"Introduction to Systems Analysis" by F.I. Peregudov and F.P. Tarasenko (in Russian)
Systems analysis — (1) in practical terms, systems analysis is a set of methods for investigating or designing complex systems, for searching for, planning, and implementing changes intended to eliminate problems; (2) methodologically, systems analysis is applied dialectics, as it implements the ideas of dialectical materialism in application to specific practical tasks whose distinguishing feature lies in the necessity of identifying the causes of their complexity and eliminating those causes; (3) in procedural terms, systems analysis is distinguished by its interdisciplinary and transdisciplinary character and by the involvement of informal, heuristic, and expert methods as well as empirical and experimental methods, and — where possible and necessary — rigorous formal mathematical methods.
Modern systems analysis is an applied science aimed at identifying the causes of real difficulties facing the "problem owner" (usually a specific organization, institution, enterprise, or team) and at developing ways to address them. In its advanced form, systems analysis encompasses direct, practical interventions aimed at improvement. The systemic character of inquiry should not appear to be some innovation or latest achievement of science. Systemicity is a universal property of matter, a form of its existence, and hence an inherent property of human practice, including thinking. However, any activity can be more or less systemic. The appearance of a problem is a sign of insufficient systems thinking; the solution of a problem is a result of greater systemicity.
The scientific and technological revolution introduced the concepts of large-scale and complex systems, which present unique challenges. Addressing these challenges led to the development and accumulation of various techniques, methods, and approaches, which were eventually generalized into technologies for overcoming quantitative and qualitative complexities. Across different fields, these technologies and their theoretical foundations acquired distinct names: in engineering — "design methods," "methods of engineering creativity," "systems engineering"; in military and economic matters — "operations research"; in administrative and political governance — "the systems approach"; in applied scientific research — "simulation modeling," "experimental methodology," and so forth.
On the other hand, theoretical thought at different levels of abstraction reflected the systemic nature of the world and the systemic character of human cognition and practice: at the philosophical level — dialectical materialism; at the general scientific level — systemology, general systems theory, organization theory; at the natural science level — cybernetics; with the development of computing technology, informatics and artificial intelligence emerged.
By the early 1980s, it had already become evident that all these theoretical and applied disciplines form, as it were, a single stream, a "systems movement." Systems thinking had become not only a theoretical category but also a consciously recognized aspect of practical activity. Since large-scale and complex systems had, of necessity, become objects of study, management, and design, a generalization of methods for investigating and influencing them was required. A certain applied science had to emerge, serving as a "bridge" between abstract theories of systemicity and living systems practice. And it did emerge — at first, as we have seen, in different domains and under different names, but in recent years it has taken shape as a science that has come to be called "systems analysis."
The distinguishing features of modern systems analysis follow from the very nature of complex systems. Having as its goal the elimination of a problem or, at a minimum, the identification of its causes, systems analysis draws upon a broad spectrum of methods to this end, utilizing the capabilities of various sciences and practical spheres of activity. Being in essence applied dialectics, systems analysis attaches great importance to the methodological aspects of any systems investigation. On the other hand, the applied orientation of systems analysis leads to the use of all modern scientific methods — mathematics, computing technology, modeling, field observations, and experiments.
In the course of investigating a real system, one usually has to confront a wide range of problems; it is impossible for a single person to be an expert in all of them. The way out is to ensure that the person undertaking systems analysis has the education and experience necessary to recognize and classify specific problems and determine which specialists should be consulted for the continuation of the analysis. This places special demands on systems specialists: they must possess broad erudition, flexibility of thinking, and the ability to engage people in work and to organize collective activity.
Systems analysis arose in response to the demands of practice, which confronted us with the need to study and design complex systems and manage them under conditions of incomplete information, limited resources, and time constraints. To this day, debates continue about whether systems analysis can be considered a science, an art, or a "technological craft." The applications of systems analysis to problems connected with "sociotechnical" and "social" systems — that is, systems in which people play the decisive role — are debated particularly intensely. In solving such problems, not only questions of constructing and using models prove essential, not only heuristic searches for solutions to ill-structured, incompletely formalizable problems, but also purely psychological aspects of human relationships, which further "distances" systems analysis from "pure sciences" such as physics and mathematics.
