2026 ASEE Annual Conference & Exposition

This Methods/Theory Work-in-Progress (WIP) paper examines how computational tools for engineering and design, while optimized for efficiency and convergence, can inadvertently constrain early-stage ideation where creativity depends on ambiguity and reinterpretation. Drawing on research in design cognition and external representations, we frame sketching as a creative language that supports thinking through incomplete and evolving representations. We introduce a conceptual framework of productive misinterpretation, defining it as an interpretive dynamic through which ideas evolve when representations are reinterpreted rather than prematurely stabilized. The framework distinguishes productive from degenerative misinterpretation and proposes design principles for ideation-supporting systems, including preserving ambiguity, delaying stabilization, enabling reinterpretation, and maintaining user agency. We then examine implications for AI-supported ideation, arguing that generative systems should function as sensemaking partners that sustain interpretive exploration rather than as tools that collapse possibilities through single convergent outputs. Finally, we outline a research trajectory for operationalizing misinterpretation episodes through interaction data and exploring implications for creativity-supporting technologies, assessment, and learning analytics.