INTRODUCTION

The influence of popular media in American culture has been widely studied (e.g., Wartella & Reeves, 1985). In particular, rock and roll music is one of the many elements of popular media to be analyzed as an integral part of society (e.g., Curtis, 1987), and it also has been acknowledged as an important channel for the communication and expression of adolescent values, conflicts, attitudes, and emotions (Leger, 1980; Santiago, 1969; Davis, 1985; Wells & Hakanen, 1991). More specifically, Fulton and Owen (1987-88) suggested that media proliferation has altered the ways death is portrayed in society, and that it is within this context that adolescents express their fears and frustrations about it.

Thrush and Paulus (1979) published the first structured analysis of death concerns in popular music and concluded that "... the vast majority of the songs address themselves to death and dying in a manner acceptable to 'median taste'". Teenage "coffin songs" (those ending with the death of one or both teenage lovers) also have been analyzed from the conceptual framework of death as a form of adolescent rebellion (Denisoff, 1983). Attig (1986) provided the most recent and thorough review of death themes in rock and roll lyrics, and he suggested the necessity to "... record the evolution of adolescent music, with particular emphasis on the presence of death and death-related themes, as it moves from its classic period into and through the 1980s".

A comprehensive analysis of death themes in rock music requires a methodology more precise than those employed in previous studies. First, a broad segment of the rock era should be reviewed. Second, there must be a clear and consistent definition of "death song." Third, some of the central features of death songs need to be uniformly coded.

The current investigation, using a rigorous methodology, addressed four questions: (1) How prominent has death been in rock and roll music? (2) How has death been depicted in selected demographic terms, e.g., who dies, how do they die, and what relationships are depicted? (3) What death-related attitudes and behaviors are portrayed in the lyrics of rock and roll death songs? (4) Have there been any trends in the prominence and content of death songs?

METHOD

Two steps were taken to operationally define popular rock and roll death songs. One step was to designate Top 40 "singles" as the record domain because they receive a greater amount of airplay than do album tracks. Another reason for using this set of songs was that the Recording Industry Association of America has indicated that adolescents and young adults make up the largest segment of the Top 40 record-buying public (1989). The Billboard Book of Top 40 Hits (Whitburn, 1991) was selected as the definitive source for the chart history of the 9,311 singles that reached the pop music charts from January 1, 1955 to July 31, 1991.

The second step was to identify "death songs," using the following three-part definition: (1) one or more persons in the song clearly had to have died or the death(s) had to be imminent (going to occur within a few days); (2) the deceased (or soon to be deceased) had to (a) be identified either by name or as a member of a small, identifiable group, (b) be the singer, (c) be in a clearly defined relationship with the singer, or (d) be essential to the meaning of the song, i.e., without mention of the deceased, the song's story line would significantly change; and (3) the deceased could not be a ghost or an apparition.

This strict definition excluded songs that (a) focused on reflections about death or other existential concerns, (b) described fantasies or dreams about people who were not actually dead, or (c) predicted global holocaust. Instrumental versions of various songs also were excluded.

Three broad classifications for death songs also were identified: (a) songs about everyday, common people who either were fictitious or unknown to the general public; (b) songs about celebrities or public figures; and (c) novelty songs, which were defined as those which tended to be humorous or treated death lightly.

Each author independently reviewed all song titles in the Billboard book and selected those songs that met the predetermined criteria of a death song. When both authors were uncertain about a song's lyrics, others who knew the song were consulted or the lyrics were obtained and analyzed.

Five facts about each record were collected during the review process. One, related to song popularity, was the highest position a record reached on the chart. A second, related to death-song chronology, was the date each record entered the chart. The remaining facts focused directly on song content, and included cause of death, gender of the deceased, and relationships described by the lyrics.

The authors constructed an 11-category descriptive code for cause of death and a 10-item descriptive code for relationships (to classify story line interactions). To prevent group deaths from skewing the analysis, single-instance causes that killed single-gender groups were counted only once in either category, e.g., a shipwreck killing 29 men was counted once as a cause of death and once as a male death.

Song titles and lyrics also were analyzed to identify attitudes toward death, responses to death, and references to the tangible realities of death. Subcategories within each of these topic areas were developed during this analysis.

Inevitably, a number of songs containing death-related elements were excluded because they clearly did not meet the operational definition used in this study (e.g., "And When I Die" by Blood, Sweat and Tears). Also, several songs were excluded because listeners would require information in addition to the lyrics to know that they were death songs (e.g., "Everything I Own" by Bread and "Knockin' on Heaven's Door" by Bob Dylan).

RESULTS

A total of 90 different death-related songs was identified. There were 69 songs dealing with the deaths of common people, 13 songs about the deaths of celebrities or public figures, and eight novelty songs containing death-related themes. In addition to the 90 songs used for the majority of the analyses, 12 other versions of these songs, recorded by other than the original artists, were included only in the calculations relating to song popularity. All songs used in this study are listed in the Appendix.

The analysis of chart history showed that death-related songs were substantially more popular than other songs reaching the Top 40. While 25.5% of death-related songs reached the No. 1 spot, only 8.6% of other songs reached that position. Additionally, while 57.8% of death-related songs reached the Top 10, only 36.2% of other songs made it that high on the charts.

Death songs about common people

This type of song was introduced late in the 1950s (seven songs), peaked in the 1960s (35 songs), declined slightly in the 1970s (26 songs), and experienced a precipitous drop-off in the 1980s and early 1990s (eight songs). Chronologically, 96% of the songs reaching the No. 1 position and 81% of the songs reaching the Top 10 did so prior to 1975.

Various causes of death were noted. Murder accounted for 24% of the deaths identified, other violent ends such as executions and shooting of criminals accounted for 24%, accidents caused 16%, and suicides caused 9%. War, illness, and disasters were infrequently cited, but natural causes were mentioned the least. Cause of death was unspecified in 23% of the songs, but in many of these songs, natural causes were implied.

Violence of one kind or another was the leading cause of death in each decade through the 1980s. The only other dominant cause of death/decade was accidents, which accounted for 29% of the deaths in songs from the 1960s. Where gender of the deceased could be determined, 70% were males, a percentage that was fairly consistent over time.

Analysis of the relationships described by the lyrics indicated that peer-romance was present in 30% of the songs, although nearly one-third of these songs were about individuals murdering their lovers. Family relationships were described in 22% of the songs, with husband-wife ties present one-third of the time and parent-child ties present two-thirds of the time.

In 16% of the songs, no relationship was discernible, but in nearly two-thirds of these songs, random murder was evident. Another type of relationship, friendship, appeared in 12% of the songs. Additionally, adversarial situations in the Old West were described in 7% of the songs, and pending executions were described in another 6%. Finally, co-workers were portrayed in 3% of the songs, people's lives were chronicled in 3% of the songs, and teenage rivalry between strangers was depicted in 1% of the songs.