Learning and imagination in construction

Learning is considered a good strategy for solving the persistent problems with defects in construction. But conditions in construction are not conducive for learning. This increases the risk that experience produces lessons that are false or limited. We illustrate the inherent difficulties of making sound assessment in even simple, ordinary situations on a construction site. Building on this appreciation of high complexity we develop a perspective on the experience of construction defects as probabilistic events. From this perspective we suggest that what should he learned from experience is not only what happened but also what might have happened; and that the challenge is not to make sound assessment of defects but to finds ways of making the assessments more likely to become sound by subsequent events.