International Journal of Communication 20(2026) Drawing the Faces of “Perpetual Foreigners”
Drawing the Faces of “Perpetual Foreigners”: A Multimodal Narrative Analysis of #StopAsianHate on Instagram
WEIYU ZHANG[1]
National University of Singapore, Singapore
Racism against Asians has been a historical problem around the world. The Internet turned into a battleground for racist and antiracist discourses and activism during the pandemic. As such, discursive engagement is often organized around hashtags such as #StopAsianHate, hashtag activism has become one of the key manifestations of the movement. Based on the affective narrative theory, this study aims to understand how multimodal messages including texts and visuals can be used to derive narratives and affects. By examining 46,621 Instagram posts published from January 1, 2020 to November 1, 2022, this study finds #StopAsianHate was short-lived and single-peaked and featured almost an equal amount of positive and negative sentiments. By also analyzing posts with more than 100,000 total interactions (N = 92), this study further delineates multiple affective narratives derived at the iconic, grammatical, and collective levels through layered interpretation. The findings suggest all hashtag movements are equal, pointing to the need to analyze plot twists and ugly feelings in affective narratives using multimodal data and methods.
Keywords: affective narratives, hashtag activism, multimodal analysis, plot twists, StopAsianHate, ugly feelings
Weiyu Zhang: [email protected]
Date submitted: 2025-03-27
Racism against Asians has been a historical problem around the world, starting from the “yellow peril” (Tchen & Yeats, 2014) in the 19th century, when Asian workers migrated to Europe and North America. Anti-Asian racism in the United States has particularly deep roots, taking both sporadic and structural forms, from violent attacks like the 1871 Chinese Massacre (Dorland, 1894) to Japanese internment camps during World War II (Ng, 2001). In the late 20th century, Vincent Chin, a Chinese American, was beaten to death in Detroit by two White autoworkers who blamed Japanese imports for job losses, and the perpetrators received no jail time (Wu, 2002). Turning to the 21st century, South Asian, Muslim, and Sikh Americans faced a surge in discrimination, profiling, and violence after 9/11 (Ahluwalia & Pellettiere, 2010). Despite a wide-ranging diversity in ethnicity, language, culture, and tradition (Kabir & Ha, 2025), Asians have often shared the experience of discrimination and prejudice based on their appearance, country of origin, spoken language, and other cultural distinctions (Lee, 2015).
Recent years have seen a spike in anti-Asian racism, especially following the breakout of COVID-19. One of the most disastrous was the Atlanta shooting incident, in which six Asian women were killed by one White man. Stop Asian American and Pacific Islander (AAPI) Hate (2022), a nonprofit organization, found that from March 19, 2020 to March 31, 2022, a total of 11,467 hate incidents against AAPI persons were reported. One in five AAPIs experienced a hate incident during the two-year period. A BBC news report (Clements, 2021) stated that “UK police data suggests a rise of 300% in hate crimes towards Chinese, East and South East Asians in the first quarter of 2020 compared to the same period in 2018 and 2019” (para. 15). A national survey of 2,003 Asian Australians (Kamp et al., 2022) found that 40% of participants experienced racism during the COVID-19 pandemic, and among them, 66% experienced racism online.
The Internet has turned into a battleground for racist and antiracist discourses and activism since the pandemic. Individuals, groups, nongovernmental organizations (NGOs), and businesses (Dong et al., 2023) have joined “the first massive online social movement for AAPIs” (Lee & Jang, 2023, p. 21) by engaging in social media advocacy. As such discursive engagement is often organized around hashtags such as #StopAsianHate (#SAH) or #AsianLivesMatter, hashtag activism has become one of the key manifestations of the movement. Social media platforms such as Twitter (Kabir & Ha, 2025; Kim & Kim, 2023; Lee & Jiang, 2023; Ofori-Parku & Moscato, 2018; Shahin & Hou, 2025; Wang et al., 2024) and TikTok (Jacques et al., 2023; Lee & Lee, 2023) emerged to allow individuals to express their creative counternarratives organized around the hashtags. Instagram is another popular social media platform, known for its visual affordances (Leaver et al., 2020). Prior studies confirmed its prominent status in hashtag activism, such as facilitating impulsive, symbolic, and affective expression (Adi et al., 2018); constructing the political activist subject (Dumitrica & Hockin-Boyers, 2023); and influencing users’ political identities (Li, 2022).
This study aims to understand how multimodal messages including texts and visuals can be used to demonstrate movement narratives and affects in #SAH on Instagram. By examining 46,621 Instagram posts published from January 1, 2020 to November 1, 2022, this study describes the evolving patterns of sentiments found. In addition, by analyzing posts with more than 100,000 total interactions (N = 92), this study further delineates multiple affective narratives derived at the iconic, grammatical, and collective levels through layered interpretation. Both theoretical contributions to affective narratives and methodological contributions to multimodal analysis are discussed. Practical implications about #SAH and the hashtag movement are provided.
Affective Narratives in Hashtag Activism
Social media has long played a significant role in social movements. From the Arab Spring and the Occupy Movement to #MeToo and #BlackLivesMatter, the last two decades have witnessed social media’s shift from one communication tool to the major movement ground, in coordination with street protests and other movement activities. Hashtag activism is one of the emerging forms of social movement. Yang (2016) defined the phenomenon clearly: “Hashtag activism happens when large numbers of postings appear on social media under a common hashtagged word, phrase or sentence with a social or political claim” (p. 13). Existing studies have moved from contesting hashtag activism as slacktivism (Christensen, 2011) to understanding how online activism is organized (e.g., Guo & Liu, 2022) and scaled up (e.g., Aydin et al., 2022; Mundt et al., 2018). Social movement theories identified frame alignment (Snow et al., 1986), a process of aligning diverse interests of different movement participants, as one of the important factors that contribute to the success of a movement.
Regarding #SAH, existing scholarship has largely focused on identifying frames and interests through topic modeling (Kim & Kim, 2023; Lee & Jiang, 2023; Shahin & Hou, 2025; Wang et al., 2024), with one exception that traced the network evolvement over time (Wang et al., 2025). These Twitter-based studies revealed that posts primarily emphasized calls to action, political participation, and raising Asian American visibility, but often lacked concrete plans for resource mobilization. Two studies found that the Asian-Black solidary in Twitter hashtag activism is questionable: While Kim and Kim (2023) observed the “lack of the spillover or reciprocity effect on solidarity” (p. 6112), Shahin and Hou (2025) demonstrated that “the most prominent topic in our corpus was driven by a tweet about an Asian woman being assaulted by a Black ‘suspect’” (p. 12). Based on these existing findings, the study tries to further answer the question about how these diverse discussions are (mis-)aligned in #SAH and why they might contribute to the rapid demise of #SAH.
