Philipe Barsamian is an independent researcher based in Montréal. His work studies how systems behave when contradiction and instability are not treated as pathologies to eliminate, but as conditions to organize.
Developing a framework for stabilization under feedback, his work explores how representational frameworks become workable, how they break, and how they may change while remaining identifiable across revisions. The central question is how systems sustain coherent structure through both constraint and transformation. What does it mean for a system to represent itself?
Currently, his research combines formal and empirical methodologies. Formally, he examines questions of canonical normalization in paraconsistent logics, type-theoretic models of identity and plasticity, and the incomplete architecture of self-referential systems. Empirically, he investigates sequential reasoning in large language models, where convergent and exploratory regimes emerge according to problem structure.
Across these contexts, his work focuses on building concepts and procedures that can be shared, tested, and revised—ways of keeping contradiction operational without forcing resolution.
Recent &
Forthcoming
-
Folding Contradiction: Toward Canonical Normalization in Paraconsistent Logic (Presentation at the World Congress on Universal Logic, Peru, 2025)
-
Contradiction as Cognitive Resource: Two Regimes of Sequential Reasoning in Large Language Models (Presentation at the World Congress on Universal Logic, Peru, 2025)
-
The Fold as Dialogue: Toward a Recursive Formalism (Presentation at the University of Konstanz, Germany, 2026)
Shimmer: Anthony McCall and the Opacity of Light (Essay co-written with Maegan Beck, ESPACE art actuel, n° 143: Light / Lumière, 2026)