Abstract

A phenomenological exploration of the distinction between a relational formal ontology (also called a process ontology) and a classical formal ontology (also called an object ontology) for modelling physical phenomena that exhibit relationally-mediated holism, such as phenomena from quantum physics and biosemiotics. Whereas a classical formal ontology is based on mathematical objects and classes, a relational formal ontology is based on mathematical signs and categories. A relational formal ontology involves nodal networks that are dynamically sustained through signalling. Nodal networks are systems of constrained iterative processes (dynamical nodes) that have individual semiotic agency within a matrix of determinate possibilities (a semiotic scaffolding). The nodal networks are hierarchically ordered and exhibit characteristics of deep learning. Clarifying the distinction between classical and relational formal ontologies may help to clarify the role of interpretative context in physics (eg. the role of the observer in quantum theory), the role of signalling in biological systems and the role of hierarchical nodal networks in computational simulations of learning in generative artificial intelligence (AI). Two experiments are conducted to explore the application of key principles of relational ontology in AI, namely, complex pattern abduction involving hierarchies of categories and progressive determination through placeholder signs.