Structure

The framework provides a database of emotional categories, where each entry is defined by a set of descriptive and technical fields. Below is a detailed explanation of each field’s meaning and function.

Core Identifiers

• ID: A unique, non-editable identifier for each emotional category (e.g., SB-001).
• Name: The primary name for the category, often a Primary / Secondary pair (e.g., “Euphoria / Ecstasy”) to add nuance.
• Emoji: A simple, recognisable icon for quick visual identification.
• Associated States: A list of related feelings and synonyms (e.g., for “Peace / Serenity,” this includes “Calmness, tranquility, contentment”).

Qualitative Descriptors

• Description: A concise, human-friendly sentence capturing the essence of the emotion.
• Musical Function: The intended use of music in this category (e.g., “To create a relaxing, meditative atmosphere”).
• Inner Experience: The subjective feelings and thoughts that may emerge in the experiencer as a result.
• Outer Expression: The observable, external behaviours and body language associated with the emotion.
• Lyrical Themes: Common topics and narratives found in the lyrics of songs in this category.
• Group: A high-level family for the emotion (e.g., “High Energy Activation,” “Transcendent States”).

Psychological & Affective Dimensions

• Quadrant: The emotion’s position within the two-dimensional circumplex model, combining Valence and Energy (e.g., “I Pleasant Activation”).
• Valence: A score from 0.0 (highly unpleasant) to 1.0 (highly pleasant).
• Energy: A score from 0.0 (low activation/calm) to 1.0 (high activation/excitement). This is analogous to “Arousal” in academic models.
• Dominance: A score from 0.0 (submissive/powerless) to 1.0 (dominant/in-control).
• Psychological Description: A technical breakdown of the emotion based on its core affect, cognitive appraisal, and physiological response.
• Psychological Mechanisms: The specific psychological pathway through which music induces the emotion in a listener. Based on established models of music cognition (e.g., Juslin & Västfjäll), this field explains the underlying ‘how’ and ‘why’ connecting the musical features to the emotional outcome.

Musical Characteristics

This suite of fields describes the typical sonic building blocks of music that evokes the specified emotion.

• Tempo: The speed of the music (e.g., “Fast, 125-140 BPM, Allegro to Presto, Stable”).
• Rhythm: The groove and temporal organisation (e.g., “4/4 time, Propulsive four-on-the-floor kick drum”).
• Harmony: The chords and progressions (e.g., “Overwhelmingly diatonic and major-key”).
• Melody: The shape and movement of the tune (e.g., “Anthemic and soaring contour”).
• Timbre: The sound quality or “colour” of the instruments (e.g., “Bright, Stacked sawtooth-wave synths”).
• Texture: The layering and density of the sound (e.g., “Homophonic and driving, with all instruments rhythmically locked to the melody”).
• Dynamics: The variation in loudness (e.g., “Vast dynamic range, built around extended crescendos”).
• Structure: The large-scale form and organisation of the piece (e.g., “Features extended build-up sections… culminating in an explosive ‘drop”).

Version Attributes

• Version Introduced: The framework version when this emotion first appeared.
• Version Modified: The last framework version when its meaning or placement changed in a significant way.
• Status: Whether the emotion is “active”, “deprecated”, or “superseded”.
  • active: Emotion is current and should be used for new tagging.
  • deprecated: Emotion should not be used for new tagging, but existing tags need not be immediately changed. Keep the row and ID for historical compatibility.
  • superseded: Emotion has a direct, recommended replacement(s). Provide pointers via companion columns (e.g., SupersededByIDs) in the dataset and via the public mapping file.
• Superseded By IDs: IDs of newer emotions that have replaced this one (if any).
• Supersedes IDs: IDs of older emotions that this one has replaced (if any).
• Migration Notes: Guidance for retagging or interpreting changes when moving between framework versions.

Versioning Convention

The framework will evolve over time, sometimes in small ways (such as clarifying descriptions or adding new emotions) and sometimes in larger ways (such as moving emotions in the VED space or replacing them). To make this clear and to help with compatibility, the framework uses a simple versioning system.

Framework version is written as a single major number (1, 2, 3, …).

Emotion IDs never change. Once created, an ID such as SB-004 is permanent. If an emotion is replaced or merged, the old ID is marked as deprecated and new IDs are introduced. This keeps all previous tags valid for historical use.

Major version increases only when there are breaking changes:

• an emotion is repositioned in the VED space; OR
• an emotion is split or merged; OR
• a field changes in a way that alters meaning.

Minor edits (spelling, clearer wording, formatting) do not change the version.

When tagging music, the Emotion ID (e.g. SB-004) is all that is needed.

The major version number is tracked in the dataset itself and in release notes, but does not need to be written into every tag.