Introduction

Regardless of their content, social media influencers aim to grow and maintain an audience to ensure their platforms are marketable and profitable. We found there are several linguistic techniques these influencers adopt to build connections with their fans. In particular, they replied to their followers’ comments under their videos to facilitate connection with their audience, and, when doing so, utilized rapport-building linguistic features.

In general, men and women have been understood to communicate differently in the process of forming connections. We wanted to further investigate differences across genders in connection-building communication in the context of online social platforms.

These concepts guided our research, as we examine the linguistic differences of male and female influencers in their written responses to followers’ comments online. 

Background

Due to our scope of interest in influencers’ platform maintenance, we examined existing literature on gendered and computer-mediated communication.

The dominance model suggests that female language use reflects male dominance in society (Lakoff, 1975), whereas the difference model proposes that differences in language between men and women reflect different cultures of conversation (Tannen, 1990). Despite both styles serving the same communicative function, women use rapport-oriented conversation, which is more emotional, while men use report-talk, reporting fact-based information and competing for hierarchy in conversation (Tannen, 1990). Further work has been done on gendered communication differences to see what linguistic features can be attributed to men: in homosocial contexts, men use expressions like “dude” or “bro,” as their way of performing male expectations, indexing their heterosexuality to promote heterosexuality as their preferred orientation. (Van Herk, 2018, p.109).

Since we are looking at social media influencers, we also looked at computer-mediated communication (CMC) studies to see how these patterns reflect in a modern online context. The content of CMC messages by females is more expressive than males, reflecting a female’s social role of being emotionally expressive and collaborative, as mentioned in the Tannen’s model (Fox et. al., 2007, p. 395). Regarding CMC-specific linguistic features like emoticons, studies suggest that females use emoticons as a means of expressing solidarity, support, positive feelings, and gratitude––reinforcing the existing stereotype that females are more emotional than males (Wolf, 200, p. 827). Literature on text-based punctuation in online messages suggests that digital cues, such as excessive punctuation and capitalization, increased the bonding of female friendships (Sherman et. al., 2013). These cues were made frequently by young women to convey emotion in their text-based conversations.

Our main research question is the following: “To what extent does the gender identity of YouTube fitness influencers affect the digital linguistic expressions they use to establish solidarity with their followers?”

In our research, we observed that females used more frequent expressions than males across four different features we examined; however, by narrowing down on more specific sub-categories under the features, we found that even though males and females used different linguistic expressions, both male and female fitness influencers were using different tools to pursue the same purpose of establishing solidarity with their fans.

Methods

We conducted our study by analyzing the computer-mediated communication used by ten female and ten male YouTubers. We narrowed our sample choice by selecting YouTubers who belong to the fitness industry, create fitness content for YouTube, and speak English. Our sample was categorized into two sections: influencers with less than one million subscribers (Table 1) and influencers with over one million subscribers (Table 2). To avoid bias, data were collected from each influencer by randomly selecting ten interaction-based comments from two randomly selected workout videos.

We categorized our data into four linguistic features based on the most salient differences we observed among the comments of male and female fitness influencers. Our data was analyzed based on the frequency of emojis, exclamation points, capitalization, and pet names. In terms of emojis, we looked at the types of emojis that were being used differently by males and females. The specific emojis were grouped into three distinct categories: facial expressions, gestures, and non-human symbols (fire, stars, sweat, etc).

Results and Analysis

First, we looked at the use of emojis in YouTube comments from female and male influencers. Symbol and facial expression emojis were popular for women, using a variety of faces such as 🤪 and hearts 💖.

Men also used emojis frequently but the specific emojis they used differed from what females used––males instead opted for symbols like 🔥 or 💦. 

Overall, women used more expressive facial emojis along with many gesture emojis (Figure 1). In general, women used facial expression, symbol and gesture emojis more frequently than men.

Second, we analyzed punctuation by examining the use of exclamation points in influencers’ responses to comments. Women tended to use exclamation points in most of their replies, often using several exclamation points in a row. In contrast, males did not use exclamation points as frequently, and when they did, they only used 1-2 per comment.

