‘Not With Us - Against Us’: ‘We’ vs. ‘They’ in the Transformation of Black Lives Matter Participants’ Collective Identities in Online Interactions with All Lives Matter, 2013-2014

Full Text

Introduction In the realm of social movements, language serves as a key mechanism for articulating and shaping collective identity. Pronouns, often perceived as simple linguistic markers, are in fact loaded with ideological significance. The pronouns “we” and “they” serve as markers of inclusion and exclusion, solidarity and opposition, often determining who belongs to a collective and who is perceived as the enemy or the other. In digital protest discourse, where communication unfolds rapidly and interactively, these pronouns become key instruments in negotiating symbolic boundaries and articulating group membership. This article explores how “we” and “they” function in the construction and contestation of collective identity in the online discourse of the Black Lives Matter (BLM) and All Lives Matter (ALM) movements during the years 2013-2014. By analyzing the frequency, referents, and collocational patterns of these pronouns in tweets posted during this period, the study traces how group boundaries were drawn, redefined, and rhetorically charged in response to shifting political dynamics. BLM’s rise in 2013 marked a transformative moment in the landscape of American racial justice movements. Initially sparked by the acquittal of George Zimmerman in the shooting death of Trayvon Martin, BLM gained momentum on social media platforms, where it articulated a collective identity centered on justice, accountability, and the affirmation of Black lives.[7] In contrast, the ALM movement arose as a counter- slogan that offered a universalist framing of equality but was widely perceived as diluting or deflecting the specific demands of BLM [Goodman et al. 2024]. Before the rise of counter- protests, BLM discourse constructed “they” primarily in reference to institutional actors - police, law enforcement, and political authority. However, with the emergence of ALM, this symbolic structure became more complex. “They” began to include not only state actors but also ideological adversaries from within civil society. The rhetorical opposition between these movements was not merely ideological; it was fundamentally linguistic. The framing of identity - who is “us” and who is “them” - became a central battleground, with both sides asserting moral and political legitimacy through language. Collective identity plays a crucial role in protest mobilization. As Melucci [1996] argues, it is not a fixed trait but a dynamic process - a constantly evolving sense of belonging shaped through shared experiences and sustained collective action. In protest movements, this identity is often formed in contrast to a perceived out- group or “other,” frequently represented by institutional power [Gulevich 2019]. The emergence of counterprotest movements introduces further complexity to this process. Counterprotests - defined as reactive mobilizations opposing a protest movement’s goals - typically arise in ideological and spatial proximity to the movements they contest. They are aimed at preserving the status quo asserting their own collective identities in opposition to those of the protestors [Wood 2020; Inata 2021]. As Zald and Useem [1987] describe, protest and counter- protest exist in a “mobilization- demobilization tango,” where both sides respond to shared social triggers, reshape their goals, and renegotiate collective identities under pressure. This dynamic process of identity construction is closely tied to discourse. According to Melucci [1996] and Wodak [2012], identity is formed not only through shared beliefs but also through the symbolic resources movements draw upon - language, stories, and rhetorical devices. Pronouns such as “we” and “they” are among the most frequent and ideologically potent elements of protest discourse. As Gordon et al. [1993] and Pennebaker et al. [2003] have shown, pronoun use is a reliable marker of group affiliation (in- group), boundary- setting, and perceived legitimacy. The social identity model of collective action (SIMCA) provides a useful framework for understanding these dynamics. It suggests that perceived injustice, group efficacy, and identity salience are core psychological drivers of protest engagement [van Zomeren et al. 2008; Thomas et al. 2012]. Recent psychological studies