Algorithmic Simplicity and Relevance

Jean-Louis Dessalles
  • January 2013, Springer Science + Business Media
  • DOI: 10.1007/978-3-642-44958-1_9

Defining what "relevant" means

What is it about?

Relevant events are those that appear unexpectedly simple. For instance, coincidences are (too) simple because one of the coinciding events can be used to describe the other one, thus sparing information. This definition of relevance is both intuitive and formal.

Why is it important?

Defining relevance is a fundamental quest, not only in the human sciences, but more broadly in any scientific enterprise. The definition of relevance presented in this paper offers a sound scientific account of an essential human phenomenon. It has far-reaching consequences, as it offers a predictive theory. It can be applied to the study of human conversation, to the understanding of human rationality, but also to the design of artificial systems.


Jean-Louis Dessalles (Author)

I hunted for years for a way of capturing what people regard as "relevant". First approximations involved notions like improbability and problematicity. I was lucky enough to come upon a definition that correctly separates what is relevant from what is not (with appropriate graduality in between). I hope that this paper will help people share my enthusiasm about this find. More on

The following have contributed to this page: Jean-Louis Dessalles