Special Track on Semiotic-Pragmatic Intelligence

Scope & Description

While Large Language Models (LLMs) represent a milestone in syntax and distributional semantics, a significant gap remains between statistical fluency and genuine pragmatic competence. Current systems excel at predicting linguistic sequences but often lack a foundational grasp of the layered dimensions of meaning: the relationship between signs and objects (Semiotics), the influence of context on intent (Pragmatics), and the formal categorization of these concepts within structured knowledge systems (Ontology).

The central ambition of the SPN-NLP track is to move beyond purely statistical word-prediction toward "meaning-aware" systems that understand context, intent, and cultural convention. We invite contributions that draw on theoretical frameworks—including Peircean, Greimassian and Saussurean semiotics, Gricean and post-Gricean pragmatics, Speech Act Theory, and Relevance Theory—to inform the design, evaluation, and interpretability of intelligent language systems.

A key focus of the track is the synthesis of these linguistic theories with computational structures. This includes the development of Pragmatic Ontologies and neuro-symbolic architectures that can map fluid human nuances into actionable knowledge. By integrating the history of linguistic thought with empirical AI research this track seeks to foster AI that genuinely respects human communicative norms and provides deeper, more reliable interpretative capabilities.

We encourage submissions that are technical, theoretical, philosophical or historical, provided they contribute to the overarching goal of bridging the gap between formal linguistics and intelligent system applications.