The lyrics retrieved from Genius.com provide a rich look into the language used in mainstream music throughout the last sixty years. There is a lot to analyse, so the page is divided into three sections. The first section goes over the differences and similarities between genres. Next comes a section exposing the trends of each decade, and lastly, the reader is invited on a deep dive into the linguistic proficiency of the most successful artist in modern times.
Investigating each genre
For the text part, we have chosen the following genres to inspect:
pop, rock, rap, folk, blues, country, uk, funk and r&b
These genres were chosen as they have been popular through time, cover a large portion of the world of music and show exciting differences and trends worth looking into.
Pop, in particular, is an interesting genre. It is defined as a genre of popular music (they are different, although often used interchangeably) that started developing in the mid-1950s. Back then, pop music encompassed rock and roll, and they were essentially synonymous until the ’70s, at which point pop changed to become more accessible to wider audiences (source: wikipedia).
Let’s look at what genres we listened to throughout the last six decades. The plot below shows the ratio between genres as a % of the total amount of genres. Note that not all genres are included, but only a pre-selected list of genres.
We can clearly see that over 50% of all music on the Billboard ‘Hot-100’ chart was pop, with rock starting out on second and R&B and blues coming in on third and fourth. In 1990 R&B would take over as second, but more notably, rap would become the third most represented genre on the list, rising from almost 0 the decade before. The popularity of rap would increase until it, in 2010, overtook pop music as the most represented genre on the ‘Hot-100’ chart. In 2020, we also see trap music overtake pop.
Now, let’s inspect the happiness of the different genres:
We can see that most genres tend to have the same sentiment. All genres of music we have identified fall between an average sentiment of 5.4 and 5.7, and as one would probably expect, rap and trap are at the lower end of that spectrum. However, all genres are still above the average sentiment of all the words in the Hedonometer data. The hedonometer score is assigned from a range of (1) sad to (9) happy. Many profanities fall in the range between 4 - 6, because of the uncertainty of the context they are used in.
Let’s take a closer look at the important words for each genre. We do this by calculating a term frequency for inverse document frequencies (TF-IDF). What is that? Basically, it measures how important and unique a word is to a document in a collection of documents (corpus). The TF-IDF score becomes larger when a word appears frequently in the document, and it becomes smaller when the word is common in the corpus, e.g. appears in many of the other documents. (for more, see: explainer notebook or wikipedia). TF-IDF is used instead of simply the term frequency because many words are common across the entire corpus, and will not show much, except the most commonly used words.
Using the words deemed important by the TF-IDF, we can create a representation of the scores with wordclouds:
Some interesting trends we can see:
The pop genre seems to be about love, heartbreak and partying typically. We see words such as broken, breaking, fear and return but also words such as promised and amor. And then, of course, there are the party words: chorus, party, fame etc.
Word
TF
Word
TFIDF
1.
know
0.0100
chorus
0.00009
2.
love
0.0099
miscellaneous
0.00007
3.
oh
0.0077
broken
0.00006
4.
like
0.0077
party
0.00006
5.
got
0.0072
breaking
0.00006
6.
time
0.0070
breathe
0.00005
7.
go
0.0066
rainbow
0.00005
8.
one
0.0063
happen
0.00005
9.
na
0.0062
nigga
0.00005
10.
see
0.0062
spoken
0.00005
Folk music is often about cultural or national identities. The words seem to suggest some wasted opportunity, with words such as annihilation, squandered and pursuit valued highest.
Word
TF
Word
TFIDF
1.
know
0.0085
annihilation
0.00031
2.
like
0.0079
squandered
0.00025
3.
time
0.0065
ragged
0.00022
4.
wa
0.0064
sunlit
0.00021
5.
love
0.0063
birch
0.00021
6.
one
0.0057
knowed
0.00021
7.
come
0.0056
canal
0.00021
8.
go
0.0055
suppertime
0.00018
9.
say
0.0052
pursuit
0.00018
10.
day
0.0052
bojangles
0.00018
Blues originates from the deep south of the United States. Important words are layla, yakka and enriched, although all the words seem to have the same importance. This could be explained by the low number of blues songs - only 149.
