Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

"More Than Words": Linking Music Preferences and Moral Values Through Lyrics

Vjosa Preniqi, Kyriaki Kalimeri, Charalampos Saitis

TL;DR

This work addresses whether moral values can be inferred from the lyrics associated with a listener’s favorite artists. It combines topic modelling, moral valence (MoralStrength), sentiment (VADER), and emotion (NRC) analyses of lyrics with regression models to predict Moral Foundation scores, leveraging data from the LikeYouth Facebook app (1,386 users) and top-5 artist songs. The findings show that lyrical features, especially those capturing moral content and topics, can predict moral values, with Binding foundations being more predictable ($r \in [0.20,0.30]$) than Individualising ($r \in [0.08,0.11]$), while demographics contribute modestly. The results have implications for psychologically aware music recommendation and targeted messaging, and point to future multimodal extensions combining audio and lyrics for improved inference of moral worldviews.

Abstract

This study explores the association between music preferences and moral values by applying text analysis techniques to lyrics. Harvesting data from a Facebook-hosted application, we align psychometric scores of 1,386 users to lyrics from the top 5 songs of their preferred music artists as emerged from Facebook Page Likes. We extract a set of lyrical features related to each song's overarching narrative, moral valence, sentiment, and emotion. A machine learning framework was designed to exploit regression approaches and evaluate the predictive power of lyrical features for inferring moral values. Results suggest that lyrics from top songs of artists people like inform their morality. Virtues of hierarchy and tradition achieve higher prediction scores ($.20 \leq r \leq .30$) than values of empathy and equality ($.08 \leq r \leq .11$), while basic demographic variables only account for a small part in the models' explainability. This shows the importance of music listening behaviours, as assessed via lyrical preferences, alone in capturing moral values. We discuss the technological and musicological implications and possible future improvements.

"More Than Words": Linking Music Preferences and Moral Values Through Lyrics

TL;DR

This work addresses whether moral values can be inferred from the lyrics associated with a listener’s favorite artists. It combines topic modelling, moral valence (MoralStrength), sentiment (VADER), and emotion (NRC) analyses of lyrics with regression models to predict Moral Foundation scores, leveraging data from the LikeYouth Facebook app (1,386 users) and top-5 artist songs. The findings show that lyrical features, especially those capturing moral content and topics, can predict moral values, with Binding foundations being more predictable () than Individualising (), while demographics contribute modestly. The results have implications for psychologically aware music recommendation and targeted messaging, and point to future multimodal extensions combining audio and lyrics for improved inference of moral worldviews.

Abstract

This study explores the association between music preferences and moral values by applying text analysis techniques to lyrics. Harvesting data from a Facebook-hosted application, we align psychometric scores of 1,386 users to lyrics from the top 5 songs of their preferred music artists as emerged from Facebook Page Likes. We extract a set of lyrical features related to each song's overarching narrative, moral valence, sentiment, and emotion. A machine learning framework was designed to exploit regression approaches and evaluate the predictive power of lyrical features for inferring moral values. Results suggest that lyrics from top songs of artists people like inform their morality. Virtues of hierarchy and tradition achieve higher prediction scores () than values of empathy and equality (), while basic demographic variables only account for a small part in the models' explainability. This shows the importance of music listening behaviours, as assessed via lyrical preferences, alone in capturing moral values. We discuss the technological and musicological implications and possible future improvements.
Paper Structure (10 sections, 4 figures, 5 tables)

This paper contains 10 sections, 4 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Background
  3. Data Collection
  4. Lyrics Content Analysis
  5. Topic Modelling
  6. Moral Valence
  7. Sentiment and Emotion Analysis
  8. Experiments and Results
  9. Conclusion
  10. Acknowledgements

Figures (4)

  • Figure 1: Spearman's Rank correlations of MFT super foundations with demographics, artist likes and lyrics features. We report only ones that were significant at $p\leq.01$. "Artist liked no" refers to the number of artist Likes per user.
  • Figure 2: Top 4 individual feature contributions (via SHAP values) for the five basic moral foundations from experiment EX8 (see Table \ref{['tab:independent_features']}). The higher the SHAP value, the more the feature contributes to the prediction model.
  • Figure 3: Individual (top 8) and grouped feature contributions (via SHAP values) for the two superior moral foundations from experiment EX8 (see Table \ref{['tab:independent_features']}). The higher the SHAP value, the more the feature contributes to the prediction model.
  • Figure :