Language and Power in Politics: A Gender Stereotype Game

Sarah Thomas, Emma Greene, Cameron Brewer, Jamie Dela Cruz

With 2020 fast approaching, everybody has their eyes on the many candidates running for president, calling into attention how they frame particular issues to gain public support. The mixed-gender debates within the Democratic party raise the question of how this new dynamic will affect future political conversations. However, it’s no secret that women have a harder time making themselves heard, with their gender inspiring the public to maintain traditional stereotypes about them.

Existing gender inequalities, or sexism, persist in language, and can be maintained through the speaker and their audience (Suleiman & O’Connell 2008). In this context, the relationships with a candidate to other candidates and the public reflect a power dynamic that women must handle to assert their own place in the political sphere. To understand how these candidates navigate mixed-gender debates, we looked at one of the Democratic primaries, paying special attention to what language tactics they used.


The United States is a melting pot of individuals with unique backgrounds, cultures, and ideals. However, this diversity is not adequately reflected in the country’s political realm. One such identity that is not equally represented is that of women; despite making up over half of the United States’ population, women represent a mere 23.6% of congressional office (Rutgers 2019). Gender politics is essential to help explain why women are so underrepresented in elective offices.

Every speech, conversation, and debate is thoroughly analyzed, critiqued, and judged by the press and public. With the upcoming Presidential election, the Democratic party’s mix-gendered political debates create a distinct dialogue between candidates and with their audience.

Politicians are intentional with their language, having a team of writers work with them to decide how they want to speak about a particular issue. In the context of gender, some linguistic functions are perceived as linked to one gender more than the other; this association gives rise to potential stereotyping of politicians based on how they speak, not their politics (Suleiman & O’Connell 2008). Communication differences and gender communicative patterns have been found to link specific traits with gender (Grebelsky-Lichtman & Katz 2019).

The present study seeks to better understand how gender communication structures of women and men in politics compared to one another in terms of usage and relate these findings to press and public reaction.

Background information

In a study that looked at how much women and men spoke, it was found that even though women speak less than men, men lashed out when they believed there were more women talking than them. This discovery emphasized the idea that men truly dominated public talk and pushed back when women tried to gain equal footing (Mooney & Evans 2015). Multiple studies have found a relationship between patterns in communication differences and gender, which are summarized in the following table (Grebelsky-Lichtman & Katz 2019).


The project design investigated how politicians spoke in a mixed-gender debate, specifically the October 15, 2019 Democratic Primary Debate.

Using the video recording  and online transcript  provided by CNN and the Washington Post, the candidates were considered individually and compared with each other. He or she was analyzed for feminine and masculine identifiers, based on Grebelsky-Lictman and Bdólach’s verbal gender communicative accountability framework, and received a mark for each unique use of the aforementioned linguistic features (2019).

After averaging how frequent each gender and each individual candidate used communicative identifiers, the speaking styles of the top and bottom-rated candidates of both genders were examined to determine whether there was a relationship between communicative patterns and public appeal.

The purpose of analyzing political candidates in the context of gender-oriented communication is to answer the following questions:

    1. How do the gender communication structures of female and male American politicians compare?
    2. What, if any, masculine communicative structures are most commonly used by females? What, if any, feminine communicative structures are most commonly used by males? In these instances, are structures that are used equally by both females and males more gender-neutral?
    3. Are a candidate’s polling numbers related to which opposite gender communicative structure they use?


From New York Time’s Democratic Polling Data, we found Elizabeth Warren to be the top female candidate and Joe Biden the top male candidate. A total of 61.7% of Elizabeth Warren’s speech patterns were considered masculine, while Joe Biden’s speech patterns had a ratio of 70.3% that were masculine-identified. When comparing the usage of specific identifiers, Warren favored emotional reference, personal example, display of solution and practical economic issues. Biden favored assertive and direct speech, pertinacity, and speech in first person, singular.


When comparing Castro, the lowest polling male candidate, to Biden, we found that his main masculine identifiers consisted of action demand, direct speech, display solution, and speech in first person singular. Similarly, Biden’s consisted of activity, direct speech, and action demand. When examining their feminine identifiers, Castro primarily used concrete examples, personal examples, and emotional reference, whereas Biden used hesitant speech, passive speech, and questions.

Klobuchar, the lowest polling female, primarily used accusatory speech, rationality, and display solution as her masculine identifiers, while Warren used display solution, activity, and practical economic issues. Both used emotional reference and personal examples as their primary feminine identifiers.

Discussion and conclusions

With this being the fourth debate in the race for the Democratic nomination, the public is familiar with each candidate’s communication style by now. Interestingly, it can be seen from the following graph that Elizabeth Warren’s likability, or public appeal, was projected to dramatically increase, which it did: she began with 5% support, and following the October debate, it rose above 20%. 

In terms of feminine communication structures, Warren focused on emotional reference and personal examples over five times more often than Biden. It is clear that despite Biden’s use of identifiers, they were commonly used as a pause to gather his next thought, not as part of his argument. With feminine speech being associated with more passive and hesitant behavior, we suspect that using personal example and emotional reference are effective identifiers that don’t hinder perceived competence. However, it is also significant to note that because Warren utilized masculine identifiers such as significant opinion or action demand, they were used more as persuasive, support gathering tools.

Warren dominated over all other candidates, including Biden, with her constant usage of masculine identifiers, such as display solutions and references to practical economic issues. Biden dominated in more aggressive and assertive language, using direct speech, pertinacity and speech in first person singular. While Biden can use aggression to show strength and leadership, when a woman uses aggression, the public reaction is widely different. Aggression in female communication patterns is portrayed more often emotionally unstable outbursts or unlikeability, or even “shrilly.” 

Despite their drastic differences in polling success, Biden and Castro had extremely similar masculine identifiers: assertive speech and action demand. What seems to truly set these candidates apart is their use of feminine identifiers; Castro’s feminine identifiers functioned more emotionally and anecdotally (concrete examples, personal examples) and Biden’s were more passive (hesitant speech, passive speech). This implies that males are judged more on their usage of feminine speech than masculine speech. Also, there is a question of whether men are judged more for their feminine identifier use than women. Specifically, while Castro’s main feminine identifiers were identical to Warren’s, they proved to be much less effective for Castro than for Warren.

When comparing the top and bottom-rated female candidates, Klobuchar also employed the same feminine identifiers as Warren; however, their masculine identifiers differed. Warren exhibited a more calm demeanor, focusing on practical masculine identifiers (activity, practical economic solutions, display solution). Klobuchar’s reliance on accusatory and emotionally-charged masculine identifiers proved less effective in the public’s eyes.

What set the top and bottom candidates of each gender apart was their use of identifiers that were typical of the opposite gender. It makes sense that the public will more closely analyze communicative patterns that are atypical of a candidate’s gender when forming an opinion about them.

Will we ever have a female president? After the unexpected upset of Donald Trump over Hillary Clinton in the 2016 Election, strong opinions formed regarding the deciding factor that led to Trump’s victory. Many people claimed the result was due to sexism in the U.S., while others summed it up to Hillary’s lack of likeability.

What if these things are just extensions of how each candidate chose to present themselves? Hillary often used very aggressive masculine identifiers, as Trump did, but what really separated them was the perceptions the public had of both of them. For women, as shown in one of Trump’s tweets, aggression and anger is viewed as a sign of weakness, while aggression in male candidates is validated and expected.

This adds to a common theme in which the effectiveness of identifiers is very context specific, depending on the gender of the candidate, their intended audience, etc. These linguistic double standards add up, and although the message the candidates are trying to convey, much of their impact on the audience has to do with the way they present their messages, and their personal identity that impacts the potential voters.


More on the October 15, 2019 Debate:

         Want this article in powerpoint form? We’ve got you.

         Watch the Debate on CNN 

         NPR’s Fact Check 

         The Fourth Democratic Debate in 6 Charts

2020 Presidential Race from the Democratic Side:

          NY Times: Which Democrats Are Leading the 2020 Presidential Race? 

          NPR: Tracking the Issues in the 2020 Election 

         Statistics of Women in Elective Office


Referenced Journal Articles:

Grebelsky-Lichtman, T., & Bdolach, L. (2017). Talk like a man, walk like a woman: an advanced political communication framework for female politicians. The Journal of Legislative Studies, 23, 275–300. doi:

Grebelsky-Lichtman, T., & Katz, R. (2019). When a man debates a woman: Trump vs. Clinton in the first mixed gender presidential debates. Journal of Gender Studies, 28, 699–719. doi:

Suleiman, C., & O’Connell, D. C. (2008). Race and gender in current american politics: A discourse-analytic perspective. Journal of Psycholinguistic Research, 37(6), 373-389. doi:10.1007/s10936-008-9087-x

Mooney, A., & Evans, B. (Eds.). (2015). Language, Society and Power: An Introduction. (4th ed.).

Read more

Exploring the Difference in Filler Word Frequency between Non-Native English Speakers and Native English Speakers

Clayton Puckett, Nicole Fonacier

Typically, when thinking about filler words, the immediate interpretation is that they’re a result of bad habits. Yet the purpose of filler words differs depending on the setting, and its frequency varies from speaker to speaker. In both informal and formal speech, filler words can be used to begin or continue streams of thought, assuage discomfort in silence, and allow time to process information. If filler words are used excessively, it can either negatively impact the credibility of the speaker or it can help string together words. This raises the question of why we use filler words, who is more likely to use them, and whether or not using them is indeed such a horrible thing to do. To answer these questions, we conducted a study focusing on the differences in filler word frequency between non-native and native English speakers. Participants were asked to answer a series of questions that would encourage the usage of verbal fillers through memory recollection and impromptu thinking; the conversations were recorded and the number of filler words used were then tallied as a proportion to the number of total words spoken. We hypothesized that non-native English speakers will use filler words less frequently in their responses due to a more conscious awareness of fluency. The results from our data supported this hypothesis: on average, native English speakers used about 4 more filler words for every 100 words spoken when compared to the non-native English speakers in the study. This suggests that the frequency of filler words could possibly be influenced by comfort levels in practicing a language and whether that language is the individual’s native language.

