The behavioural diversity seen in biological systems is, at the most basic level, driven by interactions between physical materials and their environment. In this context, we investigate the a-life properties of falling paper systems, in which different paper shapes are dropped into free fall and their behaviours observed. These systems have a simple embodiment but highly complex interactions with the environment. Using a synthetic methodology, i.e. understanding by building, we explore how morphology can be used to program certain interactions into the dynamics of a free-falling V-shaped paper. We demonstrate that morphology can encode a stochastic hierarchy of possible behaviours into the system. This hierarchy can be described by a set of conditional switching probabilities and represented in a morphological ‘state machine’. We draw a parallel with developmental processes, showing how these can emerge from interaction with the environment. Next, we demonstrate how Bayesian optimisation can be used to optimise morphology in response to a fitness function, in this case minimizing falling speed. Bayesian optimisation allows us to capture the system stochasticity with minimal sampling. By manipulating non-living raw materials such as paper, we are able to analyse how morphology can be used to control and program interactions with the environment. With this bottom-up approach we ultimately aim to demonstrate principles that turn materials into agents that show non-trivial behaviours comparable to those of living organisms.