While men did use exclamation points, they did not use them to the same extent as women. Women used exclamation points more frequently, totaling over 100 times throughout comments compared to just over 25 by men (Figure 2). The number of comments in which these features appeared was identical for male and female. Females typically used digital cues including excessive punctuation to better convey emotion online.

Next, we considered influencers’ usage of capitalized words within sentences. We found that female influencers were more inclined to capitalize individual words or phrases when replying to comments. Women often capitalized words of encouragement like “good job” yay” or “yesss”. Conversely, men rarely utilized the capitalization of words.

The capitalization of words (particularly for emphasis) was used by females at a higher rate than their male counterparts. Whereas women did this over 100 times throughout our data, men did not even reach a count of 5 (Figure 3).

Finally, we looked at the pet names influencers used when responding to comments. Female fitness influencers often used words such as “girl,” “queen,” and “babe” to address their followers, whereas males used terms like “man,” “buddy,” and “mate” to address their followers in a similar supportive fashion. Figures 5 and 6 display the overall comparisons of the four linguistic features’ rate of appearance in 10 comments written by the male and female social influencers.

We observed that females used pet names more frequently than males, but the difference was not as large (Figure 7). 

 

Discussion and Conclusions

There are four key insights that summarize our research results.

First, by taking a closer observation at the types of emojis male influencers used, we found that males usually used bicep emojis whereas females did not use them at all. Females generally used more emojis across all emoji categories. However, by narrowing down to a more specific sub-parameter within symbolic emojis, we observed that males were in fact using a different tool to strengthen their relationship with their predominantly male fans. 

This insight suggests that in CMC, both males and females likely strive to establish solidarity with their fans but through different linguistic tools. On social media platforms, influencers of all genders are driven by financial motivations to attract viewers by crafting themselves as more responsive and supportive than their competitors.

Second, the nature of the male influencers’ comments was more action-oriented than that of female influencers’ comments. For example, males often posted comments such as “Keep going” and “Well done!”, whereas females often posted gratitude-expressing, emotional comments such as, “Thank you!!!” and “ILY MY QUEEEEN”.

This second point illuminates that most of these interactions took place between same-sex followers and influencers. This phenomenon could be attributed to the fact that the fitness objectives of the videos were inherently geared to target the followers of the same sex as the influencers. For example, we noticed that the majority of the videos created by female fitness stars tend to have titles such as, “Intense Glute Workout”, but males posted videos with titles like “Build a Bigger Chest”. These fitness videos align the body areas that respective genders tend to visually prioritize when developing their body muscles. In general, females are more self-conscious of their leg and glute areas whereas males typically focus on building the size of their upper body.

Third, we observed a dichotomy of ‘emotional vs calm’ which is a digital adaptation of the ‘report vs rapport” model. In particular, male influencers’ average length of comments is significantly less than female influencers’ average length of comments. The male influencers’ responses were calmer than those female influencers. We think this ‘emotional vs calm’ dichotomy is a formal theorization of what computer-mediated gender communication can look like in the context of our digital influencer study. Our research also invites further studies by future socio-linguistic scholars interested in the intersection between gendered communication and online social media platforms.

Lastly, with male influencers specifically, we observed that those with a smaller following responded more frequently to comments than those with a larger following. Once reaching a certain level of popularity (over 1 million), males responded less frequently. We theorize this is because males use more feedback only before their platform grows to a certain level of popularity. Given that our study only examines 20 social influencers on YouTube, we’d like to invite future researchers to conduct more studies in these areas.

In short, our research study shows that females are predominantly more expressive than males across all 4 linguistic categories, but males have more frequently used bicep emojis in particular. In addition, even though females’ responses were visibly more expressive in terms of the frequency and variety of emoji usage, both males and females were pursuing the same purpose of establishing solidarity with their fans, by using different tools.

We argue that the Tannen model is being applied differently in the context of computer-mediated communication and the nature of social media, as social influencers, whether male or female – are in positions to appeal to their general audience. We also propose the “emotional vs calm” dichotomy observed from gendered communication in online platforms and invite further research to be done in this area.

Among several, one limitation of our research is that we did not incorporate the nature of the followers’ comments that the influencers responded to. In general, we observed that most viewers’ comments were positive, grateful, and supportive. This research invites future studies to undertake how the responses would look different towards comments that are hateful or negative. In addition, more studies on how non-famous males and females differ in their digital communication on social online platforms are needed.