Background Information

We began by defining filler words as any word or utterance that takes away from the main message of the speaker and paid particular attention to two categories: filled pauses and discourse markers. Examples of filled pauses include “like”, “um”, “uh”, “okay”, and “so” (Laserna, Seih, Pennebaker, 2014). These are what most people are familiar with, since filled pauses tend to be more common in everyday conversation and more tempting to use in stressful situations such as interviews or presentations. Discourse markers are words such as “I mean”, “you know”, and “like.” They differ in that there is meaning to its placement. The purpose of discourse markers is to connect and organize what is being spoken rather than serving as something to simply fill a gap in speech. If your friend invites you to hang out on the week you have finals to which you respond “you know, that does not sound like a bad idea since I have been studying so hard and am feeling ready”, that initial “you know” is the discourse marker. It was intentionally placed as a sign of agreement. Further research suggests different causes for verbal interpolations where the individual utilizes filler words as a means to shape identity (Duvall, Robbins, Graham, & Divett, 2014). This happens when fluency in a language is being used as a means to fit into a culture. People seeking to achieve this are more inclined to eliminate filler words from their speech to better identify with the more dominant speech community. Other research perceives filled pauses as cues for the creation of more complex ideas (Watanabe, Hirose, Den, & Minematsu, 2008). By assessing and comparing the frequency of filled pauses used by native and non-native English speakers, we can begin to draw conclusions regarding the usage, purpose and likely user of the filler words within verbal speech.


Data was collected from 10 current UCLA students: 5 native English speakers and 5 non-native English speakers. All participants were found in UCLA’s residential buildings. With the participant’s consent, each conversation was recorded and later used to tally the filler words as well as the total word count. A total of eight to nine questions were asked to the participant (see Figure 1). 

Figure 1: Questions asked to the Participants

To ease the participants into the study and create a comfortable environment, warm-up questions asking general information preceded the actual assessment questions of the survey. We hypothesized that filler words will be used less frequently among the non-native English speakers given that their basis of learning and the experience they may have with the English language is likely to have originated from a more formal setting. To elaborate, native English speakers do not have to fear about their fluency and are therefore not as aware of their filler word usage. The tallied data as rations was then translated to percentages in order to better compare the usage of filler words between the two groups.


It’s important to note once again that the frequency of these filler words was not recorded as a raw qualitative count, rather as a ratio indexing the usage of filler words within the user’s speak. This is a vital part of this data recording process as it levels the playing field and virtually eliminates the possibility of large ambiguities between shy and talkative participants for example. Without this adjustment, the results would be largely skewed as some participants would naturally give longer answers than others, therefore further increasing the likelihood of using filler words within their responses. After the results had been calculated, our hypothesis proved to be correct, although not by a massive margin. The percentages of filler words for the native English speakers were as follows: 15%, 14%, 15%, 6.9%, 5.8%. The mean is therefore 11.34% (see Figure 2).

Figure 2: Frequency of Filler Words – Native English Speakers

Alternatively, the percentages of filler words for the non-native English speakers were as follows: 14%, 5.6%, 3.2%, 12%, 3.8%. The mean is therefore 7.72% (see Figure 3). The difference between these two averages is close to 4% which is a statistic previously referred to stating that on average for every 100 words spoken, native English speakers used about 4 more filler words in comparison to non-native English speakers.

Figure 3: Frequency of Filler Words – Non-Native English Speakers



There are a variety of factors that could be improved throughout this experiment. Although the research question is strong, the lack of time and resources have limited us in the ability to carry out extensive research. It has been fascinating to see the results of this experiment coincide with our hypothesis, but given such a small sample size, the results could easily result in an opposite manner if we were to repeat this experiment. We still stand by our hypothesis, but to truly strive for conclusive data, we would have to enlarge our sample size and take into account a number of other factors. There are a number of factors that can additionally skew data that would either prove too hard to control or would need to be drowned in a large number of participants. These factors include influential features such as the participants educational history learning the English language as well as the participants receptiveness to language learning itself. To further articulate, the participants educational history learning the English language would include not only include the number of years that s/he has been learning it, but also the environment in which it was learning or taught in, the number of years being immersed in an English speaking society, and the frequency of English social interaction that the participant experiences. Alternatively, the participants receptiveness to language learning as a process may be a factor of how many languages they speak and how capable they are at adapting to new environments and their ability to pick up new languages.


The design of this project proved worthy of our efforts, and as we finalize our findings and reflect on the work we have conducted, we are able to critique our methods. If this experiment were to operate on a large scale, it could benefit greatly from a large sample size. With a reconfiguration of data recording methods to accommodate for the masses of participants, this experiment would prove to be conclusive as outliers are adjusted for and the results transition towards statistical relevance.


Duvall, E., Robbins, A., Graham, T., & Divett, S. (2014). Exploring filler words and their impact. Schwa. Language & Linguistics, 11, 35-49.

Laserna, C. M., Seih, Y., & Pennebaker, J. W. (2014). Um . . . who like says you know: Filler word use as a function of age, gender, and personality. Journal of Language and Social Psychology, 33(3), 328-338. doi:10.1177/0261927X14526993

Watanabe, M., Hirose, K., Den, Y., & Minematsu, N. (2008). Filled pauses as cues to the complexity of upcoming phrases for native and non-native listeners. Speech communication, 50(2), 81-94.

Read more

Feminine Stereotypes: a Closer Look at The Princess and the Frog

Tasha Bierling, Sophia Maxson, Angela Ramirez, and Daniel Walsh

The Princess and the Frog, the diversity jewel in Disney’s crown, might not be as progressive as you’d think. Pink frilly dresses, big blonde hair, bows and sparkles, and an over-the-top, ditzy personality–it’s a stereotype we’ve all seen before in movies and tv. The creators of The Princess and the Frog took this well-known “dumb blonde” stereotype to another level with the character Charlotte La Bouff, and in doing so, perpetuated a stereotype to their viewers, many of whom tend to be very young.

It’s clear from her appearance that the New Orleans princess enjoys the feminine things in life. Charlotte visibly conforms to many misogynist stereotypes in both her appearance and her viewpoint. However, the focus of our research is to investigate whether her over-the-top feminine portrayal carries through in her speech as well. We have identified several linguistic features through other research that sound stereotypically feminine: uptalk, repetition, interjections, and rapid speech.

If Disney has employed these markers in the stereotypical portrayal of a female character, then they have presented a platform for impressionable youth to internalize these unsaid judgements. We are curious as to whether these aspects are more common in Charlotte’s speech than in that of the other main characters, Tiana and Naveen, whose presentations are less traditionally feminine.


In Disney’s The Princess and the Frog, the ditzy side character Charlotte is depicted as an overemphasized representation of femininity.

She is the beautiful, blonde princess, and her character carries all its demeaning perceptions: being unintelligent, emotional, spoiled, superficial, and socially incompetent. Her personality and appearance often exaggerate these characteristics–she has a high-pitched voice, often wears puffy pink dresses, and her main goal in life seems to revolve around “finding her prince.” The way she speaks also reflects this feminine stereotype.

Comparing how often Charlotte uses these features to that of the other characters could reveal how the film uses language to stereotype Charlotte as superficial and stupid. For instance, by contrasting Charlotte with Tiana, we see that Tiana’s personality and appearance are much less stereotypically feminine.

Tiana is portrayed as responsible, practical, intelligent, and grounded, opposing Charlotte’s over-the-top personality. This is important because society often stereotypes individuals similar to Charlotte, either by making them seem dumb or criticizing this representation of women. Furthermore, studying the portrayal of these stereotypes in Disney movies is especially important, since Disney’s audience is young and impressionable.

We theorized that specific aspects of Charlotte’s linguistic performance emphasize the stereotypes that go along with her character: to the audience, she is just an airhead. We will explain why it was so easy for Disney to use these aspects to box Charlotte in under “Background.”

Our goal is to analyze how specific aspects of her speech help her sound more stereotypically feminine, that is, dumb, emotional, and socially unaware. To do so, we will focus on her use of uptalk, rapid speech, repetition, and interjections. We will compare how often she uses these linguistic aspects with two other characters: Prince Naveen, a male, and Tiana, a female who displays more masculine characteristics, such as being hardworking, practical, and independent.

We will focus our study on Charlotte because she is a side character included for comedic purposes, and therefore the stereotype is exaggerated,  as well as to contrast Tiana’s character. As a result, we expect these stereotypes to be plentiful and more noticeable in Charlotte’s language than in that of the other characters.



Charlotte plays a rich, white princess in 1920’s Louisiana. She is an upper-class woman who displays acts of significant extravagance, lack of awareness, egotism, and shallowness, which are representative of a typical wealthy woman of this time. Her exorbitant wealth acts to demean her in a story about the American Dream, and her indoctrination in 1920’s gender roles helps categorize her as shallow.

If her appearance, worldview, and speech patterns weren’t enough to define her as a feminine stereotype, you only have to look at what she actually talks about: every time we see her, Charlotte is obsessing about a man, a prince. Disney defines her as “girly” by giving her dreams of a little girl: to marry a prince and become a princess.

Target population

We will be analyzing the linguistic patterns of Charlotte, Tiana, and Prince Naveen, and how they might factor into the overall portrayal of feminine expression in the Disney film.

Charlotte’s depiction as unintelligent suggests that women who hold similar characteristics are unintelligent as well.

A push for the audience to draw unfavorable conclusions towards individuals who hold the same linguistic and characteristic patterns as Charlotte causes prejudice. Disney designates Charlotte as a stereotype as a way to further emphasize the fact that feminine women like her only serve a role as side characters.       

Linguistic aspect

The linguistic aspects we will be investigating are uptalk, rapid speech, repetition, and interjections.

Uptalk refers to when your vocal tone rises at the end of a phrase, as it would in a question. It is often seen as expressing inferiority, lack of confidence, etc. It can be seen across cultures or dialects and usually occurs in a group that society puts into an inferior position (Guy, 1986). Uptalk has been recorded in teenagers and the working class, but it is the most identifiable as being feminine and is generally understood as a female trait.

Rapid speech refers to Charlotte’s tendency to speak faster than normal, or at least more quickly than Prince Naveen and Tiana. It has been observed in numerous scientific studies that men believe that women talk more frequently and faster than they actually do. Studies such as Cutler (1990) show that women are often perceived as talking faster and more often than they really do in everyday conversation.

Also, repetition refers to Charlotte’s tendency to state the same words or phrases in quick or immediate succession. Repetition is often used for emphasis or to imply desperation or urgency. Thus, Charlotte’s use of repetition could highlight her portrayal as being dramatic.

Finally, interjections are words that are thrown into sentences that seem to be unnecessary or lack meaning, such as “Oh!” or “Eek!”. They are often an exclamation and are usually related to an emotional reaction. Since women are often stereotyped as being overly emotional, Charlotte’s frequent use of interjections further characterizes her this way.



In this experiment, we quantified the linguistic aspects we found to be related to feminine stereotypes.

