Synchronizing Development Teams with Drifting Software Systems: A Decomposability-Based Theory of Evolutionary Mechanisms

View at publisher
Authors: Amrit Tiwana, Hani Safadi
Cited by: 0
References: 58

Abstract

System evolution challenges persist as IS scholarship remains entrenched in static models inadequately capturing how systems drift over time, while siloed team and architecture studies obscure their dynamic interplay. This fragmentation obscures the mechanisms—the “how”—through which coevolving team and technical architectures shape evolutionary outcomes. Our theory, building on Simon’s decomposability idea, traces systems’ evolution to the dynamic syncing of drifting technical and team architectures. We theorize two mechanisms of system evolution—module evolution (internal changes) and architecture evolution (interconnection changes)—each differentially catalyzed by system and team architecture; their syncing simultaneously activates both mechanisms. Econometric analysis of 1,303 open-source systems encompassing 10.1 million evolutionary events across 4.8 billion lines of code supports the proposed ideas. Our contributions are threefold. First, we reconceptualize system and team architectures as temporally dynamic properties, shifting the discourse from static snapshots to dynamic evolutionary trajectories. Second, we introduce module evolution and architecture evolution as mechanistically-distinct pathways differentially catalyzed by system and team architecture, enlarging the discourse from magnitude to the content of change. Third, we show how dynamic system-team syncing sustains systems’ evolution by continuously restoring alignment as inevitable architectural drift erodes their congruence. Overall, our mechanistic theory underscores the sociotechnical dynamics through which architectural syncing drives systems evolution.

Fields of Study & Subjects

Software Engineering Research Software Engineering Techniques and Practices Advanced Software Engineering Methodologies

Publication details

Journal / source
Journal of Management Information Systems
Volume
43
Issue
1
Pages
67–108
Publication date
2 January 2026
Language
EN