Content uploaded by Tanvi Kamble
Author content
All content in this area was uploaded by Tanvi Kamble on Aug 10, 2023
Content may be subject to copyright.
“The Times They Are-a-Changin”: The Effect of
the Covid-19 Pandemic on Online Music Sharing
in India
Tanvi Kamble1[0000−0003−2605−5854], Pooja Desur1[0000−0002−0741−9141] ,
Amanda Krause2[0000−0003−3049−9220] , Ponnurangam
Kumaraguru1[0000−0001−5082−2078] , and Vinoo Alluri1[0000−0003−3689−1039]
1International Institute of Information Technology-Hyderabad, India
{tanvi.kamble,pooja.desur,vinoo.alluri,pk.guru}@iiit.ac.in
2James Cook University: Townsville, QLD, Australia
amanda.krause1@jcu.edu.au
Abstract. Music sharing trends have been shown to change during
times of socio-economic crises. Studies have also shown that music can
act as a social surrogate, helping to significantly reduce loneliness by
acting as an empathetic friend. We explored these phenomena through
a novel study of online music sharing during the Covid-19 pandemic in
India. We collected tweets from the popular social media platform Twit-
ter during India’s first and second wave of the pandemic (n=1,364). We
examined the different ways in which music was able to accomplish the
role of a social surrogate via analyzing tweet text using Natural Language
Processing techniques. Additionally, we analyzed the emotional connota-
tions of the music shared through the acoustic features and lyrical content
and compared the results between pandemic and pre-pandemic times. It
was observed that the role of music shifted to a more community fo-
cused function rather than tending to a more self-serving utility. Results
demonstrated that people shared music during the Covid-19 pandemic
which had lower valence and shared songs with topics that reflected tur-
bulent times such as Hardship and Exclusion when compared to songs
shared during pre-Covid times. The results are further discussed in the
context of individualistic versus collectivistic cultures.
Keywords: Musical emotions ·Online Music Sharing ·Covid Pandemic
·Social Surrogacy ·Lyrics
1 Introduction
The Covid-19 pandemic has significantly impacted everyday life with multiple
state and nation-wide lockdowns around the world. Long isolation periods, in-
creasing rates of unemployment, and with hundreds of thousands catching the
virus daily, the pandemic caused an unprecedented socio-economic crisis [29].
India in particular had one of the highest Covid-19 infection rates and is the
2 Kamble and Desur et al.
Fig. 1: Our study focuses on analyzing the context of music sharing from tweet
text and the content of music sharing using acoustic features and lyrical themes.
second worst affected country3in terms of reported Covid-19 cases and deaths
[11, 30]. These distressing times paired with months of isolation periods had peo-
ple searching for coping mechanisms and proxies for physical social interactions.
Social media provides a constant means of communication with the outside
world having a network reach much larger than any physical one. It enables
users to keep in touch with their friends and family through posts, updates,
and messages [5]. As the pandemic limited in-person interactions, social media
use, enabling people to meet their social needs, was at an all time high [33].
The social media platform Twitter affords a space to share thoughts and mood
states, especially through music. The music that is shared on twitter servers
various functions, be it to either promote favorite artists or as a way to express
feelings about the music shared, amongst others. From an evolutionary point
of view sharing music has shown to help in social bonding, building a sense of
community, and convey emotional states [22].
Music can play the role of an empathetic friend by acting as a social surrogate
[28] and be used as a coping mechanism. It can elicit several emotions including
feelings of being connected to others and being understood and can help boost
mood when one is feeling down [18]. When users share music online along with
how it made them feel and how it has helped them, comparisons can be made
as to how music played the role of a social surrogate before and during the
pandemic. It is possible that the kind of music one listens to or wants others to
listen to during times of crisis can convey the coping mechanisms used by people
to some extent. It has been seen that COVID-19 restrictions have led to lifestyle
3Data observed from https://www.worldometers.info/coronavirus/countries-where-
coronavirus-has-spread
The Times They Are-a-Changin 3
changes including change in trends in music consumption. People streamed songs
from their balconies more during the initial lockdowns [15], exploring new styles
and groups of music [3], and there was an increase observed in the listening time
[4, 8]. Past work has looked at music sharing online [36] but as per our knowledge,
work in this sphere has not been done during times of crisis. Furthermore, no
studies have examined why music is shared online or what need it fulfills by
sharing.
Recent times witnessed a slow rise in studies investigating music trends dur-
ing Covid-19 [16, 10, 32, 17, 12, 13]. While no study has looked into online music
sharing, they do provide insight into music consumption trends. A study on
German media consumption during the pandemic showed that media (including
music, books, movies amongst others) induced nostalgia during Covid-19 func-
tioned as a way to cope with social stress (fear of isolation) during lockdown
periods [34]. Another study on European countries found that music consump-
tion on Spotify changed in terms of nostalgia during the pandemic [35].
Another study on popular music in the UK and the US during the pandemic
demonstrated a negative trend in valence of lyrics and higher reference of inter-
personal dependence in lyrics [24]. However, there are differences in how countries
consume and associate with music [26, 19]. Individualistic cultures such as UK
and US use music as a tool for self expression. On the other hand, collectivistic
cultures, which include Asian countries like Japan, India, and China, use music
typically to add positivity to their lives [26]. Emotional connections to music and
coping mechanisms are different for individualistic cultures where people are self-
sufficient and achievement-oriented as compared to collectivistic cultures where
people are interdependent and family-oriented [19].