    1. We counted the number of times each character uses interjections.
    2. We counted the number of times each character repeated words or phrases to measure repetition.
    3. We counted the amount of words spoken per second by each character to measure rapid speech.
    4. We counted the average number of times each character used uptalk in a given situation.

We contrasted each variable’s occurrence in Charlotte’s speech with Tiana and Naveen’s, whose language patterns tended to remain relatively consistent throughout the dialogue.


As illustrated in the first graph, Charlotte uses significantly more repetition and interjections on average than both Tiana and Naveen. On average, Charlotte used 1.4 repeated phrases and 4.3 interjections per scene, while Tiana used 0.8 repeated phrases and interjections per scene, and Naveen used the least of both, with no repeated phrases and an average of 0.67 interjections per scene.


As illustrated in the second graph, Charlotte has a higher rate of words per second used in comparison to the rates depicted by Naveeen and Tiana. As Charlotte speaks at a rate of 2.20 words per second, while Tiana and Naveen speak at rates of 1.80 and 2.10 words per second. In other words, proving that there is a faster rate of speech in Charlotte’s speech patterns.


As illustrated in the third graph, nearly 60 percent of Charlotte’s lines were examples of uptalk. This percentage was almost double that of both Tiana and Naveen’s total percent of uptalk lines, which were both around 30 percent. This data proves that Charlotte uses uptalk much more than the other characters, likely as a result of her overly feminine portrayal.


Many baseless standards and stereotypes are placed onto women’s speech. We see a prime example in the character Charlotte, whose personality is drawn completely from the stereotypes of groups she is a member of. We see this especially in her frequent use of uptalk, repetition, rapid speech, and interjections compared to the other characters. For these reasons, it is conclusive that Charlotte’s portrayal is indicative of Disney’s corporate misogyny and exploitation of linguistic stereotypes.



Cutler, A., & Scott, D. R. (1990). Speaker sex and perceived apportionment of talk. University of Sussex.

Guy, G., Horvath, B., Vonwiller, J., Daisley, E., & Rogers, I. (1986). An Intonational Change in Progress in Australian English. Cambridge University Press, 15(1), 23–51.

Read more

Freshmen versus Transfer Students: Who’s More Sociable?

Zilana Aikebaer, Monica Campbell, Jenny Kim and Roselinda Kuoch

The study was performed in hopes to explore the difference in the function and purpose of the usage of phatic talk, especially in the aspect of socialization during interaction amongst two large new streams of populations at UCLA: Transfer and Freshmen students. The sample population was randomly selected at the study lounges where the two groups are most concentrated and interviewed students from each group with the same set of questions that allows enough flexibility for the students to express their feelings, opinions, or to interact with the interviewer as in day-to-day conversation. The interview was recorded to further perform statistical analysis on turn-taking, time taken with each question, and of other stylistic aspects of the interviewee’s talk. Results show significant differences and there can be many potential explanations and causations for the differences. The analytical results could reflect the difference in the sociability of the two groups and the likelihood for students of one group potentially use phatic talk as a tool to build their social networks while the other group demonstrates less tendency for such behavior. Although there is no clear evidence for such correlation, possible connections between phatic talk and sociability for the two groups are revealed in the results of our study.


It is said that college is the time where you will meet your lifelong friends. When you socialize in college, you open a door that allows you to develop relationships that can support you in academic and emotional ways. A very common type of talk used to begin the start of any relationship is a linguistic term called phatic talk. 

Phatic talk, or simply in layman’s term, small talk, is a form of communication that has a social function rather than an informative function. In other words, the phatic talk focuses on establishing a mood of sociability. Instead of centering the conversation on heavy and controversial topics, phatic talk considers lighter and more enjoyable topics such as the weather, sports, or common interests. 

The demographics in college students are categorized by what year in college they are in, first-year, second-year, and so on. This study particularly focuses on the usage of phatic talk between first-years and transfer students. Comparing the differences in communication between these two subject groups is since they come from different educational backgrounds in which transfers have 2 or more years of the college experience and first-year students don’t have any college experience at all. The differences in experience and thus differences in maturity can result in different levels of phatic talk used amongst the two groups.

Background Info

The identity category we will be focused on students at UCLA, first-year students and transfer students will be the focal point of the research. The purpose of this study is to show that short and light communicational exchanges that we may often think don’t serve a purpose, actually do carry social importance. By analyzing the language aspect studied in the phatic talk, or small talk, through first-year students and transfer students, we can look at how phatic communication can serve to maintain a social presence. This will be recorded specifically by comparing quantitative observations of first-year students and transfer students such as the number of turn-taking, pauses, speaking rate, and length of sentences and conversations measured through time.


In order to collect data from our two chosen demographics, we had two of our teammates look for participants in lounges of dormitories on the UCLA campus. The majority of transfer students were found in Rieber and the majority of first-year participants were from Dykstra. Ten subjects from each group for a total of twenty students altogether were interviewed. 

To gather samples, we voice recorded the small talk interviews with the consent of all those who participated. For each set of conducted interviews, five preselected questions were asked to establish more uniformity in our data. We aimed to choose questions that were more open-ended to give participants personal agency in how much or how little they spoke. 

Questions Asked

    1. What made you choose UCLA?
    2.  What’s your major?
    3. What do you plan to do with it?
    4. How’s the transition into college been for you?
    5. Can you give me directions to Cafe 1919?

Tip: How To Skip The Small Talk and Connect With Anyone

The two data collectors were careful in trying to keep a consistent demeanor across each case of interviewing in order to keep more consistency in the interpersonal task at hand. 

After collecting all the audio samples, each team member listened to and analyzed the samples for specific characteristics: these included the amount of words, filler words, turn-taking, and amount of time needed to answer the question. Each characteristic was recorded quantitatively by how frequently it showed up in each audio sample or by their respective amounts (like for word count or time needed to answer a question). The amounts for each speech category were recorded and averages were calculated for each question asked between the two demographic groups. We employed statistical tests to analyze our results and used a two-tailed T-test to test for a significant difference between the two groups. 


With the raw data collected, we calculated the averages for first years and transfers across each question within each category. 

From our statistical analysis, we found there to be significant differences in almost all the questions asked between the two groups across the various characteristics analyzed. Significant difference tests were done across each question for each linguistic trait. For the example of fillers (words such as uh, um, er, hm) the p-values that were calculated for each question were below the 5% significance level. This pointed to a significant difference in the number of fillers used by freshmen and transfer students. So overall, we found that the number of conversational behaviors was different across the two groups. 

As seen by our graphs across the different questions and categories, it appears that quite drastic differences exist between freshmen and transfer students. However, what appears to be a difference is not always enough to conclude on, therefore we use our two-tailed t-test to verify that the differences noted were significant.  

Numbers of Turn-Taking


Time Taken for Two groups to answer 5 questions


Average Response Count
Filler Words Count Between Two Groups


Pauses Count


Average Response Count


Discussion and Conclusions

Due to the statistical tests used, we as a team could only state that differences exist in the quantity of our chosen conversational characteristics between freshmen and transfer students here at UCLA. As the saying goes, correlation does not infer causation, we could not pinpoint reasons as to why these differences were found between the two student groups. However, we looked into possible hypotheses that could explain the results we found. Some possible explanations include differences in maturity/ difference in life stage, as the statistical results show a significant difference between the two groups when asked a task-oriented question. Which could be due to the older group has had more experiences with similar situations, and also can gather their thoughts to provide a concise answer. Based on the interviewers’ experience with the two groups of students, the difference in their maturity was revealed during the interaction. Most Freshmen were not clear on their career goals and what they ultimately will do with their majors, as career and job are not a major consideration for their age group. Transfers, on the other hand, were specific with their career goals and are in majors that align with their future goals as they will be going into working fields soon after they graduate. Other than maturity/difference in life stage, sociability could also contribute to the results of our study. College is a relatively new environment for Freshmen who just completed high school while Transfer students have had experience with the college for at least two years, which could lead to great curiosity and the desire for Freshmen to explore such a new environment and to meet new people. Also, the high school environment is very different than community college in the students’ sociability within the schools. Students at community colleges can range from the age of 18-45, with such an age difference, the students at community college are less likely to socialize at school and are adapted to the independent way of studying without having many affiliations with school or other students. On the contrary, high school is concentrated with students within the same age group thus makes it more feasible for students to build connections and thus socialize within the school setting. There are other possible reasons for their sociability and can vary on individual levels; however, such a large statistical difference between the two groups indicate there is an undeniable difference in the way the two groups communicate with others and perform verbal tasks regardless of the underlying contributing factors that can potentially account for the differences.     


Beukeboom, C. J., Tanis, M., & Vermeulen, I. E. (2013). The language of extraversion: Extraverted people talk more abstractly, introverts are more concrete. Journal of Language and Social Psychology, 32(2), 191-201. Retrieved from

Gravano, A. (2009). Turn-Taking and Affirmative Cue Words in Task-Oriented Dialogue. N.p.: Columbia University. Retrieved from

Hudak, P. L., & Maynard, D. W. (2013). An interactional approach to conceptualising small talk in medical interactions. N.p.: pubmed.Retrieved from

Valencia, D. (2009): ‘No Offense guys: Some ambiguous functions of small talk and politeness in workplace discourse’, in LCOM Papers 1, The University of Hong Kong, 17- 32. Retrieved from,%20rev/2009%20vol1/2_Diego_Valencia.pdf

Wardle, M., Cederbaum, K., & Wit, H. d. (2012). Quantifying talk: developing reliable measures of verbal productivity.  N.p.: pubmed. Retrieved from

Pappano, L. “Lost, Alone and Not a Freshman.” The New York Times, 23 Apr. 2006.,

Read more

Queer Speech: Real or Not?

Samantha Dao, Audrey Harrison, Sonia Hauser, Elizabeth Rutkevich

Have you ever thought, “Wow. That person sounds so gay.”? Maybe it’s because of the way the person speaks –his/her pitch is higher/lower than a straight person’s, the pitch at the end of his/her sentence is higher than the rest, or they have a melodious/creaky voice. But is there actually a difference between the way queer and straight people speak or is it just a stereotype? Is language used as an identifier of sexuality?

We were interested in these questions, but specifically if there’s a difference between queer and straight women’s speech. Therefore, we did an experiment, in which we asked 20 women, 10 straight and 10 queer, to tell us about a time in which they almost died and analyzed their speech to determine if a difference exists.

