THE THEORY OF REPRESENTATIONAL COMPOSITIONALITY IN CONFRONTATION WITH THE THEORY OF OBJECTIVITY: compression, distinction, simple semantics, and informational transcendence in the relations between artificial intelligence, cognition, and modal ontology

This article presents a critical–propositional analysis of Eric Elmoznino, Thomas Jiralerspong, Yoshua Bengio, and Guillaume Lajoie’s A Complexity-Based Theory of Compositionality / Towards a Formal Theory of Representational Compositionality, in confrontation with the Theory of Objectivity (TO). The analyzed work proposes a formal and quantitative definition of representational compositionality grounded in algorithmic information theory and Kolmogorov complexity, according to which a representation is compositional when it can be re-described through recombinable discrete symbolic sequences and a simple semantic function. The present study examines the possible compatibilities and modal tensions between this theory and the axioms of the Theory of Objectivity, especially the principles of distinction, boundary, composition, observation, and transcendent information. It argues that representational compositionality can be read, from the perspective of TO, as a technical-operational bridge between artificial intelligence, cognition, and modal ontology. The analysis articulates the article with the phenomenic elements, Inductive Effects, the cosmogonic theorem, and the cosmological Eras of the Theory of Objectivity. A central proposal of the article is the concept of objective compositionality, understood as the degree to which a representational or phenomenic structure can be decomposed into prior elements, distinguished by boundaries, recombined through logical tracks, observed by relational regimes, and capable of producing transcendent information. In this sense, the study expands the algorithmic notion of compositionality into a broader ontological framework, in which intelligence, representation, and meaning are interpreted as relational productions of objectivity. This analytical text received analytical support from ChatGPT. Keywords: Theory of Objectivity; representational compositionality; objective compositionality; Kolmogorov complexity; algorithmic information theory; artificial intelligence; cognition; modal ontology; phenomenic elements; Inductive Effects; transcendent information; semantic function; symbolic representation; out-of-distribution generalization; Guillaume Lajoie; Yoshua Bengio.