- Caleb Rosebaugh & Lior Shamir -
Department of Computer Science, Kansas State University, Manhattan, Kansas State, United States
Popular music lyrics exhibit clear differences between songwriters. This study describes a quantitative approach to the analysis of popular music lyrics. The method uses explainable measurements of the lyrics and therefore allows the use of quantitative measurements for consequent qualitative analyses. This study applies the automatic quantitative text analytics to 18,577 songs from 89 popular music artists. The analysis quantifies different elements of the lyrics that might be impractical to measure manually. The analysis includes basic supervised machine learning, and the explainable nature of the measurements also allows to identify specific differences between the artists. For instance, the sentiments expressed in the lyrics, the diversity in the selection of words, the frequency of gender-related words, and the distribution of the sounds of the words show differences between popular music artists. The analysis also shows a correlation between the easiness of readability and the positivity of the sentiments expressed in the lyrics. The analysis can be used as a new approach to studying popular music lyrics. The software developed for the study is publicly available and can be used for future studies of popular music lyrics.
automatic quantitative text analytics; basic supervised machine learning; popular music lyrics
Received: February 14, 2022
Revised: April 18, 2022
Accepted: April 29, 2022
Published: July 3, 2022
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-SA 4.0) license (https://creativecommons.org/licenses/by-sa/4.0/).
The style and topics of popular music lyrics have clearly changed over time. Common topics and lyrics style used in the early days of popular music are substantially different from the lyrics of modern popular music, and different popular music artists express themselves by using their unique lyrics styles. For instance, while notable early songwriters focused mostly on love and romance (
Lyrics can exhibit substantial differences between popular music genres (
With the digitization of data and the availability of lyrics in digital formats, several studies used computer analysis for the purpose of analyzing song lyrics. For instance, in the computer science community certain efforts have been invested in automatic classification of popular music lyrics (
The ability to analyze lyrics by using quantitative methods has enabled new paradigms of studying popular music that were not possible in the pre-information era. For instance, the ability to quantify sentiments expressed in thousands of songs has shown the change of sentiments expressed in popular music lyrics over time (
Popular music is one of the most influential forms of art, with impact not merely on leisure, but also on politics, society, and culture. For instance, the dominant societal movements during the 1960s were largely enabled by popular music, and popular music artists played impactful social leadership role during that time (). As music was pivotal in these movements, the cultural and social changes they accomplished could have been different in the absence of popular music (). The Vietnam war era was also heavily impacted by popular music, demonstrating the impact of popular music on political issues (). On the opposite side of that scale, popular music has been widely used by fascist movements (
This study uses a data science approach to identify quantitative stylistic elements that can differentiate between songwriters and provide cues and new knowledge about the unique lyrics styles that characterize different songwriters. That is done by first computing a large collection of quantitative text measurements from each song. These quantitative elements can be used to analyze differences in lyrics written by different songwriters. The computational approach allows to measure stylistic elements used by songwriters, and its automatic nature allows the analysis of very large datasets of songs to extract new knowledge about popular music.
As briefly described above, substantial work has been focused on the ability to classify lyrics automatically, or to measure specific lyrical elements. The approach proposed here makes use of a large number of text measurements applied in concert, therefore allowing to identify text elements that exhibit differences between songwriters in a data-driven manner. That is, the user does not need to make a specific hypothesis or focus on a specific text measurement but can rely on the method to analyze and identify specific text elements. Such measurements can show differences between different songs, albums, songwriters, genres, and more. Instead of using measurements that can discriminate between the songwriters and associate a song with its creating artist automatically, the analysis here is based on using intuitive measurements that can be interpreted manually and can therefore expand the knowledge regarding popular music. Although these text measurements are intuitive to understand, they are difficult to analyze by manual reading of the lyrics. Therefore, the proposed approach provides a new way to analyze and study lyrics that was not practical in the pre-information era.