This article attempts to examine this question through the concept of affective narratives. Narrative is a classic concept that has gained much attention in literature studies and, increasingly, social sciences. In literature studies, narrative is a form that accounts for the progression from beginning through middle to end (Yang, 2016). Film studies examines how filmic language (Braudy & Cohen, 2004) differs from verbal language in demonstrating narratives, such as that film editing allows the rearrangement of time and space and alters meanings. Other than sequencing events to establish meanings, recent political science works argue that narratives are deeply intertwined with emotions in collective actions (Frederick, 2014). Narratives not only contain emotional content (e.g., usage of sad words) but also shape emotional orientation toward the specific narrative (e.g., a sad story). Drawing on these conceptual traditions, this study defines affective narratives as a multimodal form that accounts for the progression of events to establish meanings and express emotions.
Hashtag activism is distinctive in its use of social media affordances, particularly the ability to employ a keyword to organize content and connect dispersed users. Yang (2016) observed that hashtags develop a narrative form that unfolds over time, weaving individual posts into a collective story that invites interaction and participation. Beyond facilitating narrative agency through creative and communal hashtagged postings, hashtags serve, Papacharissi (2016) argued, as connective mechanisms enabling networked publics to “come together and/or disband around bonds of sentiment” (p. 308). Hashtags thus afford not only storytelling but also streams of affective expression. Affective publics—defined as “public formations that are textually rendered into being through emotive expressions that spread virally through networked crowds” (Papacharissi, 2016, p. 320)—emerge through this interplay of narrative and emotion. Affects do not simply accompany narratives, but actively shape them, orienting interpretations of events and influencing collective responses. As Jasper (2019) highlights, emotions are constitutive of protest trajectories, motivating action and sustaining engagement. Hashtag activism fuses these processes: It narrates events while embedding emotional cues that invite empathy, anger, grief, solidarity, or other emotional responses. Narratives in hashtagged posts thus function as affective scripts, transforming individual emotional reactions into shared, networked experiences that strengthen collective identity and mobilize participation over time.
Tracking the Development of Affective Narratives
As streams of affects and unfolding stories, the first materiality of affective narratives shows an evolution over time. Narratives around hashtags should change in volume, thematic focuses, and emotive tones. There is a liveness or episodic feature to hashtag activism (Zulli, 2020). Social media affordances such as time stamps and the live streaming of posts provide empirical grounds on which stories can unfold. In addition, networked publics are connected via hashtags, but not necessarily sharing the same collective frame. We should expect that, as streams, there are multiple—often alternative to what we can find on traditional mass media—stories being told. Lee and Lee (2023) found that through hashtagged video-making on TikTok, “Asian/American women extend the discussion on anti-Asian racism to include their gendered and raced experiences, and challenge racism in affective and evocative ways” (p. 1). Moreover, these narratives do not necessarily share the same emotive tones. We should expect that some stories are happier or sadder than others. Although feelings are mixed, the intensity of the feelings should be an indicator of the level of engagement in hashtag activism.
Tracking emotive tone in hashtag movements complements volume trend by uncovering the affects driving online narratives. Jasper (2019) emphasized that emotions are central to collective action, influencing individual participation and shaping movement dynamics. Emotions expressed in online narratives—such as fear, outrage, grief, and solidarity—play a critical role in sustaining activism. Studies on #BlackLivesMatter, for example, showed that emotional content in tweets (e.g., anger and solidarity) not only amplifies engagement online but also correlates with on-the-ground protests (Field et al., 2022). Similarly, research on #MeToo indicated that emotional sharing fosters digital bonding and collective identity, enabling long-term mobilization (Xiong et al., 2019). Based on these considerations, this study starts by proposing the following research question:
RQ1: How did the sentiments of Instagram posts containing the hashtags #StopAsianHate and #AsianLivesMatter evolve over time?
Identifying Multimodal Affective Narratives
The concept of affective narratives has a focus on affects. Prior studies confirmed that users’ emotional reactions often mirror the verbal sentiments expressed in social media posts (Giuntini et al., 2019; Sumner et al., 2020). Dictionary-based sentiment analysis tools (e.g., Natural Language Toolkit [NLTK]) identify the expressed affects through counting the number of words that are labeled as certain emotions. This operational definition of affects allows us to capture not only the mixed feelings humans often express but also the intensity of the expressed emotions. The higher the count of words that contain one certain emotion, the more intensively the emotion is expressed. Compared with texts, visuals are particularly powerful in evoking emotional reactions because they are intuitive and appeal to humans’ automatic responses (Bleiker & Hutchison, 2008). Taking a multimodal approach has several advantages, such as understanding how mixed feelings can be conveyed, discovering the resonance and discordance between different types of affective expression, and identifying potential affective publics. However, current studies are limited in judging visuals’ sentiments. Other than color tones (e.g., bright colors may suggest happiness) and facial expression (e.g., an open mouth may suggest laughing, thus happiness), there are few visual features that can be used to clearly identify emotions.
Prior studies have tried to identify affective narratives in hashtag activism in terms of written texts such as tweets. However, the popularity of visual platforms such as Instagram and TikTok has urged researchers to expand the study to visual or multimodal analysis (Highfield & Leaver, 2016). To achieve this end, researchers tried to use the quantitatively inclined content analysis (Adi et al., 2018; Clever et al., 2023; Xie et al., 2023), the qualitatively inclined discourse analysis (Lee & Lee, 2023; Oh, 2023; Turvy, 2023), or a combination of both (Trillò et al., 2021). Visual features were defined as platform affordances (e.g., Instagram Live or Instagram Story; Li, 2022), visual formats (e.g., color, pixel, sketches vs. photos; Peng, 2018), visual content (e.g., featuring people, scenery, or festivals; Aramendia-Muneta et al., 2021), or features of visual objects (e.g., facial expressions of human objects in the image; Peng, 2018). Although these are all necessary visual components that collectively constitute the narrative, a simple aggregation of the features does not equal a story. Visuals have to work together with texts, interface features, and contextual cues to derive narratives. With the multimodal background in mind, the study explains how affective narratives can be demonstrated.