In this study we focus on India, a culturally rich country and that has a deep
relation to music. A study of 3,000 internet users showed that 80 per cent of
internet users called themselves as ‘music-lovers’ [14] in India. Despite being one
of the largest countries in terms of population, music sharing trends have not
been studied. Work has been done on the evolving Indian music industry [1, 7,
20] but a focus on the trends of the overall population is lacking. Moreover, on
average, an Indian spends 19.1 hours a week listening to music which is higher
than the global average of 18 hours [14]. Thus, it is important to consider how
and why users share music online in India. A large Twitter user base of 23 million
Indians provides opportunities for a large scale study. A study on India could
further enhance the comparison of the function of music between collectivistic
and individualistic cultures.
This paper aims to analyse online music sharing of Indians during the pan-
demic. To this end, we take a two-pronged approach to analyzing tweets posted
during this time. We first analyze tweet text to understand the role music plays
as a social surrogate via NLP techniques. Subsequently we analyze the musical
content being shared by examining emotional connotations derived via acoustic
features and lyrics. Additionally, we examine lyrical themes shared during the
pandemic. We compare all of the above with pre-pandemic times to identify
changes/trends.
4 Kamble and Desur et al.
Fig. 2: Pipeline of extracting emotion, lyrical themes, and the function of mu-
sic from tweets. Red circles represent BERT word embeddings and the purple,
yellow, and blue circles symbolize the different embeddings obtained from social
surrogacy statements belonging to different categories.
2 Methodology
2.1 Dataset
Using the Twitter API4, we collected tweets that were geotagged as India that
contained a Spotify URL for tracks during the first and second wave of the pan-
demic (see Figure 2). As per the WHO 5dashboard of Covid-19 cases, the peak
of Wave-1 and Wave-2 in India was recorded on September 14, 2020 and May
3, 2021 respectively. Judging by the steepness of both the peaks, we considered
a period of two months around the Wave-1 peak and one month around the
Wave-2 peak as our Wave-1 and Wave-2 time periods respectively. To compare
this with a control group, we collected similar tweets during the same period of
months in 2019 (referred to as Control-1 and Control-2 respectively) in order to
avoid temporal music sharing differences that can occur as a result of seasonal
trends. We collected a total of 1,364 tweets that were posted during both waves
of the pandemic. One tweet can have multiple track URLs shared in them, but
each song of the tweet is added separately to the dataset. We limited our analysis
to tweets in English. A total of 54.3% of the tweets collected during Wave-1 and
48.2% of tweets collected during Wave-2 had English tweet text. Detailed statis-
tics about the dataset are given in Table 1. From the Spotify URL, the name
and artist of the song was retrieved using the Spotify ID. With this information,
we collected lyrics for each track, using the LyricsGenius and azlyrics API.6
4https://developer.twitter.com/en/docs/twitter-api
5https://covid19.who.int/region/searo/country/in
6https://lyricsgenius.readthedocs.io/en/master/
The Times They Are-a-Changin 5
Table 1: Summary of dataset statistics across Wave-1 and Wave-2 of Covid-19
and their Control groups -
Data Group Tweets with
Spotify URL
Tweets with
English Songs
Songs with correct
English Lyrics
Songs
Rejected
Wave-1 (July-November 2020) 808 416 323 87
Control-1 (July-November 2019) 607 271 204 54
Wave 2 (April-June 2021) 556 317 263 94
Control-2 (April-June 2019) 351 177 155 48
2.2 Tweet Analysis
From each tweet, we examined the context and the content of the music associ-
ated with it. Context of the music shared refers to the tweet text accompanying
it. It gives a glimpse into why a user shared that particular track. We used this
text to consider if the music may have played a role of a social surrogate. On
the other hand content refers to both the musical and lyrical features of the
song. We looked at the emotional connotations of the content by extracting the
acoustic features from the music and sentiment and themes from the lyrics.
Context Analysis In order to capture the functions played by music, we used
a set of 30 statements formulated in past work [28] as demonstrations of music
functioning as a social surrogate.7These include statements such as “It reminds
of certain periods of my life or past experiences” or “I can identify with the mu-
sicians or bands”. These 30 statements were originally collected as a result of a
survey where participants described how media plays a role of a social surrogate
in their lives and was then adapted to music [28]. These statements belong to 7
overarching categories – Company, Reminiscence, Shared Experiences, Isolation,
Understanding Others, Culture, and Group Identity. The category Company de-
scribes the role of music in helping to feel less lonely and providing comfort.
Reminiscence is when music elicits feeling of nostalgia through a person or ex-
perience. Shared Experiences covers how music helps people to feel understood
and identify with the music/artists. The category Isolation involves feelings of
wanting to isolate socially and not talk to others but finding solace in music.
Understanding Others includes how music brings about feelings of belonging
and understanding others and the world. When music helps to connect to one’s
culture and allows people to express cultural uniqueness the surrogacy role falls
under the category Culture. Music that helps people to identify to a subculture
and belong to a particular social group the role falls under the category Group
Identity. Examples of tweets which came under the above categories include “Re-
minder to listen to this song when feeling underconfident” (Shared Experiences)
or “Something that reminds me of my childhood” (Reminiscence).