The dataset used in this study contained text files of lyrics of 89 artists, such that each artist in the dataset is represented by at least 100 songs. The total number of songs was 18,577, with an average of 209 songs per artist. The artist with the lowest number of songs was Guns N’ Roses with 101 songs, and the artists with the highest number of songs was Neil Young with 576 songs. The distribution of the songs of the different artists as well as the year in which the songs were released are shown in
The lyrics files were retrieved from AZlyrics.com. The lyrics files were originally in HTML format and were converted to plain text (ascii). The files were formatted such that the files only included the lyrics, and no other information such as the name of the artist, album, or year of release that were included in the file. Naturally, it is impractical to analyze all popular music artists. Artists that were selected in this study are all well-known influential popular music artists that also had lyrics for at least 100 songs. The artists were also selected such that they represent multiple genres and eras of popular music. Genres included rock, pop, hip-hop, soul, R&B, and heavy metal, as primary genres of popular music. The vast majority of the musicians were mostly active as influential songwriters during the 1960s through the 1990s. The list of musicians excludes more recent artists, as scholarly work has not yet been done by the scientific community to fully analyze and characterize these artists, making it more difficult to associate the lyrics elements with qualitative analysis of their lyrics style. For that reason, most artists used in this study were active in previous decades, while artists that were mostly active in the third millennium are not widely represented in the dataset. The full list of artists and the years during which they were active is provided in Appendix A.
The lyrics were analyzed using Udat (
Readability. Automated Readability Index (
Average sentiments. Sentiments expressed in the lyrics, as computed by automatic sentiment analysis (). Each song is assigned with a sentiment value from 1 to 5, such that 1 reflects very negative sentiment, 3 is neutral, and 5 reflects very positive. The sentiment analysis works by using 215,154 labeled phrases and 11,855 sentence that make a parse tree analyzed by a deep recurrent neural network (RNN). While the analysis process is not necessarily intuitive, the output is a simple score that reflects the estimated sentiment expressed in the text.
Differences in sentiments. Sentiment difference is measured by the difference between the most positive sentence in the lyrics and the most negative sentence in the lyrics. That measurement reflects the variations in the sentiments expressed in the lyrics. Songs that include negative and positive sentences have a higher sentiment difference. Songs where the different sentences are more consistent in the sentiments, they express have a lower sentiment difference.
Punctuation characters. The method measures the frequency of the use of different punctuation characters. That is the fraction of characters such as ‘?’, ‘!’, ‘:’, etc. in the total number of characters in the lyrics.
Word length. Word length statistics includes the word length mean, word length standard deviation, and histogram of the length of the different words in the text. That measurement reflects the use of words of different lengths in the lyrics.
Lyrics length. Statistics of the distribution of the length of the lyrics, including the mean and standard deviation of the number of words in the song.
Sounds. Diversity of sounds using the Soundex algorithm (
Word diversity. Diversity of the words appearing in the lyrics. Determined by the number of unique words divided by the total number of words. A song that uses the same words repetitively will have a lower word diversity score.
Parts of speech. Frequency of different parts of speech. The parts of speech are tagged automatically by using the CoreNLP Natural Language Processing library (
Topics. Frequency of different topics mentioned in the lyrics. The topics are determined by a predefined thesaurus that covers a range of different topics and words that identify them.