When applying narratives to identity (a salient dimension of the Stop Asian Hate movement), Somers (1994) argued that social life is storied and our sense of agency is derived from social narratives (i.e., narrative agency). Three dimensions of narrativity can be identified: ontological narratives that define the perception of ourselves, public narratives that relate to cultural and institutional stories, and metanarratives that are grand stories about the time we are embedded in. Somers’ (1994) typology shows narratives can be understood as both individual and collective; the key linkage that allows this individual-collective juxtaposition in hashtag movements is social media affordance. Inspired by Somers (1994), this study examines three levels of multimodal representation of narratives. First, the iconic level refers to composition, colors, position, objects, symbols, and their relational coexistence. This level is termed semiotic or iconic (Rodriguez & Dimitrova, 2011), often focusing on what we can see within one image or one scene in a video. Second, the sequential characteristic of narratives gives them a structure, or a grammar, as Cohn (2013) put it. Visual grammar does not always follow a linear fashion, but rather embraces a dynamic process. This dynamic process results in particular configurations, or “plots” that “weave together a complex of events to make a single story” (Polkinghorne, 1988, p. 19). Third, individual stories can be expressed in parallel with or without connections. One significant feature of social media affordances is the connective mechanisms provided by interface features such as hashtags and the interactive mechanisms such as comments. Therefore, it is possible to examine the interactions among multiple narratives as well as the evolution of the narratives in relation to the larger context. This level of collective narrative demands an analytical awareness of historical events and cultural environment. This study thus proposes the following research question:
RQ2: How do the most popular Instagram posts containing the hashtags #StopAsianHate and #AsianLivesMatter demonstrate affective narratives in a multimodal manner?
Methods
Data Collection
Using Meta’s data platform CrowdTangle, two hashtags (#StopAsianHate, #AsianLivesMatter) were used to search through Instagram public accounts. The keywords have been shown to be the most popular two hashtags used by the movement (Wang et al., 2024). A search that used six hashtags (#StopAsianHate, #AsianLivesMatter, #AsiansAreHuman, #EndRacism, #StopAAPIHate, and #StopAsianHateCrimes) showed that the volume of posts only increased about 10%. The keyword search generated 47,707 Instagram posts published from January 1, 2020 to November 1, 2022, including the Atlantic spa shooting tragedy in March 2021. The time period starts from two months before COVID-19, which has led to a significant increase in anti-Asian hatred, was declared by WHO as a pandemic. The time period ends in the month when the search was conducted. The distribution plots of both post and engagement frequencies[2] show the volume of activity has decreased to roughly the same level as early COVID-19 days. This almost three-year duration covers a series of significant events that have marked the development of the #SAH movement. This relatively longer time span allows this study to track the evolution of the movement and discover long-term patterns that take time to emerge. More detailed results and interpretations about the evolving volume can be found in Appendix A. After cleaning up duplicates and empty posts, 46,621 posts were left.
The search only included verified accounts on Instagram, which are associated with authenticated public figures, organizations, and media outlets whose identity has been confirmed by the platform. Including verified accounts helps ensure data reliability and reduces the risk of analyzing content generated by bots or fake accounts, which can distort findings on hashtag activism (Ferrara, 2020). Automated or malicious actors can compromise data quality and were thus excluded. Other than this requirement, this study did not impose any search criteria such as country of origin, language, or branded content because the #SAH movement is multicultural, transnational, and cross-sectoral.
Data Preprocessing
As the first step of data preprocessing, filtering is essential for removing irrelevant, duplicate, or low-quality posts; enhancing data precision; and minimizing noise. Stieglitz and Dang-Xuan (2013) highlighted that rigorous filtering protocols are critical for capturing authentic and meaningful discourse patterns on social platforms. Therefore, the analysis first removed numbers, URLs, punctuation, words whose length is less than one letter, and the search keywords themselves. Further cleaning was done to improve the effectiveness of dictionary-based analytical tools. Using Gensim and NLTK libraries in Python, the analysis tokenized the texts to divide them into words and removed all stopwords. The analysis then reduced all words to their root forms by stemming and lemmatizing.
Sentiment Analysis
In this study, the National Research Council (NRC) lexicon was used to identify specific emotions expressed in posts. The NRC Emotion Lexicon provides a list of English words as well as their associations with eight basic human sentiments: anger, fear, anticipation, trust, surprise, sadness, joy, and disgust (Mohammad & Turney, 2013). This list of specific emotions provides more nuanced sentiment analysis than simple positive/negative/neutral categorizations provided by other popular tools such as Vader (Bonta et al., 2019; Isnan et al., 2023). A post can contain multiple categories of sentiments, and the number of words that are associated with each sentiment can serve as a measure of intensity. Van Atteveldt et al. (2021) found that dictionary-based approaches like NRC, despite some limitations, remain a valuable option for large-scale text analysis where interpretability, transparency, and computational efficiency are considered. Their study highlights that while machine learning methods can outperform dictionaries under ideal conditions, their performance strongly depends on large, domain-specific training data sets. In contrast, well-established dictionaries such as NRC can provide stable and interpretable results across diverse domains, particularly when dealing with informal content, as is common in Instagram activism posts.
Multimodal Narrative Analysis
While computational methods allow us to have an overview about the evolving hashtag movement, in-depth understanding about personal stories and collective narratives can only emerge through qualitative analysis. Moreover, visual analyses through computational methods have been limited to the visual format and features of visual objects’ level, which are not able to detect the nuanced procedure of visual storytelling. Qualitative methods are still the most appropriate ones for such tasks. Among thousands of posts, the ones that have gone viral present opportunities to understand which personal stories contribute the most to collective narratives. The study thus focuses on Instagram posts that attracted more than 100,000 total interactions (N = 92) and conducted a multimodal narrative analysis. Selecting Instagram posts with 100,000 interactions (likes, shares, comments) or more ensures the analyzed content reflects high-visibility and high-impact narratives within the #SAH movement. According to Sprout Social (a social media management and analytics platform; Sheikh, 2022), “100,000+ likes or views and 1,000+ comments are considered to be viral” on Instagram (para. 58). Focusing on viral content is based on the following reasons. First, prior research emphasized that social media attention is unevenly distributed, with a small proportion of posts accounting for the majority of visibility and engagement (Shahin & Hou, 2025). Second, social media algorithms further promote the visibility of high-engagement posts; consequently, viral content is most likely to shape public discourse, set the agenda, and influence audience perceptions (Klinger & Svensson, 2015). Focusing on this threshold allows researchers to capture narratives with the greatest reach, which are most relevant for understanding how multimodal storytelling—through images, videos, and captions—circulates and mobilizes large-scale participation.
This study’s approach to multimodal narrative analysis was informed by two traditions: One is the visual narrative analysis commonly seen in political science and international relation studies (Freistein & Gadinger, 2022), and the other is the critical technocultural discourse analysis (Brock, 2018). While the former has a specific emphasis on the visual narratives that are often in contestation, the latter has its strength in incorporating interface analysis, a component necessary for social media analysis. Both methods share the sensitivity to cultural contexts, making them suitable for the #SAH movement, where the Asian diaspora culture (e.g., racial discrimination, model minority, inter-Asian tensions) is salient (Lee & Lee, 2023; Oh, 2023).