Word
TF
Word
TFIDF
1.
know
0.0097
layla
0.00039
2.
oh
0.0096
yakka
0.00035
3.
love
0.0083
enriched
0.00035
4.
got
0.0082
flaying
0.00035
5.
baby
0.0082
goanna
0.00035
6.
like
0.0076
conveniency
0.00035
7.
yeah
0.0073
alluded
0.00035
8.
go
0.0069
pled
0.00035
9.
time
0.0068
seeped
0.00035
10.
na
0.0068
scroungy
0.00035
R&B stands for rhythm and blues. It encompasses a lot of genres, and today a lot of rap and electronic dance music is classified as R&B as well. The important words are mostly slang terms used in African American communities, which makes sense as it originated in black communities in the 1940s.
Word
TF
Word
TFIDF
1.
know
0.0097
nigga
0.00024
2.
love
0.0089
shawty
0.00015
3.
oh
0.0083
shorty
0.00015
4.
baby
0.0080
wit
0.00013
5.
got
0.0079
hoe
0.00012
6.
yeah
0.0077
crib
0.00012
7.
like
0.0075
playa
0.00011
8.
na
0.0070
pussy
0.00011
9.
get
0.0067
booty
0.00010
10.
time
0.0066
dick
0.00009
Country is a genre of music often associated with western cowboy music, living on a farm and driving tractors. The words deemed important reflect this, with hillbilly and tailgate ranking highest, followed by a bunch of other terms often associated with farming. It is definitely safe to say that the country genre lives up to the stereotype.
Word
TF
Word
TFIDF
1.
know
0.0084
hillbilly
0.00021
2.
like
0.0082
tailgate
0.00021
3.
love
0.0078
tractor
0.00018
4.
got
0.0070
porch
0.00018
5.
time
0.0065
redneck
0.00016
6.
one
0.0062
floorboard
0.00013
7.
go
0.0061
hank
0.00013
8.
get
0.0059
gravel
0.00012
9.
wa
0.0059
bocephus
0.00012
10.
yeah
0.0055
southern
0.00012
Rock originated in the ’50s and ’60s and took the world by storm, and it spawned a myriad of sub-genres. Rock seems to be a mix of pop, country, and some other genres, as it contains much of the same words. The many variations of ‘break’ implies that hardships in life is a common topic.
Word
TF
Word
TFIDF
1.
know
0.0092
broken
0.00007
2.
love
0.0078
breathe
0.00007
3.
like
0.0075
tailgate
0.00006
4.
got
0.0071
fear
0.00006
5.
time
0.0067
sailor
0.00006
6.
oh
0.0066
redneck
0.00005
7.
go
0.0063
breaking
0.00005
8.
get
0.0060
escape
0.00005
9.
one
0.0060
southern
0.00005
10.
say
0.0058
floorboard
0.00005
Rap is a music genre primarily developed by urban black communities in the United States. It has a certain vocal rhythm, almost more like a spoken song, compared to more traditional genres of music. The tone of rap is harsher than in other genres, with many derogatory names for women being a mainstay of the genre. Furthermore, often rappers are said to have beef with one another, which is reflected in their songs.
Word
TF
Word
TFIDF
1.
like
0.0056
nigga
0.00065
2.
got
0.0055
hoe
0.00046
3.
know
0.0052
dawg
0.00037
4.
get
0.0051
rapper
0.00031
5.
yeah
0.0044
bitch
0.00028
6.
ai
0.0044
pussy
0.00027
7.
go
0.0042
dick
0.00027
8.
back
0.0038
opps
0.00024
9.
make
0.0038
beef
0.00023
10.
see
0.0038
wit
0.00022
Funk is a music genre that originated in the 1960s. The TF-IDF suggests that funky musicians simply love to funk it up, with top words being funk, funky and funkin.
Word
TF
Word
TFIDF
1.
know
0.0096
looka
0.00042
2.
oh
0.0089
funk
0.00024
3.
love
0.0085
maceo
0.00022
4.
got
0.0085
funky
0.00018
5.
get
0.0080
funkin
0.00016
6.
yeah
0.0079
wit
0.00015
7.
baby
0.0076
jab
0.00012
8.
like
0.0073
aflame
0.00012
9.
na
0.0072
maganoo
0.00012
10.
time
0.0069
karat
0.00012
UK, or “grime” as it is often called, originated in the 2000s. British communities primarily develop it, and as expected, the most characteristic words for the genre are British slang and ad-libs. Specifically, cah means because, blud means friend and paigon means opponent.