The complete list is provided in
The numerical text content descriptors allow the identification of specific differences between the way different songwriters express themselves through their lyrics. Additionally, it allows certain machine learning tasks by applying machine learning algorithms to the numerical content descriptors computed from the text. Udat implements the Weighted Nearest Distance (WND) algorithm (
According to WND, the association between a song and a musician is determined by the musician
where T is the set of songs used for training,
where
The main goal of Udat is not necessarily to perform automatic classification, but to extract knowledge from the data. The WND algorithm is used for its ability to estimate the similarity between different classes (
The Udat method described above can perform several data science tasks related to analysis of text data, including supervised and unsupervised learning, profiling similarities between text classes, outlier detection, feature selection, and identification of specific text elements that can differentiate between text classes (
While numerous previous studies focused on automatic classification of lyrics by their musician or songwriter (
The machine learning algorithm described above was applied to associate between lyrics and artists automatically. That is, the songs of each artist are separated into training and test songs. Eighty songs of each artist were used for the machine learning algorithm to “learn” the style of the lyrics, and the remaining 20 songs of each artist were used to test whether the machine learning algorithm can associate the lyrics with the artist automatically. The machine learning system first “learns” from the 80 training songs of each artist. After the training, the machine learning algorithm attempts to associate each of the 20 test songs with their artist. The accuracy of the algorithm is measured by the number of songs that the algorithm was able to associate correctly with their artist, divided by the total number of songs that the algorithm attempts to associate with an artist, including the songs that were associated with the wrong artist.
The results of the experiment showed that 12.3% of the songs were associated by the algorithm with their correct artist. While that is not necessarily very high accuracy, with 89 artists in the dataset a random guess of the artist would lead to accuracy of 1/89, which is ~1.1%. That shows that even if the machine learning algorithm cannot associate between the lyrics and the artist in all cases, the fact that it makes the association in higher rate than random guess shows that there is information in the quantitative elements that can show patterns that are unique to a certain artist. That is done without identifying certain words or pattern of words that are used differently by different songwriters, but merely with the quantitative measurements. It is important to mention that associating lyrics with an artist is a complex cognitive task, and even humans might find it difficult to associate lyrics with the relevant creating artist, unless they are already familiar the specific lyrics.
To understand the specific differences between the lyrics of popular music artists, the specific measurements were compared.
The figures show that musician with higher readability index include Bob Dylan, Alanis Morrissette, Elton John (Bernard Taupin), Bruce Springsteen, and the bands Yes, Rush, and Traffic. Hip-hop artists such as Gill Scott Heron and the Beastie Boys are ranked with a high readability index, as well as the rapper Dr. Dre (Andre Romelle Young). Hip-hop and rap are musical styles that allow longer sentences, and consequently more complex lyrics. It has been proposed that rap allows to express deeper ideas regarding social issues or personal experiences (
Artists with lower readability index include Guns N’ Roses, U2, Van Halen, and Ringo Starr. Like Traffic lyrics being characterized by different measurements that indicate on complexity of the lyrics, Starr’s lyrics showed several unrelated measurements that indicate on simplicity. These measurements will be discussed later in this section. Among the popular music artists whose lyrics are the easiest to read is also Neil Young, who is considered a notable lyricist (;
It is interesting that some of the artists with the lower readability index are also some of the artists that their lyrics express the most positive sentiments. The correlation between the readability of the lyrics and the sentiment expressed in the lyrics was tested by correlating the readability index and the sentiment expressed in the 18,557 songs. That was done by using the Pearson correlation, which was 0.21. With 18,557 songs, the probability to have such correlation by chance is lower than 0.00001. That provides quantitative evidence that songs that are less positive also tend to use longer words and sentences, making them somewhat more difficult to read.
The frequency of very positive sentences is not necessarily a reflection of the overall positivity of the lyrics, as very positive sentences and negative sentences can be used by the same artists and even in the same song. The data shows that while different artists express very positive sentences differently, that cannot be associated with a certain genre or lyrics style. For instance, Pink Floyd and Billy Joel tended to express a darker lyrics style and often focused on political and social topics in their songs (;
The diversity of sounds in Steely Dan lyrics can be associated with the different thematic tones used as a dominant aspect in the lyrics (;
Gender expressed in popular music has been a topic of interest in popular music research (
Similarly,
The gender-based use of gender-related terms changes between different artists. For instance, Imagine Dragons use gender-related terms infrequently regardless of whether the terms are related to men or women. On the other hand, The Clash use men related terms in high frequency, while women-related terms are far less frequent in their lyrics. For the Rolling Stones and the Beatles, however, women terms are more frequent than men-related terms. Both bands have been focusing in their lyrics on topics related to love and romance. For instance, it was only until their sixth studio album that the Beatles included their first notable song that did not focus on love and romance - “Nowhere man”. The graphs do not show a clear difference between different genres, and musicians within the same genres show substantial differences in the way they use gender-related terms.