Specifically, the multimodal narrative analysis in this study applied the “layered interpretation” (Freistein & Gadinger, 2022) in three steps. First, the study analyzed the visuals in terms of their composition, atmosphere, symbolism, and relations to other information formats such as texts (e.g., image caption, post texts, hashtags, comments) and interface features (e.g., Instagram image, album, or video). Extra attention was paid to the cultural connotations of the visual components (e.g., races of the featured people, cultural symbols). Second, the study interpreted the narrative conveyed in one Instagram post by zooming in on plots (i.e., establisher, initial, peak, release), voices (e.g., first-person vs. third-person), roles (hero, villain, victim, etc.), relations among roles, and genres (tragedy, romance, sarcasm, etc.). Multiple visuals were interrelated in this step to further infer the narrators’ motivation and the affect they want to express and evoke. Finally, multiple narratives were analyzed in terms of both their contestation and coordination. This step paid special attention to the collective level and the longitudinal perspective by relating the emergence, disappearance, and interactions of certain narratives to historical events and cultural environment. Reports from the Stop Asian Hate organizations, news reports from both mainstream and community media, and academic writings on Asian American history and culture were consulted to provide such a contextualized understanding.
Results
The result shows that both positive and negative emotions were expressed with great intensity. One of the most surprising findings is that trust was the most intensively expressed sentiment. Trust words had close to 60,000 mentions in March 2021, followed by anger and fear words (around 50,000 mentions each). Joy, sadness, and anticipation words had similar levels of intensity (in the range of 40,000 mentions in March 2021). Disgust (20,000) and surprise words (18,000) had the lowest mentions in March 2021. This shows that even in the darkest moments, the affects expressed through Instagram verbal messages were highly mixed. The almost equal distribution of positive and negative sentiments implies that the #SAH movement was not merely expressing grievance and the participants in the movement stayed trustful and hopeful despite the tragic incidents.
The Viral Content Authors
As the #SAH movement was short, the majority of the most popular Instagram posts were created in March 2021. Five of 92 were posted before and 12 posts after March 2021. The last post that had over 100k interactions appeared in August 2021. From then on, no Instagram posts have captured this much attention. Among the most popular posts, the authors included celebrities, news media, brands, artists, activists, NGOs, and ordinary participants. Justin Bieber’s simple image with the words “STOP ASIAN HATE” against a black background received the highest number of interactions. Following him, we see celebrities such as Jackson Wang, Damian Lillard, Sandra Oh, Arden Cho, Anne Hathaway, Amber Liu, Ellen DeGeneres, Victoria Beckham, Rihanna, and more. Less well-known names such as Aimee Song, Eugene Lee Yang, Mark Ruffalo, and Awkwafina sometimes received higher numbers of interactions than household names. Brands such as Starbucks, National Football League, Netflix, Rockstar Games, Pokémon, and Nike featured prominently. Instagram’s own official account also joined. News media including The New York Times, Bleacher Report (sports news), E! News, Now This News, CNN, and Complex Networks reported the incidents and showed support. Support from brands, celebrities, and sports figures was frequently among the most popular posts. Although these accounts were influential because of their large followings, their posts were often too simplistic to be able to derive meaningful narratives. Community media, artists, social media creators, and ordinary participants tended to provide genuine accounts of both personal and collective stories.
The Categorization of Affective Narratives
Stories at the post level can be categorized into the following: abuse, fight back, family, calling-outs, awakening, and interracial stories. The most widely circulated #SAH stories on Instagram are abuse incidents, from deadly cases, bloody attacks, and sexual and racist assaults in daily life to microaggressions. One common feature of the deadly and bloody abuse stories is the video footage from surveillance cameras. This footage not only provided hard evidence about these severe and random attacks but also confirmed the objectiveness in a post-truth era. There is no room for debate in such cases, demonstrating that anti-Asian hate is real.
For assaults and microaggressions that Asians experience in daily life, most of the personal sufferings were written in texts because video or image evidence was lacking. This might explain why one video on microaggressions, captured by a Korean Twitch streamer, went viral. In this exceptional video, five incidents of microaggressions were edited into one piece. In two such incidents, passersby tried to grab her phone when she was livestreaming, almost hitting her. In the other three, men either spoke “ching chong” or pulled their eyes in front of her camera (see Figure 1). Although in all three cases, she protested by saying, “What are you doing?” or “This is racist,” the men just carried on as they had been or looked as if they were not doing anything wrong. The last scene showed her almost in tears, as if she was defeated by the repeated racist aggressions. The viral version of the video did not come from the streamer’s own account, but was reposted by Wear the Peace, an ethical clothing brand. A frame saying “Asian hate has been a problem” was added on top of the video and sad piano music dubbed onto the original live sound (Wear the Peace, 2021). However, the comments show that this video, unlike the surveillance footage, was highly contested. Within the top 20 comments, one said, “So no one cares. ( ---- __ ---- ) <--- Chong” (Wear the Peace, 2021). Another commenter wrote, “Asian hate is real💔but THIS video is fake 💯” (Wear the Peace, 2021). Yet another stated, “If she’s gonna be getting sad kid go back to her country no ones gonna change” (Wear the Peace, 2021).
Figure 1. An example of a story about microaggression against Asians (Wear the Peace, 2021).
Fight-back stories include both immediate fight-backs to physical assaults and long-term coping strategies. The most reacted to fight-back story features a 76-year-old Chinese woman who managed to reverse the situation on her attacker, resulting in injuries severe enough to require him to visit a hospital. The comments were overwhelmingly condemning the White attacker and questioning why the grandma was not getting medical attention. A Filipino former professional boxer, who became a politician, posted, “Fight me instead” in English, Tagalog, Chinese, and Korean (Pacquiao, 2021). In the poster, his face was juxtaposed onto a combination of multiple victim pictures. In addition to physical fight-backs, coping strategies are shared in narrative formats. One widely circulated video is a studio-made short film portraying an Asian salon owner smartly recording racist assaults and protecting his family business. These stories have clearly identified the heroes and villains, mimicking the classic “good guy versus bad guy” narrative.
Family stories traced back to the history of Asian immigration and the racial discrimination that fills this history. One such story was a collage of family photos showing the creator’s grandparents and great-grandparents who came from Japan to Brazil (personal communication,[3] August 17, 2023). Texts written in Portuguese contextualized the post as a condemnation of anti-Asian prejudices during the pandemic and emphasized the creator’s pride in being Asian. Another story documented an Asian veteran’s speech during which he revealed his chest scars from serving in the U.S. military. This scene was put at the beginning of the video, followed by his full speech that condemned the insults about Asian Americans not being patriotic enough (personal communication, August 17, 2023). A sad story was told when the son of a dead victim posted about it on Facebook. The killed Korean woman turned out to be a single mother of two sons. The son was asking if “anyone knows anybody that had the misfortune to be in a situation like this that could inform me on how to go on” (personal communication, August 17, 2023). The mixture of pride, anger, and sadness made the family stories almost the most tragically emotional to read.
