2.1 Digital Activism: History and Contextualization

2.1.1 Labels and Definitions

Digital activism, also referred to as clicktivism and hashtag activism, has become increasingly popular especially with online communication gaining a central role in our society. The term hashtag activism was first observed in 2011 in a news article, and was used to describe the increasing use of instances of activism that employed by a hashtag (Reference Goswami, Bala, Kaur and RastogiGoswami, 2018). Reference Jackson, Bailey and WellesJackson et al. (2020:38–39) posit “that this online activism leads to material effects in the digital and physical sphere,” so for this reason it deserves to be recognized with a status of its own and to be given as much value as more traditional forms of mobilization. It is also notable that while the label clicktivism derives from an action, clicking, that is generally associated with the use of the computer, hashtag activism delimits the scope of this practice to the use of the hashtag symbol (#). Both these labels, used as synonyms of digital activism, are constraining, as will become evident in the following paragraphs.

Scholars interested in digital activism have yet to agree upon an exact definition of this phenomenon, as thoroughly discussed by Castillo-Esparcia et al. (2023). Nonetheless, the scholars seem to align with Reference ÖzkulaÖzkula’s (2021) view which suggests that digital activism should be considered “on the basis of its practices and not as a phenomenon originating from its relations with a specific technology” (Castillo-Esparcia et al., 2023). In truth, the variety of labels attached to this practice hints at the fact that the descriptor “digital activism” should be considered more as an umbrella term, given that this label does not only apply to activism done through the use of contemporary social media or computers, but more generally, it can be extended to activism done through all types of technological devices, from mobile phones to social networking sites. Figure 1 tries to summarize the different labels that are discussed in this section. In addition to using online and open-source applications specifically dedicated to activism, a large number of digital activists rely on already existing commercial platforms like Facebook, Twitter, Blogger and YouTube (Reference JoyceJoyce, 2010: 3–4), and more recently applications like WhatsApp, Instagram, TikTok or Snapchat. Early scholarship reflecting on the role of the internet in social movements focused on a variety of different types of production infrastructures such as bulletin boards, email, listservs, websites and electronic forums, that today might seem obsolete, or even be unknown by a big part of society. These, in fact, were the preferred platforms used by the first mobilizations that utilized an online platform, like the Zapatistas or the World Trade Organization protesters in Seattle (Reference Jackson and CloudJackson, 2018:7), as will be discussed in more detail later in this section. As a point in fact, Reference Jackson and CloudJackson (2018:8) posits that digital activism “does not and cannot exist in a bubble but rather responds to the larger media and activist ecologies of specific historical and technological moments.” Among the variety of digital activism instances, some have been explored in more detail than others. Mobile phone activism (Reference Cullum and JoyceCullum, 2010), for example, due to the privacy of the technology, presents several ethical impediments that make it a difficult domain for analysis from a linguist’s perspective. One among the many difficulties is the unwillingness of the users to provide the researcher with a sample of the – often very private – conversations. Similarly, ethical issues can be encountered in synchronous video-mediated oral conversations, such as Skype or Facetime calls, or the more recently popular Zoom and Teams calls. This type of data should not be collected without the users’ agreement and knowledge of the research purposes (Reference Xie, Yus, Haugh, Kádár and TerkourafiXie and Yus, 2021: 457). This adds a further layer of complication to the data collection, as the researcher might need to rely directly on the participants for the collection (for example, if the researcher cannot join the meeting themselves and the participants have to take on the responsibility of recording and passing on the registration).

Figure 1 Labels related to digital activism.

In line with the idea that activism online can be done in a variety of ways, Reference JoyceJoyce (2010) proposes “digital activism” as the most exhaustive and inclusive definition for this type of practice. This label can be seen as such because it not only includes an array of infrastructures or environments in which digital activism can take place, but also different means by which activism is carried out in the digital world. Another label that is frequently used as a synonym of digital activism, for example, is cyberactivism, which, by only referring to activism done on the internet, excludes the use of SMS (acronym for “short message service,” which sends messages by mobile phone without using an internet connection) or other instant messaging platforms.