 

References

Bamman, D., Eisenstein, J. and Schnoebelen, T. (2014), Gender identity and lexical variation in social media. J Sociolinguistics, 18: 135-160. https://doi.org/10.1111/josl.12080

Fox, A. B., Bukatko, D., Hallahan, M., & Crawford, M. (2007). The Medium Makes a Difference: Gender Similarities and Differences in Instant Messaging. Journal of Language and Social Psychology, 26(4), 389–397. https://doi.org/10.1177/0261927X07306982

Herk, G. V. (2018). Gender. In What Is Sociolinguistics? (pp. 96-116). Wiley Blackwell.

Lakoff, Robin. 1975. Language and woman’s place. New York: Harper Colophon Books.

Sherman, L. E., Michikyan, M., & Greenfield, P. M. (2013). The effects of text, audio, video, and in-person communication on bonding between friends. Cyberpsychology: Journal of Psychosocial Research on Cyberspace, 7(2), Article 3. https://doi.org/10.5817/CP2013-2-3

Tannen, D. (1990). “Put Down That Paper and Talk to Me!”: Rapport-talk and Report-talk. In You Just Don’t Understand: Women and Men in Conversation (pp. 74-95). HarperCollins.

Wolf, A. (2000). Emotional Expression Online: Gender Differences in Emoticon Use. CyberPsychology & Behavior, 3(5), 827–833. https://doi.org/10.1089/10949310050191809

Background

At UCLA, undergraduate students who have excelled in certain STEM classes can return to those courses and help teach current students as part of the Learning Assistant program. As part of the Learning Assistant program, undergraduate Learning Assistants attempt to foster an inclusive learning environment where students can better understand the topics covered in their respective courses. They are trained to promote an interactive environment, even through Zoom, as well as use a Socratic questioning style which encourages students to think independently (Talbot et. al 2015). As seen in this video, Learning Assistants are encouraged to let students figure out the answer themselves rather than feeding it to them. Promoting interaction and encouraging independent thinking, according to the program, force students to think more critically about the content they have to learn, and, as a result, develop a more solid understanding compared to if one were to just passively teach it (Furtak and Kunter, 2012).

However, a problem potentially arises when looking at the interactions between Learning Assistants and students themselves. In order to promote inclusivity, LAs are encouraged to act with kindness and politeness towards the students (Sellami et al., 2017). Creating a warm, supportive atmosphere has been shown to improve with facilitating student learning (Jardine et al., 2020). However, the kinds of polite statements which the Learning Assistants use may differ based on their gender, as well as potentially even the gender of the student who they are addressing. This may potentially lead to differential learning outcomes based on gender due to different levels of support from the Learning Assistants. This study sought to answer one major question: do LAs tend to use a certain kind of politeness more than another when students identified as the same gender as themselves?

Methodology

This study collected data from ten Learning Assistants for an organic chemistry class, five male and five female, who agreed to participate. These Learning Assistants recorded their discussion sections, which were held over Zoom, and sent them as attachments in emails for this study’s analysis.

When analyzing the conversations that Learning Assistants had with students, this study looked for two main kinds of politeness: positive and negative politeness. Each of these was categorized in the following ways (Brown and Levinson, 1987):

Positive Politeness: Words of affirmation towards a person who is part of an in-group or considered a friend. A very casual and friendly statement with the intent of encouragement or support.

Negative Politeness: A more formal and avoidant strategy used to maintain distance between the user and the target. There is more restraint with negative politeness and, occasionally, apologetic wording. It may involve hedging, which is used to add uncertainty or decrease assertiveness to a particular statement.

When analyzing each of the Zoom recordings, this study categorized Learning Assistant-student interactions based on the LA’s gender, the student’s gender, and whether the student’s provided answer was correct or incorrect. The number of instances of each kind of politeness was tallied for each recording. This was then divided by the number of minutes the LAs spent talking to the students in order to obtain frequencies for each kind of politeness. The frequencies for each LA were averaged, and then analyzed as data below.

