The different roles are not exclusive. It is possible for a given actor or group of actors to be system actors as well as system designers, as the case in many participatory design processes.
A system is not constructed once and for all. Carayon (2006) argues that many people no longer work for a single organisation or only with people from their own organisation. She stresses that system constructor's work across organisational, geographical, cultural and temporal boundaries. This implies that system design is a continuous process, first, because all parts of a system are not specified and constructed at the same time within a common organisational, geographical and cultural setting. Therefore design decisions are spread over time in a process that can be described as ‘sequential attention to goals’. This may not be because the goals are conflicting, but more often because of the complexity of the design process in combination with the distribution of authority within and among organisations. Consequently, decisions on the physical structure of the transformations process are taken independently of decisions regarding organisation and management of the process. Attention must therefore be paid to how one set of decisions influences and delimits other decisions.
Secondly, design of a system must be understood as a process of constant redesign, where actors involved reconstruct and modify the system in the course of their daily activities. This occurs on the basis of the learning processes derived from handling the many deviations from ‘the normal’ procedure assumed by the designers (Weick and Quinn, 1999). This second modification implies that over time system actors inevitably also play the role of system re-designers.
These aspects of system design can be illustrated by an example from a study of outpatient units in a major Danish hospital (Broberg and Edwards, 2012). These units are not constructed once and for all. Outpatient units include patients with a range of illnesses and the health care professionals who treat these illnesses. However, a constant state of development exists in such a system as patients change, illnesses change, and the knowledge and skills of the health care professionals develop. Essentially, over time, the system grows in scope in response to patients' needs and as new treatment options arise. But viewed from a productivity perspective, this system is becoming less and less effective, since the same processes are used to treat new illnesses. At the gynecology ward, an analysis showed that treating a standard patient required 71 activities. These could be reduced to 17 activities by reducing a significant number of administrative tasks. Over time, the process for the standard patient had developed to encompass treatment of increased scope without redesigning the process. From a treatment perspective, it is important for a dynamic and knowledge-intensive environment such as an outpatient unit to have the flexibility to improve and expand processes to accommodate changes in patient mix. It is of equal importance, however, to take one step back at regular intervals in order to analyse the processes. The above example from the gynecology ward illustrates an extreme situation where processes had grown organically. Such a system had gained flexibility but lost efficiency.
Weick and Quinn (1999) address the question of how to handle such a situation by turning Lewin's (2000) recommendation upside down. Lewin presented a model for organisational change as a process to unfreeze present practice, change the practice, and then refreeze the new practice. Based on an organisational understanding that emphasizes the constant flux caused by variations in performing transformation processes, Weick and Quinn suggest that it is first necessary to freeze the present situation in order to be able to analyse it, then introduce changes, and finally unfreeze the process to allow it to develop until a new opportunity to examine the situation evolves.
6. The character of knowledge
In system design and redesign, the many issues to be addressed are characterized by different levels of complexity. To handle this situation is a challenge for the person responsible for the system design process.
Snowden and Boone (2007) propose a framework model, the Cynefin framework, which focuses on recognition of different types of cause and effect relationships (see Table 2). Often participants in a system design process can have different understandings of the relationship between cause and effect. This might be caused as well by different perception of the system as by local political issues. The aim of the framework is to support development of a common understanding of the complexity of a given challenge and subsequently how to act. The framework comprises five contexts, which are characterized by the different type of cause–effect relations recognized by the participants.