Word
TF
Word
TFIDF
1.
know
0.0094
mum
0.00023
2.
love
0.0082
cah
0.00016
3.
oh
0.0074
greaze
0.00016
4.
like
0.0073
uk
0.00014
5.
got
0.0066
transmission
0.00014
6.
time
0.0064
arsehole
0.00013
7.
go
0.0062
blud
0.00012
8.
see
0.0057
paigons
0.00012
9.
never
0.0057
krishna
0.00012
10.
ca
0.0055
cuh
0.00011
Music through the decades
We now take a look at the music through the decades. We start by looking at the sentiment through the years.
We see much the same trend as with decades - higher sentiment values in the early decades than in the later. This is probably explained by the rise of rap and hip hop from the ’90s and onwards. Let’s also try to look at the sentiment more closely:
We can see much the same trend here as in the above plot. However, it is clear that there are spikes in sentiment, going both up and down. However, it is important to note that almost all songs are still above the average sentiment from labMT, which could indicate that generally, we just prefer to listen to happier music since all the songs we are investigating are only taken from the ‘Hot-100’ chart.
To look at the lyrics used through the years, we split our corpus by release year instead of by genre. Once again, we compute the TF-IDF for each decade, and that leaves us with the following wordclouds:
The words which according to the TF-IDF score define each decade are vastly different when dividing the decades according to the two rows in the above figure. The ’60s, ’70s and ’80s all contain words which are completely normal words which everyone might use in their everyday life. Some perhaps more expressive and expressive than ordinary speech, but still real words. Some quite romantic words like tenderly and gentleness are also used. The wordclouds of the ’10s and ’20s are the polar opposite to this. One has to search the clouds quite extensively to find words which appear in the dictionary. All the important words are ad-libs used for rhythm in rap songs and slang. The ’90s and ’00s mark the transition between the two extremes. It is evident that as rap emerged, calling out artist names became more common. Further description of the wordclouds can be seen below.
The 1960s were quite revolutionary for music. Rock was becoming more evolved, and artists were beginning to release more albums than singles.
Looking at the defining words of the decade, a lot of people seemed to enjoy partaking in the watusi dance. Furthermore, songs were affectionate, using words such as tenderly and fickle.
Word
TF
Word
TFIDF
1.
love
0.0122
watusi
0.00011
2.
know
0.0103
tenderly
0.00009
3.
oh
0.0083
looka
0.00007
4.
go
0.0069
sighin
0.00007
5.
got
0.0069
hully
0.00006
6.
like
0.0068
rovin
0.00006
7.
come
0.0067
billow
0.00006
8.
one
0.0066
fickle
0.00005
9.
baby
0.0065
twine
0.00005
10.
time
0.0064
doggone
0.00005
The 1970s are probably best known for the rise and popularity of disco. If you simply look at the TF-IDF scores of the lyrics, you might believe that nigger was the most defining word of the decade. However, five different songs in 1970 had the word appear in them (in mostly provoking context), and it does not appear in any other decade. Another interesting word is doggone, whose modern counterpart you might be more familiar with - damn.
Word
TF
Word
TFIDF
1.
know
0.0099
nigger
0.00006
2.
love
0.0098
doggone
0.00005
3.
got
0.0079
gentleness
0.00005
4.
oh
0.0078
toad
0.00004
5.
like
0.0073
unkind
0.00004
6.
time
0.0071
salina
0.00004
7.
get
0.0065
thoughtful
0.00004
8.
come
0.0063
softness
0.00004
9.
go
0.0062
crowing
0.00004
10.
na
0.0061
marianne
0.00004
The 1980s had the rise of electronic dance music and modern rock. Most of the important words from this decade appear to be quite normal at a glance. Apparently, musicians really liked jellybeans in the ’80s!
