A dataset hand-compiled by Sam Raker, based on lyrics taken from A-Z Lyrics Universe, with additional fact-checking based on listening to the songs.

This is a JSON-formatted dataset of things that happen to women in songs performed by the Decemeberists. Note: use --jsonArray flag if you're going to be importing this data into mongodb.

dec_women_fates.json is the (more) human-readable version. dec_women_fates_min.json is the same data, minified with json-minify. The .dat folder is for use with dat, which you should check out 'cause it's pretty cool.

Explanation: The name of the album
Type: string

Explanation: The name of the song
Type: string (unique)

Explanation: The female characters in the song
Type: array of woman objects

Explanation: General notes for the song
Type: string or null

Explanation: If capitalized, the woman's name. If uncapitalized, the woman's title or another helpful identifier.
Type: string

Explanation: What happens to the woman in the song.
Type: string or null

Explanation: An integer code corresponding to the severity of the woman's fate

• -1 : something basically positive
• 0 : null/none/neutral
• 1 : something unfortunate or harmful, but non-fatal
• 2 : death from natural causes
• 4 : murder
• 8 : rape
• 12 : rape + murder

Type: integer

Explanation: Notes on the character
Type: string or null

This is very much an alpha dataset. Please comment on/pull-request/fork it to make improvements.

Since the general idea with the code values is that they can be added together to represent multiple misfortunes befalling a character, multiple songs describing the same character in the same position would only have code values > 1 the first time, and code = 0 subsequently, until something else happens. Thus, Margaret in The Hazards Of Love is kidnapped in "The Abduction of Margaret" and remains in captivity until "The Hazards of Love 4 (The Drowned)." There are several intervening songs that describe or reference her captivity, but since it's 'the same' kidnapping event, she gets fate: null and code: 0 for those songs.

• I decided that characters referenced only as 'you' should be counted as women, absent any countervailing context clues (e.g., "The Soldiering Life"). Is this too hand-wavy?
• I tried to include all female characters referenced at all, including all the wives in "16 Military Wives" and all Margaret's fellow maidens in "The Hazards of Love 1". Should the dataset be restricted solely to 'major' characters?
• Eventually, I should probably go back and do the same thing for male characters in the songs.