sciences, various organizations focusing upon the arts, and the very crucial institutions of highly generalized law, with their articulations to ethics.
The social system is, thus, a very complex entity. As an organization of human interests, activities, and commitments, it must be viewed as a system and in functional perspective. This is the key to its lines of organization, its modes of differentiation, and its integration. Such a system may be considered as both- structure and process, in different aspects and for different scientific purposes. Structurally, we have suggested that there is a double basis for systematizing differentiation and variation: that internal to the primary social system itself and that involved in its relations to its primary environments, as analyzed with reference to the general system of action. Processually, the categories of analysis must follow from and integrate with those of structure. I suggest that, given the central position of language as definitive of human society, the more differentiated and specialized symbolic media of interchange constitute the master scheme for the systematic analysis of social system processes.
Ayeh, A. J. 1956 The Problem of Knowledge. New York: St. Martins. → A paperback edition was published in 1962 by Penguin.
Chomsky, Noam 1957 Syntactic Structures. The Hague: Mouton.
Chomsky, Noam 1964 Current Issues in Linguistic Theory. The Hague: Mouton.
Grossman, Gregory 1962 The Structure and Organization of the Soviet Economy. Slavic Review 21:203-222.
Jakobson, Roman; and Halle, Morris 1956 Fundamentáis of Language. The Hague: Mouton.
Marshall, T. H. (1949) 1964 Citizenship and Social Class. Pages 65-122 in T. H. Marshall, Ciass, Citizenship, and Social Development: Essays. Garden City, N.Y.: Doubleday. → The essay is based on the Marshall lecture delivered at Cambridge in 1949. The collection of essays was first published in England in 1963 as Sociology at the Crossroads.
Mead, George H. 1938 The Philosophy of the Act. Univ. of Chicago Press. → Published posthumously.
Parsons, Talcott 1960a Durkheim’s Contribution to the Theory of Integration of Social Systems. Pages 118-153 in Kurt Wolff (ediior), Émile Durkheim, 1858-1917: A Collection of Essays With Translations and a Bibliography. Columbus: Ohio State Univ. Press.
Parsons, Talcott 1960a The Principal Structures of Community. Pages 250-279 in Talcott Parsons, Structure and Process in Modern Societies. Glencoe, 111.: Free Press.
Parsons, Talcott 1963a On the Concept of Influence. Public Opinion Quarterly 27:37-62. → A comment by J. S. Coleman appears on pages 63-82; a communication by R. A. Bauer, on pages 83-86; and a rejoinder by Talcott Parsons, on pages 87-92.
Parsons, Talcott 1963b On the Concept of Political Power. American Philosophical Society, Proceedings 107:232-262.
Parsons, Talcott 1965 Unity and Diversity in the Modern Intellectual Disciplines: The Role of the Social Sciences. Dssdalus 94:39-65.
Smith, Adam (1776) 1952 An Inquiry Into the Nature and Causes of the Wealth of Nations. Chicago: Encyclopaedia Britannica. → A two-volume paperback edition was published in 1963 by Irwin.
Whitehead, Alfred N. (1929) 1957 Process and Reality: An Essay in Cosmology. New York: Humanities.
Political systems analysis attempts to delineate the fields of political science and political action, to give them coherence and order, to define their properties and guide research, as well as to integrate relevant findings. It seeks to isolate the arena of politics as an independent system from the remainder of society. In one sense this has been done by students of politics from the very beginning of political thought. The current efforts are distinguished by a more self-conscious approach and by the more refined technical tools that are available.
From the systems perspective, societies and other social groups tend to be seen as relatively persistent entities functioning within larger environments. These entities qualify as systems because they are considered sets of interdependent elements or variables, which can be identified and measured. Systems have distinguishable boundaries setting them off from their environments, and each has a tendency toward a state of equilibrium, i.e., the system tends to maintain itself through various processes whenever it is disturbed, either from within or without its boundaries. Each system tends to be structured in accordance with certain invariant problems characteristic of all social systems. Internal differentiation takes place, with specific structures and processes being developed to handle specific kinds of problems, and, as a result, various subsystems will evolve, such as the economic system, the political system, social stratification, etc. In the case of political systems, the major task and function, or contribution to society, is that of selecting societal goals, mobilizing resources for their attainment, and making societal decisions.
Precursors and sources . The concept of systems has come to political science only recently. In adopting it, political scientists have drawn upon the work of other social scientists, as well as biologists, physical scientists, and engineers. The physiologist Cannon, in The Wisdom of the Body (1932), influenced social scientists and particularly gave currency to the use of the term “homeostasis” to describe a crucial property of biological systems and a highly suggestive one in the study of social systems. The writings of Bertalanffy on biology and general systems analysis (1949; 1950) have been of considerable importance. Few political scientists have turned to these writings directly, but there have been important linkages through the more widely known work of some sociologists, such as Parsons (1951; 1958), Romans (1950), and Roethlisberger and Dickson (1939). Parsons, especially, has developed the technical idea of a social system, and it has been through his work that several political scientists have come to employ the approach in the study of politics. Almond (1956; 1966), Easton (1957; 1965a; 1965b), and Mitchell (1962) have based much of their conception of the political system on Parsons’ own expositions in the realm of politics (1960).
Although the idea of a system of economic behavior or action has been prominent in economics almost from the very beginning of that science (in the work of Quesnay and the physiocrats, later of the neoclassicists, as well as that of mathematical economists such as Walras, Pareto, and Cournot, in which the notion of system was honed to the precision and elegance of the calculus), economists have not exerted the most direct influence on political science. For one thing, political scientists have usually lacked the necessary mathematical sophistication, and for another, few political scientists have been exposed to homologous social processes or action. The conception that political behavior could be specified in terms of a sharply delimited number of quantitatively linear variables has found only gradual acceptance in political science. However, since the 1950s the application of economic conceptions and models has been rigorously attempted (Arrow 1951; Downs 1957; Black 1958; Buchanan & Tullock 1962; Riker 1962).
Finally, one must note the contribution of certain students of organizations, especially Chester I. Barnard (1938) and Herbert A. Simon (1947). The historical closeness of public administration to political science made these men’s work familiar to many political scientists and thereby prepared the ground for viewing politics in terms of organizations or, more broadly, social systems: inputs, outputs, equilibrium, resources, support, etc. From all these sources have come elements of the current developments in systems analysis as it applies to politics.
The political system
Structure of the system . The most traditional and conventional problem of political science has been that of describing and accounting for the internal structure of political systems. The term “structure” is generally applied to those patterns of power and authority which characterize relationships between rulers and ruled—relationships which are more or less enduring and therefore more or less predictable.
The unit of analysis for these power relationships is usually “role,” a concept developed primarily in social psychology and widely used in sociology. While the concept has a variety of meanings, basically it refers to those norms which prescribe and proscribe behavior in specific settings, relationships, and functions. Power and authority roles pertain to the acquisition, maintenance, and employment of power and thereby constitute the building blocks from which the polity is constructed. Political roles are concerned with the making of decisions in the name of society and the performance of actions which achieve or implement these decisions and allocate scarce values and costs. The set of these roles and the behavior which stems from them make up the political system.
In analyzing the internal structure of a polity, the political scientist describes these roles and the persons taking them as they engage in interaction. The content of the roles, their origins, changes, influences on behavior, etc., have been questions of interest for political scientists. Likewise, political scientists have analyzed the outcomes of various types of structures, such as the allocation of values and costs, the effective achievement of collective goals, and the maintenance of peace and order.
In order to characterize structures and distinguish them, political scientists have employed a variety of concepts and tools. Traditionally, the chief basis of classification has been the distribution of power among the members of the system.
This single base line has not proved adequate for descriptions of political systems, and even valid and reliable statements about the distribution of power are an insufficient basis upon which to compare political systems. Many crucial similarities are missed when analysis proceeds on a single dimension. Systems analysts have therefore devised more inclusive sets of variables and at the same time insisted that they be measurable. Parsons has advanced a set of concepts called the “pattern-variables” (1951), to make possible more complete analysis of social and, accordingly, political structures. Similarly, Almond (1956; 1966) has suggested a classification scheme of structures that is based on the following basic dimensions of the political system: (1) the degree of differentiation; (2) the extent to which the system is “manifest,” or “visible”; (3) the stability of the functions of various roles; and (4) the distribution of power. A fifth possibility concerns the “substitutability of roles.”
The list of these properties of political systems is likely to be extended and refined, and many actual political systems will be reclassified. Some of the recent research efforts are tending in the direction of “classifying” according to numerical scales of the cardinal type, which allow much more accurate placement of various systems within larger classification schemes. Political scientists will probably be interested in exact measures of specific properties, rather than the simple presence or absence of a property. They will not be concerned just with identifying a particular norm or institution but will want to achieve accurate measurements of its specificity, explicitness, flexibility, universality, and mode of operation. Political systems will no longer be thought of as unique congeries of attributes but will be regarded as systems which have more or less of some set of properties. Profiles of systems will be constructed, allowing comparison which will be theoretically more complex and empirically closer to reality.
Boundaries . A system is generally thought of as being distinct from its environment or as being self-contained and therefore having observable boundaries. Analysis seeks to determine both the members of the system—if they are concrete individuals—and the analytically distinct units of action which characterize the system. In the former instance we speak of those who actually are regarded as formal members, or citizens. In the second case, we speak of the actions that go into making up the political system of behavior, not in terms of concrete individuals but of the segments of their behavior which are politically relevant. In doing so, we arbitrarily assign boundaries to the political system. We then consider certain activities as political and temporarily ignore all others.
Boundary exchanges—inputs and outputs. Once the existence of bounded systems and subsystems is postulated, analysis must also account for the relationships that are to be found between systems. Common sense suggests that few, if any, actual social systems are either completely isolated or closed to specific kinds of external influences. As a consequence, systems analysis must be concerned with detecting relationships across boundaries— the problem of “inputs” and “outputs.” No firm agreement yet exists among political scientists as to the appropriate labels or concepts to be used in designating such matters. Easton (1957; 1965a), for example, sees the inputs from society and its various subsystems to the polity as consisting of “demands” and “support,” while Almond and Cole-man (1960) further divide and specify the inputs to the polity: “political socialization,” “recruitment,” “interest articulation,” “interest aggregation,” and “political communication.” Easton calls the output of the polity “decisions” concerning the authoritative allocation of values. Almond and Coleman, on the other hand, describe these outputs as “rule making,” “rule application,” and “rule adjudication.” Mitchell (1962) uses the terms “expectations and demands,” “resources,” and “support” to specify inputs; and “system goals,” “values and costs,” and “controls” to describe the political outputs.
While these several concepts vary, they are quite similar in the connotation of what is being exchanged across the boundaries of the polity and other subsystems. When Easton, Almond, and Mitchell speak of demands and of interest articulation and aggregation, they are referring to empirical phenomena concerning who demands what, from whom, how, when, and with what consequences for the participants and the system. These are all operational problems of research which can and have been treated empirically and even quantitatively. Put another way, systems analysts expect to measure inputs and outputs, so that comparisons can be made between political systems throughout the world. The objective is to establish minimal ratios of these exchanges, as a basis for predicting stability of systems and their capacity to achieve goals and to provide minimal levels of civic satisfaction for their members.
System processes . The “ultimate” concern is not simply to describe structures but to describe and account for the internal functioning of systems. At present we know far more about institutions than we do about political behavior, both in its individual and aggregate forms. Systematic information about the basic means or processes by which inputs in various systems become transformed or converted into political outputs is scant. For the most part, the major problem of inquiry seems to have been how politics allocate scarce values, but the formulation was not put in quite these terms until Easton’s work appeared (1953). Others also contributed: Dahl and Lindblom (1953), for example, elaborated four basic sociopolitical processes by which societies rationally calculate and control their collective actions: the price system, hierarchy, polyarchy, and bargaining. These four processes—not all purely political—are presumably found in varying combinations in all societies. Because they are found in different combinations and environments, they produce different results, all of which can be evaluated.
Other, more purely theoretical and mathematical attempts to handle the allocation problem are being made by the application of game theory and notions derived from welfare economics. Among these efforts is Anthony Downs’s attempt (1957) to use economic theory to account for the behavior of citizen and politician in a democracy. As yet these formulations have not been widely used by political scientists as theories of actual allocation outcomes in large-scale political systems.
In addition to the primary interest in allocative problems, political scientists, influenced by systems analysis, have been increasingly concerned with matters affecting the stability of systems and, more especially, with political socialization and other support inputs. A number of studies deal with the means whereby societies and polities assure loyalty and stimulate public participation. These studies are stimulated not only by theoretical developments in political science but also by problems in the real world of politics, such as the dramatic and often cruel means of control used in twentieth-century totalitarian states and the difficulties of establishing stable regimes in the newly developing states.
Other problems of the polity provide focal points for still other political processes and research endeavors. They include the means by which polities achieve collective goals from diverse individual demands and integrate their memberships. Much traditional work on leadership, power, bureaucracy, and control is relevant here. But the processes have not been well incorporated into the systems frame of reference.
The polity, as a system, can be assumed to have set functional problems, or exigencies, which are similar to those of all other social systems. It, too,-
tends to develop an internally differentiated set of substructures and processes to cope with each of its problems. In this view, first formulated by Parsons (1951; Parsons & Smelser 1956) and adapted by Mitchell to an interpretation of the American polity (1962), the internal processes of the polity are analogous to those of the larger social system, namely: goal attainment, adaptation, system maintenance and tension management, and integration. To these may be added the problem of the allocation of roles, resources, and values and costs. Identified as political processes, then, are the actions which members of a polity take with respect to the meeting or handling of each of these universal situations. A major task of empirical political science will be to clarify these processes; a major task of normative politics will be to improve the performances.
While much of the writing employing systems concepts, ideas, and hypotheses has been programmatic and theoretical, a growing number of empirical applications may be noted. These applications range from full-scale efforts at systems analysis to partial studies of very restricted forms of behavior and tests of but small segments of the more general scheme. They range in subject matter from international politics to intranational affairs. The international studies have been directly concerned with the structure of systems. Kaplan (1957) and Kaplan and Katzenbach (1961), for example, deal with the structure of international systems and norms in such systems. Likewise, Liska deals (1957) with the structure of international politics, equilibrium, and (1962) with coalitions or alliances within what must be described as an implicit systems framework. And Rosecrance (1963) considers nine distinct historical international systems within an explicit systems framework.
