Designing and Engineering for Emergence: A Challenge for HCI Practice and Research

Steven Alter
2010 AIS Transactions on Human-Computer Interaction  
a two-dimensional framework for exploring the scope and challenges of HCI that combines a social/ technical dimension and a behavior dimension that emphasizes differences between engineered and emergent behavior in sociotechnical systems. This framework is used to reflect on possible differences between the scope of a definition of HCI in those articles and the scope of the topics identified in the extensive survey of HCI literature reported by Zhang and colleagues (2009). Implications include
more » ... he possibility that future HCI research and theorizing may find significant opportunities related to "designing for emergence," or even "engineering for emergence." WHAT IS THE SCOPE OF HCI? This research commentary was motivated by an invitation from the AIS Transactions on Human-Computer Interaction (THCI) to share views on Future Directions for HCI Research. It responds to research commentaries on that topic by Benbasat (2010) and Lyytinen (2010) and to two articles in Volume 1 of the AIS Transactions on Human-Computer Interaction, an introduction of THCI as a new journal (Galletta and Zhang, 2009) and a survey of HCI research . The survey examined articles in eight selected journals and in other selected sources from 1990 to 2008. It classified 693 of 2302 IS research articles in the sample as HCI articles, roughly 30% of the IS articles. If the sample of IS articles is representative, HCI research comprises around 30% of IS research. While I never considered myself an HCI researcher, the broad definition of HCI used by Galletta and Zhang (2009) and Zhang and colleagues (2009) made me wonder about how my research fits into HCI and how HCI fits into IS in general. Galletta and Zhang (2009) said that THCI addresses IS issues and concerns, but with a "specific focus on the history, reference disciplines, theories, practice, methodologies and techniques, new developments, and applications of the interaction between humans, information, technologies, and tasks, especially in the business, managerial, organizational, and cultural contexts" (p. 8). That rather expansive view of HCI topics overlaps substantially with definitions of other topics related to IS in general rather than HCI in particular: The definition of sociotechnical system in Lyytinen and Newman (2008) also refers to interactions: "any organizational system viewed as a multivariate system consisting of four interacting and aligned components -task, structure, actor, and technology" (p. 613). The nomological net for IS research proposed by Benbasat and Zmud (2003, Figure 2 , p. 187) covers topics related to humans, information, technologies, and tasks in business contexts. It includes IT managerial, methodological and technological capabilities, the IT artifact, usage, impact, and IT managerial, methodological, and operational practices, where the IT artifact is "the application of IT to enable or support some task(s) embedded within a structure(s) that itself is embedded within a context(s)" (p. 186). Clarifying Zhang and colleagues' (2002) original definition of HCI, Zhang et al. (2005) said "HCI issues and concerns involve all possible interactions between a user and a system during its lifecycle, including the development stage, use in context, and the impact of such use on individuals, organizations, society, and future systems development" (p. 519). The work system method also focuses on interactions between people, information, technologies and tasks. A work system is "a system in which human participants and/or machines perform work (processes and activities) using information, technology, and other resources to produce specific products and/or services for specific internal or external customers. An IS is a work system whose processes and activities are devoted to processing information..." (Alter, 2008, p. 451). The work system life cycle model describes how work systems change over time through a combination of planned change (explicit projects with initiation, development, and implementation phases) and unplanned change (adaptations and experimentation) (Alter, 2008) . The overlap between the definition of HCI and the four statements above revolves around interactions between people, information, technology, and tasks in business and organizational contexts. Some of those interactions involve systems in operation. Others involve system life cycles. The combination of those topics probably covers much more than the 30% of IS research implied by Zhang et al.'s (2009) classification of 2,302 articles. The lack of clarity about the boundaries between IS in general and the subfield of HCI within IS led me to try to develop a framework that would make it easier to describe the boundaries of HCI and to identify topics related to IS in operation and IS development that are and are not part of HCI. Topics near the boundary and links between topics inside and outside of the boundary might be fruitful areas for future HCI research. often partial and provisional and which require bridging, integration, and articulation in order for them to work together." The relative absence of HCI research in the right side of Figure 3 related to emergent behavior seems to imply that past HCI research in the spirit of Orlikowski and Iacono's premises is relatively rare. Attention to the distinction between engineered and emergent behavior may help in thinking about future HCI research that addresses emergence in the context of interactions between people, information, technologies, and tasks. Such research would provide a counterbalance to the widespread attention to precise specifications of process and information requirements (e.g., topics in typical systems analysis books and numerous articles about the quality and use of UML, BPMN, and other specification formalisms).
doi:10.17705/1thci.00018 fatcat:x2zcp55bhzeanoxtznw2miuax4