The definition of digital activism comes in many forms and has been attempted by several scholars. Reference JoyceJoyce (2010) provides one that seems general enough to include all the range of activities and behaviors that can fall under the umbrella term: “digital activism are ‘practices’ – habitual activities that occur within a particular context and have certain effects” (Reference JoyceJoyce, 2010:2). To be noted here is the choice of words made by Joyce in defining digital activism as a series of practices. Reference JoyceJoyce (2010), in fact, suggests that this definition can include a variety of different behaviors: it can be activists’ campaigns with a social or political goal; it can be the spread of news stories that contain updates, images, or videos of an ongoing in-presence protest; or it can be the personal experience of joining or supporting someone else’s campaign via social media or other platforms. Another attempt at defining digital activism can be found in the scholarship of Athina Reference KaratzogianniKaratzogianni (2015:1) who describes digital activism as “political participation, activities and protests organized in digital networks beyond representational politics.” All in all, scholars seem to agree that there are two fundamental aspects to digital activism; one, that these actions must be oriented toward some sort of political or social change; and two, that they need to be mediated by some kind of technology. Technology and social change are also at the core of the notion of data activism, a label suggested by Reference Milan and Van-der-VeldenMilan and Van-der-Velden (2016), which also falls under the overarching definition of digital activism. Here we can identify two different approaches: on the one hand, the implementation of practices of resistance that contrast control by the government or by corporations through the use of data (see, for example, Reference Kaun and TreréKaun and Trerè, 2020); and on the other, the use of data employed to achieve social justice (the Cambridge Analytics scandal [Reference Bennett and LyonBennett and Lyon, 2019] is a good example of this second approach).

While digital activism has been largely valued positively for its potential to reach a much wider audience and provide a space to marginalized groups who find it hard to voice their needs and concerns, a few labels related to this practice hint at a negative characterization. Rather than viewing digital activism as a way to use hashtags to create social change, people who engage in digital activism have also been identified as “slacktivists” or “armchair activists.” This has to do with the degree of agency that activists have when engaging in activism and how this activity is still very much associated with more active, hands-on, or factual actions. The different degrees also account for considerably diverse agency on the part of the activist. From a more passive way of being an activist, such as changing a profile picture to black or adding a banner over it, to using digital technologies to regularly campaign for given issues (see, for example, Twitter campaigns by major associations such as Stonewall, or others, such as Mermaids, that have more recently moved to TikTok),Footnote ² to enabling mobilization of activists such as the World Trade Organization protesters in Seattle. These are three examples of different types of activism that involve a different level of engagement and therefore agency, and that are each valid in their own way. Nonetheless, despite the validity of each individual’s choice of action and the great difference between each of the types of activism just described, digital activism, at times, has been regarded as a less valid form of activism or put erroneously in competition with its live/on-site counterpart (Reference Jackson, Bailey and WellesJackson et al., 2020:39). Others have argued that digital activism is merely “click activism,” a practice that Reference Scholz and JoyceScholz (2010:27) has defined as “a kind of liberal catharsis during lunch break that gives participants the impression that they have done something about the issues, when, in fact, their online action has no offline effect at all.” In fact, he continues, “very few who join political Facebook groups become involved in long-term political campaigns” (Reference Scholz and JoyceScholz, 2010: 27). Contemporarily, we find in literature scholars who have argued in favor of less active forms of digital activism. Reference VieVie (2014), as a matter of fact, posits that even seemingly insignificant responses can be a form of resisting discriminatory behaviors against marginalized groups. Vie brings to the fore the example of the Human Rights Campaign (HRC) Marriage Equality logo – where the Facebook page logo of the organization was changed to a memetic image, shows a supportive environment, and draws awareness to important causes and should not be dismissed as a form of “slacktivism.”

Despite these negative judgements, digital activism has, in the end, been recognized as a successful tool for digital mobilization, able to keep attention high in relation to given issues on mainstream media, and has been considered especially effective among young activists to push change forward, whether it is through the creation of a persuasive slogan like #Transrightsarehumanrights or by simply using a powerful name such as #GeorgeFloyd (Reference Jackson, Bailey and WellesJackson et al., 2020).

The literature currently available on the intersection between the use of the internet and the development of social movements mainly stems from a variety of theoretical frameworks and methodological approaches that meet at the crossroads between the field of communication and feminist, queer, postcolonial, intersectional and critical race frameworks. The scholarship around digital activism has critically addressed how social movements have created their networks online, how the digital technology has been used to do so, by means of which infrastructure they do so and what role the internet has played in the development of social movements that question and challenge the current political status quo (Reference Jackson and CloudJackson, 2018:13). The essential feature that runs across all of these questions, like a fil rouge, is language. Without language there is no communication; however, this topic seems not yet to have found the interest of linguists.