confirm that the use of first- person plural pronouns in online discourse reflects heightened group identity salience and predicts engagement in collective action [Adam- Troian et al. 2021]. However, most existing studies focus on pronoun frequency alone, without examining how the referents of these pronouns shift over time, particularly in moments of ideological contestation with counter- movements. This study extends previous research that employed BERT-based topic modeling to examine the thematic transformation of BLM online discourse following the emergence of the ALM countermovement [Koncha 2025]. That analysis revealed the introduction of a second prominent out- group - ALM participants - alongside the original institutional target, indicating a shift in the discursive boundaries of collective identity. Building on that foundation, the present study adopts a more granular linguistic approach by focusing on discursive microstructures - specifically, the use of first- and third- person plural pronouns and their collocational environments. Through collocation analysis of PMI (Pointwise mutual information)-scored trigrams and manual coding of referents, this research investigates how the semantic roles of “we” and “they” evolved across 2013 and 2014 in response to the discursive presence of ALM. The analysis examines both the frequency and semantic framing of pronouns, enabling a comparative assessment of referential change before and after the emergence of ALM. The aim is to trace how collective identity is linguistically constructed, destabilized, and reconstituted in response to ideological opposition within protest- counter-protest dynamics. Data This study employs publicly available posts authored by participants in the Black Lives Matter (BLM) movement on the social media platform X (formerly Twitter[8]) as its primary data source. Such user-generated content is particularly well suited to an investigation of how protestors themselves negotiate and renegotiate collective identities in response to competing discourses, given that neither BLM nor its countermovement, ALM, operate as formally structured organizations with stable leadership or codified manifestos. Instead, mobilization tends to occur at the local level and often originates with the spontaneous utterance of a single participant. Our temporal scope extends from July 12, 2013 - the date on which the hashtag #BlackLivesMatter first appears in the Twitter archive - through December 31, 2014, by which time the #AllLivesMatter counter-hashtag had achieved sufficient diffusion to serve as a point of comparison. Data collection was conducted via the Apify web-scraping platform,[9] necessitated by restrictions on individual access to Twitter’s official API. To mitigate temporal imbalances in posting volume, a random sample of 100,000 tweets was drawn, ensuring equal selection probability across the study period. Sampling criteria encompassed the primary hashtags #BlackLivesMatter (2013-2014) and #AllLivesMatter (2014 only), as well as a set of co-occurring hashtags identified by the Pew Research Center - namely #icantbreathe, #ferguson, #ericgarner, #blacktwitter, #mikebrown, #tamirrice, and #shutitdown - which collectively reference high-profile instances of police violence. Prior to performing collocation analysis, all non-integral hashtags - those not embedded as part of the tweet’s lexical content - were removed. Prior to analysis, the textual data underwent a series of preprocessing steps: all text was converted to lowercase; punctuation marks and non- alphanumeric symbols were removed to reduce noise in the collocation analysis. Stopwords were removed with one important exception: prepositions and pronouns were retained, as they play a critical role in the discursive construction of group boundaries and were central to the focus of this study. Methods To examine the discursive construction of collective identity in protest discourse, a trigram- based collocation analysis was conducted on the sampled tweets, focusing on sequences that include first- person plural pronouns (“we,” “our,” “us”) and third- person plural pronouns (“they,” “their,” “them”). Trigram analysis, which captures three- word sequences, allows for a more precise understanding of the linguistic and rhetorical contexts in which pronouns are embedded, thereby revealing how in-g roup and out- group identities are constructed and communicated. Prior