To get the context of the music shared, we first preprocessed the text of each
tweet to remove links, hashtags, and emojis. In order to filter out tweets that did
7Refer to Appendix 1
6 Kamble and Desur et al.
not represent the function of music as a social surrogate, we manually removed
tweets that fell under the Non social surrogacy category. These tweets consisted
of keywords/strings such as “mood”, “stream this song” or “listen to this”. A total
of 502 tweets were identified as showing roles of social surrogacy of which 150
and 165 belonged to time periods Wave-1 and Wave-2 respectively and 111 and
76 belonged to Control-1 and Control-2 respectively. Only this set was used for
further tests. An automated approach was used to categorize each of these filtered
tweets into one of the seven social surrogacy categories. The sentence embedding
of the resulting text of the tweets was calculated by using a BERT transformer.8
Similarly, the sentence embedding for each of the 30 statements was calculated.
Cosine similarity which was used as a distance metric to represent semantic
similarity was calculated between each tweet and each of the 30 statements in
the embedding space. A tweet was allocated into the corresponding category of
the most similar statement embedding (provided the similarity was greater than
0.5). Thus, a tweet was only mapped to either one or zero of the seven social
surrogacy categories. Some examples of tweets along with the allocated groups
are shown in Table 2. This automated approach provided a way to observe the
ways music served as a surrogate during the pandemic without the intrusion of
human bias. It also allowed us to investigate which categories of surrogacy such
as Reminiscence or Isolation were more prevalent than others when users shared
music online during the considered time periods.
Table 2: Examples of tweet text categorized into Social Surrogacy Categories
using BERT embeddings
Tweet Text Most Similar Surrogacy Statement Category
song i remember from childhood It reminds me of certain periods of my
life or past experiences Reminiscence
the joy of discovering music thats
totally you I can identify with the musicians or bands Shared Experiences
Dedicated to the Nocturnal I want to isolate myself from my surroun-
dings Isolation
Content Analysis In some cases, the song name returned from the Spotify
API9did not match the song in LyricsGenius API owing to gibberish lyrics. We
weeded out such songs (n=236) from the pool of English songs (n=1,181). This
happened for songs which were remixed-versions or were sung live and hence
such songs were removed.
Emotions We examined the emotional connotations of the music shared in
two ways - by using the acoustic features of the songs and by performing a senti-
ment analysis on the lyrics. To obtain the acoustic features, we used the Spotify
8https://huggingface.co/docs/transformers/model_doc/bert
9https://spotipy.readthedocs.io/en/2.19.0/#
The Times They Are-a-Changin 7
API in order to extract valence and energy of the song which provides insight to
its emotional connotation. Valence is indicative of the pleasantness/positiveness
of the track while energy is self-explanatory. We then performed sentiment anal-
ysis on the lyrics using a lexicon and rule-based sentiment analysis tool called
VADER.10 It is used in grammar-free texts like social media.
Lyrical Analysis We performed topic modelling on the lyrics using DICTION
software11 to extract the topics of the songs that were shared during the different
data groups. DICTION is a language analysis software that uses dictionaries to
determine the topic(s) of a given text. There are 40 topics 12 each of which has
a dictionary of words associated with it where no two dictionaries have the same
words. It calculates the frequency of the words from the text to determine the
topics. We decided to use DICTION as it is a good choice for topic modelling for
free-grammar text like songs and poem since it has a word to word mapping [24,
6, 21,2] Further details about DICTION and the custom lists have been discussed
in the appendix. We ran DICTION on the lyrics of the songs belonging to Covid-
19 and control periods for all the 40 topics, and divided the frequency of words
by the total number of words to normalize the scores.
2.3 Statistical Tests
The contextual information and musical content was compared between the fol-
lowing time periods: Wave-1 (n=323) versus Control-1 (n=204), Wave-2 (n=263)
versus Control-2 (n=155), and Covid-19 as a whole referred to as Wave-1 +
Wave2 (n=586) versus pre-Covid period of 2019 referred to as Control-1 +
Control-2 (n=359).
Context-wise, a frequency table was created for comparing observed frequen-
cies which were the Covid-19 periods and the expected proportions which were
the corresponding control groups. A chi-square goodness of fit test [23] was per-
formed to observe if the proportions were significantly different.
Content-wise, we used the non-parametric Mann Whitney U test (MWU) to
examine the difference between valence and energy (the acoustic features) across
the conditions.
For the lyrical analysis, MWU tests were also performed on results of DIC-
TION analysis between each of the above mentioned time periods. The Benjamini-
Hochberg procedure was used to account for running multiple statistical tests.
The results of these tests are reported in the next section.
10 https://github.com/cjhutto/vaderSentiment
11 https://dictionsoftware.com/
12 40 topics include 31 dictionary based variables, five master variables which are a
combination of the dictionary based variables and four calculation based variables.
The last four variables rely on calculations such as word count, word size rather than
dictionary matches. Details about the variables are given in Appendix 2
8 Kamble and Desur et al.
3 Results
3.1 Context Analysis
The percentage of total tweets lying in each social surrogacy category during
Wave-1 and Control-1 are shown in Figure 3 . The chi-square goodness of fit test
was significant (p=0.028), suggesting that the proportions of tweets amongst
the social surrogacy categories were significantly different between Wave-1 and
Control-1. Post hoc multiple z-test comparisons were done to observe which cat-
egories showed significant differences in proportions. Two of the social surrogacy
groups – Reminiscence and Group Identity had a significant decrease in the
proportion of tweets falling into these categories during Wave-1. On the other
hand, the chi-square test was non-significant across the distribution of propor-
tions when comparing Wave-2 and Control-2, or Covid-19 as a whole (Wave-1
+ Wave2) and pre-Covid-19 period (i.e Control-1 + Control-2).