Word
TF
Word
TFIDF
1.
know
0.0100
glancing
0.00005
2.
love
0.0098
temperamental
0.00005
3.
time
0.0078
marketplace
0.00004
4.
got
0.0074
untried
0.00004
5.
oh
0.0072
jellybean
0.00004
6.
like
0.0072
sightless
0.00004
7.
one
0.0065
trouper
0.00004
8.
go
0.0064
outgrown
0.00004
9.
say
0.0063
frantic
0.00004
10.
get
0.0063
oho
0.00003
The 1990s truly saw the rise in hip-hop/rap. Many of the words here are slang terms, used mainly in rap.
Word
TF
Word
TFIDF
1.
know
0.0076
cristal
0.00009
2.
love
0.0066
quik
0.00008
3.
like
0.0064
dank
0.00008
4.
time
0.0060
phillie
0.00007
5.
got
0.0059
floss
0.00007
6.
get
0.0053
buckwild
0.00007
7.
make
0.0053
betta
0.00006
8.
see
0.0053
representin
0.00006
9.
na
0.0053
ballers
0.00006
10.
go
0.0052
rump
0.00006
The 2000s were indeed a mixed year. All genres saw a healthy consumption, but particularly teen pop and rap got increased participation. This is the decade where slang truly started taking over the lyrics of the music we listen to with words such as swag, shorty and its variation shawty. Also some artists and producers names appear. For example, luda and cris and darkchild.
Word
TF
Word
TFIDF
1.
know
0.0074
crunk
0.00013
2.
like
0.0069
luda
0.00013
3.
got
0.0063
shorty
0.00011
4.
get
0.0058
cris
0.00010
5.
go
0.0055
shawty
0.00010
6.
love
0.0054
swag
0.00009
7.
see
0.0053
darkchild
0.00008
8.
na
0.0052
konvict
0.00008
9.
yeah
0.0052
dro
0.00007
10.
one
0.0051
titty
0.00007
The 2010s had increased popularity in a hushed style of vocal delivery (dubbed whisperpop) as well as a steep rise in traditional instruments: Ukuleles, banjos, mandolins and bongos from indie-rock bands. Most notably, however, is probably the growth of hip hop, dominating most charts. Ad-libs played a massive role in the ’10s, and we saw the introduction of social media sites such as Instagram to the vocabulary of the artists.
Word
TF
Word
TFIDF
1.
like
0.0070
wraith
0.00028
2.
know
0.0066
skrrt
0.00027
3.
got
0.0063
ayy
0.00022
4.
yeah
0.0057
brrt
0.00019
5.
get
0.0056
instagram
0.00017
6.
go
0.0051
thot
0.00016
7.
na
0.0047
swag
0.00013
8.
love
0.0046
maybach
0.00013
9.
time
0.0045
hunnid
0.00012
10.
make
0.0044
bae
0.00012
Although the 2020s are still young, we can see a clear trend in ad-libbing. The words getting the largest TF-IDF scores are ad-libs, and they have become a huge part of our daily listening. Through the lens of a grandparent the words understandably may seem like gibberish.
Word
TF
Word
TFIDF
1.
like
0.0064
opp
0.00069
2.
got
0.0061
skrrt
0.00048
3.
know
0.0060
opps
0.00047
4.
yeah
0.0057
ayy
0.00035
5.
get
0.0055
brrt
0.00034
6.
ai
0.0047
baow
0.00033
7.
go
0.0047
grrah
0.00031
8.
wa
0.0045
wraith
0.00030
9.
ca
0.0042
draco
0.00030
10.
one
0.0042
hunnid
0.00029
We investigated some select words from these and made a dispersion plot to see how their use changed through the decades.
Let’s go through some of the words: Swag: The popularity of swag happened in the ’00s and ended around the ’10s. Few people still use it today, but the amount has declined steeply. Shawty: A popular term for a young woman. Prevalent in rap, hip-hop and r&b music. Boogie and funky: Boogie was a popular genre of music at the end of the disco era of the ’70s. On the dispersion plot it appears as if funk and boogie had a short revival in the ’90s as well. Darling and bitch: Two words often used to describe women, but with very different meanings. It is interesting to see how their popularity is almost the opposite, with darling being phased out as bitch is being phased in. In a sense they display the transition from old-school pop and soul to modern pop and rap. Drug: Singing about drugs is a staple in rap music, and it is not surprising to see its popularity increase through the ’90s. Skrrt: A popular ad-lib used in most rap/trap songs. Nigga: Often used in rap music, which saw a steep increase in popularity from its inception in the ’90s, which is denoted as ‘the golden age of rap’ and has become a staple in most rap music.