to analysis, the remaining text was then preprocessed through tokenization, and all trigrams containing the target pronouns in any position were extracted. Functional words were retained in order to preserve meaningful grammatical constructions and reflect natural discourse patterns. To measure the strength of association within these trigrams (PMI) was calculated for each trigram containing a pronoun. PMI is a widely used statistical measure in corpus linguistics that quantifies the extent to which words co- occur more frequently than would be expected by chance, based on their individual frequencies. The PMI of a trigram (w1, w2, w3) was computed using the formula: P w w w PMIw w w1, 2, 3 log2 1, 2, 3 P w1P w 2P w 3 where P(w1, w2, w3) is the observed probability of the trigram occurring in the corpus, and P(w1) * P(w2) * P(w3) are the individual probabilities of each word in the corpus. Higher PMI scores indicate stronger, non- random associations among the words in the trigram. To ensure reliability and mitigate noise, only trigrams with a minimum frequency of five occurrences were included in the final analysis. The resulting PMI scores were used to identify the most strongly associated trigrams for each pronoun, allowing us to trace shifts in semantic framing and rhetorical strategy over time. This method enables a fine- grained analysis of how collective identity is linguistically encoded in protest discourse, offering empirical insight into the evolving narratives and oppositional boundaries constructed by movement participants. Results Table 1 presents the frequency and proportional distribution of first- and third- person plural pronouns in tweets associated with the Black Lives Matter (BLM) movement across 2013-2014. The first- person plural set includes “we,” “our,” and “us,” while the third- person plural set includes “they,” “their,” and “them.” In both years, the pronoun “we” is the most frequently used among first- person forms, accounting for 60.49% of such pronouns in 2013 (n = 2,944) and 56.78% in 2014 (n = 7,267). The use of “us” increased notably in 2014, rising from 18.31% in 2013 (n = 891) to 25.91% (n = 3,316), while “our” declined proportionally from 21.20% to 17.31%. Overall, the total number of first- person plural pronouns rose sharply from 4,867 in 2013 to 12,798 in 2014, reflecting a strengthened emphasis on collective identity and in- group cohesion over time. Among third- person plural pronouns, “they” remains the dominant form, comprising 58.76% in 2013 (n = 3,428) and 54.24% in 2014 (n = 4,480). The relative use of “their” increased substantially - from 21.94% to 27.98% - suggesting a growing tendency to attribute possession, actions, or values to the out- group. In contrast, “them” remained relatively stable across both years. The total use of third- person plural pronouns decreased in proportion to first- person pronouns, from 5,834 in 2013 to 8,260 in 2014. These trends indicate a discursive shift toward greater internal cohesion (“we” discourse) in 2014, while references to external groups (“they” discourse) became somewhat less dominant in relative terms - potentially reflecting the consolidation of in- group identity and the emergence of new forms of opposition, such as ALM. Table 2 presents the frequency with which specific in- group referents appear in tweets that also contain the pronoun “we” in the year 2013. The referents include terms explicitly related to the movement (“blm,” “blacklivesmatter,” “black lives matter”), broader racial identifiers (“blacks”), and participant descriptors (“protesters”). Table 1 The Frequency of First- and Third- Person Plural Pronouns in BLM Tweets in 2013-2014 Pronoun Frequency_2013 Frequency_2014 We 2.944 (60.49%) 7267 (56.78%) Our 1.032 (21.20%) 2215 (17.31%) Us 891 (18.31%) 3316 (25.91%) Total 4.867 12.798 They 3.428 (58.76%) 4480 (54.24%) Their 1.280 (21.94%) 2311 (27.98%) Them 1.126 (19.30%) 1469 (17.78%) Total 5.834 8.260 Source: compiled by V. Concha. The term “blacks” is the most frequently occurring referent, appearing in 6,287 tweets, of which 580 also include the pronoun “we.” This suggests that early in the movement, in- group identity was primarily articulated in terms of racial belonging rather than affiliation with the movement’s formal name. “Blacklivesmatter” appears in 952 tweets (118 with “we”), and “blm” in 1,036 tweets (100 with “we”), indicating some early adoption of the movement’s