Investigations of national politics and cross-national or comparative political units are concerned with data on the inputs, especially those of demands and support within nation-states. Among the more notable studies are those of Almond and Coleman (1960) which, based on the previous theoretical work of Almond and of Easton, explored the political life of Asia, Africa, the Middle East, and Latin America. Works on single nation-states include Apter’s studies of Ghana (1955) and Uganda (1961). On the Soviet Union there is the volume of Bauer, Inkeles, and Kluckhohn (1956), which is explicitly based on the theory of social systems. Although these works place varying emphases on input-output categories, all are premised on the general idea of the polity as a boundary-maintaining, interdependent, and equilibrated system. An explicit attempt to use the input-output categories on a nation-state is Mitchell’s study of the United States (1962). The work of Karl Deutsch (1963) also indicates a concern for the kind of data that are required by comparative systems analysis. His data inventories relate mainly to the support input, resources for goal attainment, and the problems of integration. Many other such inventories will be needed before reliable generalizations about large numbers of nation-states can be made (see Russett et al. 1964).
At the level of specific institutions and behavior within single nation-states, systems analysis has been applied primarily to American data, with a special emphasis upon legislatures. Young’s work on Congress (1958) seems to have been influenced by systems notions. The case study by Freeman (1955) of congressional-administrative bureau relations is explicitly based on systems concepts. The major study to date on legislatures is that of Wahlke and his associates (1962). In this study of four American state legislatures, systems notions are employed with rigor and sophistication. For the most part, the study is one of role structures and orientations of the memberships rather than of system processes or the outputs. However, a variety of intersystem relationships, such as those of legislators with their constituencies, as well as with interest groups, are considered. Generally, however, the analysis is static. Typical of a more limited inquiry into legislatures is Fenno’s article (1962) and book (1966) on the Appropriations Committee of Congress as a small system. The focus of the inquiry is integration, its processes and outcomes with regard to that crucial committee.
Still other systems-based investigations and studies are Parsons’ own interpretation of elections within the systems framework (1959) and Eulau’s study of elections as a system process (1962). Easton’s work on political socialization and the support input (Easton & Hess 1961) grows directly out of his earlier development of the systems approach to politics. These various studies are all indicative of a growing trend in political science to cast its theoretical and empirical work in terms of the system framework. Even so, not all political scientists observe the trend with equanimity,-for there are some consequential problems involving theoretical and research points that require answers.
Theoretical and research difficulties
The critique of systems analysis, regardless of the source, is really three-pronged: some maintain that it has certain crucial methodological weaknesses; others claim that it is not suitable for empirical research; and still others suggest that it betrays and perpetuates a conservative bias.
The methodological criticisms are the most serious and the least likely to be resolved. For they do not raise problems subject to empirical test but, rather, question postulates about reality. Some maintain that systems analysis is misleading when it assumes that reality “really” consists of systems. In this view societies consist of far more individual and isolated events than systems analysis is capable of handling. In other words, not all variables in the supposed system are immediately affected by the disturbance of one element. Because there may be no “reciprocity,” there can be no maintenance of the system. The extent to which the elements of a society or polity are interdependent is questioned; interdependencies are matters, it is said, for empirical investigation, not for axiomatic treatment.
A second misgiving or criticism of the systems approach concerns the “boundaries” of the system. The point of the critics is that one cannot speak of systems unless one can identify boundaries or state the variables that constitute the system. The defense is that a system is an abstraction and can be specified in an arbitrary decision. There has been some clarification and refinement, so that boundaries may be empirically described and located. Thus, one may speak of different types of boundaries in terms of their permeability. Schoef-fler (1955), for instance, has characterized economic systems as (1) mechanically closed; (2) stochastically closed; (3) semiclosed (mechanically and stochastically); (4) conditionally closed; or (5) essentially open. All of these types are based on a scale of probabilities concerning the possibilities of outside influences, ranging from no influence to complete and continuous influence of outside variables. While these more or less logically derived system boundaries help to clarify the situation considerably, they do not easily lend themselves to operational definitions and empirical research. The most convenient and conventional specification of system boundaries in research is in terms of the membership, such as that of a formal association or organization. But this convenience, although often used, skirts many methodological and theoretical difficulties.
A third criticism has to do with the concept of equilibrium. There is more critical literature on this aspect of systems analysis than on any other single point. In brief, the claim is made that the concept “equilibrium” cannot be operationally defined, except perhaps in the context of economic behavior. This criticism follows logically either from the belief that systems do not in fact exist in society or from the assumption that the variables which constitute them are not linear and therefore cannot be expressed so that a state of equilibrium could be calculated or even identified. As the notion of equilibrium has not been used, in other than a very crude sense, as an analogy or metaphor in political science, criticism has not been marked. David Easton, the only political scientist who has devoted serious attention to the matter, is critical, even though his own formulation of the political system lends itself to the use of equilibrium treatment with respect to the exchanges of inputs and outputs with other segments of society. It is likely that as systems analysis is more frequently employed in political science, there will be increasing concern with the utility and shortcomings of the concept of “equilibrium.”
A fourth difficulty with systems analysis concerns the problems of boundary exchanges, or the various sets of inputs and outputs between a system and its environment or between subsystems of a larger system. So far, the use of such inputs and outputs in political science and sociology has been minimal. These elements of systems analysis can be fairly readily identified and measured, but this has seldom been done in actual research. Nevertheless, it can be pointed out that identifying these multiple exchanges is by no means a purely scientific procedure, nor are many of these presumed exchanges readily susceptible to empirical testing. What is exchanged in physical and organic systems appears to be considerably simpler to identify and measure and certainly more plausible to the conventional modes of thought.
Systems analysis is likely to be used for a long time to come, but the exact forms in which it will be cast are less certain. At present, two distinct variations may be detected: The first is the “structural-functional” approach, found mostly in sociology and comparative political studies of nation-states, which stems from organic analogies used in biology; the second—at once more mechanical and more mathematical—is found in economics and international relations analyses. The former school has tended to present a more complex and empirical picture of political systems, whereas the second variant has tended to develop abstract models of very delimited numbers of variables, which deal with a more restricted set of problems, generally centered about conflict. There are important differences in the style, methods, and data of the two approaches. But improved techniques of research will undoubtedly permit a gradual in tegration (Deutsch 1963), inspired perhaps by the historical example of the gradual unification of biology and physics.
William C. Mitchell
Almond, Gabriel A. 1956 Comparative Political Systems. Journal of Politics 18:391-409.
Almond, Gabriel A.; and Coleman, James S. (editors) 1960 The Politics of the Developing Areas. Princeton Univ. Press.
Almond, Gabriel A.; and Powell, G. Bingham Jr. 1966 Comparative Politics: A Developmental Approach. Boston: Little.
After, David E. (1955) 1963 Ghana in Transition. Rev. ed. New York: Atheneum. → First published as The Gold Coast in Transition.
Apter, David E. 1961 The Political Kingdom of Uganda: A Study in Bureaucratic Nationalism. Princeton Univ. Press.
Arrow, Kenneth J. 1951 Sociai Choice and Individual Values. New York: Wiley.
Barnard, Chester I. (1938) 1962 The Functions of the Executive. Cambridge, Mass.: Harvard Univ. Press.
Bauer, Raymond; Inkeles, Alex; and Kluckhohn, Clyde 1956 How the Soviet System Works: Cultural, Psychological, and Social Themes. Russian Research Center Studies, No. 24. Cambridge, Mass.: Harvard Univ. Press. → A paperback edition was published in 1961 by Random House.
Bertalanffy, Ludwig Von (1949) 1960 Problems of Life: An Evaluation of Modern Biological Thought. New York: Harper. → First published as Volume 1 of Das biologische Weltbild.
Bertalanffy, Ludwig Von 1950 An Outline of General System Theory. British Journal for the Philosophy of Science 1:134-165.
Black, Duncan 1958 The Theory of Committees and Elections. Cambridge Univ. Press.
Buchanan, James M.; and Tullock, Gordon 1962 The Calculus of Consent: Logical Foundations of Constitutional Democracy. Ann Arbor: Univ. of Michigan Press.
Cannon, Walter B. (1932) 1963 The Wisdom of the Body. Rev. & enl. ed. New York: Norton.
Dahl, Robert A.; and Lindblom, Charles E. 1953 Poii-tics, Economics, and Welfare: Planning and Politico-Economic Systems Resolved Into Basic Social Processes. New York: Harper. → A paperback edition was published in 1963.
Deutsch, Karl W. 1963 The Nerves of Government: Models of Political Communication and Control. New York: Free Press.
Downs, Anthony 1957 An Economic Theory of Democracy. New York: Harper.
Easton, David 1953 The Political System: An Inquiry Into the State of Political Science. New York: Knopf.
Easton, David 1957 An Approach to the Analysis of Political Systems. World Politics 9:383-400.
Easton, David 1965a A Framework for Political Analysis. Englewood Cliffs, N.J.: Prentice-Hall.
Easton, David 19656 A Systems Analysis of Political Life. New York: Wiley.
Easton, David; and Hess, Robert D. 1961 Youth and the Political System. Pages 226-251 in Seymour M. Lipset and Leo Lowenthal (editors), Culture and Social Character: The Work of David Riesman Reviewed. New York: Free Press.
Eisenstadt, Shmuel N. 1963 The Political Systems of Empires. New York: Free Press.
Eulau, Heinz 1962 Ciass and Parti/ in the Eisenhower Years: Class Roles and Perspectives in the 1952 and 1956 Elections. New York: Free Press.
Fenno, Richard F. 1962 The House Appropriations Committee as a Political System: The Problem of Integration. American Political Science Review 56:310— 324.
Fenno, Richard F. 1966 The Power of the Purse: Appropriations Politics in Congress. Boston: Little.
Freeman, J. Leiper 1955 The Political Process: Executive Bureau—Legislative Committee Relations. New York: Doubleday.
Homans, George C. 1950 The Human Group. New York: Harcourt.
Kaplan, Morton A. 1957 System and Process in International Politics. New York: Wiley.
Kaplan, Morton A.; and Katzenbach, Nicholas Deb. 1961 The Political Foundations of International Law. New York: Wiley.
Liska, George 1957 International Equilibrium: A Theoretical Essay on the Politics and Organization of Security. Cambridge, Mass.: Harvard Univ. Press.
Liska, George 1962 Nations in Alliance: The Limits of Interdependence. Baltimore: Johns Hopkins Press.
Mcclelland, Charles A. (1955) 1961 Applications of General Systems Theory in International Relations. Pages 412-420 in James N. Rosenau (editor), International Politics and Foreign Policy: A Reader in Research and Theory. New York: Free Press. → First published in Volume 12 of Main Currents in Modern Thought.
Mitchell, William C. 1962 The American Polity: A Social and Cultural Interpretation. New York: Free Press.
Parsons, Talcott 1951 The Social System. Glencoe, 111.: Free Press.
Parsons, Talcott 1958 Some Highlights of the General Theory of Action. Pages 282-301 in Roland A. Young (editor), Approaches to the Study of Politics. Evanston, 111.: Northwestern Univ. Press.
Parsons, Talcott 1959 Voting and the Equilibrium of the American Political System. Pages 80-120 in Eugene Burdick and Arthur J. Brodbeck (editors), American Voting Behavior. Glencoe, 111.: Free Press.
Parsons, Talcott 1960 Structure and Process in Modern Societies. Glencoe, 111.: Free Press.
Parsons, Talcott; and Smelser, Neil J. 1956 Economy and Society: A Study in the Integration of Economic and Social Theory. Glencoe, 111.: Free Press.
Riker, William H. 1962 The Theory of Political Coalitions. New Haven: Yale Univ. Press.
Roethlisbergeh, Fritz J.; and dlCKSON, wILLIAM J. (1939) 1961 Management and the Worker: An Account of a Research Program Conducted by the Western Electric Company, Hawthorne Works, Chicago. Cambridge, Mass.: Harvard Univ. Press. → A paperback edition was published in 1964 by Wiley.
Rosecrance, Richard N. 1963 Action and Reaction in World Politics: International Systems in Perspective. Boston: Little.
Russett, Bruce M. et al. 1964 Worid Handbook of Political and Social Indicators. New Haven: Yale Univ. Press.
Schoeffler, Sidney 1955 The Failure of Economics: ADiagnostic Study. Cambridge, Mass.: Harvard Univ. Press.
Simon, Herbert A. (1947) 1961 Administrative Behavior: A Study of Decision-making Processes in Administrative Organization. New York: Macmillan.
Wahlke, John et al. 1962 The Legislative System: Explorations in Legislative Behavior. New York: Wiley.
Young, Roland A. 1958 The American Congress. New York: Harper.
The modern use of the concept of system as a distinctive method of analysis has the following characteristics: (1) the system to be investigated is explicitly distinguished from its environment; (2) the internal elements of the system are explicitly stated; (3) there are relationships between the elements of the system and between the system and its environment that are explicitly stated; (4) where these relationships involve deductions, the canons of logical or of mathematical reasoning are employed; and (5) assertions concerning the relationships between the system and the real world are confirmed according to the canons of scientific method (see Ashby 1952).
The application of the systems concept to non-mechanical systems—and all social systems are nonmechanical—involves special problems. Whereas the independently measurable equalities of mechanical systems may provide highly general explanatory frameworks, nonmechanical systems require theoretical explanations adjusted to the special features of the system. Indeed, the use of the systems approach in the social sciences directs attention to differences in the subjects studied and in the explanatory theories that are employed to account for them. This last consideration will constitute the criterion for distinguishing between kinds of systems. Where the explanations of the behavior of two systems are derived from the same theory, the systems will be treated as being of the same kind; where the explanations are derived from different theories, the systems will be considered to be of different kinds. A new theory might, of course, produce different distinctions.
To illustrate, it will later be shown how the “balance of power” theory, which predicts rapid shifts of alliances and limited wars, can be used as at least a partial explanation of the rigid alliances and the almost unlimited nature of World War I. The pre-1870 and the post-1871 systems are considered to be similar kinds of systems operating under different circumstances. However, the loose bipolar system of the post-1945 period is regarded as different, because its behavior cannot be deduced from the assumptions employed in the “balance of power” theory, and indeed there are also different structural elements in the system.
In some cases, however, it may be possible to account for actual historical behavior by adjusting the parameters of either of two theoretical systems. And it may not be clear which adjustment gives greater explanatory power or which theoretical system comes closer to serving as an analogue for the historical situation. In such a twilight zone, individual judgment and utility are the decisive criteria.
The use of the concept of system means that attention is directed toward a specified group of interacting variables. The system, therefore, can be distinguished from its environment, and it can be compared for explicit features with other similar systems; in addition, the behavior of the system can be explained by a distinctive theory that in turn can be used to explain more complex historical situations.
The concept of an international system was first explicated in 1957 in Kaplan’s System and Process in International Politics-, by 1961, the term “international system” had become sufficiently common to serve as the title of a special issue of World Politics.
