The limits of systems-making organizations

Abstract

This paper provides an ontology of organizational structures intended for software-intensive systems-making. An archetype is presented and scaled in stages to support larger, more complex projects. Artifacts at each stage are identified and analyzed. The first artifact analyzed is the engine of scaling, specialization. Organizations that have been scaled for the purpose of developing large complex systems will be shown to themselves become large and complex, in part for matching the problem to be solved but also for self-management. Specialization first occurs with the separation of the role of designer into designers and managers whereby the designer's role and activities can be abstracted into a limited set of objects to be manipulated by managers. Design is transformed into a model of design. The problem (or challenge) for managers is of normalization of dysfunction, whereby the model becomes both mistaken for and more valued than the object modeled. The result of having transformed essential roles and activities into objects in a hierarchical organization will be shown to (one) increase the positive effects on managing large-scale organizations, (two) increase the negative effects on design outcomes and (three) prevent the logical association of item one with item two. Disassociations incurred through specialization and normalization of dysfunction will be shown to be counter-productive in that they create and sustain high process but low sense organizations. Conclusions will be presented on the nature of large organizations, on the nature of their risk evaluations, and on the inevitability of dysfunctions and pathologies to lead not only to project failure but also to immediately forget the same