label, though its usage is not yet dominant. The full phrase “black lives matter” is less common (67 tweets, 16 with “we”), likely reflecting emerging variation in how the movement was referenced. Finally, the word “protesters” appears in 243 tweets, with 18 of those also containing “we,” suggesting that the protestor identity was not yet central to the way participants framed their collective self in 2013. Overall, this table reflects a relatively diffuse in- group identity, grounded more in racial and demographic categorization than in formalized or collective political identity. Table 2 In- group Referents of “We” in BLM Tweets (2013) Referent Number of tweets containing the exact referent Total times “we” appears within those tweets Blm 1.036 (12.07%) 100 (12.02%) Blacklivesmatter 952 (11.09%) 118 (14.18%) Black lives matter 67 (0.78%) 16 (1.92%) Blacks 6.287 (73.23%) 580 (69.71%) Protesters 243 (2.83%) 18 (2.16%) Total 8.585 832 Source: compiled by V. Concha. Table 3 presents the number of tweets in which key in- group referents co- occur with the pronoun “we” during 2014. These referents include both movement identifiers and broader social categories used by BLM participants to articulate collective identity. Table 3 In- group Referents of “We” in BLM Tweets (2014) Referent Number of tweets containing the exact referent Total times “we” appears within those tweets Blm 14 (0.05%) 2 (0.09%) Blacklivesmatter 22.425 (76.64%) 1.946 (87.54%) Black lives matter 1.702 (3.66%) 138 (6.21%) Blacks 1.788 (6.11%) 90 (4.05%) Protesters 3.332 (11.39%) 47 (2.11%) Total 29.261 2.223 Source: compiled by V. Concha. The most prominent referent is “blacklivesmatter,” appearing in 22,425 tweets of which 1,946 also contain the pronoun “we.” This represents a dramatic increase compared to 2013 (see Table 3) and indicates that the full movement name - rather than its abbreviation or broader racial terms - became the primary label around which collective identity was organized. The phrase “black lives matter” appears in 1,702 tweets (138 with “we”), while “blm” occurs only 14 times (2 with “we”), suggesting that the informal abbreviation lost salience as the movement’s discursive identity consolidated. The racial referent “blacks” is mentioned in 1,788 tweets, with 90 co- occurrences of “we,” a marked decrease from the previous year. The term “protesters” appears in 3,332 tweets (47 with “we”), reflecting an increasing identification with protester status rather than with racial identity alone. Together, these data indicate a clear shift in in- group framing: from a broader racial identity in 2013 to a more politicized, movement- centered identity in 2014. The growing prevalence of “blacklivesmatter” as an in- group referent suggests that participants increasingly defined themselves not only by shared demographic characteristics, but by active membership in a collective struggle articulated through the movement’s formal discourse. This supports the broader finding that BLM’s identity construction became more unified, purposeful, and movement- specific over time. Table 4 details the out- group referents most commonly associated with the third- person plural pronoun “they” in tweets authored by Black Lives Matter participants during the year 2013. For each referent, the table provides two metrics: the number of tweets in which the exact referent appears, and the total number of times “they” occurs within those tweets. The most frequent referent is “police,” appearing in 10,398 tweets, within which the pronoun “they” appears 729 times. Other law enforcement- related terms also feature prominently: “law enforcement” (6,412 tweets / 141 “they”), “cop” (250/14), “cops” (299/39), “officer” (357/31), and “officers” (262/34). Additional institutional terms such as “authority” (483/13) and “authorities” (816/11) are also present but appear with lower frequency. This distribution confirms that in 2013, the image of “the other” constructed in BLM discourse was overwhelmingly institutional. The dominant referents are all agents of the state - particularly policing institutions - indicating that the protest movement’s collective identity was formed in opposition to systemic structures of authority, especially those perceived as enforcers of racial injustice. Table 4 Out- group Referents of “They” in BLM Tweets (2013) Referent Number of tweets containing the exact referent Total times “they” appears within those tweets