utterancewhat

Utterancewhat is a term used in linguistics and human–computer interaction to refer to the information content that an utterance communicates beyond its literal wording. It includes explicit content, intended messages, contextual inferences, and practical implications for the listener or a conversational agent. The concept is informal and not tied to a single formal theory, but it is often contrasted with the surface meaning of what is said.

Origins and scope: Utterancewhat emerged in mid-2010s discussions of pragmatics and NLP as a label for the pragmatic layer that connects form to meaning. It appears in conference notes and design discussions as a flexible concept rather than a fixed framework.

Examples: The utterance "It's getting late" may convey an implicit request to end the meeting, or a suggestion to leave, depending on context. In a chat system, "Can you handle this task?" might activate utterancewhat about user capability, not just a yes-no interpretation.

Applications and models: In natural language understanding, utterancewhat informs intent recognition, disambiguation, and dialogue management by emphasizing context and goals. Some approaches propose formal models to quantify the gap between literal content and utterancewhat, while others treat it as a heuristic guiding system responses. The term remains nonstandard and varies by community.

Relation and critique: Utterancewhat overlaps with implicature, presupposition, and speech-act theory. Its subjective nature makes measurement challenging across speakers and settings, leading to caution about overgeneralization in cross-domain applications.