Our objective is to support reasoning tasks in heritage architecture with graphics enabling analysts to visualise and share their understanding of how, from a given set of information, alternative scenarios or evolution can be inferred. The paper comments on the nature of the cognitive processes in historical sciences, and on factors that need to be weighed when interpreting sets of information. Visual solutions are proposed, and illustrated on real cases in Kraków Poland. They help spotting where alternative explanations should be considered in order to avoid unjustified assumptions and certitudes on the evolution of artefacts. The contribution expects to demonstrate that reasoning on uncertainties in historical sciences can be fruitfully backed up by concepts and practices from the infovis community.