Police 10.398 (53.94%) 729 (72.04%) Cop 250 (1.29%) 14 (1.38%) Cops 299 (1.55%) 39 (3.85%) Officer 357 (1.85%) 31 (3.06%) Officers 262 (1.36%) 34 (3.36%) Law enforcement 6.412 (33.36%) 141 (13.93%) Authority 483 (2.51%) 13 (1.28%) Authorities 816 (4.23%) 11 (1.09%) Total 19.277 1.012 Source: compiled by V. Concha. Table 5 outlines the referents associated with the third- person plural pronoun “they” in Black Lives Matter tweets from 2014. It includes both traditional institutional actors and, significantly, members or symbols of the emerging All Lives Matter (ALM) counter- protest movement. For each referent, the table reports the number of tweets containing the term, along with the total number of times “they” appears within those tweets. As in 2013, institutional actors remain prominent referents: “police” appears in 24,284 tweets, with 1,075 occurrences of “they”; “law enforcement” appears in 9,142 tweets (228 “they”), and other related terms such as “cop” (2,168/93), “cops” (5,000/414), and “officers” (2,806/117) also feature prominently. These figures indicate the sustained centrality of law enforcement as a perceived adversarial force. However, a significant development in 2014 is the appearance of new referents corresponding to the countermovement. The terms “alm” (5,814 tweets / 102 “they”), “all lives matter” (1,107/71), “alllivesmatter” (3,417/183), and “counterprotesters” (198/15) collectively signal the emergence of a discursive out- group distinct from formal authority. These counter- movement actors are represented as ideological opponents of BLM, thereby expanding the conceptual boundaries of “the other.” This diversification of referents marks a significant shift in the collective identity narrative. While in 2013 the out-g roup was composed almost exclusively of institutional power structures, by 2014 it had broadened to include societal actors who actively contested the movement’s message. This supports the study’s central assumption: the emergence of counter- protest movements reconfigured the discursive boundaries of opposition in BLM discourse, transforming “they” from a reference to institutional authority alone into a broader category encompassing both structural and ideological adversaries. Table 5 Out- group Referents of “They” in BLM Tweets (2014) Referent Number of tweets containing the exact referent Total times “they” appears within those tweets Police 24.284 (51.13%) 1075 (55.56%) Cop 2.168 (4.56%) 93 (4.81%) Cops 5.000 (10.53%) 414 (21.40%) Officer 935 (1.97%) 30 (1.55%) Officers 2.806 (5.91%) 117 (6.05%) Law enforcement 9.142 (19.25%) 228 (11.78%) Authority 1.237 (2.60%) 27 (1.40%) Authorities 1.925 (4.05%) 57 (2.95%) Total 47.497 1.935 Alm 5.814 (55.18%) 102 (27.49%) All lives matter 1.107 (10.51%) 71 (19.14%) Alllivesmatter 3.417 (32.43%) 183 (49.33%) Counterprotesters 198 (1.88%) 15 (4.04%) Total 10.536 371 Source: compiled by V. Concha. Table 6 displays the most salient three- word expressions (trigrams) containing first- and third- person plural pronouns in tweets from 2013, ranked by Pointwise Mutual Information (PMI). The trigrams containing the pronoun “we” emphasize collective resistance and moral stance, with high-PMI phrases such as “we take stand” (PMI: 28.6), “we become silent” (PMI: 22.0), and “we must fight” (PMI: 16.9). These expressions suggest a narrative of active engagement and solidarity in the face of perceived injustice. Similarly, trigrams with “our” like “plaguing our black” (PMI: 20.8) and “our black community” (PMI: 18.8) foreground shared racial identity and communal victimization, reinforcing the in- group boundary. In contrast, trigrams containing “they” and “their” in 2013 often depict the out- group in passive, indirect, or ambiguous terms. Examples such as “they are victims” (PMI: 15.2), “they just concealing” (PMI: 14.8), and “authorities say they” (PMI: 17.9) reflect a discourse that does not consistently frame “they” as institutional antagonists, but rather includes a mix of referents - some potentially sympathetic or undefined. Similarly, “their” appears in constructions like “blaming their governors” (PMI: 25.9) and “change their voting” (PMI: 20.1), which suggest political critique but do not clearly target a specific institutional actor. Overall, these patterns indicate a less sharply defined out- group at this stage of the movement, with “they” encompassing both institutional and broader societal references, often