Types of international systems . Kaplan specified six types of international systems: the “balance of power” system, the loose bipolar system, the tight bipolar system, the universal system, the hierarchical system, and the unit veto system. (These types are not exhaustive but are intended to permit useful comparative analysis.) The systems have the following characteristics in common: they contain sets of essential rules, they share other internal elements (for example, types of actors, capabilities of actor, information factors, and transformation rules), and they all have boundary conditions. The systems also include specifications of conditions for three types of equilibrium: First, the essential rules are in equilibrium in the sense that a change in one rule will produce a change in at least one other rule. Second, a change in the set of rules will produce a change in other system characteristics and vice versa. Third, the system is also in equilibrium with its environment; changes in the system will change the environment and vice versa.
“Balance of power” system. In the “balance of power” system the only essential actors are nation states with very large military and economic capabilities. It is a system without role differentiation, and unless there are at least five essential national actors it is likely to prove unstable. If there are five or more such nations, then they are more interested in preventing the elimination of other nations as essential factors, in order to preserve them as future potential allies, than they are in dividing up their capabilities and resuming the competition within the system with a smaller number of actors. They do have an interest in acquiring a margin of security, by gaining more than an equal share of the capabilities of the system, and will therefore form alliances and go to war. But the wars will be limited and the alliances will shift rapidly and will tend to break up when the war is won, particularly if one of the victorious actors tries to eliminate one of the defeated foes. For similar reasons coalitions will tend to form that are directed against actors who threaten to predominate or who have organizational or ideological advantages that might produce dominance. Any essential national actor will be an acceptable role partner, for only thus can a national actor optimize the probability that he will be in a winning coalition or that he will not suffer too severe a defeat if he is in a losing coalition. This system is quite stable. [See Balance of power.]
Loose bipolar system. The loose bipolar system is role differentiated, having different kinds of actors, such as nations, blocs of nations, bloc leaders, bloc members, nonbloc members, and universal organizations. Here stability is increased if the leaders of the blocs possess nuclear-weapons systems and if they are the only actors who possess them. Alliances in this system tend to be based on long-term interests, even where immediate interests might at times argue for independent action. Wars would tend to be total rather than limited were it not for the destructiveness of nuclear weapons and the mediation efforts of nonbloc members and universal actors. This system tends to be less stable than the “balance of power” system.
Unit veto system. The unit veto system is a system of national actors or of bloc actors in which each actor possesses an extensive nuclear-weapons system. The unit veto system does not tend to produce alliances. It tends to keep low the probability of war but to give rise to tensions that might make for relative instability. It is less stable than the loose bipolar system.
The “balance of power” and loose bipolar systems models have had empirical counterparts. The unit veto system could develop out of the existing international system largely as a consequence of the diffusion of effective second-strike nuclear-weapons systems. The following three models rest upon more extensive changes from past international systems, and one of the models, the tight bipolar system, would require conditions running counter to existing developments.
Tight bipolar system. The tight bipolar system would have many features in common with the loose bipolar system except that the role of the uncommitted nation would wither away and the role of the universal organization would largely be atrophied. It would be a system of very high tension.
Universal system. The universal system would arise if a number of important political powers were transferred to a universal organization, perhaps as a consequence of the nuclear arms race or because the actors fear the dangers of the escalation of disputes. This system would require a reorientation on the part of the member actors, giving high priority to collective and international values. The universal organization would also require capabilities greater than those of any one member actor.
Hierarchical system. The hierarchical system might arise as a consequence of changes in scale in international organization, or as a consequence of the very successful functioning of a universal system, or as a consequence of conquest or dominance by a single actor. In the first two cases the resulting system would probably be a federal and democratic one. In the last case it would probably be an authoritarian system.
Other possible systems. Roger Masters (1961) has constructed a model for a system of nuclear blocs. He concluded that the characteristics of such a bloc system would probably resemble in many ways those of the “balance of power” system.
Other variations of international systems in the nuclear age have also been discussed. In one system catalytic war was a possibility (Burns 1959), and in another possession of nuclear-weapons systems by many states was consistent with international stability (Burns 1960). The latter study was based on the assumption that if one state suffered a surprise nuclear attack its retaliatory strike would make the aggressor extremely vulnerable to attack by still a third nuclear state. Thus, in effect, the decision to attack would disarm a state if it still had to face other nuclear enemies. The symmetry of this situation would seem likely to produce peace.
Although the six main systems described above were analyzed in terms of their equilibrium conditions, it is not assumed that equilibrium is either more likely than disequilibrium or even inherently desirable. It is assumed that under equilibrium conditions actors are motivated to maintain the equilibrium.
The systems can, however, undergo certain transformations, and some examples of these follow. A “balance of power” system, for instance, might be transformed into a unit veto system if most of the major states acquire sufficiently large second-strike nuclear systems. In such a case a technological change at the system’s boundary would produce changes in its essential rules. If, on the contrary, scale factors restrict viable nuclear establishments to two national actors, and particularly if one of them possesses national institutions capable of international extension, the blocs of a bipolar system might arise. Other requirements in the system might then produce a universal organization, such as the United Nations, having a differentiated role function in the system. The “balance of power” system would then have been transformed into a loose bipolar system.
Prediction and explanation
International systems can be viewed norma-tively; that is, their essential rules can be regarded as optimal rules of statecraft under the conditions specified for the system and the assumptions concerning the motivation of actors. Alternatively, they can be viewed as making predictions concerning actor behavior if the conditions specified by the theory hold and if the decision makers for the actors are rational, informed, and free to act on the basis of external considerations alone. Systems can also be used as aids in predicting what might happen if the conditions specified by the theory do not hold or if decision makers (statesmen) act contrary to the essential rules of the system.
If one uses the theory either to predict or to explain, then factors not included in the central theory must be taken into account. This is the problem of engineering the theory (see Knorr & Verba 1961, pp. 6-24). As this engineering occurs, the system is brought closer to the complex reality, but it loses generality.
Two examples will show how a theory can be used for purposes of explanation. The “balance of power” theory, for instance, predicts that alliances will be short-lived, based on immediate interest, and neglectful of existing or previous alliance status. It is not predictive of which actors would be in which alliance. Nicolson’s historical study of the Congress of Vienna (1946) shows the many accidental features that produced the individual alliances during the congress. Yet the series of shifting alliances as a whole is congruent with the theory.
The rigid alliance systems of the European great nations between 1871 and 1914 and the relatively unlimited nature of World War i seem, superficially at least, inconsistent with the prescriptions of the “balance of power” theory. If one recognizes, however, that Prussia’s seizure of Alsace-Lorraine created in France a public opinion that was ineluc-tably revanchist (as Bismarck probably foresaw that it would), then this parameter change is seen to be consistent with the developments that followed. As long as Germany was unwilling to return Alsace-Lorraine to France, France would be Germany’s enemy. Thus, France and Germany became in time the poles of rigidly opposed alliances and would not enter the same coalition, regardless of other, common interests. The chief motivation for limitation upon conflict in the theoretical “balance of power” system is the need to maintain the existence of other essential actors as potential future allies. However, since neither France nor Germany perceived the other as even a potential ally, neither had any incentive to limit its war aims against the other.
Rigor of systems theory . International systems theory is heuristic. It contains formulations that so far can be related to empirical systems only with great difficulty, primarily because appropriate empirical information has not yet been collected but also because the criteria for confirmation are inadequately developed. However, the theory directs research to some important questions; for instance, is the kind of shifting of alliances called for by the “balance of power” theory actually found in historical “balance of power” systems? Is a minimum of five essential actors really required for stability in such systems?
The international systems theory is also nonrigorous in that the conclusions are derived from the premises in a fashion that is no more than plausible. The internal mechanics of the systems are not completely understood, and some of the more important questions have not been analyzed. For instance, what difference does it make in a “balance of power” system to move from five to six, or from six to seven, essential actors? Does it make any difference if one, or more than one, of the essential actors is motivated by hegemonic ambitions? To answer such questions, one must use techniques that were not available in the late 1950s, when the theory was developed. Instead, insights from game theory were applied to broader questions concerning the general stability of the system.
Applications and related approaches
Theories of the “balance of power” and the loose bipolar systems have been applied to international law. Kaplan and Katzenbach hypothesized (1959; 1961) that wars would tend to be more limited under the “balance of power” system than under the loose bipolar system and that the laws of war would tend to be more closely observed. The rule of nonintervention in the internal affairs of other sovereign states would be more closely adhered to in the “balance of power” system. In the loose bipolar system recognition of states would require more political qualifications and would be less likely to be decided according to neutral rules of law. The rise of bipolarity has also been related, by hypothesis, to the functional transfer of sovereignty from the state to supranational organizations. Hoffmann has also related systematic changes in international law to changes in international political systems (Knorr & Verba 1961, pp. 205-237).
The “balance of power” and loose bipolar systems have also been related to problems of foreign policy. In applying the “balance of power” model it has often been found possible to ignore the role of internal characteristics of states in the formation of alliances, except where there are important deviant actors. In applications of the loose bipolar model, however, problems of bloc cohesion and identification may make it necessary to consider the domestic features of the actors’ regimes (Kaplan 1962).
These derived applications of the theory involve additional variables and less rigorous reasoning than the central propositions of systems theory. Thus, their conclusions are even more tentative.
Size and stability in “balance” systems. Scholars tend to agree that a lower bound of five essential actors is necessary for stability in a “balance of power” system. Below this bound the value of maintaining a defeated state is not plausible enough to counterbalance alternative motivations. However, there is some disagreement about the effects of increasing the number of actors above five. A. L. Burns (1957) regards five as the point of greatest stability, whereas Kaplan (1957) believes that stability increases as the number of actors increases to an upper bound, unspecified as yet. Burns believes that there is an inherent tendency to decrease the number of actors and no tendency to increase the number. Kaplan believes there is an inherent tendency to maintain the number and some tendency to increase it.
Burns’s analysis, based on a set of explicit propositions, gave rise to two important concepts: balancing and deterrence. The propositions are based on the concepts of pressure, interaction opportunities, security, military effectiveness, and allocation of attention. Normally, nonnuclear weapons encourage balancing and nuclear weapons make for deterrence.
Multipolar power systems have also been analyzed from the standpoint of mathematical models (Deutsch & Singer 1964). The analysis explicitly takes arms races into account, as well as interaction opportunities and allocation of attention. This study also concludes that five is a minimal bound for stability in a “balance of power” system and that such a system may be stable for several hundred years at least. But the authors note that no historic “balance of power” system has in fact lasted longer than several hundred years. Thus, they say, such systems, although more stable than bipolar ones, are inherently unstable. On the basis of chance alone, a four-to-one rather than a three-to-two coalition is likely to occur at some point, and such overwhelming strength in one coalition is likely to lead to the destruction of the system.
In assuming that a four-to-one coalition will be destructive of the system the authors overlook the instability of the victorious coalition, a factor that tends to assure the maintenance of the system (Kaplan et al. 1960). In the machine model discussed below, several series of four-to-one wars have occurred in which the system remained stable. These runs were continued through several hundred war cycles and were discontinued only when it became clear that further runs would not affect the results.
Instability could be produced, of course, if one programmed the machine to behave as if it lacked information or were irrational. However, other hypotheses could also be employed to achieve this end. For instance, changes could occur in the scale of activity, or political movements that cut across national boundaries could develop. It is at least plausible that such factors—and not coalition dominance—operated in the real world to produce instability in some historic “balance of power” systems and therefore that external, rather than internal, factors were responsible for the collapse of some historic “balance of power” systems.
Does the presence or the absence of the arms race from some models lead to significant differences between them? Possibly, but the incentive for war and for the spoils of war as it operates in the “balance of power” system is likely to produce most of the consequences of the arms race and will do so under the condition ordinarily assumed to maximize instability—that is, military conflict. In any event, this problem can be resolved on the basis of tests.
Parameter-oriented systems . A number of writers treat international systems more from the stand point of the parameter or boundary conditions that influence them than from the standpoint of system structure and functions. Quite often these writers do not make this distinction explicitly.
George Modelski (see Knorr & Verba 1961, pp. 205-237) distinguishes international systems according to whether they are based on agrarian or on industrial systems. Wars in agrarian systems, he finds, tend to be less destructive than wars in industrial ones. To some extent the reasoning behind this is compelling. The technologies of industrial systems permit more destructiveness and greater social control. They also provide more resources that can be made available to the state for the purpose of fighting or otherwise pursuing its objectives. Thus, Modelski’s position is reasonable as á statement of a tendency, although inaccurate with respect to many historical systems. His proposition, however, is not integrated into a theory or used to suggest modifications of the behavior of one or more types of international system.
In his analysis of communism as an international system, Modelski (1960) stresses structural and behavioral features of the system and is thus able to investigate how ideological and organizational factors specific to communism affect the relationships of communist states in a larger world system.
Quincy Wright (1955), drawing on Talcott Parsons’ work, has established a number of analytic fields, covering values and capabilities. After states are located in these fields, it might be possible to make a number of statements concerning trends in their behavior, but problems of measurement are enormous here, as are problems of relating the variables covered by the fields. This work does not distinguish systems so much as it aspires to measure state characteristics; thus, it is not of the same genre as international systems theories.
Kenneth Boulding (1959) has attempted to define national images and to relate them systematically to other characteristics of international systems. National images, he states, are impregnated with valuational distortions that produce national responses inconsistent with peaceful equilibrium. This view of images could perhaps be assimilated to the informational factor in Kaplan’s systems, and these systems could then be tested under various conditions of misinformation.
Boulding (1962) has also formulated a theory of conflict and defense, based on Hotelling models, in which strength diminishes with distance and in which equilibrium occurs where strengths are equal. These systems models do not take into account the fact that diminishing strength is neither linear with distance, nor continuous, nor transferable between weapons systems. Boulding’s theory does distinguish between conventional and nuclear weapons. In his nuclear systems, strengths do not diminish with distance, although most empirical studies show that they do (Wohlstetter et al. 1954). This also holds true for strategic nuclear missiles and for strategic bomber forces.
Methods for studying international systems
A number of difficult methodological questions arise in the study of international systems. These questions concern both the theories of the systems and the relationships between theories and empirical reality. To help overcome some of these problems, games and machines have recently been used to simulate both theories and reality, and some comparative analyses of historical systems have been attempted.
Simulations of reality . Harold Guetzkow and his associates (1963) have conducted simulation experiments in which they adapted the small-group methods of organization specialists and social psychologists to the simulation of international systems. In these simulations, individuals are used to represent the executive branch, the Congress, the United Nations, foreign countries, the press, and so forth. They interact within a specified framework. This technique is useful in suggesting hypotheses, and it might possibly replicate some important aspects of reality. The experiment, however, is probably too complex for complete analysis, and some of the variables, including the calculations made by the individual role players, cannot easily be related to international factors. Moreover, the representation of the domestic political process as an intervening variable in the international process is much oversimplified.
