An AI-Supported Methodology for Analyzing Deflections and Misalignments in Human Interactions

Area of Science:

• Computational Linguistics
• Psycholinguistics
• Communication Studies

Background:

• Conversational breakdowns offer insights into interpersonal dynamics.
• Existing methods for analyzing miscommunication lack systematic approaches.
• Integrating information theory and dialogism provides a novel framework for meaning analysis.

Purpose of the Study:

• To introduce Speech Act Deflection and Misalignment Analysis (SADMA), an AI-assisted methodology.
• To provide a systematic tool for identifying conversational misalignments.
• To offer objective insights into subjective meaning-making in conversations.

Main Methods:

• Grounded in a "meaning-as-a-response" framework combining information theory and dialogism.
• Analyzes utterance-response pairs to detect deflection points in conversational trajectories.
• Leverages a Large Language Model to identify speech acts and conversational goals.

Main Results:

• SADMA successfully identifies patterns of relational conflict and miscommunication.
• The AI-assisted methodology provides objective insights into subjective meaning-making.
• Application to literary analysis reveals conversational breakdown dynamics in "Private Lives".

Conclusions:

• SADMA offers a systematic and objective approach to analyzing conversational breakdowns.
• The methodology has broad applications in psychology, social science, and literary studies.
• This AI-assisted tool enhances the understanding of interpersonal dynamics through communication analysis.