framed in less confrontational terms. These findings suggest that in 2013, BLM discourse was primarily centered on affirming in- group solidarity and moral agency, while representations of the out- group remained diffuse and relatively subdued. The pronoun “we” functioned as a vehicle for mobilization and collective self- definition, whereas “they” lacked a singular referent, indicating that the movement’s rhetorical confrontation with external actors had not yet fully crystallized. Table 6 High-PMI Trigrams Containing First- and Third- Person Plural Pronouns (2013) Pronoun Trigram Frequency PMI We We take stand 42 28.616 We become silent 68 21.981 We must fight 7 16.879 We do injustice 7 16.581 Brutality we gotta 42 15.629 Us For following us 8 15.161 One of us 5 10.205 Us police brutality 6 8.755 Of us blacks 7 8.694 Join us on 12 5.315 Our Plaguing our black 57 20.804 Our black community 57 18.809 Custody are our 5 16.900 Our lives matter 7 16.83318778 Racism in our 5 7.786 They They are victims 18 15.192 Authorities say they 5 17.914 They just concealing 13 14.754 Say they arrested 5 14.069 They dont know 5 13.627 Their Blaming their governors 12 25.899 Change their voting 6 20.067 Racism since their 9 13.442 As their role 7 17.466 Do their job 5 15.226 Source: compiled by V. Concha. Table 7 provides the trigram collocations for 2014, again focusing on sequences that include plural pronouns. Trigrams with “we” reflect a stronger and more cohesive collective identity, often drawing on historical or movement- oriented language. Notable examples include “we shall overcome” (PMI: 21.5), “united we stand” (PMI: 19.1), and “movement continues we” (PMI: 16.0). These expressions indicate a shift from reactive indignation to proactive affirmation and legacy- building. The phrase “we savages blacklivesmatter” (PMI: 15.0), though ambiguous, points to a possible reclamation or critique of racialized stereotypes, suggesting increasing discursive complexity. Table 7 High-PMI Trigrams Containing First- and Third- Person Plural Pronouns (2014) Pronoun Trigram Frequency PMI We We shall overcome 5 21.518 We revolt simply 8 21.449 We planted evidence 10 20.592 United we stand 289 19.119 We needed change 19 17.023 We savages blacklivesmatter 37 15.009 Movement continues we 11 15.966 Us Tweet blasting us 16 21.564 Join us tomorrow 5 17.327 Us vs them 16 15.971 Views on us 31 15.418 Held against us 8 15.166 Our Sign our petition 6 20.478 Our streets march 6 15.588 Our justice system 11 14.863 Together for our 8 12.402 Of our skin 8 11.606 They Rolling they hating 47 24.621 Rights they enjoy 17 18.932 Racist they seem 78 17.872 Whites believe they 44 17.634 They don care 25 14.436 Their Their patrol car 8 21.436 Abuse their power 6 17.501 Blood on their 26 14.931 Their bullets icantbreathe 5 14.329 Their icantbreathe commentary 6 13.683 Source: compiled by V. Concha. In contrast, trigrams with “they” in 2014 show a clear movement toward direct confrontation and ideological differentiation. High-PMI examples include “rolling they hating” (PMI: 24.6), “racist they seem” (PMI: 17.9), and “whites believe they” (PMI: 17.6). These expressions attribute hostile intent and racial bias to the out- group, which by this point includes both institutional actors and members of the ALM countermovement. “Their” appears in combative contexts as well, such as “abuse their power” and “blood on their,” reinforcing the accusatory tone. The trigram collocation analysis reveals a clear shift in the discursive construction of collective identity and opposition within BLM tweets from 2013 to 2014. In 2013, trigrams with “we” emphasized collective moral stance and solidarity, while “they” and “their” referred to a loosely defined out- group, often portrayed in passive or ambiguous terms. This suggests an early- stage movement focused more on internal cohesion than direct confrontation. By 2014, however, “we” trigrams reflected a more unified and historically rooted identity, while “they” trigrams became more explicitly adversarial, attributing hostility and racism to both institutional actors and ALM participants. This evolution indicates a growing polarization in discourse and a redefinition of out- group boundaries in response to counter- movement activity. Discussion This study set out to examine how the pronouns “we” and “they” were employed in BLM discourse during the formative years of the movement, and how their semantic