Philosophical Smoke Signals: Theory and Practice in Information Systems Design

Although the gulf between the theory and practice in Information Systems is much lamented, few researchers have offered a way forward except through a number of (failed) attempts to develop a single systematic theory for Information Systems. In this paper, we encourage researchers to re-examine the practical consequences of their theoretical arguments. By examining these arguments we may be able to form a number of more rigorous theories of Information Systems, allowing us to draw theory and practice together without undertaking yet another attempt at the holy grail of a single unified systematic theory of Information Systems.

💡 Research Summary

The paper “Philosophical Smoke Signals: Theory and Practice in Information Systems Design” tackles the long‑standing gap between information systems (IS) theory and practice by reframing the problem in philosophical terms rather than by pursuing a single, all‑encompassing systematic theory. The authors begin by noting that IS has historically been positioned between two distinct academic cultures: computer science, which emphasizes mathematics, algorithms, and engineering, and IS itself, which integrates organizational, managerial, and technical concerns. Attempts to graft a “scientific” foundation onto IS by mirroring computer science have been critiqued because they ignore this contextual divergence.

The core contribution of the paper is the introduction of two orthogonal philosophical axes that shape IS design thinking. The first axis is epistemological, contrasting rationalist and empiricist positions. Rationalism treats knowledge as the product of reason; it assumes that designers can fully articulate their mental models (representations) and translate them into explicit, unambiguous descriptions. In this view, description and representation are interchangeable, and design consistency is readily achieved. Empiricism, by contrast, sees knowledge as derived from observation and experience; it acknowledges tacit knowledge that designers cannot fully verbalize. Consequently, representations evolve with each new experience, and descriptions lag behind, creating a perpetual mismatch. The paper illustrates these positions with concrete software‑design artifacts: an entity‑relationship diagram and a UML class diagram of a juice‑plant control system. In the empiricist scenario, the designer’s mental image of a cooking tank changes as experience accrues, forcing continual revision of the description. In the rationalist scenario, the description is fixed first, and the representation is derived from it, preserving stability.

The second axis is ontological, contrasting realist and anti‑realist perspectives on reality. Realism posits that an objective reality exists independent of the designer, allowing a direct mapping between a model (representation) and the world. Anti‑realism argues that reality is inseparable from perception; any model is inevitably a construction shaped by the observer’s mind. Modern “seamless” design approaches attempt to blend these views, seeking a single model that simultaneously captures logical (representation) and physical (reality) aspects. The authors argue that such blending often neglects tacit knowledge (a problem for realism) or sacrifices design consistency (a problem for anti‑realism).

Rather than insisting on a monolithic theory, the authors advocate a pluralistic, context‑sensitive framework. Designers should deliberately select the epistemological and ontological stance that best fits the current phase of a project. For phases where strict consistency and verification are paramount (e.g., safety‑critical modules), a rationalist‑realist combination is advisable. For exploratory phases where user requirements are fluid and tacit insights dominate, an empiricist‑anti‑realist stance better accommodates change. By making the “description‑representation‑reality” triad explicit, the paper provides a meta‑model that helps researchers and practitioners diagnose which philosophical assumptions underlie their methods, and to switch or combine them as needed.

The paper’s metaphor of “smoke signals” captures the idea that theoretical arguments send signals that must be interpreted in the messy, real‑world context of IS design. The authors conclude that embracing multiple, well‑articulated philosophical positions—rather than chasing a single unified systematic theory—offers a pragmatic path to bridge theory and practice, manage tacit knowledge, and maintain design coherence. This approach promises to enrich future IS research with both methodological rigor and practical relevance.