Analysis

We can use the technique of discourse analysis (Janks, 1997) to analyze the conversation in Example 1 between a male Learning Assistant and a male student. 

Transcription Key:

MLA indicated Male Learning Assistant

MST indicates Male Student

FST indicates Female Student

(time) indicates a pause for [time] seconds

- indicates a short, untimed pause

: indicates a stretched syllable

(.h) indicates an inhalation

Underline indicates positive politeness phrases


Italics indicates negative politeness phrases

For the purposes of confidentiality, the actual names of the students and LA will be replaced in conversations with MST/FST/MLA, respectively.

In Example 1, the male Learning Assistant uses four instances of positive politeness with the male student. When looking at this conversation with the idea of a student’s perceived levels of support in mind, one can see how this student would feel very supported. The casual language which the Learning Assistant uses, such as “dude,” as well as the phrases of encouragement such as “You got it!” and “Exactly!” would most likely make this student feel as if his answer was validated and that he belonged in such a learning environment. The four instances of positive politeness, marked using an underline, reaffirm the student that he is correct. The Learning Assistant’s action of making this student more comfortable in the discussion section may, in turn, make him feel more confident about his ability to succeed in the class.

Now let’s take a look at an example of a conversation where the same male Learning Assistant interacts with a female student:

Much like the male student from the previous example, the female student in Example 2 also answered the male Learning Assistant’s question correctly. However, the Learning Assistant’s response towards her answer was not as comforting and casual as it was with the male student. For example, he only used two instances of positive politeness compared to the four he used in the previous example. Additionally, he now uses two instances of negative politeness, indicated in italics, when asking her to further explain her answer.

Similar overall trends to the two examples above are seen when looking at the combined data of this study. This data is pictured in the two graphs below:

There are two major points of interest in Figure 1. First, when looking at each interaction, every correct response yielded more frequent positive politeness than negative politeness. However, more importantly, one can notice that male LA/male student and female LA/female student interactions have more frequent use of positive politeness than the other two kinds of interactions with different LA/student genders. This finding supports the idea that positive politeness is more commonly used when an LA interacts with a student of the same gender.

The results for incorrect responses in Figure 2 are not as conclusive as they were for correct responses. Once again, this data shows that positive politeness is used more frequently when a student and LA have the same gender compared to when they have different genders. More importantly, however, is the fact that female/female interactions had more instances of negative politeness than female/male interactions here. This finding disproves the idea that negative politeness is dependent on gender in interactions. Unlike positive politeness, which was higher in same-gendered interactions for both examples, negative politeness does not have any trend with relation to gender.

Conclusion

After analyzing the data, the study ultimately concluded two things. The first conclusion reached was that positive politeness is used more frequently when Learning Assistants interact with students who share the same gender as them. In addition to this, however, the study also concluded that the frequency of negative politeness is independent of the Learning Assistant and student’s genders.

The finding that positive politeness is higher in same-gendered interactions may have implications with regard to perceived student support: students may feel more supported if they interact with LAs who identify with their same gender compared to LAs who do not. Increasing awareness of this issue amongst LAs may allow them to be cognizant of the existence of such a bias, and therefore, would help them know to be just as supportive towards differently-gendered students. Similar future studies should include non-binary participants in order to extend the results past the gender binary.

 

References

Brown, P., & Levinson, S. C. (1987). Politeness: Some Universals in Language Usage. In Politeness: Some universals in language usage (pp. 311-323). Cambridge: Cambridge University Press.

Furtak, E. M., & Kunter, M. (2012). Effects of autonomy-supportive teaching on student learning and motivation. The Journal of Experimental Education,80(3), 284-316. doi:10.1080/00220973.2011.573019

Janks, H. (1997). Critical discourse analysis as a research tool. Discourse: Studies in the  Cultural Politics of Education,18(3), 329-342. doi:10.1080/0159630970180302

Jardine, H., Levin, D.M., & Cooke, T. (2020). Group Active Engagement in Introductory Biology: The Role of Undergraduate Teaching and Learning Assistants.