The ignorance and denial of racism against Asians were called out using various narrative forms. An Asian TV personnel wrote, “Asian women are your punchlines” to call out mainstream media’s reluctance to report the Atlanta shooting as a racist attack (personal communication, August 17, 2023). Her Instagram post comprised three screenshots of her tweets, making them read like three pages of a poem (see Figure 2). In a post starting from “how the media’s depiction of Asian women harms them” (Diet Panda, 2021), six video clips of films, TV series, and music videos were shown along with written texts to explain how the racist media depictions perpetuate dangerous stereotypes of Asian women as exotic beings. The post ended with screenshots from Asian females’ fight-back tweets, reinforcing the argument that media depiction normalizes the hypersexualized stereotype that contributes to racist violence. Other than media, fashion brands, public figures, and even American police were called out for their racist advertisements, racist comments, and reluctance to frame the Atlanta shooting as a racial hate crime. Images and videos of their advertisements, speeches, and press releases were provided as evidence, accompanied by explanations to contextualize the interpretations.
Figure 2. An example of the callout story (personal communication, August 17, 2023).
Calling-out stories target others. Awakening stories target Asians themselves. One Korean American’s awakening posts appeared three times among the top posts. Two posts showed the videos of his interviews with Bloomberg and CNN, and a third one was an article posted on TIME Magazine’s website. In all three posts, the singer-songwriter reflected deeply on Asian Americans being the “perpetual foreigners” who internalized anti-Asian racism. The “model minority” myth both trivialized the racist attacks on Asians and made Asians believe they had to be silent about these abuses to “fit in.” Resonating with the awakening narrative, there were artist drawings that give faces to the Asians, especially the Asian women, who are usually too fearful to be seen. In one such drawing (Figure 3), the first image shows a masked woman with long black hair and eyes closed; surrounding her are hands holding Post-It notes filled with racist comments. In the second image, her eyes are wide open, and “#StopAsianHate” is written on her mask. An Instagram influencer who reposted the images wrote alongside them, “I can’t even begin to put into words how I’m feeling about the most recent attacks against Asians. But staying silent is not an option!” (Wang, 2021).
Figure 3. An example of the awakening story (Wang, 2021).
The last storyline refers to the interracial relationships that revealed the unique challenges #SAH faces: the interracial tension between African Americans and Asian Americans (Kim & Kim, 2023; Shahin & Hou, 2025). One post appeared twice in the top list and included a screenshot of a tweet published by Candace Owens, an American conservative political commentator who is an African American female. The tweet said, “The #1 violent offenders against black people are other black people. The #1 violent offenders against Asian Americans are also black people. But both #BlackLivesMatter and #AsianLivesMatter are campaigns dedicated to stomping out white supremacy because, clown world” (prageru, 2021). The Instagram account that reposted it is a conservative nonprofit organization. Nevertheless, the comments on the posts were mixed with support and objection. Under some other #SAH posts, comments that repeated this storyline were made. This story complicated #SAH by making the targets of change confused. White supremacy did not appear to be the cause. This story appealed to a sarcastic feeling of surprising “truth.”
To answer RQ2, the study finds that in the #SAH movement, the stories go beyond presenting the abuses that elicit anger and sadness. By asking for or showing support, positive feelings such as trust and anticipation are expressed. The stories highlight fight-backs from the minority with joy and call out media, brands, and public figures with disgust. The family stories bring out strong feelings of pride and sorrow. The awakening stories reveal deep self-reflection on the longstanding anti-Asian hate. However, the interracial stories shift the focus on White supremacy to the interracial tension between African and Asian Americans, eliciting a feeling of surprise. Three collective-level narratives can be further identified: first, a narrative of grievance and support, which positions the narrators as victims or shows their supporters; second, a narrative of identity and community, which manifests the reflective agency of the narrators as a group rather than individuals; and, last, a narrative of conflicts and confusion, which includes narrators who take a third-person observer position. These narratives are supported by a mixture of multimodal expressions, including surveillance footage, livestreaming clips, studio-made short films, family photos, poems, artist drawings, and other traditional media forms. These materials were almost always edited or sequenced into a meaning-making process. Through video frames, captions, descriptions, and comments, the raw content was given a certain lens of interpretation. However, inconsistencies between modes of expressions (e.g., comments challenging the narrative in the posts) or across media platforms (e.g., racial representations in old films interpreted as perpetrating stereotypes) blur the storylines.
Plot Twists and Ugly Feelings
Two plot twists are found under the identity and conflict narratives, further complicating the metanarrative of the #SAH movement. The identity narrative mostly reaffirms pride and loyalty to one’s Asian identity. However, the self-reflection story goes beyond reaffirmation and questions Asian Americans’ fear, compliance, or even conspiracy in the long history of anti-Asian hate. The self-reflection story implies that Asians themselves are part of the problem because they are too complaisant or even complicit. The second plot twist is found in the interracial story that singles out African Americans as the most serious cause of anti-Asian hate. As both plot twists center on racial minorities in the country, the plot twists may confuse the majority of American society and lead to at least a pause in taking concerted actions. Plot twists in #SAH are achieved in a multimodal manner—texts, visuals, and platform affordances are all present. Textually, stories with numbers and arguments are written; visually, a narrator with a face that is part of the victim group tells the stories; affordance-wise, interaction figures such as the number of likes and comments bring visibility to these posts, and cross-platform connections amplify the plot twists’ spread.
Moreover, the affects expressed in plot twists are not strongly positive or negative, but rather mixed with other affects that are weak. Ngai (2004) called them “ugly feelings,” the minor negative feelings that are not “grander passions” such as anger or joy, but “generally unprestigious feelings” such as irritation or anxiety (p. 66). Such feelings suggest an “obstructed agency” and provoke a sense of passivity, which indicate the actual material inequalities that are structured. The sarcasm and surprise seen in the interracial story are diagnostic of the situation: Anti-Asian hate exists across the American society, not just among the majority race. The self-reflection story is presented in rather rational language, but subtle affects such as shame are scattered here and there. For instance, in one interview, the self-reflection author cited his own experience of “almost wanting to apologize to his attacker,” showing he is ashamed of his own reaction (personal communication, August 17, 2023). This ugly feeling again is indicative of the discursive traps of Asian Americans being the model minority. However, neither sarcasm nor shame leads to productive actions, at least not clearly shown in the stories themselves. The interracial story attributes it to the “clown world,” and the self-reflection story admits there is no solution the author can offer.
