Work

Cultural Cognition and Creative Practice of Traditional Music Recontextualization in Human AI Co-Creation Environments

Year: 2026
Type: article
Abstract: Conventional music is being increasingly embedded into generative artificial intelligence contexts, where fast melodic transformations would help recontextualize while losing modal integrity, motivic unity, and culturally grounded meanings. A symbolic human-AI collaborative music-making system was built based on an efficient Transformer model, culturally interpretative reranking of cultural features, and interactive recording. A within-subject counterbalanced experiment with seventy-two participants covered unaided creation, general AI generation, and culturally constrained AI generation. The assessment included a 15-question cultural-cognition test, five-expert cultural appropriateness ratings, symbolic feature analysis, MIDI edit distance, completion time, cognitive load, and perceived agency. Simulation results suggested that culturally constrained AI led to greater cognition gains, better modal and motivic preservation, more culturally appropriate rating, and shorter completion time while having greater creative agency compared to general AI. Cultural cognition gain was 2.31 ± 1.26 points, expert cultural appropriateness ratings were 5.62 ± 0.58, and completion time was 29.6 ± 4.8 min. These trends reflect how culturally constrained AI reranking can link computation of music generation with culturally informed cognition and keep meaningful human participation.
Author: Binru Zou
Language: English
Subfield: Music
Open Access status: hybrid
OpenAlex ID: W7235467036
FWCI:
Cites: 11
Cited by:
Related to: 10