Fig. 3: Distribution of tweets into social surrogacy categories during Wave-1 and
Control-1 (covid and control in legend respectively)
3.2 Content Analysis
Emotions Figure 4 displays distributions of valence and energy derived from
acoustic features for different conditions. Results of the Mann Whitney U tests
revealed that valence of music shared during Covid (Wave-1 and Wave-2 com-
bined) was significantly lower (U=613715, p=0.006) than pre-Covid time period.
The Times They Are-a-Changin 9
Similar results were observed when comparing Wave-1 with Control-1 where va-
lence was significantly lower during Wave-1 (U=231695, p=0.037) while energy
was found to be higher (U=229345, p=0.018). No significant differences were
observed in either valence or energy of music shared when comparing Wave-2
with Control-2.
(a) (b)
Fig. 4: Comparison of acoustic features between different periods. (a) Valence
is significantly lower during Covid than pre-Covid. (b) Valence is significantly
lower during Wave 1, while Energy is higher as compared to the Control-1.
Sentiment analysis on lyrics as shown in Figure 5 was also performed. During
comparison only the time periods of Wave-2 against Control-2 showed differences
that were statistically significant (U=33920.5, p=0.0005) according to the Man
Whitney U test. During Wave-2 the lyrics of the songs had a lower mean senti-
ment (0.24) as compared to the Control-2 (0.39).
Fig. 5: Sentiment Analysis of Lyrics of the songs according to the different time
periods. Mean Sentiment of Wave-1 was higher than Control-1 but it was vice-
versa for Wave-2 and Control-2.
10 Kamble and Desur et al.
Lyrical Themes We observed that different lyrical themes were shared in dif-
ferent time periods. Table 3 summarises the results of the Mann-Whitney U tests
conducted on the various DICTION categories.
Wave-1 demonstrated an increased sharing of music with lyrical themes sig-
nifying Exclusion (U=33874.0, p=0.005) and Hardship (U=34698.5, p=0.041),
when compared to Control 1. Similar pattern was observed when comparing
the Covid-19 time period to pre-Covid-19 time periods where topics of Exclu-
sion(U=103193.0, p=0.049) and Hardship (U=102221.0, p=0.05) were shared
more as well. The topic of Motion was shared more in Wave-1 than Control-1
(U=38425.5, p=0.04) but was shared less in Covid-19 group as a whole than pre-
Covid times as a whole (U=100560.0, p=0.018). Wave-2 witnessed an increase in
music with lyrical themes representing Communication ( U=12557.0, p=0.012)
when compared to Control-2. Another observation to be made was that the
values of Satisfaction increased in the songs shared during Wave-1 when com-
pared with Control-1 (U=38066.5, p=0.026) but the opposite happened when
we observed Wave-2 and Control-2.
Table 3: DICTION variables for different time periods that show significant
differences with p < 0.05 in Man Whitney U Test (* p<0.01). The arrows indicate
an increase or decrease in the songs shared in the time period.
Wave-1 (vs. Control-1) Exclusion* ↑, Satisf action ↑, Har dship ↑, M otion ↑
Wave=2 (vs. Control-2) Communication ↑, Satisf action∗ ↓
Wave-1 + Wave-2 (vs. Control-1 + Control-2) Hardship ↑, Exclusion ↑, M otion ↓
4 Discussion
Our study examined the effect of the Covid-19 pandemic on the type of music
that was shared online via Twitter in India. Overall music sharing seems to be
different during Covid-19 and pre-Covid-19 times. The number of music related
tweets posted by users to share a certain song on Spotify increased during Wave-
1. This increase in music sharing online could be attributed to higher rates
of social media usage during periods of isolation in the pandemic [31], or to
how music sharing builds a sense of community [27] which was needed during
the pandemic. A portion of the increase should be credited to the fact that
Spotify was released in India during 2019, and as its popularity began to grow,
the number of users sharing Spotify links increased subsequently. Apart from
the increase in number of tweets there were several differences observed in the
context and content as discussed below.
We examined how music might function as a social surrogate during the
pandemic. Tweets which fell under the categories Company (in which people
use music to feel less lonely and find comfort) and Shared Experiences (which
covers music experiences that help people to feel understood) made up 41% of
The Times They Are-a-Changin 11
the total tweets containing music collected during the pandemic. During Wave-
1, music acted as a social surrogate through the role of Group Identity to a
lesser amount then pre-pandemic times. Group Identity involves feeling a sense of
belonging to a particular social group on a smaller scale, such as identifiying with
a particular artist or associating oneself with a fan group. India, a collectivistic
community, has shown to put a great importance on community, and as the
pandemic brought about distressing times, the sense of community tended to
expand, which is reflected by more music shared that was not targeted towards
a specific subgroup. This result suggests that perhaps people may have felt a less
pressing desire to separate into smaller subgroups and preferred to experience
and share music with a larger, more diverse demographic to fulfil their need for
community.
Similarly the proportion of tweets falling into the category Reminiscence
which is more self directed (reminds one of their past experiences or a person),
also decreased during Wave-1. While past work [34] has demonstrated that media
associated with a sense of nostalgia was consumed more during the Covid-19,
this study was done on a individualistic culture (Germany). India has shown a
different trend where the role of self-involved social surrogacy categories reduced
during the pandemic. Our interpretation is that music was shared to foster a
sense of greater community rather than for personal preferences.
