Cultural evolution of emotional expression in 50 years of song lyrics

Scripts and data associated to the preprint: "Cultural evolution of emotional expression in 50 years of song lyrics" by Charlotte Brand, Alberto Acerbi, and Alex Mesoudi

This repository contains the R scripts and the data to reproduce the analysis described in the preprint.

The main analysis script is analysis_final.R

Here is the content of the various folders:

data

The processed datasets used in the analysis:

• billboard_analysis.csv
• mxm_analysis.csv

Results

A summary of the models' results:

• Results_fin

The plots of the results:

• bb_neg_full.png
• bb_pos_full.png
• mxm_neg_full.png
• mxm_pos_full.png

And the script to generate them:

• predPlots.R

LIWC

The lists of emotion-words used to process the original data:

• negemo.csv
• posemo.csv

The same lists, but without swear words (as used for analyses in supplementary material):

• negemo_no_swear.csv
• posemo_no_swear.csv

The non-stemmed lists (expansions of the original LIWC 2007, that contains wildcards):

• negemo_non_stemmed.csv
• posemo_non_stemmed.csv

collaboration_processing

An R script used to disambuigate collaborations in the original data (see the preprint from the details):

• disambiguate_collaborations.R