Our hypothesis was that there would be no significant differences in the phonetics, or in the way sound in speech is produced, except that queer women would have a bit more roughness or creakiness in some parts of their casual speech. We also believed that queer women would speak at a lower pitch than straight women. However, after getting the results, we found that our hypothesis wasn’t entirely correct. Can you guess what part of our hypothesis was proven wrong?

Language is so powerful and can let people know a lot about ourselves. Beyond what you literally say, the way you speak can cause people to form opinions about you, or what you are implying can be interpreted differently by different people. Stereotypes even in language are constantly shaping the way people see us (Waksler 2001), so how can we use language to align ourselves with our identities instead of projected ones other people place on us?

We perform gender, sexuality, and identity constantly, so how is language used to perform these things? We did a little experiment to see if there is a difference in the ways that queer and straight women cultivate their own identities through language. We concocted a study involving 20 UCLA undergraduate women from ages 18-24, half being straight and the other half being queer. We asked participants to tell us about a time they almost died. This could be funny and dramatized or serious, whatever they wanted, in hopes to get them to speak casually, how they would outside the space of an interview.

We compared how many times each group used vocal fry, the creakiness found in speech, typically in vowels (for example in this video clip from 0:00-0:31), and upspeak, the rising in voice at the end of sentences (as demonstrated in this video from 0:00-1:03), and rated overall pitch, the highness or lowness of voice, from 1-5 (you can learn about high and low pitch in this video). We were curious about pitch because some studies say it differs between the two groups (Barron-Lutzross 2015) while others say there are no significant differences (Waksler 2001). Vocal fry and upspeak are stereotypical of women’s speech as a whole.

Before doing the actual experiment, we thought queer and straight women would have similar usage of upspeak because they belong to the same gendered social category, so parts of their identity formations and expressions would be similar (Valocchi 2005). However, we thought queer women would use more vocal fry and have a deeper overall tone than straight women to differentiate themselves and construct distinct identities from straight women.

Background Information

There is a cultural awareness and focus on the way gay men speak. The stereotypically effeminate “gay voice” is characteristic of a higher pitch, slight lisp, and hyper-enunciation. The documentary Do I Sound Gay? questions where the “gay voice” comes from and some possible explanations of this phenomena (Thorpe 2014). The documentary concludes that gay men’s voices are a complex mix of the presence of female role models, group identification, and the persistent association with the aesthete stereotype (similar to a dandy, aesthetes are a Wildean, intellectual, effeminate man). Here is a clip from The New York Times of Thorpe, along with a linguist featured in his documentary, describing the specifically male “gay voice.”

What, then, does a lesbian or queer woman’s voice sound like? Are these categories distinct enough to draw phonetic conclusions? In what ways are gender and sexuality intertwined in the way we speak, complicating intra-gender distinctions? To answer philosophical questions about gender we looked to Judith Butler’s seminal Gender Trouble. Butler proposes the idea of “performativity” in relation to gender–all gender is a performance and reliant on external interaction. Others pick up gender cues from the ways we dress, speak, and move. So how do queer and straight women express their identities and gender differently? Or is the category of woman unifying enough to surpass sexuality?


In order to find out if there was a difference in speech between queer and straight women, 10 UCLA undergraduate women from ages 18-24 in each category were asked to tell about a time in which they almost died. This question was to make sure that they would speak their mind without worrying who they were talking to since the purpose of the experiment was to compare casual talk. Participants were friends and acquaintances and were asked to be recorded before the interview took place. The recordings were analyzed for upspeak, vocal fry, and pitch. Since each interview differed in the length of time, vocal fry and upspeak were analyzed by times used per minute. Pitch was determined on a scale of 1 to 5 with 1 being the lowest and 5 being the highest, and it was evaluated by the interviewer.


There was no significant difference in the average amount of upspeak used per minute between the two groups with 4.56 and 4.6 upspeak used per minute in queer and straight women’s speech respectively. (Fig. 1). However, as seen in figure 1, queer women used slightly more vocal fry (3.5 usage per minute) than straight women (2.82 usage per minute). There was no significant difference between the pitch of both groups as the pitch was measured to be 2.95 and 3.05 for queer and heterosexual women, respectively (Fig. 2).

Figure 1. Analysis of average vocal fry and upspeak used per minute shown in blue and red bars respectively. The data was analyzed from speech of 10 queer and 10 straight women.

Figure 2. The graph shows the average pitch from 10 queer and 10 straight women. Pitch was measured from a scale of 1 to 5, with 1 being the lowest and 5 being the highest.

Discussion and conclusions

The results of the study revealed a ubiquity of phonetic linguistic performance among women regardless of their self-identified sexualities. The lack of significant differences in pitch and upspeak occurrence, and the predicted, slight variation in vocal fry occurrence points to the similarities in the phonetic patterns of all the women aged 18-24, in our immediate UCLA-range community that participated in the study. This predictable finding about the range of women’s speech within immediate speech communities and social networks carries implications that gender categories and socialization are stronger determinants of speech than sexuality is.

There are possible errors associated with our methods. Participants could still have modified their speech since it was an interview. Friends were interviewed which could have affected the speech samples, but keep in mind that the experiment was comparing casual speak, so in fact this could have been beneficial. The sample size was relatively small and involved a specific group, so the results may not be representative to all queer and heterosexual women. The evaluation of pitch was subjective, and there was possible bias as prior research was done. Lastly, recordings might not be true to the actual voice because there’s error in the recording equipment. The errors were minimized as best as possible.

For future studies involving the role of sexuality in women’s linguistic performance, non-phonetic features, such as the frequency of tag questions and filler words, could be tested for differences between queer and heterosexual women. It could also be beneficial to perform a preliminary study evaluating whether there are pre-existing perceived differences between and stereotypes about the speech of queer versus straight women.


We would like to thank the volunteers that participated in the study and our teaching assistant Madeleine Booth and Professor Daria Bahtina for guiding us through our experiment. 


Barron-Lutzross, A. (2015). “The Production and Perception of a Lesbian Speech Style.” UC Berkeley PhonLab Annual Report, 11. Retrieved from

Butler, J. (2006). “Gender Trouble.” Routledge.

Howard, G. and Thorpe, D. (Producers) and Thorpe, D. (Director). (2014). “Do I Sound Gay?” United States: Sundance Selects.

Valocchi, S. (2005). “Not Yet Queer Enough: The Lessons of Queer Theory for the Sociology of Gender and Sexuality.” GENDER & SOCIETY, 19(6), 750-770, DOI: 10.1177/0891243205280294.

Waksler, R. (2001). “Pitch range and women’s sexual orientation.” Word, 52(1), 69-77, DOI: 10.1080/00437956.2001.11432508

Read more

How Do Different Genders Speak in The Office?

Chelsea Gleason, Priscilla De Luna, Kat Dang, Briana Tena

Do males and females speak differently in a professional setting? If so, does this cause any implications? In this study, we look at the different language patterns in the popular TV show, The Office, a comedy show following the lives of workers at a desk job.

The aim of our research was to see if differences in speech establish a power hierarchy between genders in the workplace. This research was motivated by the growing number of women in higher skilled professions compared to previous decades. Thus, we developed a coding system to study the frequency of rise in pitch and use of interruptions among the characters in this TV show. We then analyzed the data and found that the speech patterns did contribute to a power hierarchy, however it was represented through men establishing dominance over other men, rather than men establishing dominance over women.

In recent decades, more women are securing higher skilled positions in the workforce compared to previous time periods in which professional work was mostly male dominated (Ziman, 2013, p. 1). There is little to no research studying the new relationships between genders in this setting, thus we conducted a study focusing on one of the fundamental aspects of interaction: language.

We completed a linguistic analysis of the American version of The Office, meaning we studied the differences in language use between individuals. The Office served as an excellent source for our study as the show has over 57 million viewers (Stern, 2018, col. 4), so the show should represent what Americans view as common speech.

To complete this study, we took a sociolinguistic approach, which relates differences in language to social factors, most notably in our case: gender. We collected our data by analyzing phonetics, which is the study of speech sounds, and via conversational analysis, which is a broad linguistic field that studies natural conversations in everyday situations. To make our data more quantifiable, we defined phonetics as the use of upspeak, which is the rising pitch at the end of a statement, or more commonly known as the “Valley Girl Accent”, and defined conversational analysis as any interruptions made between characters.

We chose upspeak and interruption because social scientists have previously found that these factors aid in creating differences in power among genders and ultimately form a gender hierarchy (Linneman, 2013, p. 83), (Schegloff, 2001, p. 289). However, there is no research indicating if this is true in a professional setting where individuals are expected to act and speak in a more polite manner.

This led us to hypothesize that the use of upspeak and interruptions in the workplace places males in a superior power position compared to females.

First, we had to research if there was any data that indicated gender inequality in the workplace, and what steps, if any, are being made to address the issue.

As of 2019, women get paid 80 cents on the dollar, compared to men, and only 79 women are promoted to a higher position compared to every 100 men (Schooley, 2019). Women in the workplace experience microaggressions verbally and behaviorally during the communications at work. These actions are both intentional and unintentional, ranging from, but are not limited to “hostile, derogatory, or negative prejudicial slights and insults.” (Schooley, 2019). 

This idea is illustrated in Nidhi Dua’s Tedx Talk, “Gender Equality at Workplace.” Dua discusses the problem of obtaining gender equality, where the problem stems from, and possible solutions. Gender equality is a series of complex issues with no single, direct cause. It is acknowledged that no single person or group can solve the problem, but they can each do their part.

In order to start positive changes in the workplace, Dua believes that “engaging with men and women workers, spreading awareness, and addressing issues that act as impediments to gender equality” were important factors. The training at the factorial level consisted of : peer to peer learning, audio-visual presentations, focus-group discussions, role plays that highlight gender stereotypes, and developing grievance handling mechanisms. Dua highlighted the development of grievance handling mechanisms because women (especially) are put through tremendous amounts of stress and pressure both in and out of the workplace. Women also tend to feel uncomfortable with speaking up and reporting their incidents and concerns.

We applied this idea and studied the grievance handling mechanisms through upspeak. “Upspeak” (or “uptalk”) refers to one’s pitch rising towards the end of a statement. This speech pattern is often compared between men and women. Upspeak’s existence can be dated back to as early as the 17th century (Gorman, 1993) and its usage is common among women. This linguistic phenomenon has the connotation of  being a “Valley Girl” accent, although it has become more common in both men and women since its “discovery” (Rutter, 2013). 