Sellami, N., Shaked, S., Laski, F. A., Eagan, K. M., & Sanders, E. R. (2017). Implementation of a learning assistant program improves student performance on higher-order assessments. CBE—Life Sciences Education,16(4). doi:10.1187/cbe.16-12-0341

Talbot, R.M., Hartley, L., & Wee, B. (2015). Transforming Undergraduate Science Education With Learning Assistants: Student Satisfaction in Large Enrollment Courses. The journal of college science teaching, 44, 24.

Introduction

TikToks are vertically filmed videos 6-60 seconds long. TikTokers play off various viral trends with repeated phrases and terms, and introduce new lexical items into users’ everyday speech. Importantly, users comment on others’ TikToks and share TikToks on other social media platforms. Also of note, TikTok users will write specific hashtags in their captions to make their videos appear on the “For You page”, the place where TikTok users discover new videos. Due to a combination of the users comments, hashtags, and sharing it is common for new words to become extremely viral through widespread use. TikTok’s powerful influence is reflected by the fact that the social media platform has over 800 million users worldwide. Thus, TikTok forms a kind of speech community with jargon through interaction between speakers. We analyzed college-aged students’ usage of the TikTok lexical item “periodt”, used to emphasize a point or signal the end of a discussion. We hope to answer, does a speaker’s gender identity and/or sexual orientation affect whether they adopt TikTok terms into real-life social interactions?

Background

The field of sociolinguistics understands linguistic variation as an “essential feature of language” through which speakers “index” aspects of their identity (Eckert 2012: 94). Specifically, gender is performance, in which speakers use linguistic features to construct and reinscribe their gender for their audience (Butler 1990: 179). For the older 18+ TikTok population, which includes our target population, using TikTok language requires more speaker agency because TikTok is not the norm, and thus TikTok language indexes aspects of identity for these speakers. Holloway and Valentine (2014) show that a teenager’s identity can become a mixture of their online and offline influences, since teenagers perceive both their online and offline identities as real spaces. Based on the studies by Eckert and Butler we see that a person’s identity is reflected in their lexicon. Our study seeks to examine whether the influence of the online spaces of TikTok will be reflected in our participant’s lexicons. Podesva (2011) analyzed the Californian vowel-shift for one LGBTQIA+ speaker and demonstrates that speakers use phonological features to index their identity. Podseva focuses on speech acoustics in various social settings. Our study is different from this in the way that we are examining our participant’s lexicons, and in the way that we are analyzing a large number of people rather than one person. The study that was most similar to ours was Banman’s (2014) which analyzed Tweets by men and women and compared gender markers such as pronouns. Banman et. al found that there are lexical items strongly associated with each gender. Our study is similar because we also seek to examine the lexical differences between gender, but also different because we are also examining their sexual orientation. To learn more about gender and other demographics in social media, check out this TED Talk by Johanna Blakley.

Methods

We collected data from 100-200 American college students (aged 18-22) through an online survey with questions about usage of TikTok lexical items as well as gender and sexual orientation. Our preliminary hypothesis was that members of the LGBTQIA+ community are more likely to know and actively use TikTok language, since many TikTok terms are associated with said community. We also hypothesized that self-identified females are more likely to use TikTok terms, because of TikTok’s origins in Musical.ly and because self-identified females are usually the pioneers of new lexical items. The word that we chose to analyze was “periodt”, since it is a word that not only has roots in the LGBTQIA+ community, but was also popularized through TikTok by the use of hashtags and viral audios used in various videos. If you’re interested, check out this TED Talk by Dao Nguyen that discusses what kinds of videos and topics go viral.

Results

We received 109 total responses. Of those, we received 80 female responses, 26 male responses, and 3 nonbinary responses. Although our percentage of female responses was disproportionate, we still found that a far greater majority of female respondents were familiar with the term.

Similarly, according to our hypothesis, when sorted by self-reported sexual identity, all of our LGBTQIA+ respondents were familiar with the term, while a smaller percentage of our heterosexual respondents were familiar with the term. At the same time, we are aware that there is further diversity within the queer community, so analyzing the entire LGBTQIA+ community as one homogenous section does weaken our analysis.