Music shared online in India during Wave-1 was more negatively valenced
but had higher overall energy values in terms of acoustic features. The valence
and energy values mirror the feelings caused by the turbulent times the country
experienced. This is in line with results from a study on UK and US where
songs with negatively valenced lyrics were found to be more popular during the
pandemic as compared to before [24]. Typically since negative valence of music is
congruent to negative valence derived from lyrics [9], these results are comparable
and the cultures seem to show a similar change in trends over the period of the
pandemic. Lyrical themes of Exclusion and Hardship were observed more in India
during the pandemic and these themes also reflect the distressing times. There
was also more Satisfaction related lyrical themes which has to do with positive
affective states, although valence was negative. This could be an attempt to try to
balance the negative mood of the music with positive words while still reflecting
the current hard times of the pandemic. This result opposes results done on
a study of individualistic culture where lyrical themes of Satisfaction were less
predominant during the first six months of the pandemic [24], albeit limited to
top charts on Spotify. Nevertheless, these results concerning the lyrical themes of
shared music highlight the similarities and differences between the consumption
and sharing of music in individualistic and collectivistic cultures.
Wave-2 was much more disastrous for India [25] than Wave-1. Music sharing
online evidenced a decrease in the sentiment of lyrics as well as a decrease in the
themes of Satisfaction. One potential explanation in Wave-1 themes of Satisfac-
tion arose by trying to build up hope but by Wave-2 the decrease showed the
opposite trend. Additionally, music shared with themes of Communication(See
Table 4) increased during Wave-2. From an evolutionary point of view, music
12 Kamble and Desur et al.
has been used as a social bonding tool and reveals a person’s emotional states
to others [22]. In this way themes of Communication in music expose a desire
to reach out and bond with others which could explain this increase. Overall
the combined duration of Wave-1 and Wave-2 brought about music sharing with
more negative valence than pre-Covid-19 times, with lyrical themes of Hardship
and Exclusion thereby mirroring the dire situation of the pandemic.
In sum, our study is the first that looks at music sharing trends online during
the pandemic. Music shared online provides a glimpse into the emotional state
of a person. Studying the text accompanying the sharing of a song online helps
reveal the role music can play in our lives and its benefits during distressing
times. While this study has shown that collectivistic and individualistic cultures
both reported negative trends in valence of music during the pandemic, Wave-1
in India reflected a glimmer of hope with the increase in lyrical themes of Sat-
isfaction (Refer Table 4). Furthermore, India as a collectivistic culture has a
tendency to put the well being of a community at large above ones own. This is
well demonstrated by the role that Group Identity and Reminescence played as
a social surrogate which decreased during the pandemic. The social surrogacy
approach which summarizes the various functions music plays as a social surro-
gate can be extended to study music shared in other countries. It can also be
extended to data collected from other social media platforms. Future research
on this topic could also benefit from a mixed methods approach, where manual
annotation of the tweets could complement the automated approach, although
it may not be feasible and scalable for a large number of datapoints. While our
study was limited to music with lyrics in English, future work can accommodate
music and tweets in other languages using advanced NLP techniques. Lastly, we
note that while we have collected the music shared on the Twitter platform, and
as such the music analysed in the present study is not a complete representation
of the music listened to by the Indian population as a whole,future work could
also explore the use of other platforms and collaborative playlists.
Table 4: DICTION variables and the topics they represent
DICTION Variable Meaning of the Variable
Communication Terms of social interaction either with one person or a
group
Exclusion Describes the sources and effects of social isolation [24]
Hardship Terms of Natural Disaster, problems faced, and human
fears.
Motion Terms of movement, speed and transit
Satisfaction Terms associated with positive affective states [24]
Acknowledgements This research was partially funded by IHUB at IIIT Hy-
derabad.
The Times They Are-a-Changin 13
5 Appendix
5.1 Appendix 1
The 30 statements that were used to model the role of music as a social surrogate
and their assigned categories as per [28] are given in table 5.
Table 5: Social Surrogacy statements corresponding to each group
Social Surrogacy Category Statement
Company
It keeps me company.
It can make me feel less lonely.
It comforts me when I’m sad.
Culture
It mirrors the history and culture of my country.
It makes me feel connected to my culture.
It is a good way to express the uniqueness of
our culture.
Group Identity
I would like to identify with a particular subculture.
It helps me to show that I belong to a particular
social group.
It makes me feel connected to all the people who like
the same kind of music.
I would like to take the artists as role models
It makes me feel connected to others.
Isolation
I don’t want to talk to anybody.
I like to have some sound in the background.
I want to isolate myself from my surroundings.
I don’t want to hear the surrounding sounds.
Reminiscence
It reminds me of the people that I used to listen to
the music with.
It reminds me of a particular person.
It reminds me of certain periods of my life or
past experiences.
It helps me reminisce.
Shared Experience
I can recognize myself in the lyrics.
The songwriter has made similar experiences as
I have.
I like to immerse myself into the lyrics.
I can identify with the musicians or bands.
I can sing along with it.
It makes me feel like somebody else feels the same as
I do.
Understanding Others
It helps me understand the world better.
It makes me feel connected to the world.
It tells me how other people think.
It makes me feel like I belong.
It helps me to understand what is going on in others
people’s heads.