The graph shows substantial differences between the degree of word repetition. The artist with the most diverse lyrics as reflected by the use of different words is Steely Dan. The lyrics of Steely Dan has attracted a relatively significant interest in the music research community (;
On the other end of that graph, Ringo Starr is characterized by low diversity of words and low readability index. That combination can be an indication of a simple approach in Starr’s lyrics style. With Starr’s positive sentiments shown in
The availability of digital platforms allows quantitative analysis that was impractical to perform in the pre-information era. For instance, even a simple task such as counting the number of words or the frequency of terms that appear in popular music lyrics is a daunting task that requires substantial labor. Computer analysis allows to quantify these text elements in a large number of songs, enabling a new approach to the studying of popular music. Here we extract several intuitive text measurements from popular music lyrics. The fact that a machine learning classifier can identify the artists by the lyric’s elements in accuracy much higher than mere chance shows that popular music artists have a certain style in their lyrics reflected by the collection of quantitative measurements used here.
Machine learning classifiers that can associate lyrics with their creating artist are not new. However, identifying the artist automatically does not necessarily lead to new knowledge about popular music. Machine learning-based document classifiers tend to work by complex data-driven rules that act as a “black box”, and therefore often do not provide the user with substantial new knowledge that the user does not already know. The study here is limited to intuitive explainable measurements of the lyrics, and therefore can be used to profile the differences and identify specific quantitative features that can reflect differences between songwriters. Because the text elements are explainable and intuitive, the analysis is interpretable, and can therefore lead to new knowledge about popular music. Such differences can provide ques for further studying of the uniqueness of lyricists or identifying similarities and influential links between popular music songwriters. The analysis shown here identified several differences between songwriters and demonstrates that songwriters are different from each other in a manner that can be quantified and measured. Lyrics elements expressed by songwriters, whether intentional or subconscious, can be identified and quantified by applying computer analysis. That approach can expand the set of tools that can be used for analyzing popular music.
The results of the computer analysis led to several observations regarding the songwriters. For instance, several mathematically unrelated measurements of Ringo Starr’s lyrics show a positive and simple lyrical style. Early Hip-hop artists such as Gil Scott-Heron or the Beastie Boys, as well as rock bands that featured improvisation-like musical pieces tended to express more complex lyrics. The analysis also showed that artists with lower readability index also tend to express more positivity in their songs. When applying the analysis to all songs in the dataset, the correlation between the positivity expressed in the lyrics and the readability index showed that lyrics that express positive sentiments are also easier to read, while less positive songs also have more complex lyrics. Although the link between sad songs and the expression of complex has been proposed (
The quantitative approach used in this study can be used for more detailed analysis, involving different artists and different research questions. Therefore, the Udat software developed for this study is available for free download (
This research was supported in part by NSF grant AST-1903823.
Alper, G. (2022). Not just derision and darkness: The interplay of lyrics and music in Steely Dan’s compositions.
An, Y., Sun, S., & Wang, S. (2017). Naive Bayes classifiers for music emotion classification based on lyrics.
Ballard, M. E., Dodson, A. R., & Bazzini, D. G. (1999). Genre of music and lyrical content: Expectation effects.
Bindas, K. J. (1993). The future is unwritten: The Clash, punk and America, 1977-1982.
Bindas, K. J., & Houston, C. (1989). “Takin’ care of business”: Rock music, Vietnam and the protest myth.
Borshuk, M. (2021). “Steely Dan at 50.”
Clements, P. (2009). Cultural legitimacy or ‘outsider hip’? Representational ambiguity and the significance of Steely Dan.
