Closing the Loop: How Semantic Closure Enables Open-Ended Evolution

What distinguishes life from non-life? In the last and current century we have developed many theoretical tools which ended up being useless in order to answer that question. Potentially the closest characterization of life can be found in the work by Howard Pattee, who proposed semantic closure—the self-referential mechanism through which symbols actively construct and interpret their own functional contexts—as a way to distinguish life from lifeless. In this talk I proposed a model to explain the evolutionary emergence of semantic closure by integrating concepts from relational biology, physical biosemiotics, and ecological psychology into a unified computational enactivist framework. By extending Hofmeyr's (F, A)-systems—a continuation of Rosen's (M, R)-systems—with temporal parametrization and multiscale causality, our model is capable of capturing critical life properties, including autopoiesis, anticipation, and adaptation. We then establish a formal equivalence between our extended (F, A)-systems and swarms of communicating automata, resolving self-referential challenges concerning the realizability of relational models. Our stepwise model traces the evolution of semantic closure, from simple reaction networks that recognize regular languages to self-replicating chemical systems with memory and anticipatory capabilities, identifying self-reference as necessary for robust self-replication and open-ended evolution. Such a computational enactivist perspective underscores the essential necessity of implementing symbol-matter transformations into unconventional computing frameworks, opening pathways to new models of computation for life, agency and cognition.