Still another method of simulating international politics has been attempted by Clark Abt (see Abt & Jaros 1961). He employs an extremely complex computer model that includes great and small nations, budgetary decision processes, complicated military postures, and so forth, as well as Kaplan’s essential rules. His model is so complex that it becomes ad hoc. Very small changes in some of his parameters might produce quite different results. All, or at least most, of the important variables have been included in the simulation, but it is difficult to know whether the computer problem has any real empirical referent despite its explicit empirical orientation.
Simulations of theories . In order to represent important aspects of the “balance of power” theory rather than to simulate reality, a table-stakes game was designed (Kaplan et al. 1960). It used players who tried to optimize particular kinds of counters. The game was designed to minimize the influence of the players’ images about international politics. And the game, indeed, can be played without any reference to international politics. It was thus hoped to use the players’ minds as a computer for the strictly strategic elements of play in lieu of programming a computer to play the game. Even so, it was by no means clear what produced the results, and they were, therefore, suggestive at best.
Some of the difficulties in using a computer to test the theory have, however, been overcome. A simple model of the “balance of power” system has been placed on a computer which in the initial phase makes decisions for five players, but the program is being generalized for three to seven players. The players, representing essential national actors, are assigned sizes, military potentials, attitudes toward risk, attitudes toward foreign imbalance, subjective estimates of military strengths (which may differ from the objective strengths), and a factor representing eagerness to form coalitions. The machine then makes decisions for the players on the basis of recursive optimization procedures, which take into account the results over one war cycle.
According to the early runs, if the players have a small distaste for foreign imbalance—that is, if they behave like the postulated “balance of power” players—there is a series of limited wars that do not eliminate any of the players. Increasing the distaste for foreign imbalance produces peace. Decreasing it produces an unstable series of wars in which players are eliminated. If one of the players likes foreign imbalance, it is necessary to increase the distaste of the other players to produce stability. Stability might also be maintained by employing a counter-deviancy factor (analogous to rule four of Kaplan’s “balance of power” system) that calls for coalitions against deviant players that strive for hegemony within the system. In the absence of such a counter-deviancy factor, if one of the players likes imbalance and if the distaste for imbalance of the other players is not increased, one or more of the players will be eliminated. The deviant player who likes imbalance will not be among the eliminated players and thus will have a selective advantage. It would therefore seem to be advantageous, even for the players with a distaste for foreign imbalance, to play as if they liked it. But in this case the system would become unstable. And having—or acting as if one had—a liking for imbalance would then create no selective advantage; but security would be lower than if all players acted as if they had a distaste for imbalance.
Alternative formulas for figuring imbalance in the system are employed to guard against the possibility that the results are artifacts of the formulas. Two methods have been found that increase the ratio of three-to-two wars: changes in the battle ratios and side-payments for latecomers to coalitions.
No effort has yet been made in the project to put more complicated international systems (for instance, systems similar to the loose bipolar one) on the computer. It is not yet known how to program for the complicated features of this system, and more sophisticated computers with more powerful memory systems may well be required. Furthermore, although an attempt had been made (Kaplan 1957) to relate a number of internal political processes and structural political elements to the theory and to indicate how these would modify it, the computer was programmed only for the simplest features of national or bloc decision making. Once it is known how the models operate in the absence of internal politics, ad hoc adjustments can be made. [See Simulation, article onpolitical processes.]
Systematic empirical research . Although one can show how theories of international systems can be used to explain real political events, it is less easy to test these theories systematically. One could attempt to test individual propositions derived from the theory—the usual method in the physical sciences, where circumstances permit. It has been suggested that both patterns of historical activity and frequency of interaction processes should be investigated (Kaplan 1957). It now appears that the frequency of interactions plays a less important role than was originally anticipated. Since the number of cases is too small, different variables assume importance in different cases, and in many cases events have no simple meanings that are discoverable apart from contextual analysis. Thus, the investigation of historical patterns of activity now appears to assume even greater importance than was originally anticipated.
Little systematic historical analysis, however, has been carried out. Richard Rosecrance’s work (1963), although interesting, is not directly useful for testing hypotheses stemming from theory, because his evidence is not collected according to a systematic set of questions. At the Ford Workshop at the University of Chicago, efforts are being made to study empirical international systems—such as the Italian city-state system and the Chinese-warlord system. These studies use a systematic theory of international politics and are therefore oriented toward a systematic set of questions.
Such historical studies should probably investigate interactions between those variables included in the theory and other important variables that affect the behavior of the specific system under investigation. In the Italian city-state system, for example, the use of mercenaries by the city-states helped maintain the equilibrium conditions specified by the theory. The mercenaries themselves had an incentive to behave consistently with the essential rules of the system, for instability would have undercut their own role in the system. In this case the explanation offered by the theory holds in general, but the particular way in which the equilibrium is maintained requires, among other things, an analysis of the interaction of the mercenary system with the city-state system.
This problem suggests the way in which empirical research and computer analysis could be combined to test, evaluate, and, if necessary, reformulate the propositions of international systems theory.
The mercenary system might have been an element that increased the stability of the historic system but that was not strictly necessary for stability. If so, one would not desire to change the theory, for a change in the theory would narrow its range of useful application. However, if investigation seemed to show that historic “balance of power” systems are stable only when some additional factor is operating, the system might be modified to include it. This factor might then be built into the computer model, and if the model is stable without it, one might investigate what other changes in the model would be needed—conceivably, the addition of motivation or information elements—to make such a factor necessary for stability. Alternatively, the computer runs could show that the specific elements in the historical systems not included in the theoretical system merely maintain the parameters of the systems at their theoretical equilibrium or merely move them from equilibrium under specific conditions.
There may—and ideally should—be a continual process of learning that relates historical studies, systems theories, and computer models. Although this methodology will not provide the kind of assurance that the “hard” sciences aspire to, it will introduce more rigor and scientific method into the study of international politics.
Morton A. Kaplan
Ast, Clark C.; and Jaros, Walter F. 1961 Design for a Strategic Model. No. BR 1354A. Bedford, Mass.: Raytheon Company, Missile and Space Division.
Aron, Raymond (1962) 1967 Peace and War: A Theory of International Relations. Garden City, N.Y.: Doubleday. → First published as Paix et guerre entre les nations.
Ashby, W. Ross (1952) 1960 Design for a Brain: The Origin of Adaptive Behavior. 2d ed., rev. New York: Wiley.
Boulding, Kenneth E. (1959) 1961 National Images and International Systems. Pages 391-398 in James N. Rosenau (editor), International Politics and Foreign Policy: A Reader in Research and Theory. New York: Free Press.
Boulding, Kenneth E. 1962 Conflict and Defense: A General Theory. New York: Harper.
Burns, Arthur L. 1957 From Balance to Deterrence: A Theoretical Analysis. World Politics 9:494-529.
Burns, Arthur L. 1959 The Rationale of Catalytic War. Research Monograph No. 3. Princeton Univ., Center of International Studies.
Burns, Arthur L. 1960 Power Politics and the Growing Nuclear Club. Policy Memorandum No. 20. Princeton Univ., Center of International Studies.
Deutsch, Karl W.; and Singer, J. David 1964 Multi-polar Power Systems and International Stability. World Politics 16:390-406.
Guetzkow, Harold S. et al. 1963 Simulation in International Relations: Developments for Research and Teaching. Englewood Cliifs, N.J.: Prentice-Hall.
Kaplan, Morton A. 1957 System and Process in International Politics. New York: Wiley.
Kaplan, Morton A. (editor) 1962 The Revolution in World Politics. New York: Wiley.
Kaplan, Morton A.; Burns, Arthur L.; and Quandt, Richard E. 1960 Theoretical Analysis of the “Balance of Power.” Behavioral Science 5:240-252.
Kaplan, Morton A.; and Katzenbach, Nicholas Deb. 1959 The Patterns of International Politics and of International Law. American Political Science Review 53:693-712.
Kaplan, Morton A.; and Katzenbach, Nicholas Deb. 1961 The Political Foundations of International Law. New York: Wiley.
Knorr, Klaus E.; and Verba, Sidney (editors) 1961 The International System: Theoretical Essays. Princeton Univ. Press. → See especially pages 6-24, “Problems of Theory Confirmation in International Politics,” by Morton A. Kaplan; pages 118-143, “Agraria and Industria: Two Models of the International System,” by George A. Modelski; and pages 205-237, “International Systems and International Law,” by Stanley Hoffmann.
Masters, Roger D. 1961 A Multi-bloc Model of the International System. American Political Science Review 55:780-798.
Nicolson, Harold (1946) 1961 The Congress of Vienna: A Study in Allied Unity, 1812-1822. New York: Viking.
Rosecrance, Richard N. 1963 Action and Reaction in World Politics: International Systems in Perspective. Boston: Little.
Wohlstetter, Albert A. et al. 1954 Selection and Use of Strategic Air Bases. Santa Monica, Calif.: RAND Corp.
Weight, Quincy 1955 The Study of International Relations. New York: Appleton.
Systems, or systemic, analysis in contemporary psychology is an attempt to relate behavior to the organizational aspects of its underlying structure. It is a way of conceptualizing the phenomena that mediate between a stimulus—or environmental event—and the behaving organism’s response to it. This article will briefly describe some considerations of general systems theory and will continue with discussions of important systemic variables in psychological analyses, some representative usage of systems analysis by psychologists, including its potential for the area of mental ability and retardation, and an evaluation.
It must be pointed out clearly that there is an alternate use of the phrase “systems analysis” in psychology. The language, concepts, and theoretical propositions devised to understand, integrate, and explain behavioral phenomena may themselves constitute systems. These systems have been subject to thorough and critical evaluation and analysis (e.g., Koch 1959). Such analyses, which treat of and transcend psychological phenomena, are not the concern of this article, which deals, rather, with systems involved in and part of psychological phenomena.
Definition and general considerations
The problems of defining and identifying systems are discussed at length in the other articles on systems analysis. It is sufficient here to quote Miller (1955): “Systems are bounded regions in space-time, involving energy interchange among their parts, which are associated in functional relationships, and with their environment. ...” A system is also considered to be a group of events that have a higher interchange of energy or a higher rate of communication among themselves than with other events (see Scott 1962, p. 97). The generality of this definition allows the several sciences the opportunity to select various types and levels of events to be treated as systems. Even so, within individual disciplines there are diverse conceptions of systems.
In psychology the assumptions made about the nature of behavioral phenomena are reflected in the conceptions of the systems that govern these events. Allport very carefully, and with considerable breadth and perspective, describes four levels of “openness” in psychological systems (1960). The lowest level of openness is exhibited by those systems whose relations with their environment consist merely of engaging in intake and output of matter and energy. Perhaps this is the lowest common denominator of all organic systems. The traditional psychophysics of Wundt and the classical and instrumental learning models of Pavlov, Watson, and Skinner consider behavioral events in such terms: behavior is primarily reflexive, stimulus-generated, and unmediated. The relationship between the behaving organism and its environment is unilateral and mechanical.
A second level of openness is attained when, in addition, systems achieve and maintain homeosta-sis. Hullian learning theory and Freudian psychoanalysis view behavioral systems at this level. The importance of the internal state of the system is fully appreciated. Behavior is considered to be adjustive and represents some form of mediation of stimuli by internal mechanisms. [See Homeostasis.]
With the addition of increasing organization among the internal components of the system, a third level of openness is reached. Gestalt theories of perception and of insight learning, stressing the tendencies to organize, to restructure, and to improve upon present status, and personality theories, such as Jung’s and those of the ego psychologists, that emphasize growth and development of the self in addition to immediate adjustment, conceive of psychological systems at this level. Behavior is not viewed as merely a mediation of the external by the internal to achieve some adjustive, tension-free state but includes, as well, natural growth and development of the mediating agency, apart from and not necessarily contingent upon external events.
Finally, at the fourth and highest level of openness, Allport places those conceptualizations of behavioral events which provide for their engaging in transaction with their environment. Allport is less explicit at this level, but he implies that the system itself is capable of acting upon its environment with appreciable autonomy and is not restricted to mere responding or reacting. Allport is careful to point out that such acts must nonetheless be viewed and understood with proper consideration of the environment in which they occur. Needless to say, the Western philosophical heritage, which encourages an “integumented” view of personality as well as a dichotomous view of man and his social world, has prevented a full appreciation of the extent of this transaction.
Airport’s use of the term “openness” is somewhat at variance with the usage of others, but his emphasis on the need to consider human behavior as a reflection of systems that are more open rather than less open (or quasi closed) can be readily appreciated. And his discussion is an interesting general exposition of various systemic analyses : ways of relating behavior to its organizational aspects. [See Personality: contemporaryviewpoints, article ona unique and open system.]
Structural and functional dimensions . The most important structural concept in systems analysis is the boundary (see, for example, Parsons 1959, p. 645). A boundary (of a system or subsystem) is often difficult to demonstrate, depending for the most part on relatively different frequencies and intensities of interchange between events. Where one set of events demonstrates greater interchange within itself than with other events or sets of events, a boundary is said to exist around it, and the set of events is considered a bounded region.
Psychologists who employ a systemic framework are primarily concerned with the nature and relationships of bounded regions within—and, to a lesser extent, with the boundaries between—systems. Boundaries can be considered in terms of the frequency with which they occur within a system; in other words, the number of parts of which the system is constructed, the dimension of differentiation.
Boundaries can be considered, in addition, in terms of the extent to which they permit interchange between regions of a system and between a system and its environment: the dimension of interdependence. (When interchange occurs with equal frequency and intensity within and between regions, by definition no boundary exists.)
The dimension of openness refers to the degree of interchange across the outer boundary of the system itself; in other words, the degree of interchange with the system’s environment.
Differentiation, interdependence, and openness are basic structural attributes. “Structural” may be a misnomer, however, in the case of openness, since openness implies some activity or response on the part of the organism and thus has functional import. Indeed, the very essence of systems analysis is its appreciation of the close relationship between structure and function.
One way of separating these two levels of discourse is to think of structure as involving a spatial dimension and of function as involving a temporal one. Even so, in systems where there is increasing organization over time, the structures themselves will become functions.
Still, one can think of function as implying the continued existence of the system over time, the perpetuation of its outer boundary, and the maintenance and growth of its inner regions.
Representative major uses
Three representative uses of systems analysis in psychology will be considered. For each, the universe of events treated as a system will be outlined, along with a discussion of differentiation, interdependence, openness, and their implications. The first two sections especially refer to works that provide considerable breadth, depth, and detail, and their treatment here is not meant to be synoptic but selective. In addition, the implications of systems analysis for mental ability and retardation will be discussed.
Open and closed minds . Rokeach and his associates, in an attempt to integrate findings in areas such as authoritarianism, conformity, yielding, resistance to acculturation, ethnocentrism, and prejudice, have devoted considerable attention to the belief-disbelief system (Rokeach 1960). Each person, in the course of his development, acquires an “organization of verbal and nonverbal, implicit and explicit beliefs, sets, or expectancies” about the world in which he lives (p. 32). He has beliefs about the physical world, about the past, about the future, about the supernatural, and about the social world. The simple statement ’Table salt is made up of sodium and chlorine” is in actuality a belief; it may well be prefaced by “I believe.”