14 Kamble and Desur et al.
5.2 Appendix 2
DICTION 7.0, as mentioned earlier, is a content analysis software which uses
dictionaries of topics to perform word to word mapping and gives a score related
to each topic. The value of each variable for a song is a float value. Even though
it is a frequency, homographs are incremented as decimals instead of ‘1’ in the
DICTION software. The 40 categories of DICTION can be divided into the
following sub-categories:
1. Dictionary based variables: Each variable has a dictionary of words as-
sociated with it. There are 10,000 words classified into a total of 35 discrete
variables. The number of words in each dictionary range from 10 to 745.
Table 6 contains a brief description of each of the 35 variables.
Table 6: 31 Dictionary Based Variables
Variable Name Variable Definition
AMBIVALENCE Words expressing hesitation or uncertainty like
confusion, mystery, etc.
ACCOMPLISHMENT Words expressing task-completion, modes of
expansion like grow, buy, employ, produce, etc.
AGGRESSION Words of human competition and forceful action,
social domination, energy, personal triumph
BLAME Terms of social inappropriateness, unfortunate
circumstances like cruel, miserly, painful, etc
CENTRALITY
Terms denoting institutional regularities and/or
substantive agreement on core values. Words
like indigenous terms, typicality, etc.
COGNITION Words referring to cerebral processes, both
functional and imaginative.
COLLECTIVES
Singular nouns connoting plurality that
function to decrease specificity like army,
crowd, country, world, etc.
COMMUNICATION
Terms referring to social interaction, both
face-to-face, mediated, social actors,
social purposes, etc.
CONCRETENESS
No thematic unity other than tangibility and
materiality. Words of sociological units,
occupational groups, political alignments.
COOPERATION
Terms designating behavioral interactions
among people that often result in a group
product like words of designations of formal
work relations and informal associations.
DENIAL Words of negative contractions, negative function
words like not, nothing, etc.
The Times They Are-a-Changin 15
Table 6: 31 Dictionary Based Variables
Variable Name Variable Definition
DIVERSITY Words describing individuals or groups of
individuals differing from the norm
EXCLUSION Terms describing the sources and effects of
social isolation.
FAMILIARITY
Words including common prepositions
(across, over, through), demonstrative pronouns
(this, that) and interrogative pronouns
(who, what), and a variety of particles,
conjunctions and connectives (a, for, so).
HARDSHIP
Containing words related to natural disasters,
hostile actions, injustice, human fears
and griefs.
HUMAN INTEREST standard personal pronouns, family members and
relations and generic terms friend, baby, human.
INSPIRATION Words of Abstract virtues deserving of universal
respect.
LEVELING TERMS
Words used to ignore individual differences and
to build a sense of completeness assurance like
everyone, absolute, each, fully, etc.
LIBERATION Terms describing the maximizing of individual
choice and the rejection of social conventions
MOTION Terms connoting human movement, physical
processes, journey, speed etc.
NUMERICAL TERMS
Words indicating numbers or numerical
operations like sum, percentage or quantitative
topics like mathematics.
PASSIVITY Words ranging from neutrality to inactivity
like inertness, compliance, docility, etc.
PAST CONCERN The past-tense forms of the verbs contained
in the Present Concern dictionary.
PRAISE Social, Intellectual, entrepreneurial, moral
and Physical Qualities
PRESENT CONCERN
A selective list of present-tense verbs like
general physical activity, social operations,
task-performance, etc.
RAPPORT
This dictionary describes attitudinal
similarities among groups of people like
terms of affinity, deference, etc.
SATISFACTION Terms associated with positive affective
states, undiminished joy, moments of triumph
SELF-REFERENCE Terms of first-person references
16 Kamble and Desur et al.
Table 6: 31 Dictionary Based Variables
Variable Name Variable Definition
SPATIAL TERMS
Terms referring to geographical entities,
physical distances, and modes of
measurement.
TEMPORAL TERMS
Terms that fix a person, idea, or event within
a specific time-interval, thereby signaling
a concern for concrete and practical matters.
TENACITY Forms of the verb ‘ to be’. These verbs
connote confidence and totality.
2. Calculation Based Variables: Four DICTION variables result from calcu-
lations rather than dictionary matches. They are calculated using a formula
which envolves data like the number of words in the text, length of each
word, occurrences of each word, etc. Table 7 explains the different Calcula-
tion based variables.
Table 7: Four Calculation based Variables
Variable Name Variable Definition
COMPLEXITY Average number of characters-per-word
INSISTENCE All words occurring three or more times that function
as nouns or noun-derived adjectives are identified
EMBELLISHMENT A selective ratio of adjectives to verbs.
VARIETY Type-Token Ratio, ratio of number of different words
in a passage to the passage’s total words.
3. Master Variables: The five master variables provide the most general
understanding of a given text. They are a combination of the Dictionary-
based and Calculation-based variables. The are formed by converting all
subaltern variables to z-scores, combining them via addition and subtraction,
and then by adding a constant of 50 to eliminate negative numbers. For
example, Optimism, which is [Praise + Satisfaction + Inspiration] – [Blame
+ Hardship + Denial] to which 50 is added standardizes the six variables.
Table 8 gives an idea about what the master variables signify.
The Times They Are-a-Changin 17
Table 8: Five Master Variables which are a combination of the Dictionary based
variables
Variable Name Variable Definition
ACTIVITY Language featuring movement, change, the implementation
of ideas and the avoidance of inertia.