Cohen, S. (2001). Popular music, gender and sexuality. In J. Street, S. Frith, & W. Straw (Eds.),
Coleman, M., & Liau, T. L. (1975). A computer readability formula designed for machine scoring.
Condit-Schultz, N., & Huron, D. (2015). Catching the lyrics: Intelligibility in twelve song genres.
Davies, P. (1990). “There’s no success like failure”: From rags to riches in the lyrics of Bob Dylan.
de Boise, S. (2020). Music and misogyny: A content analysis of misogynistic, antifeminist forums.
Dunlap, J. (2006). Through the eyes of Tom Joad: Patterns of American Idealism, Bob Dylan, and the Folk Protest Movement.
Echard, W. (2005).
Edwards, W. (2002). From poetry to rap: The lyrics of Tupac Shakur.
Everett, W. (2004). A royal scam: The abstruse and ironic bop-rock harmony of Steely Dan.
Fell, M., & Sporleder, C. (2014). Lyrics-based analysis and classification of music.
Flynn, M. A., Craig, C. M., & Anderson, C. N. (2016). Objectification in popular music lyrics: An examination of gender and genre differences.
Fox, W. S., & Williams, J. D. (1974). Political Orientation and Music Preferences Among College Students.
Freudiger, P., & Almquist, E. M. (1978). Male and female roles in the lyrics of three genres of contemporary music.
Fricke, D. (2001, December 27). “Imagine”: The anthem of 2001.
Gosa, T. L. (2017). Hip hop, authenticity, and styleshifting in the 2016 presidential election.
Hess, M. (2005). Hip-hop realness and the white performer.
Hewett, M. R. (2016). Two linguistic case studies of the craft of songwriting: “Imagine” and “Like a Rolling Stone.”
Hobson, J. (2021). A hard day’s night.
Kresovich, A., Reffner Collins, M. K., Riffe, D., & Carpentier, F. R. D. (2021). A content analysis of mental health discourse in popular rap music.
Kutschke, B. (2016). Political music and protest song. In K. Fahlenbrach, M. Klimke, & J. Scharloth (Eds.),
Lammer, J. (2016).
Logan, B., Kositsky, A., & Moreno, P. (2004). Semantic analysis of song lyrics.
Manning, C., Surdeanu, M., Bauer, J., Finkel, J., Bethard, S., & McClosky, D. (2014). The Stanford CoreNLP Natural Language Processing toolkit.
Martin, P. J. (2006). Musicians’ worlds: Music-making as a collaborative activity.
Mise, U. (2020). Rap music as resistance and its limits, two diverging cases: Sulukule and Bağcılar rap.
Mori, K., & Iwanaga, M. (2014). Pleasure generated by sadness: Effect of sad lyrics on the emotions induced by happy music.
Napier, K., & Shamir, L. (2018). Quantitative sentiment analysis of lyrics in popular music.
Nielson, E. (2009). “My president is black, my lambo’s blue”: The Obamafication of rap?
North, A. C., Krause, A. E., & Ritchie, D. (2021). The relationship between pop music and lyrics: A computerized content analysis of the United Kingdom’s weekly top five singles, 1999–2013.
Odell, M. K. (1956). The profit in records management.
Orlov, N., Shamir, L., Macura, T., Johnston, J., Eckley, D. M., & Goldberg, I. G. (2008). WND-CHARM: Multi-purpose image classification using compound image transforms.
Ortega, J. L. (2021). Cover versions as an impact indicator in popular music: A quantitative network analysis.
Petrie, K. J., Pennebaker, J. W., & Sivertsen, B. (2008). Things we said today: A linguistic analysis of the Beatles.
Ray, M. (2013).
Richardson, J. E. (2017). Recontextualization and fascist music. In L. C. S. Way & S. McKerrell (Eds.),
Rozinski, T. (2015). Using music and lyrics to teach political theory.