Despite being practiced by both men and women, what upspeak determines for each gender in society varies greatly. In a study that was conducted on the show Jeopardy! By Virginia Rutter, PhD., she found that men use upspeak when they are uncertain or unconfident with their responses to a statement or question. It was also discovered that men would use upspeak more with women when they are correcting a woman compared to a fellow man. Her study indicated that men used upspeak 22% of the time when correcting another man, but was used 53% of the time with women. It is speculated that men do this as a form of chivalry towards women.

We wanted to see if these findings correlated with our study. So, we began analysis on the TV show, The Office. The Office, contains 9 seasons with a total of 201 episodes. Since there was not much time to gather all of the data, only 20 episodes were chosen to be analyzed. This way, each group member was responsible for 5 episodes. A random number generator, set to 1-201, was used to select which episodes were used. The episodes were randomly selected to make sure that the data collected was a good representation of the entire show. To analyze each episode, a system of coding was created that focused on how each gender used uptalk and interruptions (Reed, pgs. 10-11). The code used is shown below:

When a character used upspeak or interrupted someone, the exact words they used in the conversation were recorded. To show when upspeak was used the words were either highlighted yellow or made bold and a dash line was used to signify an interruption. For example, the scene from the image below is from the episode “Frame Toby” in season 5, episode 9 of The Office. In this scene, Michael found out that Toby came back to work at Dunder Mifflin. When Michael saw Toby he immediately interrupted Toby by yelling, “No”. This interruption was classified using the number 10 from the code, since Michael interrupted a male colleague, and recorded as shown below. There is also an example below about how upspeak would be recorded. In the second example, 2 and 6 coded the scene because Michael used upspeak when talking to Dwight.

The Office- Season 5, Episode 9 “Frame Toby”

(19:52-19:48), 10

Toby: “Hi Michael-”

Michael: “Noooo god! Nooo god! Please no. Nooo.”

(13:04-12:57), 2 & 6

Michael: “You’re the bait for Toby?”

Dwight: “Mmmhm.”

Michael: “Uhh for one thing he’s not gay and if somebody were to be bait, it would be Jim or Ryan or me.”

From the data, we were able to construct two different graphs to display the visual variations found when deciphering the different phonetic codes used within The Office. When analyzing coded upspeak, we found that the ratio of upspeak used in a professional setting greatly differs between men and women.

Data shows that men tend to use upspeak at least twice the given amount a woman does within the workplace. Upspeak towards the opposite sex differed as well.  Women were more likely to use upspeak towards men in ratio to the amount of times men used upspeak towards women. This is an effect of  lack of confidence within the workplace, thus women felt the need to critique men less often to avoid coming across as too strong. Men however, were likely to use upspeak towards women as a form of peer correction or chivalry, not as a form of dominance, hence they used upspeak towards the opposite sex less often.

Data also implies that males are more likely to use upspeak towards men as opposed to the usage of it towards women. The stark difference has led us to believe that in a work setting, men often feel the need to assert their dominance and establish a visible hierarchy amongst their male coworkers than towards women.

In coded interruptions, there was a small difference between data when analyzing codes 7 and 8. Females were more likely to be interrupted by males than males were to be interrupted by females. However, in codes 9 and 10, our data displayed a stark contrast of a <1: 5. Essentially, this established the notion that males are more likely to interrupt a male coworker in the workplace than females were to interrupt a female coworker.      

In the original hypothesis, we expected men to use more upspeak and interruptions towards women than towards men. We hypothesized that this would occur in the form of chivalry or as a form of peer correction. Data results, however, differed greatly from our initial theory.

After analyzing the codes through a random draw of various scenes, we can infer that men used more upspeak than women within the workplace altogether. Males, however, were less likely to use upspeak towards the opposite sex than women. This correlates to the notion that women are more likely to use upspeak when speaking to men because they felt the need to be apologetic of their success.

Further analysis concluded that males were more likely to interrupt another male counterpart than a female counterpart. With this data we were able to see a clear sense of hierarchy and several dominant attributes within men in the workplace. As a result, men often establish more dominance over male counterparts through peer correction and interruptions than towards female coworkers.

These findings coincide with the gender distribution of characters within The Office. A larger male population within this workplace can be a correlative factor that encourages males to assert dominance over a competitive atmosphere with other men. Women, however, were less likely to interrupt the same sex or use more upspeak than men since The Office is not female dominated, therefore they were not encouraged to assert their dominance like males in this setting.


Gorman, J. (1993, August 15). ON LANGUAGE; Like, Uptalk? The New York Times Magazine. Retrieved from

Rutter , V. (2013, December 28). Men and Women Use Uptalk Differently: A Study of Jeopardy! – Sociological Images. Retrieved from

Schegloff, E. A. (2001). Accounts of conduct in interaction: Interruption, overlap, and turn-taking. In J. H. Turner (Ed.), Handbook of Sociological Theory (pp. 287–321).

Schooley, S. (2019, May 20). The Workplace Gender Gap and How We Can Close It. Retrieved from

Stern, M. (2018, December 15). Is ‘The Office’ the most popular show on netflix? The Daily Beast. Retrieved from

YouTube. (2019, February 26). Gender Equality at Workplace | Nidhi Dua | TEDxGurugramWomen. Retrieved from

Ziman, Rebecca L., “Women in the Workforce: An In-Depth Analysis of Gender Roles and Compensation Inequity in the Modern Workplace” (2013). Honors Theses and Capstones. 157.

Read more

Media Depictions of African Americans in Incidents of White-on-Black Violence

Faith Ngo, Madyllen Kung, Melissa Aguirre, Sabrina Huang

Racial inequalities have been a fundamental aspect of the underlying fabric of the United States since its conception almost 250 years ago. From brutal incidents of racialized violence to educational disparities that have continually oppressed communities of color, inequities rooted in the throngs of racism have persisted and accumulated over time. An example of such racial inequities is violent incidents in which white police officers shoot and kill unarmed African American individuals. Proof that discriminatory biases still exist today, these events have become fuel for groundbreaking social movements that are centered on uplifting the voices of oppressed communities and challenging hegemonic ideologies. 

Over the last ten weeks, we have learned about the vital role language plays in constructing and maintaining identity. Through stereotypes and “otherizing,” which have amplified the perceived differences between social groups and intensified the already vast racial boundaries, language can codify and perpetuate discriminatory biases.

As we started our project, we asked ourselves, would articles dehumanize African Americans or would they place blame on the white police officer? Would race be a salient aspect? Would there be a notable difference in the styles of language across different social identities? Or would we find a difference between various news outlets?


Language is a powerful tool that can be used to construct our understanding of the world or perpetuate traditional beliefs. As tensions between African Americans and White Americans continue to grow, it is important to recognize the ways in which language can reaffirm discriminatory biases.

With this in mind, we decided to focus specifically on the linguistic elements news articles utilize to cover incidents of white-on-black violence. In addition to being widely accessible to the public, such articles play an important role in either reaffirming or challenging prejudicial stereotypes.

This led us to our research question: In situations of white-on-black violence, how do different news outlets utilize linguistic elements to depict and characterize African American individuals as archetypes of widely-held stereotypes?


Research has shown that the media routinely associates African Americans with criminality. News outlets disproportionately report on criminal incidents which involve African American suspects in comparison to white suspects, especially if the incident involves violence (Oliver, 2013). The overrepresentation of African Americans in incidents of crime creates a stark dichotomy between the portrayal of black and white Americans by news outlets. While violent crimes perpetrated by African Americans are widely reported on, violent incidents involving white suspects are largely ignored (Johnson and Dixon, 2008). In addition to the racial disparities apparent in coverage of crime, studies have shown that specific language is used to dehumanize African Americans regardless of their role as the perpetrator or suspect. The use of “micro-insults” in descriptions of African Americans can implicitly link them with social categories that are historically viewed as “inferior” to normative social groups (Smiley and Fakunle, 2015). On the other hand, the use of “micro-invalidations” can trivialize the experiences of black individuals (Smiley and Fakunle, 2015). Such language can have alarming effects in priming audience members towards internalizing negative stereotypes of black individuals and potentially acting upon those implicit biases (Oliver, 2013). The actions which result from biases produced and maintained by mainstream news outlets — “racial microaggressions” — form the foundation of discriminatory structures that continually relegate African Americans to the bottom of the social hierarchy (Kulaszewicz, 2015).


Since the manner in which white-on-black violence is depicted largely depends on the political affiliation of the reporting news source, our methodology was designed to account for a range of varying political viewpoints. Each researcher selected two liberal sources, two conservative sources, and one moderate source based on the AllSides Top Online News Media Bias Ratings chart (Figure 1). These five articles of varying political affiliations were used to analyze the following individuals: Tamir Rice, Tanisha Anderson, Trayvon Martin, and Eric Gardner. We selected these individuals because they are figures who are representative of White-on-Black incidents of violence in the United States.

Figure 1: AllSides Top Online News Media Bias Ratings We used this chart to find articles that aligned with a political orientation. One article was selected from each column from the news outlets listed in this chart. Source:

After the articles were chosen, a Total Point System was employed to evaluate the salience of race in the article through explicit references of race. The point system was designed to ask one question: to what extent did the article make race a conspicuous and contributing factor in the white-on-black incident? The following are the six criteria of the system: 

It is important to note that an article that scored six out of six points does not necessarily indicate that it is more racially biased.

Since language also has an implicit function, the second part of our analysis involved a Guiding Question system to account for indirect references to race that could not be captured by the point system. The following are the seven criteria used in this system: 

The assessment of articles using the Guiding Question system provides insight into the discreet manner victims and assailants are framed. These questions illustrate how the language of articles can indirectly position individuals as a particular actor in larger racial narratives (this is often referred to as “interpellation”). 


Figure 2: Point totals between various news outlets across the four individuals This bar graph depicts the varying point totals of articles which cover the deaths of Tamir Rice, Trayvon Martin, Tanisha Anderson, and Eric Garner. The different colors depict the political alignment of the articles selected.

As Figure 2 shows, point totals varied across articles which covered the deaths of our sample of individuals. We found articles on Tanisha Anderson and Trayvon Martin to have disproportionately high point totals because of their existence at the intersection of multiple oppressed identities. While Anderson had a mental disability that was commonly referred to, Martin’s appearance at the time of his death was a salient component of several articles. 

Foregrounding in the lede

In our quantitative analysis of selected news articles, we focused on the journalist’s word choice throughout the article and how such words evoke a reaction from readers (Jakobson’s “conative function” of language). However, special attention was paid to the first sentence of the article, which is often referred to as the “lede”. This sentence encapsulates the who, what, where, why, and when of the situation or topic in question and helps to set the tone for the remainder of the article. Due to the position of the lede at the beginning of the article, information included here can be utilized to foreground certain elements. 

