This essay has been written, from a survey of the sociological literature on the topic. It aims to be a contribution to the current debate among social scientists on the place of technology as an explanatory variable in the study of administrative structure and behavior within organizations. This field of research has been mainly stimulated by J. Woodward's book,Industrial Organization : Theory and Practice, and Perrow's conceptual framework concerning the comparative analysis of organizations. Both authors have sought to show that administrative structure and behavior are determined by the technology variable. Woodward's contribution was based on empirical research, while Perrow proceeded from a theoretical model.In opposition to this hypothesis, which has been contradicted by several observations, the author submits that administrative structure is a multi-dimensionalvariable, and therefore cannot be a dependent variable of technology. If technology could explain administrative structure, this would mean that it could settle once and for all its institutional dimension represented by the division of labour between the two specific spheres of work organization : decision-making, and execution. In this regard, administrative structure can be substantially defined as the relationship between those two spheres of work described by Woodward through the use of such concepts as number of hierarchical levels, locus of decision-making, span of control, style of supervision, etc ...However, this definition implies a dynamic social process between two groups (higher management and subordinates) but this cannot be limited to the organization itself, because the distribution of rights and obligations between them must be recognized as legitimate by the subordinate group, and thus requires reference to the society itself. It is through this social process that structure and behavior are determined and have to be explained.To say that technology determines administrative structure and behavior over-looks this institutional dimension of organization which alone can give back to the parties their freedom of choice and action. This dimension implies that the parties can exercise a reciprocal « control » upon their actions.If this line of reasonning is sound, it should be possible to show that administrative structure is dependent on rational criteria of action. A. Chandler has used his study of the natural history of several large American corporations to show that « structure follows strategy ». To go even further, the author has reinterpreted research results of other authors, to show that structure appears to be dependent not only on strategy but also on administrative objectives.In the author's view, this line of reasoning and observation, while acceptable, contains an intellectual limitation because it glosses too rapidly over the institutional aspects of administrative structure. And some significant researches support this point of view showing that ideology, or power are related to administrative structure.By contending that administrative structure is a multi-dimensional variable, the author does not mean that each structure can be classified along a continuum between two distinct poles, one rational, and the other institutional. It is rather a question of emphasis. In other words, a social process of legitimation of authority is at work within any administrative structure.Some research results support this hypothesis, showing specifically that individuals adapt themselves to organizations according to characteristics not only of administrative structures — not of technology — but also of external factors that pertain to these structures, this being apparently true for any kind of organizations.One dependent variable used to specify the way individuals adapt to organization is job satisfaction (in the form of their general judgment of their job, or a general index of job satisfaction). This variable has been analysed by several students in relation to administrative structure, but not in relation to technology. Another variable — the quality of labour relations — has been analysed in relation to technology, but not in relation to administrative structure.Firstly, the quality of labour relations cannot be explained by technology, but by economic, politico-economic, and political factors, all of them external to the organization. Secondly, job satisfaction (as defined) has been explained either by referring to characteristics of administrative structure (what is in accord with the author's own hypothesis), or in relation to economic or political factors (such as search for more power in the society, or political radicalism). Both categories of results converge in the same direction.These results lead the author to conclude that administrative structures contain an institutional dimension whose basis is power. In other words, they can regulate the superordinate-subordinate relationships provided that they have been legkimated. But this legitimation, is always being built into the organization through administrative structure. In this sense, the analysis of power relationships as an organizational problem must take into account the relation between organization and society. The determinist use of the technology variable masks this fundamental problem.In conclusion, the author stresses the need for study of administrative structures in the light of institutional perspectives in order to pose the complex sociological problem of the evolution of industrial society. The author's own research is currently oriented toward this problem.