In addition to the set of beliefs and expectancies that the person accepts as true, he also has a set of beliefs that he rejects as false. These involve the same kinds of events or objects—the supernatural, physicality, etc.—as do the accepted beliefs. Together, the set of beliefs accepted as true and the set rejected as false—disbeliefs—constitute the belief-disbelief system. The disbelief system itself comprises a series of disbelief subsystems, each reflecting a different degree of rejection. The term “system” implies that there is an organization of parts and that the nature of this organization has implications for the behavior and functioning of the system.
Differentiation. Rokeach’s concern with differentiation is manifested in his assertion that belief-disbelief systems vary in their “degrees of differentiation or articulation or richness of detail” (I960, pp. 37-38). The amount of knowledge a person has about things he believes or disbelieves is one indication of the degree to which the belief-disbelief system or either part of it is differentiated. Differentiation within the disbelief system can be measured by the degree to which two disbelief subsystems are perceived as different. The Hearst press concept “communazi” implies no differentiation between disbeliefs, while an individual’s distinction between communism and national socialism indicates that they are differentiated by him, even though they may each be a disbelief. It is further assumed that there is generally greater differentiation in the belief system than in the disbelief system.
Rokeach is also theoretically concerned with the total number—or the range of—disbelief subsystems represented in a belief-disbelief system. While he calls this comprehensiveness or narrowness, it is essentially differentiation.
Isolation. Rokeach’s concern with interdependence is reflected in the assertion that the belief-disbelief system can vary in the extent to which a person sees his beliefs as being connected or related in some way. When two intrinsically related beliefs are seen as being in no way related, Rokeach terms them isolated. One indication of isolation is the coexistence of logically contradictory beliefs within the belief system. As examples, Rokeach points to the person who believes in democracy and who advocates, at the same time, that a government should be run by the intellectually elite. Another indication is the accentuation of differences and the minimization of similarities between belief and disbelief systems, as when, for example, a staunch advocate of psychoanalysis claims that it has nothing in common with behaviorism. Other indications of isolation are a disproportionate tendency to perceive events as irrelevant to one’s beliefs and a tendency to deny that events contradict one’s beliefs. When logically related beliefs are perceived as being related, then communication exists between them; communication is identical with the concept of interdependence.
Openness, closedness, and structure. A major goal for Rokeach and his associates is relating structural, organizational variables, such as differentiation and isolation (interdependence), to a dimension called openness-closedness. In any behavioral situation, a person’s response can be appropriate to the relevant characteristics of the situation or determined primarily by factors not related to the real demands of the situation. It is necessary, then, for people to be able to distinguish and evaluate relevant and irrelevant cues. A basic, defining characteristic of “open” minds is the ability to “receive, evaluate, and act on relevant information received from the outside on its own intrinsic merits, unencumbered by irrelevant factors in the situation arising from within the person or from the outside . . .” (Rokeach 1960, p. 57). Such irrelevant internal pressures are unrelated habits, irrational motives, a need to allay anxiety; irrelevant external pressures include rewards and punishments from external authorities.
It is explicitly suggested that all communication has a dual character: it contains information about a source, as well as information about an event or set of events. Of major importance are the relative strengths of two powerful and opposing motives served by all belief-disbelief systems: “the need for a cognitive framework to know and to understand and the need to ward off threatening aspects of reality” (Rokeach 1960, p. 67). When the need to know is dominant and the need to ward off threat is minimal, irrelevant pressures and drives are pushed aside and information is objectively evaluated; when the need to ward off threat is stronger, the need to know becomes weaker and source and substance are not properly separated.
Threat arises from experiences that make man feel alone, isolated and helpless, and anxious about his future. Accordingly, under threat, when a man cannot appreciate the duality between source and substance, he accepts a belief on the basis of irrelevant factors rather than on its inherent merits, and the relation between his beliefs will reflect this irrelevance: there will be no necessary logical consistency; his beliefs will exist in isolation from one another. Where the dual nature of communication is understood and properly evaluated, beliefs are more readily accepted for their coherence and logical consistency; there will be less isolation (greater interdependence) between them.
In addition, if a person is able to distinguish between source and substance, he will be more ready to seek information about beliefs he rejects from the sources that advocate them. The more he is unable to distinguish source from substance, the more readily he will depend on the sources that advocate his own beliefs for information about the disbelief system. The closed mind, consequently, will become even more rejecting of its disbeliefs and will know much less about them.
In this way the open mind is considered to represent a structural organization generally having greater differentiation within its disbelief system and greater communication within and between belief and disbelief systems (interdependence), while the closed mind is characterized by less differentiation within its disbelief system and greater isolation within and between belief and disbelief systems.
Methodology. From the above theoretical considerations Rokeach and his associates developed the Dogmatism Scale, to measure the degree of openness or closedness of belief-disbelief systems. The Dogmatism Scale contains items specifically related to differentiation and isolation, as well as items deemed relevant to other related characteristics, including the degree to which a person feels isolated, uncertain, and inadequate: the experiences of threat, from which arise closed systems. The Opinionation Scale, designed to assess the degree to which a person’s acceptance or rejection of others is based on how similar their beliefs are to his own, was also developed. Both of these scales were designed to eliminate the effects of specific belief contents and to stress instead the structure of the belief system.
Results and implications. A discussion of all the subtle and varied relations between performances on the above scales and aspects of behavior is precluded. However, it was demonstrated that closed systems exhibit greater difficulty than open systems in forming totally new belief systems but exhibit no differences in overcoming individual old beliefs (Rokeach & Vidulich 1960). This poorer ability to synthesize leads to greater difficulty in certain kinds of problem-solving tasks, where synthesis is crucial. Also, closed minds persist in trying to handle insoluble problems with a belief system that had formerly proved useful. They do not defect from their belief systems as readily as do open minds. This strange, seemingly maladap-tive loyalty is considered to be an indication of lack of communication across boundaries within the belief system (Rokeach et al. 1960a). Closed minds seem to perform less adequately in perceptual synthesis, or the restructuring of the perceptual field (Levy & Rokeach 1960), to be less receptive to “new music” (Mikol 1960), and to exhibit more compliance than open minds in accepting new beliefs when these are presented all at once rather than gradually (Rokeach et al. 1960&). It is pointed out that closed minds are not always resistant to change and will, in fact, change more readily than open minds when new belief systems are offered to them all at once and arbitrarily.
To the extent that circumstances demand synthesis—the ability to integrate and accept a series of new beliefs in place of old ones (”beliefs” broadly defined)—the open-minded person, i.e., the person whose belief-disbelief system comprises differentiated parts that are in communication with each other, will have an advantage over the closed-minded person.
Furthermore, Rokeach has demonstrated that authoritarians of both left and right political orientations have belief systems that are structurally similar. Both groups score relatively high on the Dogmatism and Opinionation scales. This represents an advance over the widely used California F and E scales, which are attuned primarily to a fascistic and ethnocentric authoritarianism and do not reveal the dogmatism and authoritarianism of the communist, who to as great an extent is closed-minded and dependent on limited authoritarian sources and whose belief system may be reflecting the same defenses against anxiety attributed to the traditional conception of authoritarianism. And it is consistent with the concept of implied greater rejection of disbeliefs that the closed mind is found to be more opinionated and prejudiced than the open mind.
Concrete and abstract conceptual systems . Harvey, Hunt, and Schroder present another attempt to integrate many areas of interest to psychologists and other social scientists: child development, attitude change, personality measurement, motivation, and psychopathology. Of prime concern to their work are concepts, or conceptual systems. Concepts refer to “. . . the individual’s standardized evaluative predilections toward differentiated aspects of his external world. ... A concept is a system of ordering that serves as the mediating linkage between the input side (stimuli) and the output side (response)” (1961, p. 1). Of basic interest is the structural nature of these concepts: how the person is related to objects in his world, rather than the nature and content of these objects.
It is assumed that a single concept probably never functions independently of others; accordingly, it makes more sense to discuss groupings of concepts, or conceptual systems. The totality of all of a person’s conceptual systems is termed the self system. Regardless of the size of the conceptual system, the same types of structural analyses are appropriate.
Concreteness-abstractness. The most important functional dimension of a conceptual system is the degree to which it is concrete or abstract. “Con-creteness” implies that the relation between the person and an object is determined primarily by the physical attributes of that object. “Abstractness” connotes that the relation between a person and an object is determined to a considerably lesser degree by the physical nature of the object and to a much greater degree by mediating factors that result from the individual’s experiences and activities. Among other things, abstractness involves being able to detach oneself from immediate inner and outer experiences, to switch mental sets, to grasp essential aspects of a given whole and be able to analyze its parts, and to plan ahead ideationally, all of which represent a minimal amount of stimulus determination (Harvey et al. 1961, p. 25).
Although concreteness-abstractness is called a structural attribute, the relative nature of the terms “structure” and “function” is recognized (pp. 19-20). Accordingly, in the terminology of this article concreteness-abstractness is considered a functional aspect of systems, to be related to the system’s structural aspects.
Harvey, Hunt, and Schroder consider development to represent increasing abstractness, and learning is considered to be the acquisition of concepts. Learning occurs through a process of differentiation and integration, wherein the child breaks down or analyzes his perceived environment into parts that are relevant to him and then relates these parts to each other and to his entire conceptual system (p. 4). Abstractness reflects the degree to which the processes of differentiation and integration have occurred within conceptual systems.
Structural attributes. The processes of differentiation and integration result in several important structural attributes, among them clarity-ambiguity, compartmentalization-interrelatedness, and openness-closedness. Clarity-ambiguity refers to the degree to which the concept, or relation between the person and an object or event, is composed of articulate, differentiated parts. The greater the number of these differentiated parts, the greater is the concept’s clarity. Thus, clarity-ambiguity within a concept or within a conceptual system is identical to the dimension of differentiation treated in this discussion. Compartmentaliza-tion-interrelatedness refers to the degree to which differentiated concepts within a system are connected and is identical to interdependence. Openness-closedness represents the receptivity or sensitivity of the system, much as it has been defined in this article.
Harvey, Hunt, and Schroder proceed to relate concreteness-abstractness to these structural components. Concrete functioning is assumed to reflect conceptual systems that are unarticulated, highly compartmentalized, and inaccessible to impinging events, whereas abstract functioning is an attribute of conceptual systems that are highly articulated, highly interrelated, and sensitive to impinging events.
The developmental process. Considerable detail is provided (Harvey, Hunt & Schroder 1961, chapters 4, 5) as to how training and experiential conditions influence development and shape the emergence of a high level of abstract functioning. Normally, at the very earliest stage of development, when differentiation has not yet taken place, the child’s concepts are poorly articulated, relatively ambiguous. Accordingly, his behavior is determined by the absolute (to him) physical nature of objects in the world around him: external criteria control him; rules and their goals are not distinguished; and rules have an absolutistic quality. As development proceeds, the child becomes aware of the limits of such absolutes: through the occurrence of gross differentiation, the external attributes of objects are distinguished from the internal wishes and needs of the child. At this point internal needs develop a compelling force on the child’s behavior, in much the same absolute way that external attributes categorically controlled him before. And there is considerable resistance to and hostility toward external control. It is only through increased differentiation and integration of his concepts both of external and of internal events that the young person can become aware of an increasing number of facets of his environment and thus take these into account in his behavior, giving him a basis for more abstract functioning.
For example, the child’s earliest concept of an understood parental demand or request is largely undifferentiated. It is categorical, to be complied with unquestioningly. As differentiation progresses and the parent himself is perceived as being psychologically distinct from the child, the child comes to differentiate his own needs and wishes from each other and from those of the parent, and the concept of the parental demand comes to represent these articulated components. After some initial conflict between the external and the internal requirements, the internal requirements will become ascendant for a while, and the child’s behavior will be determined by them. However, as the process of integration emerges, there will develop some synthesis of juxtaposed external and internal demands, and the concept of the parental command will contain an evaluation of its positive and negative effects on the child, as well as the positive and negative implications of the child’s acceptance or rejection of the command. Consequently, the child’s response will reflect greater mediation and more abstractness. He has “multiple alternatives” to consider as bases for his behavior.
Psychopathology. It is also pointed out that various psychopathological “syndromes” can be related to the structural organization of conceptual systems. When a person encounters experiences or events that refute his conceptual system, he experiences threat. When concepts are relatively un-articulated, as they are in early childhood, threat is likely to occur when experiences deny the absoluteness of rules or the absolute validity of a concept. Schizophrenia is seen to be an extreme means of warding off threat in the earliest, unarticulated, concretistic system. When internal and external aspects have been grossly differentiated, threat is produced by pressures toward dependency on the external. The psychopathic personality is seen as an extreme reaction to threat in the conceptual system that has grossly differentiated internal and external events and is negatively dependent on the external. Other forms of extreme reactions to threat are systematically considered, as are the abnormal behaviors reflecting less extreme reactions to threat.
The structural attribute of openness-closedness is relevant here. Where differentiation and integration either have not occurred or have occurred in limited measure, where concepts are poorly articulated, too many events in the environment are potentially refuting. Where concepts represent cate-goricals, any experiences that cast doubt on their “oughtness” are refutive. Consequently, to preserve the existing conceptual structure, the person avoids coming into contact with the threatening experiences, fails to perceive them, denies they exist, or misinterprets them. On the other hand, more abstract conceptual systems are threatened by experiences that deny the validity of “multiple alternatives” and that employ categorical evaluations and emphasize authoritarian submission. These systems are considerably more open in general and react much differently to such threats. For example, the possible validity of two different viewpoints may be asserted, or the person may engage in information-seeking behavior.
Cognitive tuning . Zajonc (1960) has provided a systematic attempt at experimentally manipulating aspects of cognitive systems. Defining a cognitive structure as “an organized subset of the given cognitive universe in terms of which the individual identifies and discriminates a particular object or event” (p. 159), Zajonc suggests the importance of the structural properties of differentiation, complexity, unity, and organization in determining the behavior of the cognitive structure.
Structural attributes. Differentiation is reflected in the number of attributes that can be utilized in identifying and discriminating objects. Unity reflects the degree to which the differentiated attributes depend on each other, theoretically identical to interdependence. Unity is mathematically and behaviorally represented by the degree to which a change in one attribute is reflected in changes in other attributes. Complexity is a more advanced treatment of differentiation and represents the number of different categories of discriminanda that are represented in the number of discriminated attributes. Organization reflects the grouping of parts and the degree to which any part or grouping dominates the whole; basically it is an aspect of unity.
Experimental findings and implications. Zajonc’s study consisted of experimentally varying situa-tional demands and measuring the ensuing differences in the structural aspects of cognitive systems. It was predicted that when subjects Ss anticipated they would be receiving information, they would exhibit cognitive systems qualitatively different from those exhibited when they anticipated they would be transmitting information.