Figure 3: An example of foregrounding in the lede The ledes included in this figure are from two articles which covered the murder of Tamir Rice. Bolded and underlined words highlight the noticeable differences between each.

Across the four individuals we studied, we observed that there were noticeable differences in language use. Liberal news outlets typically included language which positioned the African American individual as the “victim”, while conservative news outlets utilized language which portrayed the police officer(s) and their actions as reasonable. Information that supported each agenda was included in the lede, while information that undermined such portrayals was either excluded or backgrounded. 

Differences in point totals between conservative and liberal news articles

Our research also found differences in average point totals between conservative and liberal news articles (Figure 4). Although these differences were not large, it appears that conservative articles have lower point totals and conservative articles have higher point totals. Meanwhile, moderate sources had point totals that fell between the scores of liberal and conservative news outlets. However, these average point totals fell closer to those of conservative sources. 

Figure 4: Average point totals of liberal, moderate, and conservative news outlets.

Differences in guiding questions

We also found stark differences in answers obtained by way of our guiding questions. Across the twenty articles we studied, conservative articles utilized a greater degree of language that reinforced common stereotypes associated with African Americans. Conservative sources often drew upon the individual’s criminal history or appearance (e.g. wearing a hoodie) as a subtle way of shaping the individual’s character. If the officer was mentioned, it was to justify his/her actions in some way. Conversely, liberal sources often focused on the motivations of the officer and their judgment errors in interpreting the situation. The results demonstrate that in general race was made more salient in conservative sources that liberal sources as an inherent contributing factor to the situation.


Our results reveal that there is a stark and noticeable difference in the language use between conservative and liberal news outlets. While liberal publications use language to position African Americans as “victims”, conservative publications position them as the “assailant” or “instigator”. Such information draws attention to the detrimental role journalism plays in furthering and reinforcing stereotypes that support the criminality of African Americans. 

There were several limitations to our research, however. One noticeable limitation is the assumption that the political alignment of a news outlet directly corresponds with its article’s attitude towards race. There are numerous factors that may play a role in the way national publications depict race that we unfortunately did not account for in this project. Other limitations include differences in publication times (which impact how much information is available to journalists and, by extension, what particular elements are salient) and personal biases.

Regardless, our results offer insight into the way African Americans are both implicitly and explicitly discriminated against. Such biases underscore the role language can play in shaping public perceptions and encouraging prejudicial actions. It also reveals ways in which we can uproot social stereotypes surrounding historically marginalized groups and tackle harmful racial disparities. 



Johnson, K. A., & Dixon, T. L. (2008). Change and the illusion of change: Evolving portrayals of crime news and blacks in a major market. The Howard Journal of Communications, 19(2), 125-143. doi:

Kulaszewicz, Kassia E.. (2015). Racism and the Media: A Textual Analysis. Retrieved from Sophia, the St. Catherine University repository website:

Oliver, M. (2003). African American Men as “Criminal and Dangerous”: Implications of Media Portrayals of Crime on the “Criminalization” of African American Men. Journal of African American Studies, 7(2), 3-18. Retrieved from

Smiley, C.J. & Fakunle, D. (2016) From “brute” to “thug”: The demonization and criminalization of unarmed Black male victims in America, Journal of Human Behavior in the Social Environment, 26:3-4, 350-366, DOI: 10.1080/10911359.2015.1129256


Read more

I Am Who You Are Not: Insults in Films

Anthony Waller, Avery Robinson, Nicole Rasmussen, Jun Jie Li

Creativity and complexity are not often two factors that are considered when we insult; we typically go to our personal shelf of offensive phrases and let our selections do their damage. When we look at high school oriented films, however, we see that insults are a means of identity negotiation and employ creative and complex techniques that serve to compound the effect and project a strategic process of identity projection and negotiation. In this article, we will be examining how films act as a social mirror by reflecting a description of contemporary teenage culture. Specifically, we will be considering two factors that we believe to have had a significant impact on the motivation of portrayals: gender and time. Looking at several classic selections that spans the decades of the 80’s through the 00’s, we utilized a nexus and inductive approach in isolating specific linguistic elements of insults that appear most salient to our research. We conducted a series of comparative analyses of creativity and complexity parameters and extrapolated a loose correlation between gendered insults and the passage of time. From there, we will be discussing some implications of this correlation and how insulting is a process of identity prioritization and constructivism through self-isolation.

Introduction: How We Offend

For our research, we will be investigating the depiction of insults in teenagers as portrayed in films of high school settings from the 1980s, 1990s, and 2000s. Specifically, we want to take a look at how the complexity and creativity, as defined below, of insult formation are expressed across gender boundaries, and how the mechanism of that formation has evolved over the decades. Based on our preliminary observations, we are expecting to see depictions of greater structural complexity and communicative creativity in females characters over males. We also believe that there will be an inverse proportional relationship between the integration of elaboration in insult formation and the time period.


Film can often provide valuable insights into how an era sees itself (Kalinak, 2010). Its choices shed light on realities and stereotypes, and insults and derogatory language natural entry points for analysis. Insults and derogatory language have two important, interdependent functions: the attack and distancing of the other and the defense and reassertion of the self. Teenagers are at a critical stage of self-discovery, and these functions offer insight into their views of self (Goffman, 1971). Choices in insult delivery will show the teenager’s prioritization in their identity expression, therefore by analyzing teenagers’ conspicuous insult expression, we can learn a great deal of what adults think of their successors.

The basis of our first hypothesis rests on Lakoff’s features of women’s speech. According to Lakoff, women are expected to use super-polite forms e.g. indirect language or euphemisms, and avoid swear words (Mooney & Evans, 2015). Therefore, if women want to insult someone, they would need to be more creative in order to get their point across while still adhering to the conventions of what is acceptable for women to say.

Our reasoning for predicting a general decline in complexity and creativity as we get closer to the current time is due to the improvement in technology and the emergence of “text speak,” “meme culture,” and the general notion that teenage speech has become more coded and somewhat less markedly intelligent (Brinkley, 2013, Dijk, 2016). Teenagers have found ways to say more with a lot less and to make a greater use of the referential creativity (see below).

Methods: Nexus and Induction

For our research, we will be utilizing a nexus and inductive approach; we will be drawing conclusions based on data and observations that we make in teenage films. Below we have six films, two from each of the three decades of our research parameter, that we believe will be illustrative of the teenage perception:

1980’s: The Breakfast Club, Ferris Bueller’s Day Off

1990’s: Clueless, 10 Things I Hate About You

2000’s: Mean Girls, The Princess Diaries

As we watch these films, we will be observing and taking notes of specific instances of derogatory language use by teenagers, as well as creativity and complexity levels. One way we have found to quantify these measures is to check if an insult actually contains an insult or a derogatory word, or whether it contains a series of words, reliant on references and word plays, constructed to make an insult. Additionally, we will analyze word choices in terms of commonality of use, with the thought in mind that less common words constitute a more creative insult. Once we have our data on creative versus non-creative insults, we will be able to form a ratio. We will compare the ratio between men and women in the movies, between the different decades, and between men and women differences in different decades. Further methods of analysis will include measurements of length, as well as comparisons of the types of references made across our parameters. As we have mentioned previously, we predict that insult use is more creative among women, and that insult use has become less creative since the 1980s.

Definitions / Parameters

Complexity: a function of length, diction, syntax.

Length: number of words in an insult or an insult group

Diction: word choice (common/uncommon)

Syntax: construction; whether the insult is formed in a non-declarative, complex way

Creativity: a measure of tone, reference, and blatant insult word choice

Tone: insults delivered through the use of tone or body language

Reference: use of references that are contextually significant in making the meaning of an insult apparent. This can fall into two main categories:

Cultural: An appeal to cultural, epistemic domains, such as arts and history, that are predominantly apparent to the individuals.

Social: An appeal to social norms, an attempted outing of the individual from the social hierarchy from an identity perspective.

Presence of blatant insult word: whether one insults with a pre-established jab or creates the pointedness themselves.

Results: Correlations

Our data from the 1980s is from Ferris Bueller’s Day Off and The Breakfast Club. The combined data from the two films tells us that the average word length per female insult is 5.71, and for males is 12.54. 46% of the insults were syntactically significant, and only 5.6% of those which were syntactically significant were from females. Social and cultural references were 19% and 14%, respectively, with females contributing 0% to both categories. 14% of the insults included uncommon and notable lexicon, but again with 0% contribution from females. In 5% of the data we saw insults delivered through tone, all attributed to male insults.

Our data from the 1990s films Clueless and 10 Things I Hate About You were 64% female. The average word length for a female-given insult is 8.84 words, and for males it was 8.54 words. 46% of the insults given were syntactically significant, and 78% of these are attributed to females. Only 15% of insulted included uncommon word choice, and about 44% of these were given by women. In regards to references, females made up around 80% of all cultural referenced insults, 50% of socially referenced insults. 27% of the insults were delivered through the use of tone, and 88% of those were female-delivered.

Our data from the 2000s derives from Mean Girls and Princess Diaries. From these films, 94% of the insults were from females. The average female word length was 6.45, and the average male word length is 9. 46% of the total insults were syntactically significant, and 93% of those were from women. In regards to references, 6% of the insults included cultural references and 13% included social references; all of these are attributed to females. 21% of the insults contained uncommon word choice, 93% from females. In regards to insults delivered through tone, 15% of the total insults employed this method and 80% is due to females. Finally, 66% of the insults contained an actual insult word, with 95% of that being from females.

Figure 1: Direct Insults and Derogatory Words Over Time.

Our data from the 80s show us that males employed much more complex and creative insults than females at the time. Going into the 90s, the trend shifts, and the majority of our data point to women being a bit more creative and complex in their insult use than their male counterparts. Finally, in the 2000s, we see a drastic change in our results with women demonstrating much higher levels of insult creativity and complexity than men. We were off from our original predictions.  We see from our data that insults, among females, increased in complexity and creativity. Additionally, we do not see a decrease in general creativity as we moved through the decades.

Figure 2: Breakdown of Derogatory Techniques Across Decades – Syntax, Diction, Structure, Cultural and Social References.

Discussion and conclusions: Why We Offend

First, from a pragmatic perspective, why do teenagers feel the need to beat around the bush in insults? At first glance it seems rather counterintuitive, but as we have seen, they serve important linguistic functions. For one, creative and complex insults can inflict a greater amount of damage by constructing a vehicle in which the insult can be delivered in more deceptive and cognitively disorienting way. It can also be viewed as a “flex” of intellectual superiority, or as a way to make the insult less refutable, as a retort would necessitate an equal level of craftsmanship (Goffman, 1971).