Discussion and Conclusion
Theoretically, this study focuses on affective narratives in a context of social movement or collective action (Frederick, 2014; Jasper, 2019), in contrast to other discursive devices such as frames (Snow et al., 1986). It finds that a sequence of linked events with a beginning, a middle, and an end is complicated in hashtag movements such as #SAH. Prior studies have focused on competing narratives (Canevez et al., 2022) that attribute the causes of the problem to different actors. For instance, the Atlanta police’s initial narrative of a sexually driven crime attributes the cause to the sexual addiction of the killer. This study suggests that plot twists are even more problematic than competing narratives because they are embedded in the metanarrative of the movement’s beginning, middle, and end. Future research should pay more attention to plot twists or frame misinterpretation. Moreover, this study shows that narratives are deeply intertwined with emotions, but the examination of emotions in social movement has largely focused on passions, or strong positive or negative emotions. This study makes it evident that ugly feelings, or mild and ambiguous affects, also play significant roles in the narratives around social movements, especially in terms of obstructing agency and halting actions.
Methodologically, the findings confirm that affective narratives are demonstrated through multimodal forms, which are conditioned by platform affordances. However, the study finds that a visual-centered social media platform poses special challenges to movement mobilization and organization because of the apparent gaps among multimodal channels. Although visuals are central on Instagram, textual information, either overlaid on the visuals or separately presented in the posts, provides additional interpretations of the visuals. However, the visuals and the texts are not always consistent in their sentiments and stories. An apparently negative image (e.g., all black with #SAH) may accompany a largely positive textual message (e.g., showing support). A video that clearly demonstrates microaggressions may be questioned for its authenticity and seriousness by top comments. Another methodological challenge arises from the fact that the multimodal information flows among multiple platforms. Although Instagram is often understood as visual-centered and Twitter text-centered, taking screenshots of tweets and posting them as images comprised a significant portion of top Instagram posts during #SAH. However, the Twitter platform is a different ecosystem that conditions the performance of the tweets (e.g., short length). How Instagram as an ecosystem reconditions such cross-platform content remains a method question that must be explored.
Taking the lessons from #SAH, activists should strategically leverage visual storytelling to strengthen the emotional resonance and mobilizing capacity of campaigns. Since visuals on platforms like Instagram may convey ambiguous or even contradictory meanings when paired with texts, activists should aim for tighter alignment between images and accompanying narratives to minimize interpretive gaps. Storyboards, infographics, or short-form videos with clear beginnings, turning points, and resolutions can help overcome the disruptive effects of plot twists by presenting coherent storylines. Activists should also recognize that ugly feelings such as ambivalence, shame, or exhaustion are not merely obstacles, but diagnostic signals of the limits of existing narratives. Addressing these emotions directly (e.g., by creating spaces for interracial solidarity) can help prevent disengagement when momentum wanes. For policy makers, legislation such as the COVID-19 Hate Crimes Act is not sufficient to address deep, historical, intertwined racial tensions. Anti-Asian discourses and behaviors are widely spread in everyday life and need both the majority race’s and other minority races’ support to counter. Policy makers should provide resources for community-level conversations and collaborations to exchange shared experiences and build mutual trust.
This study is not without its limitations. First, relying on Instagram allows us to conduct the multimodal analysis, but limits us to just one social media platform. It is known that each platform has its own digital ecosystem that may not be shared by other platforms. Although the findings are consistent with existing studies that examined other visual-focused platforms such as TikTok, cross-platform analyses will help us better generalize the findings. Second, #StopAsianHate is a hashtag movement that has its own unique history and issue focus. Although this study demonstrates common ways of constructing affective narratives, it is expected that future research on other movements may discover more nuances. Last, this study approaches multimodal analysis as a mixture of computational and qualitative methods. As artificial intelligence tools develop, researchers can imagine further integration of human and machine intelligence in the analysis of affective narratives. There is much to anticipate with the research methodology.
In conclusion, not all hashtag movements are equal. Although many hashtags are created and made viral, the affective narratives organized through the hashtags contain vastly different meanings and express various affects. As a result, hashtag movements appeal to different types of actors and lead to diverse outcomes. This study suggests going beyond dominant narratives and examining not only competing narratives but also the competitions of meanings and feelings within the narratives. The study also demonstrates that this theoretical query is best achieved through a multimodal analysis, when the gaps among multiple channels and platforms present unique opportunities and challenges to surface the plot twists and ugly feelings that influence the development direction of a social movement.
References
Ahluwalia, M. K., & Pellettiere, L. (2010). Sikh men post-9/11: Misidentification, discrimination, and coping. Asian American Journal of Psychology, 1(4), 303–314. https://doi.org/10.1037/a0022156
Aramendia-Muneta, M. E., Olarte-Pascual, C., & Ollo-López, A. (2021). Key image attributes to elicit likes and comments on Instagram. Journal of Promotion Management, 27(1), 50–76. https://doi.org/10.1080/10496491.2020.1809594
Aydin, Z., Fuess, A., Förster, M., & Sunier, T. (2022). When birds of a feather Instagram together: Debating the image of Islam in echo chambers and through trench warfare on social media. Social Media + Society, 8(3), 1–13. https://doi.org/10.1177/20563051221115211
Bleiker, R., & Hutchison, E. (2008). Fear no more: Emotions and world politics. Review of International Studies, 34(S1), 115–135. https://doi.org/10.1017/S0260210508007821
Bonta, V., Kumaresh, N., & Janardhan, N. (2019). A comprehensive study on lexicon-based approaches for sentiment analysis. Asian Journal of Computer Science and Technology, 8(S2), 1–6. https://doi.org/10.51983/ajcst-2019.8.S2.2037
Braudy, L., & Cohen, M. (Eds.). (2004). Film theory and criticism: Introductory readings (Vol. 1974). Oxford, UK: Oxford University Press.