CERTAINITY Language indicating resoluteness, inflexibility, and
completeness and a tendency to speak ex cathedra
COMMONALITY Language highlighting the agreed - upon values of a group
and rejecting idiosyncratic modes of engagement.
OPTIMISM Language endorsing some person, group, concept or event
or highlighting their positive entailments.
REALISM Language describing tangible, immediate, recognizable
matters that affect people’s everyday lives.
References
1. Agarwal, P., Karnick, H., Raj, B.: A comparative study of indian and western
music forms. In: ISMIR (2013)
2. Anglada-Tort, M., Krause, A.E., North, A.C.: Popular music lyrics and
musicians’ gender over time: A computational approach. Psychology of
Music 49(3), 426–444 (2021). https://doi.org/10.1177/0305735619871602,
https://doi.org/10.1177/0305735619871602
3. Cabedo-Mas, A., Arriaga-Sanz, C., Moliner-Miravet, L.: Uses and percep-
tions of music in times of covid-19: A spanish population survey. Fron-
tiers in Psychology 11 (2021). https://doi.org/10.3389/fpsyg.2020.606180,
https://www.frontiersin.org/article/10.3389/fpsyg.2020.606180
4. Carlson, E., Wilson, J., Baltazar, M., Duman, D., Peltola, H.R., Toivi-
ainen, P., Saarikallio, S.: The role of music in everyday life during the
first wave of the coronavirus pandemic: A mixed-methods exploratory study.
Frontiers in Psychology 12 (2021). https://doi.org/10.3389/fpsyg.2021.647756,
https://www.frontiersin.org/article/10.3389/fpsyg.2021.647756
5. Clark, J.L., Algoe, S.B., Green, M.C.: Social network sites and well-
being: The role of social connection. Current Directions in Psychologi-
cal Science 27(1), 32–37 (2018). https://doi.org/10.1177/0963721417730833,
https://doi.org/10.1177/0963721417730833
6. Cook, S.L., Krupar, K.: Defining the twentieth century and impacting the twenty-
first: Semantic habits created through radio and song. ETC: A Review of General
Semantics 67(4), 412–434 (2010), http://www.jstor.org/stable/42579071
7. Evans, A.E.: Music in india: An overview. In: The 2016 Symposium (2016)
8. Fink Lauren K., Warrenburg Lindsay A., H.C.R.W.M.H.N.C.W.F.M.: Viral tunes:
changes in musical behaviours and interest in coronamusic predict socio-emotional
coping during covid-19 lockdown. Humanities and Social Sciences Communications
8(2021). https://doi.org/10.1057/s41599-021-00858-y
9. Fiveash, A., Luck, G.: Effects of musical valence on the cog-
nitive processing of lyrics. Psychology of Music 44 (02 2016).
https://doi.org/10.1177/0305735615628057
18 Kamble and Desur et al.
10. Gibbs, H., Egermann, H.: Music-evoked nostalgia and wellbeing during the
united kingdom covid-19 pandemic: Content, subjective effects, and function.
Frontiers in Psychology 12 (2021). https://doi.org/10.3389/fpsyg.2021.647891,
https://www.frontiersin.org/article/10.3389/fpsyg.2021.647891
11. Goel, I., Sharma, P., Kashiramka, P.: Effects of the covid-19 pandemic in india: An
analysis of policy and technological interventions. Health Policy and Technology
10 (12 2020). https://doi.org/10.1016/j.hlpt.2020.12.001
12. Granot, R., Spitz, D.H., Cherki, B.R., Loui, P., Timmers, R., Schaefer, R.S.,
Vuoskoski, J.K., Cárdenas-Soler, R.N., Soares-Quadros, J.F., Li, S., Lega, C.,
La Rocca, S., Martínez, I.C., Tanco, M., Marchiano, M., Martínez-Castilla,
P., Pérez-Acosta, G., Martínez-Ezquerro, J.D., Gutiérrez-Blasco, I.M., Jiménez-
Dabdoub, L., Coers, M., Treider, J.M., Greenberg, D.M., Israel, S.: “help! i need
somebody”: Music as a global resource for obtaining wellbeing goals in times of cri-
sis. Frontiers in Psychology 12 (2021). https://doi.org/10.3389/fpsyg.2021.648013,
