Abstract
This perspective paper explores the bidirectional influence between languageemergence and the relational structure of subjective experiences, termed qualiastructure, and lays out a constructive approach to the intricate dependencybetween the two. We hypothesize that the emergence of languages withdistributional semantics (e.g., syntactic-semantic structures) is linked to thecoordination of internal representations shaped by experience, potentiallyfacilitating more structured language through reciprocal influence. Thishypothesized mutual dependency connects to recent advancements in AI and symbolemergence robotics, and is explored within this paper through theoreticalframeworks such as the collective predictive coding. Computational studies showthat neural network-based language models form systematically structuredinternal representations, and multimodal language models can sharerepresentations between language and perceptual information. This perspectivesuggests that language emergence serves not only as a mechanism creating acommunication tool but also as a mechanism for allowing people to realizeshared understanding of qualitative experiences. The paper discusses theimplications of this bidirectional influence in the context of consciousnessstudies, linguistics, and cognitive science, and outlines future constructiveresearch directions to further explore this dynamic relationship betweenlanguage emergence and qualia structure.