LLMs and Meaning: what the current semantic challenges in LLMs highlight about Natural Language

Recent advances in large language models (LLMs) have led to increasingly strong claims about their human-level linguistic competence and, by extension, their relevance for understanding natural language and cognition. While a growing interdisciplinary literature has used behavioural tests and psychometric tools to assess LLM performance, particularly in linguistics, such comparisons raise substantial conceptual and methodological concerns. This paper examines what current semantic limitations in LLMs reveal about the nature of meaning in natural language. Drawing on empirical findings from psycholinguistics, the analysis shows that, despite fluent surface-level performance, LLMs systematically lack core aspects of semantic competence, including real-world grounding, communicative intent, and stable grammatical judgment. These limitations are then situated within broader theoretical frameworks from the philosophy of cognitive science, focusing on Miracchi’s reformulation of the Frame Problem and the Relevance Realisation framework. Together, these approaches highlight the central role of embodied agency, environmental embeddedness, and socio-cultural context in human meaning-making, and clarify why current computational architectures struggle with semantic relevance. The paper concludes by arguing for a bio-cultural conception of language, suggesting that the semantic gaps observed in LLMs are not merely technical shortcomings but reflect fundamental differences between algorithmic systems and human linguistic cognition.