Brock, A. (2018). Critical technocultural discourse analysis. New Media & Society, 20(3), 1012–1030. https://doi.org/10.1177/1461444816677532
Canevez, R. N., Karabelnik, M., & Winter, J. S. (2022). Police brutality and racial justice narratives through multi-narrative framing: Reporting and commenting on the George Floyd murder on YouTube. Journalism & Mass Communication Quarterly, 99(3), 696–717. https://doi.org/10.1177/10776990221108722
Christensen, H. S. (2011). Political activities on the Internet: Slacktivism or political participation by other means? First Monday, 16(2). https://doi.org/10.5210/fm.v16i2.3336
Clements, L. (2021, March 9). Covid in Wales: Racist incidents “take your breath away”. BBC News. https://www.bbc.com/news/uk-wales-56323775
Clever, L., Schatto-Eckrodt, T., Clever, N. C., & Frischlich, L. (2023). Behind blue skies: A multimodal automated content analysis of Islamic extremist propaganda on Instagram. Social Media + Society, 9(1), 1–14. https://doi.org/10.1177/20563051221150404
Cohn, N. (2013). Visual narrative structure. Cognitive Science, 37(3), 413–452. https://doi.org/10.1111/cogs.12016
Diet Panda [@diet_panda]. (2021, March 18). How the media’s depiction of Asian women harms them [Video]. Instagram. https://www.instagram.com/p/CMiMBh5ni9V/
Dong, C., Liu, W., & Zhang, Y. (2023). Leveraging moral foundations for corporate social advocacy combating anti-Asian racism: A computational approach. Asian Journal of Communication, 33(2), 138–157. https://doi.org/10.1080/01292986.2023.2169944
Dorland, C. P. (1894). Chinese massacre at Los Angeles in 1871. Annual Publication of the Historical Society of Southern California, Los Angeles, 3(2), 22–26. https://doi.org/10.2307/41167579
Dumitrica, D., & Hockin-Boyers, H. (2023). Slideshow activism on Instagram: Constructing the political activist subject. Information, Communication & Society, 26(16), 3318–3336. https://doi.org/10.1080/1369118X.2022.2155487
Ferrara, E. (2020). Bots, elections, and social media: A brief overview. In K. Shu, S. Wang, D. Lee, & H. Liu (Eds.), Disinformation, misinformation, and fake news in social media: Lecture notes in social networks (pp. 95–114). Cham, Switzerland: Springer. https://doi.org/10.1007/978-3-030-42699-6_6
Field, A., Park, C. Y., Theophilo, A., Watson-Daniels, J., & Tsvetkov, Y. (2022). An analysis of emotions and the prominence of positivity in #BlackLivesMatter tweets. Proceedings of the National Academy of Sciences, 119(35), 1–8. https://doi.org/10.1073/pnas.2205767119
Frederick, W. M. (2014). Narrative politics: Stories and collective action. Oxford, UK: Oxford University Press.
Freistein, K., & Gadinger, F. (2022). Performing leadership: International politics through the lens of visual narrative analysis. Political Research Exchange, 4(1), 1–19. https://doi.org/10.1080/2474736X.2022.2124922
Giuntini, F. T., Ruiz, L. P., Kirchner, L. D. F., Passarelli, D. A., Dos Reis, M. D. J. D., Campbell, A. T., & Ueyama, J. (2019). How do I feel? Identifying emotional expressions on Facebook reactions using clustering mechanism. Institute of Electrical and Electronics Engineers Access, 7, 53909–53921. https://doi.org/10.1109/ACCESS.2019.2913136
Guo, J., & Liu, S. (2022). From #BlackLivesMatter to #StopAsianHate: Examining network agenda-setting effects of hashtag activism on Twitter. Social Media + Society, 8(4), 1–12. https://doi.org/10.1177/20563051221146182
Highfield, T., & Leaver, T. (2016). Instagrammatics and digital methods: Studying visual social media, from selfies and GIFs to memes and emoji. Communication Research and Practice, 2(1), 47–62. https://doi.org/10.1080/22041451.2016.1155332
Isnan, M., Elwirehardja, G. N., & Pardamean, B. (2023). Sentiment analysis for TikTok reviews using VADER sentiment and SVM model. Procedia Computer Science, 227, 168–175. https://doi.org/10.1016/j.procs.2023.10.514
Jacques, E. T., Basch, C. H., Fera, J., & Jones II, V. (2023). #StopAsianHate: A content analysis of TikTok videos focused on racial discrimination against Asians and Asian Americans during the COVID-19 pandemic. Dialogues in Health, 2, 1–4. https://doi.org/10.1016/j.dialog.2022.100089
Jasper, J. M. (2019). The emotions of protest. Chicago, IL: University of Chicago Press.
Kabir, M. E., & Ha, L. (2025). Influencers against hate: A comparison of counter speech strategies among South Asian, East Asian, and non Asian American social media influencers. Howard Journal of Communications. Advance online publication. https://doi.org/10.1080/10646175.2025.2494271
Kamp, A., Denson, N., Sharples, R., & Atie, R. (2022). Asian Australians’ experiences of online racism during the COVID-19 pandemic. Social Sciences, 11(5), 227–248. https://doi.org/10.3390/socsci11050227
Kim, J., & Kim, J. W. (2023). How did #StopAsianHate and #BlackLivesMatter react to each other after the Atlanta shootings: An analysis of Twitter hashtag networks. International Journal of Communication, 17, 6096–6119. https://ijoc.org/index.php/ijoc/article/view/21259
Klinger, U., & Svensson, J. (2015). The emergence of network media logic in political communication: A theoretical approach. New Media & Society, 17(8), 1241–1257. https://doi.org/10.1177/1461444814522952
Leaver, T., Highfield, T., & Abidin, C. (2020). Instagram: Visual social media cultures. Hoboken, NJ: John Wiley & Sons.
Lee, C. S., & Jang, A. (2023). Questing for justice on Twitter: Topic modeling of #StopAsianHate discourses in the wake of Atlanta shooting. Crime & Delinquency, 69(13–14), 1–27. https://doi.org/10.1177/00111287211057855
Lee, E. (2015). The making of Asian America: A history. New York, NY: Simon & Schuster.
Lee, J. J., & Lee, J. (2023). #StopAsianHate on TikTok: Asian/American women’s space-making for spearheading counter-narratives and forming an ad hoc Asian community. Social Media + Society, 9(1), 1–11. https://doi.org/10.1177/20563051231157598
Li, M. (2022). Visual social media and Black activism: Exploring how using Instagram influences Black activism orientation and racial identity ideology among Black Americans. Journalism & Mass Communication Quarterly, 99(3), 718–741. https://doi.org/10.1177/10776990221108644
Mohammad, S. M., & Turney, P. D. (2013). Crowdsourcing a word–emotion association lexicon. Computational Intelligence, 29(3), 436–465. https://doi.org/10.1111/j.1467-8640.2012.00460.x
Mundt, M., Ross, K., & Burnett, C. M. (2018). Scaling social movements through social media: The case of Black Lives Matter. Social Media + Society, 4(4), 1–14. https://doi.org/10.1177/2056305118807911
Ng, W. (2001). Japanese American internment during World War II: A history and reference guide. New York, NY: Bloomsbury.