Ruth, N. (2019). “Where is the love?” Topics and prosocial behavior in German popular music lyrics from 1954 to 2014.
Salkin, P., & Crisci, I. (2015). Billy Joel: The chronicler of the suburbanization in New York.
Setiawan, A. (2013). Analysis on anti capitalism in the “Clampdown” lyric by The Clash.
Shamir, L. (2017). UDAT: A multi-purpose data analysis tool.
Shamir, L. (2021). UDAT: Compound quantitative analysis of text using machine learning.
Shamir, L., Macura, T., Orlov, N., Eckley, D. M., & Goldberg, I. G. (2010). Impressionism, expressionism, surrealism: Automated recognition of painters and schools of art.
Shamir, L., Orlov, N., Eckley, D. M., Macura, T., Johnston, J., & Goldberg, I. G. (2008). Wndchrm – an open source utility for biological image analysis.
Smith, E. A., & Senter, R. J. (1967). Automated readability index.
Socher, R., Perelygin, A., Wu, J., Chuang, J., Manning, C. D., Ng, A., & Potts, C. (2013). Recursive deep models for semantic compositionality over a sentiment treebank.
Strong, M. C. (2000).
Thrasher’s Wheat. (2004, April 14).
Tomiyama, H. (2017). Neil Young: Some complexities in his songs. In T. Connolly & T. Iino (Eds.),
Tsaptsinos, A. (2017).
Vandagriff, R. S. (2015). Talking about a Revolution: Protest Music and Popular Culture, from Selma, Alabama, to Ferguson, Missouri.
Werner, V. (2012). Love is all around: A corpus-based study of pop lyrics.
West, A., & Martindale, C. (1996). Creative trends in the content of beatles lyrics.
Whiteley, S. M. (2013). Popular music, gender and sexualities.
Williams, M. L. (2020). “Meditate, don’t medicate!” An analysis of addict rap, black men’s social issues, and J. Cole’s K.O.D. album.
Yang, Y. (2020). “Musicalization of the culture”: Is music becoming louder, more repetitive, monotonous and simpler?
Yeh, C.-H., Tseng, W.-Y., Chen, C.-Y., Lin, Y.-D., Tsai, Y.-R., Bi, H.-I., Lin, Y.-C., & Lin, H.-Y. (2014). Popular music representation: Chorus detection & emotion recognition.
Yoo, Y., Ju, Y., & Sohn, S. Y. (2017). Quantitative analysis of a half-century of K-Pop songs: Association rule analysis of lyrics and social network analysis of singers and composers.
|
|
|
|
|
| Alanis Morissette | 126 | 1991 | 2020 |
| Alice Cooper | 330 | 1969 | 2019 |
| Aretha Franklin | 310 | 1961 | 2011 |
| Bad Company | 127 | 1974 | 1996 |
| Beastie boys | 121 | 1986 | 2011 |
| Billy Idol | 111 | 1981 | 2014 |
| Billy Joel | 129 | 1971 | 1993 |
| Black Sabbath | 177 | 1970 | 2016 |
| Blue Oyster Cult | 123 | 1972 | 2001 |
| Bob Dylan | 484 | 1962 | 2017 |
| Bob Seger | 210 | 1969 | 2017 |
| Bon Jovi | 286 | 1984 | 2020 |
| Bruce Springsteen | 354 | 1973 | 2019 |
| Cheap Trick | 229 | 1977 | 2017 |
| Chicago | 275 | 1969 | 2019 |
