Rational versus soft management in complex software: lessons from flight simulation

This paper compares actual (A-type) software processes at work in flight simulation, one example of a software-intensive complex product system, with ideal, rational (B-type) processes as contained in company manuals, tools and procedures. The aim is to identify the causes and consequences of divergencies between A- and B-type processes in a complex product, and to draw implications for theory and practice. The paper also develops a simple partial model to show what A-type processes actually "look like" in practice. The evidence indicates that in response to industrial turbulence, uncertainty, technical complexity and difficulties in capturing user requirements, software engineers and project managers fall back heavily on A-type informal systems and "soft" management becomes essential to project success. Success under such circumstances, B-type rational systems are inadequate to the task at hand and soft factors such as goodwill, negotiation skills, rule breaking and informal communications become essential to project success. These findings contrast with most approaches to software engineering which try to impose highly rational processes and tend to ignore soft issues. The paper also indicates potential benefits of the rational approach, sometimes overlooked by contemporary organisational scholars, by arguing that it is the manner in which B-type processes are developed and implemented which leads to divergencies and difficulties, rather than the processes themselves which are an essential part of orderly progress. While the findings may not be relevant to more simple products and tasks, other complex product systems may well face similar divergencies and project management challenges.