But why do we insult? What do we have to gain in insulting others? From our observations, it appears to be a practice of identity projection, of a more aggressive degree, because it is forceful definition of the self via an equally forceful definition of the other. In other words, along the same line of “who am I if not myself?,” it appears that the teenage response is merely “I am not you.” This seems to suggest that identity is only salient, or more radically, only existent, through expression and a process of negotiation and prioritization with the other. Insults serve as a way to categorize and define oneself against others (Marsden, 2009).



Brinkley, A., & McGraw-Hill Education (Firm). (2013). American history : Connecting with the past (Twelve edition, Updated. Updated AP ed.). New York, NY: McGraw-Hill Education.

Dijk, C. V. (2016). The Influence of Texting Language on Grammar and Executive Functions in Primary School Children. Retrieved from

Goffman, E. (1971). Relations in public: Microstudies of the public order. New York: Basic Books.

Kalinak, K. M. (2010). Film music a very short introduction. Oxford: Oxford University Press, USA.

Marsden, E. (2009). What the Fuck? An Analysis of Swearing in Casual Conversation. Retrieved from

Mooney & Evans (2015) Language and Gender. In Language, Society and Power (pp. 108-131). London: Routledge.

Read more

“Language 1, Language 2, and The Ol’ Switch-A-Roo” Mix & Match: Bilingual Edition

Language preferences and code-mixing among UCLA bilinguals in different social settings

Shiqi Liang, Leen Aljefri, Yingxue Du and Tianyi Shao

Here at UCLA, we have a diverse student body coming from many different backgrounds, which means we do have a sizable bilingual population on campus. Bilinguals and multilinguals often find themselves navigating through different social settings that require them to speak different languages. As bilingual speakers, switching between languages is quite common for us that it almost becomes a daily routine. However, when we really carefully think about that daily routine, there are so many questions we want to ask. Do we have a preference of one language compares to the other? Do our preferences vary? How do they vary? Do we mix languages? If so, how and why do we mix languages? Do bilinguals here at UCLA have a specific language preference when it comes to discussing fluid dynamics or gossiping about the latest juicy drama? Based on our study of 47 questionnaire responses collected from UCLA bilinguals and multilinguals, we arrive at the conclusion that among them, English is predominantly preferred in academic and professional related settings as well as social settings. At the same time, non-English languages are preferred in family settings and are present in social settings as well. We observed that code-mixing, the practice of mixing different languages together, is generally avoided, except when it is used as a tool for clarification.

Bilinguals experience potential conflicts between the two cultures behind the languages they speak. Language preference among bilinguals is related to the process of acculturation and socialization. Previous studies have identified the relationship between language preference and socialization, and literature addressing the relation can provide us with insights into the subject of interest. Song (2017) addresses the relationship between second language acquisition and socialization in “Second Language Learning as Mode-Switching” through the following idea: if social relations/context changes, then people employ a different linguistic and pragmatic mode to adapt to the new social expectation. Song adds that learning a second language requires the understanding of different speaking norms, linguistic values, and the rules of grammar. Language preference among bilinguals, therefore, can indicate the preference of one social norm to another to some extent. On the other hand, the fact that language preference among bilinguals is related to socialization is further addressed through a study conducted on infants and 9-month-old children (Valji & Poka, 2014), in which the infants show no preference for one language over the other and the 9-month-old children show preference in their native languages over the non-native language. As the social situations get more complicated when the bilinguals enter adulthood, the factors that might influence them to choose one language over the other are increasingly complicated and it is reasonable to articulate a relationship between social circumstances of a specific conversation and the language preference in that specific setting.

Observing the way multilinguals communicate with individuals in predominantly monolingual community is different than observing multilinguals in their own communities. Social and linguistic characteristics of multilinguals can be more noticeable when directly contrasted to monolinguals in the same community. As a first step to understanding what it means to be multilingual in a monolingual community, it is useful to look at a small bilingual population in such a community. The main focus of this project is to study the change in language preference according to situations and the frequency of code-mixing (practice of mixing different languages in one interaction) in bilinguals. In an effort to determine if a trend exists among the bilingual population here at UCLA when it comes to linguistic behavior, we conducted a case study and surveyed a group of 47 bilinguals at UCLA.

To better illustrate exchanges and preference in the use of language among bilinguals, here is an exchange between two English-Mandarin bilingual speakers talking about. The excerpts from this conversation is to present an example of how bilingual speakers interact with each other and how code-mixing happened during the conversation. The two bilingual speakers have conversations in their native language, and case study is to record and transcribe their conversation, and analyze the part that code mix happened. Throughout the whole conversation, code-mixing happened four times when participant B’s spoke. The four code-mixing can roughly be divided into two categories based on their cause, for clarification purposes and habit of word using.

A: 不是,我是说现在就你一个人在这个...... 空间啊?

   No, I mean right now are you just alone in that...... space?

B: 现在?Right now?

   Right now? Right now?

This is where code-mixing first happened during the conversation, and the purpose of it is to clarify the meaning of the word “现(xian)在(zai)”, which means present time. However, the meaning is not clear enough, because that word can represent different length of present time, and that can make the whole sentence a different meaning. Here, the phrase “right now” appeared as a clarification, which is similar to the purpose of the next exchange.

A:我听说过,但我不清楚是治愈(Zhi Yu)的还是致郁(Zhi Yu)的?

  I’ve heard about that, but I am not quite sure if it’s a healing story or a gloomy story.

Bhealing的那种,...... 结果两个人无意间卷进了road trip, 然后慢慢变好。

   It’s the healing type, .... The two people happened to be on a road trip, and things are getting better.

In this exchange, individual B needs to use another language to clarify her sentence since the Mandarin for “healing” and “gloomy” has the same pronunciation, “zhi yu”. In the second case she chose to say “road trip” in English mostly because she want to evoke a special cultural reference not widely available in Chinese culture.

The main methodology of this research is centered around analyzing data gathered through an online questionnaire designed to generate simple yet precise responses from participants. Before answering the survey, participants would read a text that ask them to evaluate themselves and only proceed to answer the questions if they match all the requirements of what we consider to be bilingual/multilingual. There are 10 mandatory questions and 6 additional questions if the participant speaks more than 2 languages. Participants would first self-report the languages they speak (free response) then choose scenarios in which they would prefer to speak a certain language and the reasons behind that. In order to avoid half-completed questionnaires and encourage complete responses, questions that involves picking scenarios and reasons would be in forms of multiple choice instead of free response. However, if none of the options provided are satisfying, participants are free to enter their own response through the “other” option. The questionnaire itself was distributed among the researchers’ group of bilingual friends and an incentive (free boba) was provided to further encourage participation. You can find the full questionnaire here.

In the end 47 responses were gathered and subsequently analyzed. You can find our raw data and analysis here. Out of all those responses, all 47 of them indicated English as a language they speak, with the Chinese language family (Mandarin, Cantonese and Taiwanese) ranking the second most self-reported spoken language with 34 responses. But yet surprisingly, only 23.4% of participants consider English as their first language.

Chart 1.1 and 1.2: self-identified “first language” and “second language”

Chart 2: total counts of languages participants reported speaking

Chart 3: self-report race and ethnicity among participants. “/” means decline to answer.

Unfortunately, as the sample size is relatively small and might not be an accurate representation of the entire student population at UCLA, the sample selection might be biased and the conclusions derived from the questionnaires might not be a representation of the entire multilingual student population. English, Mandarin and Arabic were selected because they have the most speakers and thus could relatively better represent themselves.

Chart 4.1, 4.2 and 4.3: language preferences in different scenarios. Red-schemed bars represent social/emotional/casual settings and blue-schemed bars represent academic/professional settings.

We could see that English is predominantly used in academic and professional setting (discussing academic work, discussing homework questions with friends, etc) and often used in social settings (talking to friends), yet less often used in family settings (talking to parents). Chinese and Arabic are less often used in academic and professional settings, but more prominent in social and family settings. This is rather predictable since UCLA is mostly a monolingual community and using a non-English language to discuss academic work is regarded as a social taboo. The prominence of non-English languages in family settings could be best explained by language preference in immigrant households in general. Children would mostly speak their parents’ native language in their own household due to new immigrants’ limited English proficiency.

Surprisingly, a lot of students also choose to discuss emotional issues in English. We predicted that since English is often associated with professional and academic settings, students might prefer a language that isn’t heavily associated with cold and rigid setting to discuss emotional issues. Our best explanation for this observed pattern is that some non-English languages, such as Mandarin and Arabic, are often associated with a more reserved culture. Thus, students may feel more comfortable speaking in English.

In terms of code-switching, most students answered “depends”. Only a few answered “almost in every sentence”. Data suggest that most students don’t prefer not to mix languages too often in their daily conversations. As for reasons for mixing language, almost everyone answered “in order to avoid misunderstanding or meanings lost in translation” or some variety of the same reason.

Chart 5.1 and 5.2: code-mixing among UCLA individuals. The first chart talks about the frequency of code-switching and the second one deals with reasons behind code-switching or lack of code-switching.

The complete reasons found in Chart 5.2 are listed below, from lowest to highest frequency:

    • I don’t know how to say a word in Spanish
    • I don’t know how to say a word in Chinese so I switch to English
    • I don’t know how to say a word in Chinese so I switch to English
    • I think it’s very hip and cool to do so.
    • I don’t
    • I want to highlight a part of my identity
    • Some words lose their meanings when translated to another language, so to avoid misunderstanding I would mix the language together

As predicted, the majority of the responses confirmed the hypothesis that English would be the favored language in academic settings. Conversely, the other language by majority is likely to be spoken in more personal conversations such as speaking to family and friends.

While, UCLA is home to a large multilingual community, the general language of instruction is English. In a way, the level of English knowledge is controlled by the admission requirements. Consequently, that may play a role in justifying the preference for speaking English in academic settings. It is the university’s expectation of its affiliates, and so it is upheld by the student population regardless of multilingualism within the community itself.



Song, S. (2017). Second Language Learning as Mode-Switching. Second Language Acquisition as a Mode-Switching Process, 75–100. doi: 10.1057/978-1-137-52436-2_5

Valji, A., & Polka, L. (2004). Language preference in monolingual and bilingual infants. The Journal of the Acoustical Society of America, 115(5), 2505–2505. doi: 10.1121/1.4783066


About the Authors

Shiqi (Susan) is a second-year statistics major at UCLA. She enjoys studying human geography and drawing in her free time.

