La notion de « système » est largement utilisée dans le domaine des relations industrielles, et cela, depuis fort longtemps. Elle implique un mode de reconstitution de la réalité différent de l'approche causale linéaire sur laquelle se fonde la méthodologie de la recherche empirique. Au lieu d'accentuer les différences entre ces deux modes de réflexion, les auteurs de cet article tentent de démontrer qu'ils sont complémentaires pour l'acquisition des connaissances dans le domaine des sciences humaines.
In the field on human sciences, the researcher is constantly faced with numerous and difficult methodological problems. Moreover, the methods and techniques, on which he can count, often come from other disciplines and their application to human sciences is not without creating some difficulties.Particularly in the case of human sciences one can notice in the passed years, the proliferation of a certain literature said to be « systemic ». Such a situation interested the researchers of those fields who, generally, must cope with very complex phenomena and, consequently, are in search of methods better adapted to their problems and capable of maintaining a high degree of scientific exactness.However, a survey of this literature is worrysome. Beside the writing of which only the title is systemic, it is remarkable to notice the atmosphere of near-religion surrounding this literature: the reader has the impression to be in presence of a second« Discours de la Méthode ». This type of pro-systemic literature has raised many categorical antagonists (D.C. Phillips, 1969, 1972; F. Thayer, 1972). This literature is criticized because of its unpreciseness, its prejudice toward the scientific method, the low predictive value of the models conceived from the approach explosed in this literature. Some will even pretend that the Systems approach has nothing to offer that has not been already offered by the scientific method.Is the traditional scientific method obsolete? Is it opposed to the Systems approach? If such is the case, where can we find such an opposition? On the other hand, if there is no real opposition between those two approaches, is it possible that they could be complementary and/or have specific fields of application? Such questions are at the origin of this paper.This paper does not pretend to give a complete and definitive answer to those questions. It merely tries to gather around two main topics (The concept of « connection » and of « reconstruction of reality ») some elements of an answer scattered throughout literature. Moreover, an analysis of the complementarity of the two approaches is made with the help of a model inspired from cognitive psychology.The scientific method is born from the preoccupation of certain men facing a mode of knowledge acquisition favoring almost exclusively exploration and cristalisation, using a mode of knowledge transmission based on intellectual authoritarianism. The mode of acquisition of this era (and this was the problem facing Descartes) abandoned the rigour of the realisation and specification phases. The scientific method has had so much success (the theories were so numerous), that, for a long time, we mingled scientific method with method of acquisition of knowledge.This fashion for Systems approach anses from the necessity to take up, in a rigourous manner, the study of the first two dimensions of the cognitive process. Without a well devined System and a verifiable structure, a verification method loses its meaning.We must notice the fact that many of the supporters of the Systems approach were recruited in some new scientific fields, where phenomena are highly complex and finalised. These scientists were forced to acknowledge the limits of the scientific method.According to us, the Systems approach completes the scientific method. On one hand, it enlarges the universe of the phenomena that could be studied; on the other hand, it furnishes the descriptive tools permitting the formulation of a frame of reference for the investigated problems. The Systems approach constitutes a serious effort, while still unperfect, to develop a method of exploration. The scientific method and the Systems approach, together, constitute a research methodology which, for the first time, matches the totality of the steps of the cognitive process.
