The Sensemaking-Coevolution-Implementation Theory of Software Design

Abstract

Understanding software design practice is critical to understanding modern information systems development. New developments in empirical software engineering, information systems design science and the interdisciplinary design literature combined with recent advances in process theory and testability have created a situation ripe for innovation. Consequently, this paper utilizes these breakthroughs to formulate a process theory of software design practice: SensemakingCoevolution-Implementation Theory explains how complex software systems are created by collocated software development teams in organizations. It posits that an independent agent (design team) creates a software system by alternating between three activities: organizing their perceptions about the context, mutually refining their understandings of the context and design space, and manifesting their understanding of the design space in a technological artifact. This theory development paper defines and illustrates Sensemaking-Coevolution-Implementation Theory, grounds its concepts and relationships in existing literature, conceptually evaluates the theory and situates it in the broader context of information systems development.

The Tarpit – A general theory of software engineering

“…authors, reviewers and program committee members for the GTSE workshop series [1][2][3] and associated special issues [4]), we find contributors with an engineering-oriented perspective [38][39][40][41][42][43] as well as participants influenced by the social sciences (typically via the information systems discipline) [44][45][46][47][48]. This broad community seems to contain also differing views on epistemology.…”

“…This broad community seems to contain also differing views on epistemology. Schön, an author cited by some GTSE authors [44,46], describes the differing viewpoints as follows:…”

“…A number of relevant articles were published in these fora, including theoretical proposals by Ralph [46,169], Perry [39,170], Adolph and Kruchten [47], Erbas and Erbas [43,171] and Smolander and Päivärinta [44,172]. Another recent addition to the stream of publications on software engineering theory is an article by Dieter Rombach, also published in 2013 [5].…”

Scientific software development viewed as knowledge acquisition: Towards understanding the development of risk-averse scientific software

“…More recently, (eg., Ralph [30]) there has been work on the underlying theories of how the human agent designs a software system without following systematic methods. Ralph [30] states that the software development literature is dominated by "methods" which are, by definition, prescriptive and "…include practices, techniques, tools and phase models", and that, "… phases are also explicitly adopted in the official IEEE Guide to the Software Engineering Body of Knowledge". We found methods dominate software quality standards as well (eg.…”

“…Ralph [30] proposes an alternate process theory of software design, based on realizing perceptions, refining understandings, and recording that understanding in the technological artifact (the software), in other words a design theory that does not prescribe methods and is based on acquiring knowledge. The model we propose can be viewed as an extension of the alternative that Ralph is proposing.…”