Inspecting individual artists
41 highly influential artists by the number of songs on the Billboard ‘Hot-100’ chart are split over four genres in the tabs below. Each artist has an associated word cloud, created from their TF-IDF scores and a similarity score that indicates which other artists are most similar to the given artist, based on the lyrics in our corpus. The TF-IDF scores were calculated using all artists. Finally, the average sentiment of the artist is presented with a comparison of all artists and top artists.
One funny aspect of the word clouds is that most modern artists have their own name as one of the most significant words. This could be because they use tags in their songs, like DJ Khaled’s iconic “It’s DJ Khaled” to promote themselves. Another explanation could be that when artists collaborate on a song, they often mention each other by name. Since the entire song is attributed to the artist and not just the part they wrote/sang, these lyrics will also be part of their TF-IDF scores. Either way, it is a tendency which is not shared by the old artists. Perhaps, back in the day back in the day when the competition in the music industry was not as fierce and fewer collaborations took place, artists did not need to include their names in the lyrics to be remembered.
Let’s take a closer look at Four artists: Ariana Grande, Drake and Juice Wrld and The Beatles.
Ariana Grande
Ariana Grande is a pop singer who went from starring in a Nickelodeon tv series to topping the Billboard charts with her debut album Yours Truly in 2013. Since then, she has continued to top the charts with every new release and has won numerous awards from various outlets. Today, she has had 68 songs appear on the Billboard ‘Hot-100’ chart.
Her word cloud reflects her style of humming and vocalizing, with top words being mmm, woah, ayy and yee. Her sentiment is pretty much right in the middle of everyone.
The top similar artists are Justin Bieber, Chris Brown, The Weeknd, Drake and Rihanna. All of these have heavy pop and R&B connections, which is what Ariana Grande excels at.
Drake
Drake is one of the most famous musicians living today. He has had the highest amount of songs on the Billboard ‘Hot-100’ list, coming in at 253 songs, almost 100 more than the runner-up. He is most well known for his rap songs. He debuted with his mixtape Room for Improvement in 2006. In 2018, he released the hit album Scorpion, which had three Billboard ‘Hot-100’ number one singles: God’s plan, Nice for What and In My Feelings.
His word cloud appears to have pretty common words from the rap genre (according to the rap word cloud from earlier). The term Drizzy appears, which is a nickname that he often uses to refer to himself with.
The most similar artists to him are other rappers. Lil Wayne, Kanye West, Lil Baby, Future and Nicki Minaj are all highly respected rappers, and he has worked together with them in the past.
His sentiment is average compared to the other artists, although he is skewing a tiny bit to the lower side - a common trait amongst all rappers.
Juice Wrld
Juice Wrld was a rapper who pioneered the emo-rap and SoundCloud rap subgenres. In 2018 he released his hit single Lucid Dreams, which took the number two spot on Billboard’s ‘Hot-100’ chart. Juice Wrld has had 70 songs appear on the chart. He mainly collaborated with other rappers: Drake, Future and the trio of Lils: Lil Wayne, Lil Uzi Vert and Lil Baby. All prominent names in the rap industry.
Juice Wrld tended to sing about battling demons and doing all types of drugs, which is also reflected in his word cloud.
His sentiment is comparatively low, and his songs were often associated with turmoil, heartbreak and fragmented feelings.
The Beatles
The Beatles were an English rock band formed in 1960. They are often regarded as the most influential band of all time, and played a huge role pop musics' recognition as a proper art form. Their sentiment is comparatively on the high end, and they share similarity with other famous musicians from the same time period: Aretha Franklin, Elvis Presley, Ray Charles.
Their defining words are classic English/British terms: knickers, joob, gloucester etc.
The Beatles has had 65 songs on the ‘Hot-100’ list, and were one of the first bands to experience the hyper-fan culture, with many calling the period of time they were active a period of beatlemania.
We invite you to go through the tabs and see if you can recognise some of the artists and their similar artists or most defining words.
Note: Artist clouds may not indicate the average word usage by an artist. The word clouds are most likely skewed towards what is mainstream, as it is far from all songs by an artist which reach the top 100.