https://www.frontiersin.org/article/10.3389/fpsyg.2021.648013
13. Howlin, C., Hansen, N.C.: Music in times of covid-19 (Mar 2022).
https://doi.org/10.31234/osf.io/z94fq, psyarxiv.com/z94fq
14. Industry, T.I.M.: Digital music study (2019)
15. Lehman, E.T.: “washing hands, reaching out” – popular music, digital leisure and
touch during the covid-19 pandemic. Leisure Sciences 43, 273 – 279 (2020)
16. Levstek, M., Barnby, R.M., Pocock, K.L., Banerjee, R.: “it all
makes us feel together”: Young people’s experiences of virtual
group music-making during the covid-19 pandemic. Frontiers in
Psychology 12 (2021). https://doi.org/10.3389/fpsyg.2021.703892,
https://www.frontiersin.org/article/10.3389/fpsyg.2021.703892
17. Martínez-Castilla, P., Gutiérrez-Blasco, I.M., Spitz, D.H., Granot, R.: The
efficacy of music for emotional wellbeing during the covid-19 lockdown
in spain: An analysis of personal and context-related variables. Fron-
tiers in Psychology 12 (2021). https://doi.org/10.3389/fpsyg.2021.647837,
https://www.frontiersin.org/article/10.3389/fpsyg.2021.647837
18. Mas-Herrero, E., Marco-Pallares, J., Lorenzo-Seva, U., Zatorre, R.J., Rodriguez-
Fornells, A.: Individual Differences in Music Reward Experiences. Music Per-
ception 31(2), 118–138 (12 2013). https://doi.org/10.1525/mp.2013.31.2.118,
https://doi.org/10.1525/mp.2013.31.2.118
19. Miyamoto, Y., Ma, X., Petermann, A.G.: Cultural differences in hedonic emotion
regulation after a negative event. Emotion 14(4), 804 (2014)
20. Muthusamy, S.: Music industry in economic development of india. ANJAC Journal
of Humanities and Social Sciences, ISSN: 0976-4216 3, 1 – 9 (01 2012)
21. North, A.C., Krause, A.E., Kane, R., Sheridan, L.: United
kingdom “top 5” pop music lyrics. Psychology of Music
46(5), 638–661 (2018). https://doi.org/10.1177/0305735617720161,
https://doi.org/10.1177/0305735617720161
22. Patel, A.D.: Music, biological evolution, and the brain (2010)
23. Pearson, K.: X. on the criterion that a given system of deviations from
the probable in the case of a correlated system of variables is such that
it can be reasonably supposed to have arisen from random sampling. The
London, Edinburgh, and Dublin Philosophical Magazine and Journal of Sci-
ence 50(302), 157–175 (Jul 1900). https://doi.org/10.1080/14786440009463897,
https://doi.org/10.1080/14786440009463897
The Times They Are-a-Changin 19
24. Putter, K.C., Krause, A.E., North, A.C.: Popular music
lyrics and the covid-19 pandemic. Psychology of Music 0(0),
03057356211045114 (0). https://doi.org/10.1177/03057356211045114,
https://doi.org/10.1177/03057356211045114
25. Ranjan, R., Sharma, A., Verma, M.: Characterization of the second wave of covid-
19 in india (04 2021). https://doi.org/10.1101/2021.04.17.21255665
26. Saarikallio, S., Alluri, V., Kulmunki, I., Toiviainen, P.: Emotions of mu-
sic listening in finland and in india: Comparison of an individualistic and
a collectivistic culture. Psychology of Music 49, 030573562091773 (05 2020).
https://doi.org/10.1177/0305735620917730
27. van der Schyff, D., Silverman, M.: Music in the Community (11 2019)
28. Schäfer, K., Eerola, T.: How listening to music and engagement with other media
provide a sense of belonging: An exploratory study of social surrogacy. Psychol-
ogy of Music 48(2), 232–251 (2020). https://doi.org/10.1177/0305735618795036,
https://doi.org/10.1177/0305735618795036
29. Shang, Y., Li, H., Zhang, R.: Effects of pandemic outbreak on
economies: Evidence from business history context. Frontiers in
Public Health 9(2021). https://doi.org/10.3389/fpubh.2021.632043,
https://www.frontiersin.org/article/10.3389/fpubh.2021.632043
30. Shukla, M., Pandey, R., Singh, T., Riddleston, L., Hutchinson, T., Ku-
mari, V., Lau, J.Y.F.: The effect of covid-19 and related lockdown phases
on young peoples’ worries and emotions: Novel data from india. Fron-
tiers in Public Health 9(2021). https://doi.org/10.3389/fpubh.2021.645183,
https://www.frontiersin.org/article/10.3389/fpubh.2021.645183
31. Tsao, S.F., Chen, H., Tisseverasinghe, T., Yang, R., Li, L., Butt, Z.: What social
media told us in the time of covid-19: a scoping review. The Lancet Digital Health
3(01 2021). https://doi.org/10.1016/S2589-7500(20)30315-0
32. Vidas, D., Larwood, J.L., Nelson, N.L., Dingle, G.A.: Music listening as
a strategy for managing covid-19 stress in first-year university students.
Frontiers in Psychology 12 (2021). https://doi.org/10.3389/fpsyg.2021.647065,
https://www.frontiersin.org/article/10.3389/fpsyg.2021.647065
33. Wong, A., Ho, S., Olusanya, O., Antonini, M.V., Lyness, D.: The use of social media
and online communications in times of pandemic covid-19. Journal of the Intensive
Care Society 22(3), 255–260 (2021). https://doi.org/10.1177/1751143720966280,
https://doi.org/10.1177/1751143720966280
34. Wulf, T., Breuer, J., Schmitt, J.: Escaping the pandemic present: The
relationship between nostalgic media use, escapism, and well-being dur-
ing the covid-19 pandemic. Psychology of Popular Media (09 2021).
https://doi.org/10.1037/ppm0000357
35. Yeung, T.Y.C.: Did the covid-19 pandemic trigger nostalgia? evidence of music
consumption on spotify (08 2020)
36. Zangerle, E., Pichl, M., Gassler, W., Specht, G.: nowplaying music
dataset: Extracting listening behavior from twitter. In: Proceedings of
the First International Workshop on Internet-Scale Multimedia Man-
agement. p. 21–26. WISMM ’14, Association for Computing Machinery,
New York, NY, USA (2014). https://doi.org/10.1145/2661714.2661719,
https://doi.org/10.1145/2661714.2661719