Ngai, S. (2004). Ugly feelings. Cambridge, MA: Harvard University Press.
Ofori-Parku, S. S., & Moscato, D. (2018). Hashtag activism as a form of political action: A qualitative analysis of the #BringBackOurGirls campaign in Nigerian, UK, and U.S. press. International Journal of Communication, 12, 2480–2502. https://ijoc.org/index.php/ijoc/article/view/8068/2375
Oh, D. C. (2023). Reflexive racialization and discursive affect with the #VeryAsian hashtag. Asian Journal of Communication, 33(2), 105–120. https://doi.org/10.1080/01292986.2023.2169723
Pacquiao, M. [@MannyPacquiao]. (2021). We have one color in our Blood! Stop discriminating: LOVE AND PEACE TO EVERYONE!! #StopAsianHate [Photograph]. Instagram. https://www.instagram.com/p/CNHIk3rle_t/
Papacharissi, Z. (2016). Affective publics and structures of storytelling: Sentiment, events and mediality. Information, Communication & Society, 19(3), 307–324. https://doi.org/10.1080/1369118X.2015.1109697
Peng, Y. (2018). Same candidates, different faces: Uncovering media bias in visual portrayals of presidential candidates with computer vision. Journal of Communication, 68(5), 920–941. https://doi.org/10.1093/joc/jqy041
Polkinghorne, D. (1988). Narrative knowing and the human sciences. Albany, NY: State University of New York Press.
prageru [@PragerU]. (2021, March 23). 🔥 Don’t let yourself be so easily manipulated [Photograph]. Instagram. https://www.instagram.com/p/CMvzV-ALTwp/
Rodriguez, L., & Dimitrova, D. (2011). The levels of visual framing. Journal of Visual Literacy, 30(1), 48–65. https://doi.org/10.1080/23796529.2011.11674684
Shahin, S., & Hou, M. (2025). #StopAsianHate as hashtag activism: Provocateurs, celebrities, and fan practices of collective action against racism. Social Media + Society, 11(1), 1–15. https://doi.org/10.1177/20563051241309701
Sheikh, M. (2022, October 24). How to go viral on Instagram: Proven strategies for your brand. Sprout Social. https://sproutsocial.com/insights/how-to-go-viral-on-instagram/
Snow, D. A., Rochford, E. B., Worden, S. K., & Benford, R. D. (1986). Frame alignment processes, micromobilization, and movement participation. American Sociological Review, 51(4), 464–481. https://doi.org/10.2307/2095581
Somers, M. R. (1994). The narrative constitution of identity: A relational and network approach. Theory and Society, 23, 605–649. https://doi.org/10.1007/BF00992905
Stieglitz, S., & Dang-Xuan, L. (2013). Emotions and information diffusion in social media—Sentiment of microblogs and sharing behavior. Journal of Management Information Systems, 29(4), 217–248. https://doi.org/10.2753/MIS0742-1222290408
Stop AAPI Hate. (2022, July). Two years and thousands of voices: What community-generated data tells us about anti-AAPI hate. https://stopaapihate.org/wp-content/uploads/2022/07/Stop-AAPI-Hate-Year-2-Report.pdf
Sumner, E. M., Hayes, R. A., Carr, C. T., & Wohn, D. Y. (2020). Assessing the cognitive and communicative properties of Facebook reactions and likes as lightweight feedback cues. First Monday, 25(2). https://doi.org/10.5210/fm.v25i2.9621
Tchen, J. K. W., & Yeats, D. (2014). Yellow peril!: An archive of anti-Asian fear. London, UK: Verso Books.
Trillò, T., Scharlach, R., Hallinan, B., Kim, B., Mizoroki, S., Frosh, P., & Shifman, L. (2021). What does #Freedom look like? Instagram and the visual imagination of values. Journal of Communication, 71(6), 875–897. https://doi.org/10.1093/joc/jqab021
Turvy, A. (2023). Potholes and power: A multimodal critical discourse analysis of “Look at This F*ckin’ Street” on Instagram. Social Media + Society, 9(3), 1–18. https://doi.org/10.1177/20563051231194580
Van Atteveldt, W., Van der Velden, M. A., & Boukes, M. (2021). The validity of sentiment analysis: Comparing manual annotation, crowd-coding, dictionary approaches, and machine learning algorithms. Communication Methods and Measures, 15(2), 121–140. https://doi.org/10.1080/19312458.2018.1458084
Wang, J. [@jessicawang]. (2021, March 19). Angry? Shattered? Sad? Exhausted? [Photograph]. Instagram. https://www.instagram.com/p/CMktsQBjWul/
Wang, R., Zhang, W., & Shin, J. (2025). Mapping the evolving networks of the #StopAsianHate movement on Twitter: The role of serial participants in digital activism. Information, Communication & Society. Advance online publication. https://doi.org/10.1080/1369118X.2025.2571407
Wang, R., Zhou, A., & Kinneer, T. H. (2024). Moral framing and issue-based framing of #StopAsianHate campaigns on Twitter. Chinese Journal of Communication, 17(1), 42–60. https://doi.org/10.1080/17544750.2023.2218646
Wear the Peace [@wearthepeace]. (2021, March 23). Asian hate has been a problem [Video]. Instagram. https://www.instagram.com/p/CMu26Wmh5ZG/
Wu, J. (2002). Teaching who killed Vincent Chin?—1991 and 2001. Amerasia Journal, 28(3), 13–23. https://doi.org/10.17953/amer.28.3.tx7721541438q58x
Xie, C., Liu, P., & Cheng, Y. (2023). Praxis, hashtag activism, and social justice: A content analysis of #StopAsianHate narratives. Asian Journal of Communication, 33(2), 121–137. https://doi.org/10.1080/01292986.2023.2180529
Xiong, Y., Cho, M., & Boatwright, B. (2019). Hashtag activism and message frames among social movement organizations: Semantic network analysis and thematic analysis of Twitter during the #MeToo movement. Public Relations Review, 45(1), 10–23. https://doi.org/10.1016/j.pubrev.2018.10.014
Yang, G. (2016). Narrative agency in hashtag activism: The case of #BlackLivesMatter. Media and Communication, 4(4), 13–17. https://doi.org/10.17645/mac.v4i4.692
Zulli, D. (2020). Evaluating hashtag activism: Examining the theoretical challenges and opportunities of #BlackLivesMatter. Participations, 17(1), 197–215. https://www.participations.org/17-01-12-zulli.pdf
Copyright © 2026 (Weiyu Zhang).
Licensed under the Creative Commons Attribution Non-commercial No Derivatives (by-nc-nd).
https://doi.org/10.65476/vjdz2s05