Leen is a first-year engineering student from Saudi Arabia.

Christie is a senior majoring in theatre and has working experience of teaching bilingual children before. She enjoys listening to music and observing sunset glow and sunrise glow.

Tianyi is a senior majoring in Mathematics/economics at UCLA. She enjoys video games, music, and photography and she loves making observations about her life.

Read more

The Language of Good and Evil in the Disney Universe

Wendy Barenque, Maria Martignano Cassol, Kelli Sakaguchi, Sophia Siqueiros, Ellis Song

Every year Disney and Pixar release blockbuster hits watched by millions of children. Disney and Pixar characters have a huge impact on how children learn to view people in real life through the use of regional and foreign accents categorizing intrinsic “goodness” or “badness” (Lippi-Green, 2012). Recently, there has been a rising trend in the usage of “switch characters” in the Disney and Pixar cinematic universe. “Switch characters” are characters who are able to fake membership in the “good” character category and later reveal to not belong to this category. In this research, accent along with other linguistic variables such as pitch and creaky voice were tracked to determine if correlations exist between these linguistic variables and “switch characters” portrayals of “goodness” and “badness.” Does a “switch character” use a linguistic variable differently when portraying themselves as good rather than bad? For example, if linguistics changes do occur, do audiences begin to associate a certain pitch, accent, or creaky voice with “good” or “bad” categories of people? Specifically, we examined how the language aspects of “switch characters” changed between pre- and post- revelation scenes in nine Disney and Pixar films such as Frozen and Zootopia. Ultimately, we found a linguistic trend that may affect the audience’s perspective on movie characters. Keep on reading to see the effects these movies may unconsciously have on your associations of “good” and “bad” people!

In this project, we examined the correlation between linguistic features and a character’s group membership (as good or bad) in Disney and Pixar films. The specific characters we looked into are those we call “switch characters.” “Switch characters” are those that fake membership as one of the “good guys” but are later revealed as villains. The three linguistic features we felt were most important consisted of pitch, creaky voice, and accent.

Some important definitions:

Pitch: how high or low the speaker’s voice is.

Creaky Voice: also known as vocal fry, happens when the speaker drops their voice to their lowest natural register for emphasis.

Accent: pronunciation specific to an individual or location.

We predicted there would be a change in one or more of these features when a “switch character’s” true membership was revealed. For pitch, we compared range (high, medium, low) of the “switch characters” before and after their reveal to determine if there is a trend in pitch change in a certain direction. Similarly, we looked at the presence of creaky voice preceding and following the switch. In analyzing accents, we aimed to identify any kind of change the character’s pronunciation may undergo.

Our analysis studied the use of linguistic profiling (being able to identify social characteristics based on the language used by the speaker) used by movie makers to reinforce the goodness or badness of a character. We presumed speaker agency in pitch, creaky voice, and accent, through the lens of Speaker and Audience Design Models (Bell, 1984, p. 158). This means that we assumed that “switch characters” actively shift their language based on what group they identify with to distance themselves from or bring themselves closer to their audience.

We based our project on Lippi-Green’s (2012) research that revealed a correlation between accents and variations of standard English with villains. We expanded on her project by looking at additional linguistic variables in Disney and Pixar movies made after 1995 which we believe better represent modern society. The nine movies and characters we analyzed are Frozen (Prince Hans), Coco (Ernesto de La Cruz), Big Hero 6 (Professor Callaghan), Toy Story 3 (Lotso), The Incredibles 2 (Evelyn Deavor), Monsters Inc. (Mr. Waternoose), Toy Story 2 (Stinky Pete), Cars 2 (Sir Miles Axlerod), and Zootopia (Dawn Bellweather).

Methodology: A Sneak Peek into Film Analysis   

Here is an example from Toy Story 3. This example is representative of the methodology that the group utilized to accurately label all nine “switch characters” – pitch, creaky voice, and accent. For the purpose of data collection, a chart adapted from Soares (2017) was used to organize and uniformize character analysis. We repeated the process with all nine films and compiled the analysis into graphs included below.

This selected scene features an exchange between the hero Buzz and the villain Lotso. At the beginning of this scene, Buzz is unaware that Losto is a villain. We see Buzz requesting a group transfer to the Butterfly playroom. Things take a turn for the worse, however, as Lotso only agrees to let Buzz transfer playrooms. Click on the link to see what happens next!

Focusing on “pitch,” the group found uptalk in phrases such as, “showed initiative” and “we got a keeper.” Uptalk is a manner of speaking with a rising intonation at the end of sentences. The italics represent Lotso’s rising intonation. After Lotso’s villainous nature is revealed, uptalk disappears and we hear a deepening and leveling of pitch. Phrases such as, “family man” and “back in the timeout chair” exemplify this deepening and leveling. Therefore, the group labeled Lotso’s pre-reveal pitch as “high: (uptalk)” and post-reveal as “low/monotone.”

Focusing on “creaky voice,” the group didn’t find any phrases that employed a rough voice quality and a lowered pitch. Therefore, the group labeled Lotso’s pre-reveal and post-reveal “Creaky Voice” as “Not Present.”

Focusing on “accent,” the group agreed that Lotso’s phrases possessed the slurred speech patterns of a Southern American accent. Lotso’s Southern accent was exemplified in words containing “r’s” such as ”caterpillar.” Therefore, the group labeled Lotso’s pre-reveal and post-reveal accent as “Southern.”


We noted that eight characters changed at least one linguistic element (pitch, creaky voice, or accent) after their reveal. Our prediction based on Lippi-Green’s analysis proved true, language aspects in the Disney universe do correlate to a character’s identity as good or bad.

Charts 1-9. Linguistic Analysis of Disney and Pixar “Switch Characters” Comparison of pitch, creaky voice and accent pre-reveal and post-reveal.

The linguistic aspect that changed most was pitch, followed by creaky voice and accent. Only one character, Stinky Pete, had an accent change, settling completely into Standard American English (SAE) after the reveal as opposed to switching between Southern American and SAE. Considering that Stinky Pete employed SAE before revealing himself as a villain, we decided to view accent as not indexing goodness or badness in his character. This diverges from our initial prediction, since Lippi-Green’s study demonstrated a strong relationship between accent and intrinsic goodness and badness.

Fig 1. Linguistic Changes After Character Switch. Amount of characters that presented change in a certain linguistic after their reveal as villains.

After determining which aspects changed after the reveal (pitch and creaky voice) we analyzed exactly how these aspects changed. For eight of the nine characters, there was a drop in pitch, and only one character had a rise in pitch. It is also worth noting that some of the characters’s pitch dropped when they produced especially aggressive statements or when they mocked their villainous persona. From our data, we conclude that a strong correlation exists between lower pitch and evil personas.

Fig 2. Pitch Change. Percentage of “switch characters” that presented either a rise or drop in pitch.

The other linguistic aspect we noticed a change in was creaky voice. Six characters used creaky voice after their reveal. Of the characters that initially presented creaky voice all maintained creaky voice after reveal. One thing to note is that creaky voice is closely related to pitch, therefore a drop in pitch normally meant the addition of creaky voice.

Fig 3. Characters With Creaky Voice. The number of characters that presented creaky voice before their reveal and number of characters that presented creaky voice after their reveal.

Overall, our data supports the hypothesis that certain linguistic aspects correlate with group membership (as good or bad). However, this change seems to be mostly related to pitch and not accents as studied by Lippi-Green (2012). Drop in pitch seems to be the universal linguistic aspect in Disney and Pixar’s universe that signifies a villainous persona and a higher pitch seems to signify and contribute to blending in with good characters.


We know that Disney and Pixar movies have helped to socialize children into stereotyping and othering, based on accents in the research done by Lippi-Green (2012) and others. But do “switch characters” also contribute to this categorization in children? Through this study, we conclude that pitch, as well as the presence of creaky voice, are heavily correlated to an evil persona. So do children begin to associate these features with villains after seeing such movies?

Children tend to relate a higher pitch to brightness (Marks, Hammeal, Bornstein, 1987). This association creates a positive attitude towards a higher pitch, as shown by Banaji and Greenwald in “Into the Blindspot.” Therefore a lower pitch may imply a more negative attitude towards the person speaking. This could imply that children are wary of those with lower pitches in their speech and so, when the “switch characters” do this, it only reinforces this association.

There aren’t enough studies on children’s perception of creaky voice to conclude its influence on them. But if lower pitch implies a negative attitude, then the lowest register (creaky voice) will most likely imply one as well.

As a result, we can theorize that children notice and are affected by the changes in pitch and the use of creaky voice. However, our conclusions on the effect of the movies on the audience can only be hypothetical, as our data does not include audience responses.


In our study we analyzed how certain linguistic features (pitch, creaky voice, and accent) changed when a character switched from good to bad. The purpose of our study was to find linguistic trends in these characters.

Our data showed that most “switch characters” dropped pitch and added creaky voice when they revealed to be evil, while their accent remained constant. Looking at Marks, Hammeal and Bornstein (1987), we found that children are likely to view a higher pitch positively and theorized that Disney and Pixar movies might contribute to this phenomenon, or, at the very least, rely on it for indicating a character’s identity as good or bad.

However, we can’t make definite conclusions because our small sample size and lack of data of the audience’s response. So, we can only theorize what kind of impact these “switch characters” have on their audience and what linguistic trends are present in the Disney universe. But linguistic trends in Disney characters remains an important topic to be researched, because of the continued promotion of the dominant ideology presented in Disney and Pixar movies, especially considering the size of their audience.



Banaji, M., & Greenwald, A. (2013). Blindspot: Hidden biases of good people. New York: Delacorte Press.

Bell, A. (1984). Language Style as Audience Design. Language in Society, 13 ( 2), 145-204. Retrieved from

Girard, F., Floccia, C., & Goslin, J. (2008). Perception and awareness of accents in young  children. British Journal of Developmental Psychology, 26(3), 409-433.

Lippi-Green, R. (2012). Teaching Children how to discriminate: What we learn from the Big Bad Wolf. English with an Accent: Language, Ideology and Discrimination in the United States, 7, 101-129.

Marks, L. E., Hammeal, R. J., & Bornstein, M. H. (1987). Perceiving Similarity and Comprehending Metaphor. Monographs of the Society for Research in Child Development, 52(1),1-92.

Soares, Telma O. (2017). Animated Films and Linguistic Stereotypes: A Critical Discourse Analysis of Accent Use in Disney Animated Films. Bridgewater State University Master Theses and Projects. 53, 1-53.

Read more