Debates about the "degree of scientificity" of systems analysis stem from several reasons. The work involved in formulating problems is often underestimated. Many believe that until formal models have been constructed, the "real" work has not yet begun, and they regard the expression "to formulate a problem well is to half-solve it" as a joke. In systems analysis, attention is focused on the difficulties of problem formulation and ways to overcome them.
The most critical stage of systems analysis is the formulation of the problem situation. This stage only begins with the statement of the problem by the client. It is necessary to identify all those who will be affected by possible changes and to formulate the problems arising from those changes for each of them. The resulting set of problems, called the problematique, is the starting point for systems analysis.
Both well-formalized and ill-structured problems must be brought to a form in which they become tasks of selecting suitable means for achieving given goals. Therefore, it is first of all necessary to determine the goals. At this stage of systems analysis, what needs to be done to resolve the problem is determined (as distinct from subsequent stages, which determine how to do it).
After defining the problem, the next most important stage of the analysis is goal identification. Establishing the correct goal is more important than finding the best alternative. A less-than-optimal alternative still leads toward the goal, even if not in the optimal way. Choosing the wrong goal leads not so much to solving the problem itself as to the emergence of new problems.
In the study of systems, one has to formulate and solve both well-formalized problems expressed in mathematical terms and "ill-structured" problems expressed in natural language and solved heuristically. Overcoming complexity whose nature is connected with incomplete formalizability requires the systematic application of informal knowledge and methods. Systems analysis deliberately unites theory and practice, common sense and abstract formalization. The chief achievement of systems analysis is the development of methods for transitioning from informal to formal problems, from "black box" to "white box" models. Most of these methods are non-formalizable (in the mathematical sense), but they are sufficiently concrete and suitable for practical use and may be called not only "art" or "craft" but also technology.
Criteria are quantitative models of qualitative goals. Consequently, discrepancies between criteria and goals are inevitable, and it is very important to ensure that the transition to working with the chosen criteria actually leads to movement in the direction of the given goals.
Compiling a list of possible solutions to problems (generating alternatives) is the most intellectually demanding creative stage of systems analysis. Many methods have been developed to facilitate and accelerate this work, and command of them is essential for the systems analyst.
Algorithms for conducting systems analysis may vary. Depending on the degree of complexity of the problem being analyzed, "linear" algorithms (in the simplest cases), algorithms with loops (the more complex the system, the more loops and the more iterations are carried out in each loop), and complex "sequential" algorithms — that is, algorithms constructed in the course of the study (including those containing loops, random search, adaptation, self-organization, and so forth) — are employed.
From the practical standpoint, systems analysis is a methodology and practice of improvement-oriented intervention in problem situations. Since such interventions affect people's interests, systems analysis attaches great importance to questions in sociology, psychology, and ethics. Special methods are being developed to resolve these questions during the practical implementation of recommendations derived from a systems investigation.
If one attempts to characterize modern systems analysis once more, in a very generalized way and from a somewhat different angle, one can say that it includes such types of activity as: scientific research (both theoretical and experimental) into questions connected with the problem; design of new systems and changes in existing systems; and implementation in practice of the results obtained in the course of the analysis.
This list alone obviously deprives of meaning the debate over whether there is more theory or practice, science or art, creativity or craft, heuristics or algorithmics, philosophy or mathematics in systems research — all of these are present in it. Of course, in a specific study the ratios among these components may vary greatly. The systems analyst is prepared to bring to bear on the solution of a problem any knowledge and methods necessary for that purpose — even those he does not personally command; in that case, he is not the executor but the organizer of the research, the bearer of the goal and the methodology of the entire investigation.