Numerous researches on the representation and handling of the uncertainty, starting with the pioneering works about probability of Laplace, Bernoulli, etc. to reach a substantial increase during the twentieth century have been made, with applications in many fields (for example, psychology, decision theory, etc.), including the artificial intelligence one (in this case, related to the multi-agent systems). On the other hand, the notions of holon and informon developed by Koestler, Sulis, Alonso, Pazos et al. have shown to be a very useful way of to model evolving complex dynamic systems. In this work the relationship between holons, informons, randomness and uncertainty (as the opposed to certainty) in a holonic system that reacts to events is analyzed.