Experimental Ss were all given copies of a letter from a job applicant to a potential employer and were told to imagine what kind of person the applicant was. After reading the letter, Ss in the transmitter group were told that they would be responsible for communicating the information obtained about the letter writer to another group of persons but that before they proceeded, they were to write down what they had learned from the letter. Ss in the receiver group were told that another group of persons had detailed information about the letter writer and would communicate this to them but that before this was done, Ss were to write down what they had learned from the letter. Special forms were distributed for writing down the information. The number of characteristics attributed to the letter writer was used as a measure of differentiation; the number of groupings into which Ss were then able to arrange these characteristics was used as a measure of complexity; measures of unity and organization were based on asking Ss to list all of the other characteristics that would change if characteristic X were changed.
“Receivers” demonstrated significantly less differentiation, complexity, unity, and organization than did “transmitters.” The nature of the task, it seems, is an important determinant of the structural aspects of cognitive systems. When it is necessary to anticipate admitting all types of information to the system, fewer specifically differentiated attributes are required than when specific information must be transmitted. And the existing attributes had better represent broad, inclusive classes of differentiated attributes, rather than many, clearly distinguishable classes.
In a second, related experiment, Zajonc demonstrated that when Ss deal with incongruent information, the major differences between receivers and transmitters decrease. Transmitters exhibit somewhat less differentiation; receivers, somewhat more differentiation and significantly more unity, when dealing with incongruent, rather than congruent, information. According to Zajonc, Susceptibility and resistance to change may be thought of in terms of the strength of forces necessary to induce changes in a given cognitive structure. In order to change a highly unified structure a strong force is necessary because the components of the structure depend on one another to a great degree. Thus a change in one attribute should result in changes in other attributes. In order to change a single attribute of a highly unified cognitive structure, a force must be applied which is capable of overcoming the resistance of not only the given attribute, but of all the other attributes on which it depends. To produce a similar change in an un-unified structure, a force is required that is capable of overcoming the resistance of the given component alone. (1960, p. 163)
In this way interdependence is seen as a barrier against change that is inconsistent with the subject’s existing cognitive structure.
Mental ability and retardation . As far back as 1935 Lewin ascribed feeble-mindedness to a relative lack of differentiation and interdependence in cognitive systems (Lewin 1935). Kounin (1941) supported Lewin’s contention with experimental demonstrations that perseveration and stereotyped responses in the feeble-minded were reflections of such cognitive variables. [See Mental Retardationfor a discussion somewhat at variance with this view.]
Gochman (1962; 1966) has extended the systemic approach to show that adaptive and flexible problem-solving behavior represents differentiated perceptual-cognitive systems in children at all levels of IQ and that differentiation is a more reliable predictor of problem-solving ability than IQ scores for fifth-grade and sixth-grade youngsters with low and average IQs (56-113). The results led to the interpretation that the person whose perceptual-cognitive system is poorly differentiated may, in fact, not even be aware of a problem which confronts him or may not be aware of a sufficient number of facets of the problem to permit him to solve it efficiently or of multiple alternatives of solution. These parts of his physical world have not become parts of his perceptual-cognitive system; less of his world has relevance for him.
On the other hand, greater differentiation of the perceptual-cognitive system increases the likelihood that the aspects of the world relevant to coping with problems will become part of the individual’s own awareness. To the extent that these aspects exist in his awareness, he will be more likely to benefit from them in his dealings with the world. Accordingly, there is support for the view that mental retardation is a reflection of lack of differentiation, and normal mental ability a reflection of the acquisition of a differentiated perceptual-cognitive system. Furthermore, differentiation of perceptual-cognitive systems appears ultimately to provide a theoretical underpinning for the nontheoretically based IQ.
Evaluation Certain comparisons are warranted of the three major works discussed. Rokeach’s and Harvey, Hunt, and Schroder’s broad perspectives on the day-to-day normal functions of the human personality are considerably different from, and strongly contrast with, Zajonc’s unique and ingenious experimental manipulation in a laboratory setting, but the value of both approaches is appreciable. The differences between Rokeach and Harvey, Hunt, and Schroder are primarily those of emphasis: Harvey, Hunt, and Schroder are considerably more concerned with development, with the effects of parental training, and with psychopathology. Rokeach is more concerned with underlying similarities in specific diverse attitude contents, with the development of a definitive technique for assessing these structural determinants, and with the application of this technique in logically derived series of studies. Yet, the phenomenon that Rokeach calls a belief or a disbelief is nearly identical to what Harvey, Hunt, and Schroder call a concept; and there is considerable agreement about relevant systemic attributes.
Some substantive differences should be noted, particularly the asymmetry of Rokeach’s belief-disbelief system. While it is important to realize that some asymmetry is warranted, i.e., that the disbeliefs are not merely the opposites of the beliefs, the disbelief system is treated as an organized set of subsystems, while the belief system is not treated in this way. It is possible that future research might conceptualize the belief system as a set of subsystems. It would make sense within the systemic framework and would be as consistent with observations of real behavior: people do not have one set of beliefs; they have many sets of beliefs, each set focused on one of many different objects or classes of objects, and each of these sets can be considered a separate subsystem, subject to analysis in terms of the same dimensions that subsystems of the disbelief subsystem are. One could then enumerate the number of different things a person believed in as one way of measuring differentiation within the belief system; it makes for a considerable difference in behavior whether there is a limited or a wide range of different contents that are held to be valid.
Methodology . While the attempt at integrating a wide variety of phenomena must be appreciated and respected, there is room for some comment on the methods generally employed in systemic approaches.
There is no highly developed, widely validated, generally accepted technique or set of techniques for assessing systemic attributes. The Dogmatism Scale represents a major step toward attaining such an instrument. Harvey and his colleagues have begun to develop the Personal Opinion Scale, as well as a sentence-completion test called This I Believe. These are designed primarily to assess the abstractness or concreteness of conceptual functioning. It remains for future systematic investigations to relate these scales specifically to structural attributes and to devise scales to assess these structural attributes specifically.
One might question the general tendency of researchers to use the existence of seemingly contradictory beliefs within the same belief system as an indication of isolation or compartmentalization. This assumption is often made in cognitive investigations within the systemic framework. The contradiction may not actually exist: in the subject’s thinking there may be actual communication, and thus logical connection, between the beliefs. Furthermore, beliefs are maintained with varying degrees of strength and universality. There is considerable difference between beliefs which are implicitly prefaced by “It is universally true in every case and under all circumstances that . . .” and beliefs which are prefaced with “In most cases, it is generally true by and large that. ...” A belief of this second type can exist in the same system with others that may appear superficially to contradict it, but the contradiction is illusory and the inferred isolation of beliefs is debatable.
An attempt has been made to isolate component systemic variables in a study of school children (Gochman 1962; 1966). Differentiation and interdependence were each measured in several different ways. Differentiation was found to be a homogeneous attribute; interdependence was not. Since the process of integration and ensuing interdependence is presumed to occur later than differentiation, quite probably the fifth graders and sixth graders studied in this research were too young to manifest interdependent perceptual-cognitive systems. This approach has the further limitation that the perceptual-cognitive systems studied were devoid of content reflecting personal and social experiences.
Zajonc’s study has—within its restricted scope— the merit of successfully isolating structural variables and measuring them with considerable ingenuity.
Scott (1963) presents a detailed and comprehensive discussion of methodological problems confronting the systemic approach. He treats both the general problems of assessing cognitive behaviors and problems specific to the assessment of properties such as differentiation and interdependence, with analyses and evaluations of different research strategies employed in each case.
Validity and scope . The merits of a systemic approach are primarily integrative: many different behavioral phenomena can be understood as manifestations of variations in systemic structural variables. Whether the systemic approach has greater validity than other general theoretical orientations is difficult to assess. It has not been subjected to wide-scale experimental tests that contrast it with other theories, but this rarely occurs in contemporary psychology.
Background for the systemic models is found in the pioneer work of Köhler (1929) and Sherif (1936) dealing with perception as a reflection of both internal and external events.
Systemic models have been used with biological phenomena by Weiss (1939, p. Ill) and with physical phenomena by Bertalanffy (1950). They have proved very useful in the less tangible area of personality, as Lewin’s work exemplified (1926-1933; 1939-1947). Campbell (1958) has applied some gestalt dimensions to social phenomena in an attempt to demonstrate the presence of social systems. Scott (1962) provides an analysis of systemic properties of cognitive and social structures and a discussion of their mutual implications. Newcomb (1961) has demonstrated the effectiveness of a systemic approach in interpersonal relations. Carlson (1956), Rosenberg (1956), and Scott (1958; 1959) have employed the concept of interdependence in their work on attitude structure, consistency, and change. An attitude that is “consistent,” i.e., that reflects communication between other cognitive components, is more stable, particularly in the face of nonrational pressures for change.
Tuckman (1966) has used the developmental process outlined by Harvey, Hunt, and Schroder as a basis for the exploring and predicting of probing and self-disclosure behaviors. His study revealed that subjects at the highest level of development, i.e., those with the most complex conceptual systems, tended to be less revealing of information about themselves generally and more apt to probe information from a friend (as opposed to an acquaintance) than were subjects at lower developmental levels.
Cognitive theories of symmetry, balance, and dissonance (e.g., Festinger 1957; Heider 1958) also reflect concern with interdependence in their assertions that there are tendencies for parts of a structure to achieve relative harmony with one another and that the existence of unharmonious components is a source of tension with behavioral implications: the tension must be reduced and harmony restored.
Such a conception of tension and tension reduction provides a way of extending systemic analysis to emotional reactions, one of the most complex and least understood behavioral phenomena. A study by Price, Harburg, and Newcomb (1966) demonstrates how, with some modification, balance theories might provide a model for understanding affective processes.
While the systemic approach at present is particularly relevant to studying beliefs, attitudes, attitude change, intelligence, adaptability, and other cognitive behaviors, its potential applicability to other areas of behavior opens a broad research horizon. The development of future research in systemic analysis is dependent upon isolating component variables, measuring them by a variety of operations, and making specific predictions about their relations to behavior. Essentially, progress will reflect how well psychologists themselves have differentiated and made interdependent their conceptions of the phenomena with which they deal.
David S. Gochman
Allport, Gobdon W. 1960 The Open System in Personality Theory. Journal of Abnormal and Social Psychology 61:301-310.
Bertalanffy, Ludwig Von 1950 The Theory of Open Systems in Physics and Biology. Science 111:23-29.
Campbell, Donald T. 1958 Common Fate, Similarity, and Other Indices of the Status of Aggregates of Persons as Social Entities. Behavioral Science 3:14-25.
Carlson, Earl R. 1956 Attitude Change Through Modification of Attitude Structure. Journal of Abnormal and Social Psychology 52:256-261.
Festinger, Leon 1957 A Theory of Cognitive Dissonance. Evanston, 111.: Row, Peterson.
Gochman, David S. 1962 System Theory and Adaptability. Ph.D. dissertation, Univ. of Colorado.
Gochman, David S. 1966 A Systemic Approach to Adaptability. Perceptual and Motor Skills 23:759-769.
Harvey, O. J.; Hunt, D. E.; and Schroder, H. M. 1961 Conceptual Systems and Personality Organization. New York: Wiley.
Heider, Fritz 1958 The Psychology of Interpersonal Relations. New York: Wiley.
Koch, Sigmund (editor) 1959 Psychology: A Study of a Science. Volume 2. New York: McGraw-Hill.
KÖhler, Wolfgang 1929 Gestalt Psychology. New York: Liveright. → A paperback edition was published in 1947 by New American Library.
Kounin, Jacob S. 1941 Experimental Studies of Rigidity. Parts 1-2. Character and Personality 9:251-282. → Part 1: “The Measurement of Rigidity in Normal and Feeble-minded Persons.” Part 2: “The Explanatory Power of the Concept of Rigidity as Applied to Feeble-mindedness.”
Levy, Jacques M.; and Rokeach, Milton 1960 The Formation of New Perceptual Systems. Pages 257-269 in Milton Rokeach, The Open and Closed Mind. New York: Basic Books.
Lewin, Kurt (1926-1933) 1935 A Dynamic Theory of Personality: Selected Papers. New York: McGraw-Hill.
Lewin, Kurt (1939-1947)1963 Field Theory in Social Science: Selected Theoretical Papers. Edited by Dor-win Cartwright. London: Tavistock.
Mikol, Bernard 1960 The Enjoyment of New Musical Systems. Pages 270-284 in Milton Rokeach, The Open and Closed Mind. New York: Basic Books.
Miller, James G. 1955 Toward a General Theory for the Behavioral Sciences. American Psychologist 10: 513-531.
Newcomb, Theodore M. 1961 The Acquaintance Process. New York: Holt.
Parsons, Talcott 1959 An Approach to Psychological Theory in Terms of the Theory of Action. Volume 3, pages 612-711 in Sigmund Koch (editor), Psychology: A Study of a Science. New York: McGraw-Hill.
Price, Kendall O.; Harburg, Ernest; and Newcomb, Theodore M. 1966 Psychological Balance in Situations of Negative Interpersonal Attitudes. Journal of Personality and Social Psychology 3:265-270.
Rokeach, Milton 1960 The Open and Closed Mind: Investigations Into the Nature of Belief Systems and Personality Systems. New York: Basic Books.
Rokeach, Milton; and Vidulich, Robert N. 1960 The Formation of New Belief Systems: The Roles of Memory and the Capacity to Entertain. Pages 196-214 in Milton Rokeach, The Open and Closed Mind. New York: Basic Books.
Rokeach, Milton et al. 1960a On Loyalty to and Defection From a Belief System: An Experimental Analogy. Pages 243-256 in Milton Rokeach, The Open and Closed Mind. New York: Basic Books.
Rokeach, Milton et al. 1960b On Party-line Thinking: An Experimental Analogy. Pages 225-242 in Milton Rokeach, The Open and Closed Mind. New York: Basic Books.
Rosenberg, Milton J. 1956 Cognitive Structure and Attitudinal Affect. Journal of Abnormal and Social Psychology 53:367-372.
Scott, William A. 1958 Rationality and Non-rationality of International Attitudes. Journal of Conflict Resolution 2:7-16.
Scott, William A. 1959 Cognitive Consistency, Response Reinforcement, and Attitude Change. Soci-ometry 22:219-229.
Scott, William A. 1962 Cognitive Structure and Social Structure: Some Concepts and Relationships. Volume 1, pages 86-118 in Decisions, Values and Groups. Edited by D. Willner. New York: Pergamon.
Scott, William A. 1963 Conceptualizing and Measuring Structural Properties of Cognition. Pages 266-288 in O. J. Harvey (editor), Motivation and Social Interaction. New York: Ronald.
