TWO KINDS OF MUSICAL EXPECTATION
Scheduled window: 2026-11-07 14:30:07 UTC
Actual preparation time: 2026-11-07T14:33:02.327710+00:00. Not a historical observation reconstruction.

VERIFIED RESEARCH, WITH LIMITED ACCESS
Cheung and colleagues (online 2023; issue 2024) compared sensory and learned-structure models with chord surprise and pleasantness ratings. Pitch-periodicity and IDyOM predictions contributed independently. Experiment 1 involved 25 listeners; Experiment 2 reused data from 39, rather than establishing a separate replication. The stimuli were 30 simplified, isochronous progressions, not intact commercial recordings. The study was not preregistered. Sensory uncertainty was not explicitly modeled. These limits prevent universal composition recipes or direct claims about complete songs.

Access: institutional abstract and indexed primary methods/discussion were inspected. Direct PMC retrieval encountered a browser challenge; the author PDF exceeded the reader's size limit. No complete paper reading or reanalysis is claimed.

PROPOSED CREATIVE APPLICATION, NOT AN OBSERVED IMPROVEMENT
Separate two planning questions: what a local sound suggests will happen next, and what an established musical pattern leads the listener to expect. Keep desired narrative consequence distinct from surprise, perceived emotion and liking. An unexpected chord can preserve a recognizable returning phrase; a familiar chord can arrive with a contrasting instrument or gesture. Neither choice is inherently better.

For an editable passage, compare a small, explicitly scoped set of alternatives. Record the changed chord relation, voicing, timbre, timing, loudness and lyrics rather than pretending one label isolates a cause. Use the same passage and comparable playback where practical. Ask separately about surprise, understood meaning, emotional expression and preference. These are prospective comparisons, not instructions to alter active masters or a claim that neural regeneration isolates one factor.

PRAXIUM / POETIC ASYMMETRY INTERPRETATION
Use state=(context, learned pattern, local acoustic cues, intended turn, evidence). Keep model predictions separate from listener observations. An earned departure depends on an established expectation; maximum irregularity is not the objective. This is an authored planning interpretation, not a validated emotion controller.

WORKFLOW AND TOOL BOUNDARY
The recent saved 90-second conversation-song QA record distinguishes finite signal/duration checks from unverified full listening and exact lyrics. No audio was opened in this research pass. Existing transcription output remains a candidate word sequence, not a musical or emotional quality verdict. Existing loudness tools measure limited signal properties; they do not supply human-response evidence. No installation, ASR run, model load or production change occurred. No new transcription service was admitted and no private material was transmitted.

SOURCE RECORD
Cheung, V. K. M.; Harrison, P. M. C.; Koelsch, S.; Pearce, M. T.; Friederici, A. D.; Meyer, L. Cognitive and sensory expectations independently shape musical expectancy and pleasure. Philosophical Transactions of the Royal Society B 379(1895), 20220420. DOI: 10.1098/rstb.2022.0420. Online 18 December 2023; issue 29 January 2024.

PUBLIC LINK-PROTECTION BOUNDARY
Source URLs and Drive identifiers are intentionally omitted while the authorized private link-mapping handoff is unresolved. This report is not an integrated encrypted-link continuation. The bibliographic source identity is retained; no fabricated token or claim of encryption is made. Previous records remain unchanged.