 

 

 

Regardless of gender identity and sexual orientation, overall, the participants feel the most comfortable using the term “periodt” online with their peers, a little less comfortable when using it face-to-face with peers, and almost equally uncomfortable when using it online and face-to-face with a professor/boss. When using the term with peers, female and non-binary participants feel slightly more comfortable when using it online than using it face-to-face. From our data, male LGBTQIA+ participants clearly feel more comfortable than other groups when using the term “periodt” with their peers online and face-to-face, and slightly more comfortable when using it with a professor/boss online. But since only 2 LGBTQIA+ male participants responded, we cannot confirm that this pattern applies to most people with this gender identity and sexual orientation. Also based on the results we received, non-binary participants feel less comfortable than other groups when using the term with their peers online and face-to-face. But since only 3 non-binary participants responded, we cannot confirm that this pattern applies to most people with this gender identity and sexual orientation.

Discussion and conclusions

Based on our results, we conclude that gender identity and sexual orientation seems to affect the familiarity with newly-emerged TikTok terms, but have little effect on the usage of those terms in both online and in-person settings. We collected data on many factors that we didn’t have the time or space to explore, such as participants’ native languages and hometowns. We also asked participants which other TikTok terms they were familiar with, which in the future could be helpful to contextualize “periodt” among other TikTok lexical items. An additional factor that we didn’t have the space to explore was race. Although the word “periodt” originates from the African American community, we did not analyze whether our participant’s knowledge of the word could originate from their racial backgrounds. The final shortcoming of our study was our lack of LGBTQIA+ participants. In an ideal study, our study would have equal amounts of participants for every subdivision we analyzed, especially for the male LGBTQIA+ participants.

In further studies, we would be interested in correlating time spent on TikTok with comfort using the word. Then, we would analyze whether time spent on the app factors more into TikTok term usage than gender or sexual identity (Holloway and Valentine, 2014). A similar study could also be performed on other words that have become popular with AAVE and LGBTQIA+ roots, like “shady, tea, sis”, but on platforms like YouTube or Twitter. Also of interest, how do people, especially non-binary and LGBTQIA+ groups, acquire TikTok terms? Our study asked whether the participants knew the word, but not whether the participants learned the word from TikTok. A few more specific questions to investigate include: Is there a difference in the usage between talking with close friends and classmates? Is there a difference when talking with the same group of people, but about different topics?

Our study regarding the term “periodt” is also important for broader research and issues. The large number of participants using “periodt” despite not knowing the origins of the word is emblematic of the larger issue of erasure in western society. Despite the fact that “periodt” has its origin from AAVE, the word was co-opted by the LGBTQIA+ community during the late 1960s. “Periodt” was then popularized in mainstream culture as a word that indexes LGBTQIA+ membership, since it was popularized by shows like Rupaul’s Drag Race, Queer eye, and various YouTube content creators. Prior to TikTok, “periodt” was only popular within the LGBTQIA+ and African American communities. After “periodt” was popularized on TikTok through famous audios, and hashtags, the word had become viral and it was no longer used exclusively by the queer and African American communities. Our study shows an abundance of users who use “periodt” without actually knowing the origin of the word. Other viral phrases that were created by the black community and was co-opted by the LGBTQIA+ along with “periodt” that do not have their origins widely acknowledged would be “spill the tea, sis, yas, queen, shady”. Acknowledgement of the origins of “periodt” is important because of the U.S. erasure of black and queer history.

 

References

Bamman, David, et al. “Gender Identity and Lexical Variation in Social Media.” Journal of Sociolinguistics, vol. 18, no. 2, 2014, pp. 135–160., doi:10.1111/josl.12080.

Butler, Judith. 1990. Gender Trouble: Feminism and the Subversion of Identity. New York: Routledge

Eckert, P. (2012). Three waves of variation study: The emergence of meaning in the study of sociolinguistic variation. Annual Review of Anthropology, 41, 87-100.

Holloway, Sarah, and Gill Valentine. “Cyberkids? Exploring Children’s Identities and Social Networks in On-Line and Off-Line Worlds.” 2014, doi:10.4324/9781315011257.

Kulkarni, Vivek, and William Yang Wang. “TFW, DamnGina, Juvie, and Hotsie-Totsie: On the Linguistic and Social Aspects of Internet Slang.” 22 Dec. 2017.

Podesva, R. (2011). The California vowel shift and gay identity. American Speech, 86(1), 32-51