Sherif, Muzafer (1936) 1965 The Psychology of Social Norms. New York: Octagon.
Tuckman, Bruce W. 1966 Interpersonal Probing and Revealing and Systems of Integrative Complexity. Journal of Personality and Social Psychology 3:655-664.
Weiss, Paul 1939 Principles of Development. New York: Holt.
Zajonc, Robert B. 1960 The Process of Cognitive Tuning in Communication. Journal of Abnormal and Social Psychology 61:159-167.
Systems analysis is the process of examining a business situation in order to develop a new system, or improve the current system, for the purpose of solving a problem or taking advantage of an opportunity. In present-day usage, systems analysis is geared towards establishing a new computer system or updating an older computer system. Systems analysis, however, goes beyond computer or database programming. It involves understanding activities and interactions within an organization.
Before the development of any system can begin, a project proposal is prepared by the users of the potential system and/or by systems analysts and submitted to an appropriate managerial structure within the organization. The project proposal is the attempt to respond to or take advantage of a particular situation and is an essential element for correctly launching the system analysis. Although there are no hard and fast rules as to the form and content of the project proposal, the proposal should address the following points:
- The specifics of the business situation or problem
- The significance of the problem to the organization
- Alternative solutions
- The possible use of computer information systems to solve the problem
- The various people interested in or possessing knowledge relevant to the problem
System projects that are to be shared by a number of departments and users are usually approved by a committee rather than an individual. A project proposal is submitted to a committee that determines the merits of the proposal and decides whether or not to approve it. The
committee is made up of people from various functional areas of the organization who have an interest in the operation and information of the proposed system.
THE SYSTEMS DEVELOPMENT LIFE CYCLE
The systems development life cycle (SDLC) describes a set of steps that produces a new computer information system. The SDLC is a problem-solving process. Each step in the process delineates a number of activities. Performing these activities in the order prescribed by the SDLC will bring about a solution to the business situation. The SDLC process consists of the following phases:
- Preliminary investigation—the problem is defined and investigated.
- Requirements definition—the specifics of the current system as well as the requirements of the proposed new system are studied and defined.
- Systems design—a general design is developed with the purpose of planning for the construction of the new system.
- Systems development—the new system is created.
- System installation—the current operation is converted to run on the new system.
- Systems evaluation and monitoring—the newly operational system is evaluated and monitored for the purpose of enhancing its performance and adding value to its functions.
Looping back from a later phase to an earlier one may occur if the need arises.
Each phase has a distinct set of unique development activities. Some of these activities may span more than one phase. The management activity tends to be similar among all phases.
The SDLC is not standardized and may be unique to a given organization. In other words, the names and number of phases may differ from one SDLC to the next. However, the SDLC discussed here is, to a large extent, representative of what is typically adopted by organizations.
At each phase certain activities are performed; the results of these activities are documented in a report identified with that phase. Management reviews the results of the phase and determines if the project is to proceed to the next phase.
The first two phases of the SDLC process constitute the systems-analysis function of a business situation. The following discussion will concentrate on phase one (preliminary investigation) and phase two (requirements definition) of the outlined SDLC process.
The first phase of the systems development life cycle is preliminary investigation. Due to limited resources an organization can undertake only those projects that are critical to its mission, goals, and objectives. Therefore, the goal of preliminary investigation is simply to identify and select a project for development from among all the projects that are under consideration. Organizations may differ in how they identify and select projects for development. Some organizations have a formal planning process that is carried out by a steering committee or a task force made up of senior managers. Such a committee or task force identifies and assesses possible computer information systems projects that the organization should consider for development. Other organizations operate in an ad hoc fashion to identify and select potential projects. Regardless of the method used, and after all potential projects have been identified, only those projects with the greatest promise for the well-being of the organization, given available resources, are selected for development.
The objective of the systems-investigation phase is to answer the following questions: What is the business problem? Is it a problem or an opportunity? What are the major causes of the problem? Can the problem be solved by improving the current information system? Is a new information system needed? Is this a feasible information system solution to this problem?
The preliminary-investigation phase sets the stage for gathering information about the current problem and the existing information system. This information is then used in studying the feasibility of possible information systems solutions.
It is important to note that the source of the project has a great deal to do with its scope and content. For example, a project that is proposed by top management usually has a broad strategic focus. A steering committee proposal might have a focus that covers a cross-function of the organization. Projects advanced by an individual, a group of individuals, or a department may have a narrower focus.
A variety of criteria can be used within an organization for classifying and ranking potential projects. For planning purposes, the systems analyst—with the assistance of the stakeholders of the proposed project—collects information about the project. This information has a broad range and focuses on understanding the project size, costs, and potential benefits. This information is then analyzed and summarized in a document that is then used in conjunction with documents about other projects in order to review and compare all possible projects. Each of these possible projects is assessed using multiple criteria to determine feasibility.
The feasibility study investigates the problem and the information needs of the stakeholders. It seeks to determine the resources required to provide an information systems solution, the cost and benefits of such a solution, and the feasibility of such a solution. The analyst conducting the study gathers information using a variety of methods, the most popular of which are:
- Interviewing users, employees, managers, and customers
- Developing and administering questionnaires to interested stakeholders, such as potential users of the information system
- Observing or monitoring users of the current system to determine their needs as well as their satisfaction and dissatisfaction with the current system
- Collecting, examining, and analyzing documents, reports, layouts, procedures, manuals, and any other documentation relating to the operations of the current system
- Modeling, observing, and simulating the work activities of the current system
The goal of the feasibility study is to consider alternative information systems solutions, evaluate their feasibility, and propose the alternative most suitable to the organization. The feasibility of a proposed solution is evaluated in terms of its components. These components are:
- Economic feasibility—the economic viability of the proposed system. The proposed project's costs and benefits are evaluated. Tangible costs include fixed and variable costs, while tangible benefits include cost savings, increased revenue, and increased profit. A project is approved only if it covers its cost in a given period of time. However, a project may be approved only on its intangible benefits such as those relating to government regulations, the image of the organization, or similar considerations.
- Technical feasibility—the possibility that the organization has or can procure the necessary resources. This is demonstrated if the needed hardware and software are available in the marketplace or can be developed by the time of implementation.
- Operational feasibility—the ability, desire, and willingness of the stakeholders to use, support, and operate the proposed computer information system. The stakeholders include management, employees, customers, and suppliers. The stakeholders are interested in systems that are easy to operate, make few, if any, errors, produce the desired information, and fall within the objectives of the organization.
This phase is an in-depth analysis of the stakeholders' information needs. This leads to defining the requirements of the computer information system. These requirements are then incorporated into the design phase. Many of the activities performed in the requirements definition phase are an extension of those used in the preliminary investigation phase. The main goal of the analyst is to identify what should be done, not how to do it. The following is a discussion of the activities involved in requirements definition.
Information Needs of the Stakeholders . Analysis of the information needs of the stakeholders is an important first step in determining the requirements of the new system. It is essential that the analyst understands the environment in which the new system will operate. Understanding the environment means knowing enough about the management of the organization, its structure, its people, its business, and the current information systems to ensure that the new system will be appropriate.
The Current Information System . A comprehensive and detailed analysis of the current system is essential to developing a high-quality replacement information system. The analyst should understand and document how the current system uses hardware, software, and people to accept and manage input data and to convert such data into information suitable for decision making. The documentation should be detailed and complete. For example, the analyst should assess the quality of input and output activities that form the user's interface. In addition, the volume and timing of such activities may be documented.
The Capabilities of the New Computer Information System . Functional requirements include the necessary hardware and software configurations along with the appropriate human resources. Specific functional requirements often include the following:
- User interface requirements—the input and output needs of the user that must be provided for by the new computer information system. These needs include layouts and definitions of input and output, volume, frequency, origination of input, and destination for reports.
- Processing requirements—the activities required for converting input into output, including calculations, decision rules, database operations, and other processing operations. In addition, requirements concerning capacity, throughput, turnaround time, response time, and the system's availability time, are established.
- Storage requirements—the organization, content, and size of databases, and types and frequency of updates and inquiries. Furthermore, backup procedures and the length of time and rationale for retention of backups are delineated.
- Control requirements—the accuracy, validity, security, and adaptability requirements for the system's input, processing, output, and databases. Crash recovery and auditing requirements of the organization are further specified in this stage.
The analysis team, at the end of this phase, produces a document containing the functional requirements of the new computer information system. Additionally, the document contains preliminary schedules and a budget for the next phase. The task force or committee responsible for the project studies the document for the purpose of approving or not approving the work of the analysis team. In addition, the analysis team provides the committee with a demonstration. In essence, the analysis team walks the committee members, step by step, through the requirements definition phase. If the committee approves this phase, then the analysis team is funded and given the go-ahead to proceed to the next phase. However, if the committee does not approve this phase, then either the project is canceled or, after appropriate modifications, the analysis team resubmits a new document to the committee.
A walk-through starts with a description of the project. From this point, the analysts delineate a set of well-defined goals, objectives, and benefits of the computer information system. Following that, the budgets and staffing requirements are articulated and the plans are shared with the committee. Specific, planned tasks are compared to actual accomplishments, and deviations, if any, are noted and accounted for. The plans for asset protection and business control are reviewed with the committee members. Finally, the analysts seek the committee's approval of the objectives, plans, time table, and budget for the next phase—systems design.
In summary, systems analysis is an essential starting point in the development of computer information systems projects. An organization generally follows a development pattern set up to meet its needs. Regardless of which methodology an organization uses, the objective of systems analysis is to fully understand the current environment and future requirements of a computer information systems project.
SEE ALSO Business Process Reengineering; Data Processing and Data Management; Management Information Systems; Open and Closed Systems; Systems Design
Dennis, Alan, Barbara Wixom, Robey Roth. Systems Analysis and Design: An Applied Approach. 4th ed. Hoboken, NJ: Wiley, 2008.
Kendall, Kenneth, and Julia Kendall. Systems Analysis and Design. 7th ed. Englewood Cliffs, NJ: Prentice-Hall, 2007.
McLeod, Raymond, Jr., and George Schell Sumner. Management Information Systems. 9th ed. Englewood Cliffs, NJ: Prentice-Hall, 2004.
Valacich, Joseph, Joey George, and Jeffrey Hoffer. Essentials of Systems Analysis and Design. 4th ed. Englewood Cliffs, NJ: Prentice-Hall, 2009.
System analysis is a broad, technical area focused on the creation, enhancement, and trouble-shooting of systems for users. These can be data, information, or knowledge systems. The purpose of these systems is to provide an understanding of what is going on in a particular environment. Sensors, including radar , sonar , and satellites , for example, are components of systems that provide specific knowledge about the physical world. Sensors in the home can warn residents that someone is breaking in. Telephones are part of a system that brings police assistance when one dials 911. People have computers they use to perform a number of tasks—from writing term papers or diaries to communicating with people they have never met via "messaging" software programs. There are decision support systems that help people use computers to solve problems, and communication systems that tell people what is happening around them. System analysis is used to design, enhance, and fix problems in all of these systems.
System Analysis Methods
System analysis is creative work. The systems analyst can be considered an artist, an information scientist, and an engineer, all in one. The work begins with thinking about how to accomplish something. System analysis can be considered to have three primary functions, each of which is related to the others. First, system analysis is done to fix something that has gone wrong and to help one understand why there is a problem. Second, analysis is used to figure out how to do something more easily and less expensively as new technologies become available. Third, system analysis is done to help design a system that can accommodate future circumstances, such as anticipated events that are not being experienced now, but that might need to be dealt with in the future.
Each system has components that perform certain functions and, when put together, do a particular job, or serve a specific purpose. System analysts are trained to ensure that each component or function of a system—whether people, tools (technology), or procedures— is acting properly. If a system fails, the system analyst tries to find out how and why. The systems analyst then communicates with the designer about the factors found to be related to the failures in order to find a solution and avoid future problems. One way to consider system analysis is that it is a process that first identifies the factors that influence and lead to system breakdown, and then identifies ways to repair or avoid breakdowns.
Systems have a life cycle. They become operational, they age, and they become obsolete. Typewriters have given way to computers. People still use telephone booths but cell phones are replacing them because they allow people to make phone calls and do many other things better and faster. As new tools and technologies become available, people want to use them. The system analyst examines and studies how technology and people can be placed and used in current systems, and figures out how existing systems could benefit from all the new ideas and inventions that are coming to market. System analysis is a process for updating or retrofitting systems, replacing old technology with new, and installing new ways of doing things.
System analysis also considers situations or events where no existing system may yet be available to deal with them, such as biological warfare, or global warming, and analysts work to find ways to better address other large-scale events that affect the fabric of human life. Systems do exist that can respond to events such as hurricanes, tornadoes, floods, and health epidemics. Yet the need to anticipate unpredictable global circumstances—economic, political, medical, or environmental—demands new, creative approaches to minimize the potential damage to lives and property. System analysis is a way to explore how new situations or challenges can be met.
The Relationship Between Analysis and Design
A system analyst investigates; a system design specifies. The analyst asks questions such as who, what, when, where, why, and how. The system designer finds the best procedures, tools, and human skills to meet the needs and requirements of people and organizations. System analysis and system design work best when analyst and designer work together. The analysis component helps reduce the likelihood that design and technology will drive and influence the problem-solving process. If design precedes and directs analysis, there is a good chance that a given system may not be what the user needs or requires. Basements are full of technologically interesting gadgets that people buy and rarely use because they never needed them in the first place—no matter how intriguing the design, they were not designed specifically to meet the identified needs or requirements of the user!
Needs and Requirements
People are born with needs and requirements. A need is a state of being. Requirements are the things that meet these needs. People are hungry or thirsty; they need to feel well; they need shelter from environmental conditions and circumstances. They need to make a living; they need to feel as if their actions have some meaning. To address these varied needs, they require food, water, air, shelter, and other resources. But the resources, or requirements, applied to address these needs in a tropical environment, for example, would be inappropriate in an arctic climate. Individual human characteristics such as personality or physical limitations can also influence the appropriateness of certain resources being applied to meet these needs. Requirements to meet needs vary from situation to situation. The same logic applies to system analysts. They must learn how to match up needs and requirements so that a system will actually function effectively in its particular environment.
Human needs may be physical, psychological, intellectual, emotional, or social. They can generally be identified only through a careful process of examination and investigation. Needs and requirements can be difficult to sort out and obtain from users because dictionaries and people's language habits lead them to ignore the distinctions. But the success of any system depends on the skills of a system analyst to recognize these distinctions and gather the correct needs and requirements information before trying to engineer a system.