roles and referents shifted following the emergence of the ALM counter- protest. Drawing on the collective discourse- analytical approaches to collective identity construction [Melucci 1996; Wodak 2012], the analysis demonstrates that the discursive construction of collective identity in digital protest discourse is not static but evolves dynamically in response to external ideological pressures. In 2013, the pronoun “we” was predominantly associated with messages of moral agency, resistance, and internal solidarity. High-PMI trigrams such as “we take stand” and “we must fight” reflect an early- stage identity rooted in a shared sense of injustice. At the same time, the referents of “we” were diffuse - more strongly linked to racial identifiers such as “blacks” than to the formal name of the movement. This supports Melucci’s [1996] view of collective identity as a processual and emergent phenomenon, shaped by shared experiences rather than pre- established categories. The out- group in 2013 was primarily institutional. The pronoun “they” was largely associated with references to police, law enforcement, and authority figures - aligning with prior findings that social movements often define themselves in opposition to institutional power. However, the trigrams associated with “they” during this period often carried ambiguous or passive connotations (e.g., “they are victims,” “they just concealing”), suggesting that antagonism had not yet crystallized into a direct, personalized confrontation. This aligns with the early stages of protest mobilization, where in- group identity formation precedes the clear delineation of ideological enemies [Thomas et al. 2012]. By 2014, this discursive landscape had significantly changed. The use of “we” became more closely tied to the formal identity of the movement, with a dramatic rise in co- occurrence with the term “blacklivesmatter.” Trigram data reflect a stronger, more cohesive collective identity framed through historical references and political continuity (e.g., “we shall overcome,” “united we stand”). This suggests a transition from reactive protest to proactive self- definition, consistent with the SIMCA model’s emphasis on identity salience and perceived group efficacy as motivators for sustained mobilization [van Zomeren et al. 2008; Adam- Troian et al. 2021]. Concurrently, the referents of “they” expanded. While institutional actors remained prominent, a significant portion of tweets now associated “they” with members of the ALM counter- movement, including terms like “alllivesmatter,” “alm,” and “counterprotesters.” This diversification supports previous research on counter- protest as a catalyst for identity renegotiation [Zald and Useem 1987; Wood 2020], introducing an ideologically distinct adversary into the protest narrative. Trigrams in 2014 such as “racist they seem” and “abuse their power” illustrate a discursive shift toward direct accusation and ideological opposition, reflecting the dual nature of the out- group: both institutional and societ al. Сonclusion In sum, the linguistic patterns observed in this study reflect a clear transformation of collective identity within BLM discourse. The shift from a singular focus on institutional “others” to a dual adversarial framing that includes public counter- protesters illustrates the dynamic nature of identity construction in response to perceived threat and opposition. Language - especially pronouns - served both as a mirror of this transformation and as a mechanism for reinforcing group cohesion and oppositional clarity. This study contributes to our understanding of protest- counter-protest dynamics by highlighting the discursive mechanisms through which collective identities are reinforced, challenged, and redefined. It highlights the importance of micro- level linguistic analysis in capturing the evolving boundaries of group membership and ideological conflict in contemporary movements. Future research might build on this work by analyzing the ALM discourse in parallel or exploring these dynamics in later protest cycles.

About the authors

Valeriya Koncha

HSE University

Author for correspondence.
Email: vkoncha@hse.ru
ORCID iD: 0000-0001-6588-6752

Postgraduate student, Lecturer at the Department of Politics and Management, Junior Researcher at the International Center for the Study of Institutions and Development

Moscow, Russian Federation

References

Supplementary files