IS - Information System

ISD - Information Systems Development

PMU - Project Management Unit

p.51 At the core of prior research is the belief that information systems development (ISD) is a planned, deliberate activity - bounded in time and carried out in a systematic and orderly way... Another fundamental assumption is that ISD is a methodical process... The objective of this paper is, then, to develop an alternative perspective on ISD - a perspective that posits improvisation and emergent change rather than methodical behavior and planned change as fundamental aspects of ISD in organizations.

p.51-52 We recognize, of course, that ISD can be and often is performed as a deliberate, purposeful project with well-defined requirements, milestones and substantial technological and organizational resources. However, we want to highlight the fact that ISD also happens in a myriad of other ways and that important activities take place outside of the formal projects - in the cracks and crevices in the official project portfolio so to speak. These development processes are more emergent, more continuous, more filled with surprise, more difficult to control, more tied to the content of action and more affected by what people pay attention to than by intentions, plans and methodologies (Weick 1993a). They are grounded in the situated practices of organizational actors, and emerge out of their adaptations to and experiments with the everyday contingencies, breakdowns, exceptions, opportunities and unintended consequences that they encounter when they appropriate new technologies in their work (Orlikowski 1996).

p.52 A working definition of improvisation can be taken from jazz music, where it connotes composing and performing contemporaneously (Barrett 1998; Weick 1998). Within organizations, it can be described as the conception of action as it unfolds, drawing on available material, cognitive, affective and social resources (Cunha et al. 1999). It means that: (1) Improvisation is deliberate, meaning that it is the result of intentional efforts on the behalf of the organization and/or any of its members. (2) Improvisation is extemporaneous. It deals with the unforeseen; it works without a prior plan and without blueprints and methods (Weick 1993a). (3) Improvisation occurs during action, meaning that organizational members do not stop to analyze a perceived problem or an unanticipated opportunity and come up with a plan. Instead they develop their response by acting on the problem or opportunity, and can only judge its suitability by hindsight, not by foresight as in traditional planning (Cunha et al. 1999). (4) Finally, improvisation implies the preexistence of a set of resources, be it a 'plan of action,' tools and technologies, knowledge or a social structure, upon which variations can be built (Cunha et al. 1999; Weick 1999).

p.53 Organizational improvisation is closely linked to the concept of bricolage, i.e. the ability to use whatever resources and repertoire one has to perform whatever task one faces (Lanzara 1999, Louridas 1999; Weick 1993a). Because improvisation means to act in an extemporaneous and spontaneous way to changing needs and conditions, improvisers cannot wait for optimal resources to be deployed and have to tackle the issues at hand with currently available resources (Cunha et al. 1999; Weick 1993a, 1999). Therefore, when improvisation happens, then necessarily, bricolage will too.

p.53-59 [boring BioCorp case study... zzzz]

p.57 In summary, Jean and Stella had some vague ideas... but they did not know exactly what they wanted at the outset of the design process. It was only during the design process, through their interactions... and through their joint exploration of different solutions (prototypes) that it gradually became clear... how it should be designed.

p.59 Sensemaking. It is striking how little the key "designers" of ProjectWeb (Stella and Jean) knew about what they were doing at the beginning of the development process. They did not have a clear goal or a precise idea about what they were designing or how it would fit into people's work practices. On the contrary, it was difficult for them to make sense of the web technology and their first attempts to conceptualize the emerging system relied on comparisons with older, more familiar technologies (LAN drives and email). It was not until they started experimenting with the design that they began to develop more detailed and sophisticated technological frames (Orlikowski & Gash 1994), and mental models of the technology. It was then they discovered what they wanted to do with it. Their understanding of the technology and their design goals evolved gradually and interactively as they developed new versions of the system in close collaboration with David and Hal (from the IT department) and tested it in practice together with their fellow users in PMU.

p.60 The point we want to make is that the development of ProjectWeb was fundamentally a sensemaking process (Weick 1995). Sensemaking is a process where people strive to convert a world of experience into an intelligible and meaningful world. It "is about sizing up a situation, about trying to discover what you have while you simultaneously act and have some effect on what you discover" (Weick 1999). It is an attempt to grasp a developing situation - in this case the design and implementation of an innovative information system - in which the observer affects the trajectory of that development. Because new technologies are equivocal [JLJ - sooo Carl Weick-ish... you owe him royalties for using this Weick-word] and thus lend themselves to multiple, conflicting interpretations, all of which are plausible, the development and use of technical systems require ongoing sensemaking (Weick 1990).

At the heart of sensemaking is the idea that understanding lies in the path of action. Action precedes understanding and focuses interpretation. It was by developing "prototypes" and trying out different versions of ProjectWeb in practice that Stella and Jean (and David and Hal) began to discover what their emergent system design meant and where they were heading. This is an example of "sense-making as manipulation" (Weick 1995). Sensemaking by means of manipulation involves acting in ways that create something... that people can then comprehend and manage:

"Manipulation generates clearer outcomes in a puzzling world, and these outcomes make it easier to grasp what might be going on... Manipulation is about making things happen, so that a person can then pounce on those created things and try to explain them as a way to get a better sense of what is happening" (Weick 1995 p. 168).