The System Development Cycle
The system development cycle consists of the steps taken for the conceptualization and engineering of a system. There are several ways to represent or describe the system development cycle. One is to show the analysis process in a series of blocks in hierarchical (top-down) or horizontal line (timeline) form. Another way is to show the system analysis process as a circle of operations, namely, requirement analysis, specifications, design, implementation, testing, and maintenance. Another approach is to regard the process as representing a waterfall cycle—one step flowing into another in a continuous stream. All share a common property of sequencing. System analysis is a step-by-step procedure: each step follows or interacts with the others, and all are directed toward meeting the objectives stated by the intended user of the system.
The Conceptual Stage
Once the initial information gathering stage is completed—that is, user needs and requirements have been identified, and the parties involved in a system design process have agreed on certain parameters of time, money investment, and expected outcomes—the system analyst begins conceptualizing the problems to be solved and the possible solutions to be applied.
One of the first steps is known as event analysis. The system analyst will engage in a detective-like process of investigating the properties and attributes of an event, or of a series of events that make up the problem to be solved. For each event, the system analyst creates a model, which is a tool in the analytical process. A model provides a view, a mental and physical picture, of the total system, explaining how the various parts of the system are structured and how they work together.
A prevailing model for system analysis is to consider three functions: namely, input, throughput, and output. What goes into the system? What happens to it? What is the outcome? These three dimensions of the analysis are applied to each component or event within the system and to the total system, as well, in its final configuration.
Input refers to the data that are acquired, through human or machine means, as part of an event in which the system is engaged. At the throughput stage, these data are transmitted to a processor, which can again be human, machine, or both. The data may then be modified, organized, stored for retrieval, or used in problem-solving and decision-making activities; whatever happens to it during processing is part of the throughput function. The output of a system, or of an event within the system, is the result of the processing steps being applied to the data originally entered into the system. The system analyst's model should account for everything that happens from the time data enters the system to the point at which the end results are achieved.
Throughout the system analysis process, analysts test the ideas and conclusions that arise. Often the model of the system provides the basis for creating a prototype of what is being studied. Although prototyping is a common exercise of designers, who use prototypes to test system configurations and hardware-software specifications, it is also a method for the system analyst to refine the conclusions of the analysis before design decisions are made.
Documentation is a necessary part of system analysis. Documentation means that all actions taken in the process of analyzing a system are recorded. This provides an enduring record of everything that has taken place and all the thoughts or ideas generated throughout the process. This includes both the individual work and testing done by a particular system analyst, and the group work, such as brain-storming, that is usually part of the overall process of analyzing and designing a system. Documentation objectifies the system analysis process. Thorough documentation can help reduce the amount of guessing that goes into solving certain system problems. It helps analysts keep track of what has been tried, and when, and under what specific circumstances. It can be useful in future work on a particular system, and it is a practice through which the process and outcomes of system analysis can be improved and validated.
see also Decision Support Systems; Design Tools; Information Systems; Office Automation Systems; Systems Design.
Coad, Peter, and Edward Yourdon. Object-oriented Analysis, 2nd ed. Englewood Cliffs, NJ: Yourdon Press, 1991.
Debons, Anthony. Information Science: Forty Years of Teaching. Philadelphia: Foundation for Information Technology Education, 2000.
Miller, James Grier. Living Systems. New York: McGraw-Hill, 1978.
Osborne, Larry N., and Margaret Nakamura. Systems Analysis for Librarians and Information Professionals, 2nd ed. Englewood, CO: Libraries Unlimited, 2000.
Sage, Andrew P., and William B. Rouse, eds. Handbook of Systems Engineering and Management. New York: John Wiley and Sons, 1999.
Satzinger, John W., Robert B. Jackson, and Stephen D. Burd. Systems Analysis and Design in a Changing World. Cambridge, MA: Course Technology, 2000.
System analysis, system inquiry, or systems theory is the study of the interdependence of relationships. A system is composed of regularly interacting or interrelating parts that, when taken together, form a new whole with properties distinct from its constituent parts. Systems are seen to be structurally divisible but functionally indivisible wholes with emergent properties. Central to system analysis is the recognition that the structure of any system—the many interlocking, sometimes time-delayed, sometimes circular interrelationships among its components—is often just as important, if not more important, than the individual components themselves in determining the system’s behavior.
Systems are characterized by complexity, a set of boundaries, and the ability to regenerate. Complexity refers to a large number of densely connected parts and multiple levels of embeddedness and entanglement. A system is defined by a set of parametric conditions or boundaries that delimit it or set it apart from its environment. No system can be completely closed or else we could not perceive it; there are only varying degrees of closure set by boundaries. A system regenerates itself through the self-reproduction of its own elements and of the network of interactions that characterize them in a process known as autopoiesis. Thus an autopoietic system renews, repairs, and replicates or reproduces itself in a flow of matter and energy.
Systems can change through an evolutionary process with a tendency toward greater structural complexity and organizational simplicity, more efficient modes of operation, and greater dynamic harmony. Change is enacted through a process of feedback where information concerning the adequacy of the system, its operation, and its outputs are introduced into the system. Negative feedback signals that there is a discrepancy between what the system produces and what it should produce. It tells us that we should change something in the system so that we can reduce the deviation from the norms stated in the system’s output model. Positive feedback signals that the whole system should change, that we should increase the deviation from the present state and change the output model. Functionalism is based on this adaptation. To survive or maintain equilibrium with respect to its environment, any system must to some degree adapt to that environment, attain its goals, integrate its components, and maintain its latent pattern, a cultural template of some sort.
A system can be ordered as a hierarchy or a heterarchy. A hierarchy is a vertical arrangement of entities within systems and their subsystems. A heterarchy is an ordering of entities without a single peak or leading element, and which element is dominant at a given time depends on the total situation. Systems may be understood through holism, where attention is focused on the emergent properties of the whole rather than on the behavior of the isolated parts, or reductionism, where phenomena are understood by breaking them down into their smallest possible parts.
Several fields utilize system analysis. Cybernetics, chaos theory, and social dynamics, for example, are among the disciplines that apply system analysis. Some areas of education and environmental sustainability also utilize system analysis. The systems framework is also fundamental to organizational theory, as organizations are complex, dynamic, goal-oriented processes; in anthropological studies, notably those incorporating positive and negative feedback; and in cybernetics, catastrophe theory, chaos theory, and complexity theory, all of which have the common goal of explaining complex systems that consist of a large number of mutually interacting and interrelated parts. In biology the living systems theory of James Grier Miller is a general theory about how all living systems work, maintain themselves, develop, and change. Living systems can be as simple as a single cell or as complex as a supranational organization such as the European Union. In sociology the structural functionalism of Talcott Parsons argues that the largest system is “the action system” consisting of interrelated behaviors of individuals, embedded in a physical-organic environment with others, with each part in a social system arranged in a pattern of interpenetrating relationships influenced by a socializing culture that constitutes standards and channels for guiding actions. Societies (which are highly complex), like systems and organisms, have functional needs that must be met if the society is to survive. Parsons says that all societies have four basic needs: adaptation, goal attainment, integration, and pattern maintenance (i.e., inertia, latency, or self-maintenance).
The deterministic or restrictive nature of systems is addressed by aspects of structuralism. Structuralism rejects the concept of human freedom and choice and focuses instead on the way human behavior is determined by various structures. Thomas Kuhn, for example, notes how scientists operate under a standard praxis of “normal science,” deviating from a standard “paradigm” only in instances of irreconcilable anomalies. In political science the structural realism of Kenneth Waltz describes international politics as a systemic interaction of states within an anarchical environment. States first seek survival and are socialized by an anarchical environment to act and react based on threats to survival and to form self-help alliances with like units. The system effects described by Robert Jervis notes how political relations among states in a system, similar to biological interactions among cells and other scientific phenomena, can produce effects different from the sum of individual actions.
SEE ALSO Catastrophe Theory; Chaos Theory; Cyberspace; Functionalism; Heterarchy; Hierarchy; Models and Modeling; Parsons, Talcott; Sociology, Post-Parsonian American; Stability, Political; Structuralism; Waltz, Kenneth
Banathy, Bela. 1996. Designing Social Systems in a Changing World. New York: Plenum.
Bateson, Gregory. 1979. Mind and Nature: A Necessary Unity. New York: Dutton.
Bausch, Kenneth C. 2001. The Emerging Consensus in Social Systems Theory. New York: Kluwer Academic/Plenum.
Bertalanffy, Ludwig von. 1968. General System Theory: Foundations, Development, Applications. New York: George Braziller.
Churchman, C. West. 1968. The Systems Approach. New York: Delacorte.
International Society for System Sciences (ISSS). http://www.isss.org/
Jantsch, Erich. 1980. The Self-Organizing Universe: Scientific and Human Implications of the Emerging Paradigm of Evolution. New York: Pergamon.
Jervis, Robert. 1997. System Effects: Complexity in Political and Social Life. Princeton, NJ: Princeton University Press.
Kahn, Herman. 1956. Techniques of System Analysis. Santa Monica, CA: Rand Corporation.
Miller, James Grier. 1978. Living Systems. New York: McGraw-Hill.
Parsons, Talcott.  1967. The Structure of Action. New York: Free Press.
Parsons, Talcott. 1977. Social Systems and the Evolution of Action Theory. New York: Free Press.
Parsons, Talcott, and Neal J. Smelser. 1956. Economy and Society. Glencoe, IL: Free Press.
System Dynamics Society. http://www.systemdynamics.org/
Waltz, Kenneth. 1979. Theory of International Politics. Reading, MA: Addison-Wesley.
Weinberg, Gerald M. 1975. An Introduction to General Systems Thinking. New York: Wiley-Interscience.
During the normal course of doing business, companies engaging in e-commerce are faced with a wide variety of challenges or problems. These vary depending on many factors, including the way a company is structured to profit in the marketplace, the industry in which it operates, legal requirements for tracking or reporting information, and more. A process called systems analysis is used to create new solutions to these challenges, or to improve solutions that already are in place. Solutions normally involve the use of information systems or software applications, which are developed with specific issues or challenges in mind.
Information systems themselves can vary in form and type, and involve various elements like databases, software applications, and computer hardware. Among the different kinds of information systems are transaction processing systems (TPS), which involve the movement of data as it relates to things like payment or inventory; management information systems (MIS), which include databases that managers rely on to evaluate and organize operations; decision support systems (DSS), which are similar to MIS but focus more heavily on supporting a user's actual decision; and office automation systems (OAS), which allow employees to manipulate and distribute data with productivity tools like word processors, e-mail, electronic scheduling, and spreadsheet programs.
Systems like MIS and DSS, which helped e-commerce executives make faster or better-informed decisions, were especially valuable in the early 2000s. At that time, the pace of Internet business was very quick. Combinations of different conditions—including the impact of new technologies, customer demands, government regulations, legal issues, and pressure from investors—changed frequently and could impact the success or failure of a company or industry in a short period of time. TPS systems were another example of critical e-commerce systems. The ability to engage in fast, secure transactions and integrate them with many different areas of an enterprise (including accounting and inventory) was central to its success.
Professionals generally known as systems analysts are responsible for understanding what a company's business challenges are, how they change over time, and how these issues translate to or affect systems. More specifically, systems analysts evaluate data input, flow, processing, storage, and output as related to business challenges. To accomplish this, analysts work with an organization's management team as well as those who currently or eventually will use the solution or application and engage in information gathering and problem solving. As explained in the Journal of Systems Management, "With this information, the systems analyst, working with other MIS personnel, defines the requirements which are used to modify an existing system, or to develop a new system. The systems analyst identifies and evaluates alternative solutions, makes formal presentations, and assists in directing the coding, testing, training, conversions, and maintenance of the proposed system."
HOW SYSTEMS ANALYSIS WORKS
Ideally, system analysis begins by developing an outline or map of the organization and how it functions. According to the Journal of Systems Management, "A Map lets management see how functions are performed and an Analyst see what to do first and where to go afterward. A Map may suggest where productivity improvements might be made and even where an analysis could begin. For example, there may be duplication of effort, or effort not tied to the rest of the business or even similar effort under split management leadership."
After developing a clear picture of a business and the situation in which it operates, analysts draft and present project proposals prior to the beginning of the systems analysis process. Because e-commerce can involve many different business or operating units, especially within a large organization, system projects often require the ultimate approval of a committee, rather than just one person. Included in a good proposal are several different elements, including a detailed description or explanation of the business problem or situation; an explanation of why the problem is important to the organization; several possible solutions; how computerized information systems might be used as a part of a solution; and a list of the individuals who have an interest in or knowledge about the problem or situation.
There are six problem-solving steps involved in producing a new computer information system, collectively known as the systems development life cycle (SDLC). Each step or phase involves different tasks, which may take place over several phases. Additionally, it may be necessary to revert to a previous phase during the analysis process. Although the following is a representative overview of the systems analysis process used by many organizations, much like the general title of systems analyst, the SLDC process may vary from company to company. The first step in SDLC is preliminary investigation, in which the problem is defined and investigated. Next, characteristics of an existing system are evaluated, along with the requirements of the proposed system. A general systems design is then developed, followed by the creation of the new system. System installation, which involves replacing an existing system with the new one, then occurs. Finally, systems evaluation and monitoring take place, with the ultimate goal of enhancing performance and increasing functionality.
One potential problem with any system is that it may become more cumbersome than the problem it is trying to solve, in terms of the amount of resources required to properly maintain and develop it. A system may lead to the development of new problems that, when considered in sum, make its ultimate development a bad idea. Systems analysts often conduct feasibility studies to review factors like the problem or situation itself; costs versus benefits; the needs of users; and the resources required to provide a solution. Depending on the situation, this information may be collected in a number of different ways, including interviews with managers, customers, users, and other employees; questionnaires; monitoring or observing users of existing systems; collecting and reviewing different manuals, reports, and documents; and sometimes simulation or modeling of existing systems. The ultimate goal is to determine which solutions are worth pursuing, and of those, which hold the most potential for the organization.
Bauer, Michael W. "The Very Beginning of Analysis and Development: The Map." Journal of Systems Management, December 1992.
Farah, Badie N. Business Information Systems: Development and Implementation 2nd ed. Needham Heights, MA: Simon and Schuster, 1996.
Kendall, Kenneth E. and Julie E. Kendall. Systems Analysis And Design. 1999. Upper Saddle River, New Jersey: Prentice Hall.
Misic, Mark. "The Skills Needed by Today's Systems Analysts." Journal of Systems Management, . May/June 1996.
Schuptheis, Robert, and Mary Sumner. Management Information System. 4th ed. Chicago: Richard D. Irwin, 1998.
Stair, Ralph M. Principles of Information Systems: A Managerial Approach. 2nd ed. Boston: Boyd and Fraser Publishing Company, 1996.
"System Analysis & Design." Techencyclopedia, May 7, 2001. Available from www.techweb.com.
"Systems Analyst." Ecommerce Webopedia, May 9, 2001. Available from e-comm.webopedia.com.
Zwass, Valdimir. Foundations of Information Systems. Chicago: Richard D. Irwin, 1998.