| David Bowie | 349 | 1967 | 2017 |
| Deep Purple | 189 | 1968 | 2020 |
| Def Leppard | 179 | 1980 | 2015 |
| Electric Light Orchestra | 189 | 1971 | 2019 |
| Elton John | 419 | 1969 | 2016 |
| Eric Clapton | 357 | 1970 | 2018 |
| Fleetwood Mac | 255 | 1968 | 2013 |
| Genesis | 180 | 1969 | 2000 |
| George Harrison | 130 | 1970 | 2002 |
| Green Day | 159 | 1990 | 2020 |
| Guns N' Roses | 101 | 1987 | 2008 |
| Heart | 170 | 1976 | 2016 |
| Imagine Dragons | 106 | 2008 | 2018 |
| Iron Maiden | 183 | 1980 | 2015 |
| Jackson Browne | 189 | 1967 | 2014 |
| Jeff Beck | 120 | 1968 | 2016 |
| Jethro Tull | 253 | 1968 | 2003 |
| Jimi Hendrix | 103 | 1967 | 2013 |
| John Fogerty | 121 | 1973 | 2013 |
| John Lennon | 121 | 1970 | 1980 |
| John Mellencamp | 258 | 1976 | 2018 |
| Journey | 170 | 1975 | 2011 |
| Judas Priest | 223 | 1974 | 2018 |
| Kiss | 221 | 1974 | 2012 |
| Lenny Kravitz | 144 | 1989 | 2018 |
| Lou Reed | 243 | 1972 | 2011 |
| Lynyrd Skynyrd | 174 | 1974 | 2012 |
| Madonna | 265 | 1983 | 2019 |
| Marvin Gaye | 243 | 1961 | 2019 |
| Meatloaf | 156 | 1977 | 2016 |
| Metallica | 146 | 1983 | 2016 |
| Michael Jackson | 217 | 1972 | 2014 |
| Motley Crue | 156 | 1981 | 2008 |
| Motorhead | 266 | 1977 | 2015 |
| Neil Young | 576 | 1969 | 2019 |
| Nick Cave & The Bad Seeds | 191 | 1984 | 2019 |
| Oasis | 117 | 1994 | 2008 |
| Ozzy Osbourne | 153 | 1980 | 2020 |
| Paul McCartney | 343 | 1970 | 2018 |
| Peter Frampton | 126 | 1972 | 2019 |
| Phil Collins | 119 | 1981 | 2019 |
| Pink Floyd | 139 | 1967 | 2015 |
| Prince | 507 | 1978 | 2019 |
| Queen | 186 | 1973 | 2008 |
| R.E.M | 192 | 1983 | 2008 |
| Radiohead | 144 | 1993 | 2016 |
| Ramones | 191 | 1976 | 1994 |
| Ringo Starr | 248 | 1970 | 2019 |
| Robert Plant | 149 | 1982 | 2017 |
| Rod Stewart | 398 | 1969 | 2018 |
| Rush | 170 | 1974 | 2012 |
| Sammy Hagar | 181 | 1976 | 2019 |
| Santana | 211 | 1969 | 2019 |
| Scorpions | 222 | 1972 | 2015 |
| Steve Miller Band | 172 | 1968 | 2011 |
| Stevie Nicks | 128 | 1981 | 2014 |
| Stevie Wonder | 275 | 1962 | 2005 |
| Styx | 173 | 1972 | 2017 |
| Ted Nugent | 111 | 1974 | 2018 |
| The Beatles | 295 | 1963 | 1996 |
| The Clash | 130 | 1977 | 1993 |
| The Doobie Brothers | 132 | 1971 | 2010 |
| The Doors | 109 | 1967 | 1978 |
| The Grateful Dead | 167 | 1967 | 1989 |
| The Kinks | 402 | 1964 | 1993 |
| The Moody Blues | 209 | 1965 | 2003 |
| The Rolling Stones | 360 | 1964 | 2016 |
| The Who | 195 | 1965 | 2019 |
| Thin Lizzy | 127 | 1971 | 1983 |
| Tom Petty and the Heart Breakers | 201 | 1976 | 2014 |
| U2 Band | 224 | 1980 | 2017 |
| Van Halen | 119 | 1978 | 2012 |
| Yes | 172 | 1969 | 2014 |
| ZZ Top | 166 | 1971 | 2012 |
Source: AZLyrics.com, authors’ estimation.