The key point is that sensemaking is an active process and that action is a precondition for sensemaking: "Action is intelligence, and until it is deployed, meaning and sense will be underdeveloped." (Weick 1993c).

Improvisation. The development of ProjectWeb was not guided by a preconceived plan or a systematic method. On the contrary, it was informed by hunches rather than well-developed knowledge, it relied on 'ad hoc' solutions, and it had a strong core of experimentation and unjustified trial and error. In other words, the actors (users as well as developers) depended on improvisation and extemporaneous action in order to cope with unexpected problems, unanticipated opportunities, multiple meanings, and transient organizational requirements.

Improvisation deals with the unforeseen, it works without a prior plan and without blueprints and methods (Weick 1993a). "Improvisation is the deliberate and substantive fusion of the design and execution of a novel production" or performance (Miner et al. 2001, p. 314). It can be conceived of as a form of short-term learning where real-time experience informs novel action at the same time that the action is being taken. Much research on improvisation has focused on individuals, but improvisational action can occur at any level (individual, team, organization) and is often a collective process (Miner et al. 2001), as when Jean, Stella, David and Hal together designed the second version of ProjectWeb.

The notion of improvisation implies that attention and interpretation rather than intention and decision making drives the process of designing. From this perspective, ISD is more an act of interpretation rather than an act of decision-making. The people involved improvise to make sense of unexpected possibilities and constraints that emerge. They are never in full control of the development process, but continuously challenged by having to address the unintended effects that are so commonplace in development projects. As a consequence, people are forced to revise their sense of what is happening and what can be accomplished. These revised interpretations are what guide action, and not the initial decisions (Weick 1993a). Since the only things we can sense are enacted events that have already taken place, attention rather than intention becomes central to the design process.

p.61 Bricolage. The development of ProjectWeb was clearly a process that made do with whatever materials were at hand. Version 1, for instance, was a modified version of a program, which Jean had borrowed from the corporate library; and important elements of version 1 were again reused in version 2. Thus, ProjectWeb is a good example of the general phenomenon that "new systems are built, sometimes literally, on the ruins and with the ruins of old systems" (Lanzara 1999, p. 346). Pieces of past code become "building materials" and are used - together with available commercial software components (in this case e.g. a web server and a DBMS from Microsoft) - to construct new systems, which then becomes more or less coherent assemblies of mixed components.

In other words, the development of ProjectWeb can best be described as an instance of "bricolage," i.e. a constructive activity based on transforming and reshaping what is already in use, or creatively rearranging components to fulfill new purposes (Lanzara 1999). The French word bricolage means, "to use whatever resources and repertoire one has to perform whatever task one faces" (Weick 1993a). Invariably the resources are heterogeneous and less well suited to the exact project than one would prefer, but they are all there is. The materials are not project-specific, but, instead, they represent the contingent result of all of the previous uses to which those items have been put. The key to understand the nature of bricolage as an innovative activity is Levi-Strauss's statement that materials "are not known as a result of their usefulness; they are deemed to be useful or interesting because they are first of all known" (Levi-Strauss 1966, cited from Weick 1993a).

Bricolage is closely associated with improvisation. Improvisation increases the chances that bricolage will occur because there is less time to obtain the necessary materials and resources in advance. Bricolage and improvisation are not synonymous, however, as bricolage (at least in theory) can occur in nonimprovisational contexts (Miner et al. 2001).

In summary, the development process we have described and analyzed here differs remarkably from the orderly, structured paths that most current IS theories and methods tend to assume. Instead, we have observed a more emergent, more spontaneous, more open-ended and more continuous process involving bricolage, unjustified trial an error, small-scale practical experiments, local readjustments, and improvisations. The process has been shaped more by action than by plans, and more by attention than by intention.

As already mentioned in the introduction, we realize that this study is exploratory, that ProjectWeb belongs to a special class of information systems (WIS) and that this fact may limit the generality of our findings. Consequently, future research should engage in careful testing of our concepts and relationships in other ISD contexts.

p.61-62 Conclusion

The case of ProjectWeb provides an occasion to think more carefully about the way we conceptualize ISD. We believe it is about time to dismiss the tradition of viewing ISD as an inherently rational, methodical and orderly process. System development processes, in our experience, are never very tidy, neat or sensible, and the importance and value of system development methods in practice are vastly overestimated (Bansler & Bodker 1994, Ciborra 1998, Fitzgerald 1998, Gasson 1999, Truex et al. 2000). System developers and users find themselves in a more complex, less stable and less well-understood world than assumed by most IS researchers. They are placed in a world, which does not always make sense and over which they often have only modest control.

Thus, there is a pressing need to develop an alternative theoretical perspective on ISD - a perspective that takes the messy reality of systems development practice seriously and makes it possible to grasp its non-methodical, un-planned and fortuitous aspects. We believe that the concepts of sensemaking, improvisation and bricolage proposed here constitute a useful starting point for developing such a perspective. They offer a theoretical lens for examining how people cope with ISD in practice and explaining why methods and plans have limited value in most real-life situations.