Introduction

Recurring failures in the development of information systems (IS) (Boustred, 1997), the persistence of the "software crisis" (Gibbs, 1994; Naur & Randell, 1969), as well as the "productivity paradoxon" (Attewell, 1994; Brynjolfsson, 1993; Strassmann, 1997) indicate that information systems research (ISR) — in spite of all progress made at the methodic level — did not proceed very far in developing a methodological foundation for the understanding of these problems. Their persistence gives rise to the question of the appropriateness of today's presuppositions in IS research and practice, and provides a practical motivation for engaging in more fundamental, methodological reflections. Consequently, these exceed today's known boundaries of ISR, and extend to disciplines like philosophy of science, anthropology, sociology, and psychology.

Furthermore, a review of the publications in major outlets of the ISR community makes quite obvious that methodological considerations as well as underlying philosophical presuppositions are rarely explicated. This (prevailing) negligence of philosophical considerations has severe consequences for IS research and practice. As Collier (1994) states:

"a good part of the answer to the question 'why philosophy?' is that the alternative to philosophy is not no philosophy, but bad philosophy. The 'unphilosophical' person has an unconscious philosophy, which they apply in their practice — whether of science or politics or daily life" (p. 17).

In the early days of ISR, Gorn (1958) writes to the editor-in-chief of the Communications of the ACM (CACM) that "philosophy has served some of its most important functions to the world of communication in the society by […] critical evaluations of the fundamental concepts of the arts and sciences." He concludes the "general subject matter of the letters-to-the-editor department should include, then, the philosophy of computing" (p. 2). As can be observed in editions of the CACM ever since, Gorn's idea did not result in much change yet.

As a reason for this lack of transdisciplinarity in ISR, McFarlan (1984) states:

"…many IS researchers, although they possess strong technological skills, lack the tools and perspectives necessary for cross-disciplinary work. Further, even if they have such skills, IS scholars are often not intellectually inclined to undertake these complex studies" (p. 2).

Twenty years later, this assertion is still true, as can be seen — with some exceptions — in current curricula, conference programs, and the table of contents of many journals in the IS domain.

A cause for this dissatisfying methodological situation of ISR can be traced back in its historical development. Because ISR is rooted in disciplines like cybernetics, general management and industrial engineering, various theories, concepts, and methods as well as their underlying presuppositions are transferred to ISR, mostly without prior methodological reflection.

Some consequences of this unreflected transfer became obvious in areas such as artificial intelligence (AI) and business process reengineering (BPR). With AI's roots in cybernetics, information theory and cognitive psychology, the adopted assumptions about human nature and the human mind that become most obvious in the computational theory of mind (e.g., Putnam, 1975), led to ideas and visions concerning the possibilities of AI which, from today's point of view, would be regarded as illusionary or science fiction at best (e.g., Dreyfus, 1972, 1992). In BPR, failure can also be traced back to unrealistic assumptions about human nature. It is commonly known that about 70 percent of all BPR projects failed in one way or another (Hammer & Champy, 1993), mostly due to the negligence of social and cultural aspects of work. This negligence is rooted in the widespread mechanistic view of the human being, which has no room for the social and cultural conditionality of humanness.

For our purpose, philosophical considerations in the context of ISR are important for two reasons: they affect the nature of research itself, and they are fundamental for the understanding of information systems. Viewing information systems as tools of mediated human inquiry and trying to understand the implications of this view for the development and use of information systems requires a prior understanding of human inquiry.

As various analyses have revealed, positivism is the dominating concept of inquiry in ISR (Klein & Hirschheim, 1987; Orlikowski & Baroudi, 1991; Ridley & Keen, 1998). This epistemological position is based on presuppositions of human inquiry already being criticized for a longer time in disciplines like organization theory and sociology (Hassard, 1993; Dahms, 1994). These disciplines (should) have considerable impact on theories of design, development, and use of information systems, since social institutions provide the context for their use.

The development of postmodern concepts, within the field of organization theory since the 1980s, has led to application and further elaboration of interpretive approaches towards human inquiry. The (now) competing positivistic and interpretive approaches lead, due to their fundamental character, not only to different understandings of organization, but of information systems as well. Therefore, any discussion about information systems as tools of mediated inquiry is inevitably integrated into a discourse about epistemological positions.

Several authors have already addressed this problem (Hirschheim, Klein & Lyytinen, 1996; Iivari, 1991; Falkenberg et al., 1998). Their emphasis, however, is more on methodological considerations of IS research and development, and less on consequences of philosophical presuppositions for the understanding of information systems. Nonetheless, a trend towards more discussion of philosophical questions and an increasing critique of positivist approaches can be noticed (Fitzgerald & Howcroft, 1998).

In this chapter, we try to substantiate the need for a philosophical foundation of ISR, as a basis for a well-founded understanding of information systems. Focusing on issues of methodological pluralism, we provide a short outline of the modernism-postmodernism debate that is taking place in the social sciences. Most importantly, we further elaborate on its methodological consequences for ISR. From this we derive the need to ground research in well-defined and explicated philosophical suppositions. Referring to Kuhn's (1996) concept of a paradigm, we describe a framework for the conceptualization of the so-called "paradigms of inquiry," allowing a systematic description of existing approaches and the development of new approaches to human inquiry. Being unsatisfied with existing approaches, we also find the need for a new paradigm of inquiry. After providing an outline of our paradigm, referred to as "sociopragmatic constructivism," we describe a research agenda which serves both as a basis for the further development of our approach and as a framework for IS research and practice.