Page 76 -

IDENTIFYING THE DRIVERS OF ISD METHOD EMERGENCE 61
process developed from a literature study (Sambamurthy and Kirsch, 2000); and the method-in-
action framework that incorporates past and contemporary thinking and empirical findings about
ISD methods into one conceptual frame (Fitzgerald, Russo, and Stolterman, 2002). Common to
these frameworks is that they stress the importance of understanding the context, the formalized
method(s), the developers’ preconceptions and actions, and their interactions with other stakehold-
ers, as well as the influence that these concepts have on the ISD process. Our work builds on the
insight provided by these frameworks and models in that we draw on similar concepts and share
similar assumptions about their relationship to the emergent method. However, our framework
extends the line of thinking through a clear focus on the temporal dimension of the development
process, and the emergent method is conceptualized as a sequence of activities that unfolds over
time. To explain why emergent methods unfold differently, we draw on theoretical ideas as put
forward in Pettigrew’s contextualism (1985, 1987), and Giddens’s structuration theory (1984),
and subsequently synthesized by Walsham (1993) as well as Van de Ven and Poole’s (1995) four
process theories that specify four different process forms and drivers.
The framework constitutes an organizing structure for providing: first, a narrative account of
how the emergent method unfolded in the individual cases; second, a systematic cross-case com-
parison to explain why the methods emerged differently; and third, a generalization of analysis
results to process theory.
The object of study is the emergent method, which is defined as the unfolding development
process and the activities, and the applied method elements that constitute this process. The narrative
account of the emergent method describes what happened over time, because an event sequence
with a clear beginning, middle, and end is the core of narrative structure (Pentland, 1999). How-
ever, the event sequence is only the first important step toward understanding why this particular
pattern of activities occurred (Pentland, 1999). It is also necessary to focus on influential actors
(their roles, perceptions, social relations, and demographics), power, culture, and broader context
to generate meaningful explanations (Pentland, 1999). Thus, to explain why the processes unfolded
as they did, we draw on the key concepts of content of change, social process, and social context
(Walsham, 1993) as interlinked units of analysis that facilitate progression from surface description
to explanation (Kautz, 2004; Kautz and Nielsen, 2004; Pentland, 1999; Pettigrew, 1987).
Content of change refers to the planned and actual process and product of the development
endeavor (Kautz, 2004; Kautz and Nielsen, 2004); that is the planned and actual ISD process and
information system under development. The planned is assumed to be an expression of expectation
that shapes attention and action (Bruner, 2002), and we consider the gap between the expected
and the actual important for an initial understanding of what characterizes the content and drives
the process of change. After the initial narrative description of the emergent method, the concept
of content is therefore applied to come to understand what characterizes the change (Pettigrew,
1987). Social process focuses on the political (i.e., the distribution of power and balance between
autonomy and control) and the cultural (i.e., subcultures and the interaction between them) aspects
of ISD and helps to explain how, that is, through which mechanisms, changes to the content take
place (Pettigrew, 1987; Walsham, 1993). Social context addresses social relations, social infra-
structure, and the history of previous procedures, structures, and commitments and helps to explain
why the social process emerges as it does (Walsham, 1993). Previous application of the framework
to empirical cases shows that the social context creates the social and structural landscape within
which the social process can emerge and that the social process in turn both enables and constrains
the content of change (Madsen, 2004; Madsen, Kautz, and Vidgen, 2005, 2006).
As the last step from empirical phenomena toward conceptual understanding, analysis results
are generalized to Van de Ven and Poole’s (1995) four process theories. Van de Ven and Poole