2.2 Language, Gender, Sexuality and Online Mobilizations

Activism within the LGBT+ and feminist community is not a new phenomenon. On the contrary, it is thanks to the mobilization of the people, who fearlessly and tirelessly choose to be in the front lines of these fights to see basic human rights recognized for everyone regardless of gender identity or sexuality, that social change happens. These movements have been on the radar of scholars for years, exploring the way people organize themselves, how the protests are carried out and how linguistic and discursive practices impact on this type of communication.

As Jackson et al. (2016) point out, the technological structures of the internet have been used by individuals who identify within marginalized groups in society such as lesbian, gay, bisexual, transgender and queer activists, as much as by Black and Latino communities whose interest often lies in racial justice activism, not only to make claims of civic and social belonging, but also to build life-saving networks of support. This is true not only for social media platforms but also, for example, for the creation of other materials and spaces, as the transgender and queer communities have done on many occasions, using their understanding and experience as marginalized communities to create websites, blogs, digital zines, dictionaries, video content and cross-platform social media networks that can be positioned outside the cultural and political spaces in which their identities are disciplined and punished (Reference BaileyBailey, 2016; Reference CavalcanteCavalcante, 2016; Jackson et al., 2017). This parallel online world in which trans and queer people are able to thrive contributes to a greater attempt at a different type of politics that aims not only at shifting the mainstream narrative but also at developing collaborative and experimental political projects (Reference Hundley and RodriguezHundley and Rodriguez, 2009; Reference RawsonRawson, 2014). Thanks to their function as a new paradigm of communication (Reference Kopytowska and KopytowskaKhoshraviNik, 2017), social media are a fertile ground in which change can be engendered, and therefore a particularly relevant area of analysis when it comes to the representation of gender and sexual identities. Reference KhoshraviNik, Flowerdew and RichardsonKhoshraviNik (2017:752) highlights this very aspect of social media by suggesting that it provides a new type of communicative affordance by developing at the intersection of mass and interpersonal communication.

3.1 Introduction and Aims

In this section, I discuss a case study that focuses on the Twitter platform and serves as an example of what can and should be labeled as a study on LGBTQ+ digital activism in linguistics. It is largely acknowledged by the scientific community that Twitter has become one of the most important technologies contributing to pushing forward pressing social issues in ways that are possibly more powerful and visible than ever before in the mainstream public sphere (Reference Jackson, Bailey and WellesJackson et al., 2020:32). As mentioned in the previous section, this trend might be on the verge of changing, but at the time in which the data in this case study was collected, this statement was definitely relevant. The importance of the Twitter platform is due mainly to its ability to spread information quickly and vastly. In fact, the narratives built around and through the use of hashtags on Twitter evolve and spread much more rapidly than they would through more traditional media channels. This aspect has elevated Twitter to being one of the “major tools for disseminating information to the public in the hope of spurring particular actions or outcomes” (Reference Jackson, Bailey and WellesJackson et al., 2020:34). From the perspective of a linguist, Twitter was for a long time a very useful platform, as it could be described as a collection of language, publicly accessible, that includes metadata, available digitally and representative of a variety of genres, languages, topics, or even personalities. It served as a place for data collection not only to analyze language use or language change in time, but to observe political developments, historical events or social behaviors, and thus was a useful source for many disciplines (Reference ZottolaZottola, 2020a). Twitter is based on social interaction between users and can be employed to create ambient affiliation through the use of hashtags (Reference ZappavignaZappavigna, 2012).

This case study aims at highlighting some of the linguistic and discursive practices used by digital activists on Twitter in relation to a specific event that targeted the trans community. With this case study I hope to shed light on a set of patterns which could be considered beyond its scope and be associated, in more general terms, to digital activism practices related to gender identity and sexuality issues.

3.2 Background

I have so far argued in favor of this new communicative paradigm and have admired the positives of the cybersphere and its role within social change. But it is impossible not to observe that it is this very global, immediate and participatory nature of social media communication paradigm that has also made the cybersphere a breeding ground for the expression and dissemination of a range of exclusionary, intolerant and extremist discourses, practices and beliefs (Reference Kopytowska and KopytowskaKopytowska, 2017). This behavior seems to have been taken on not only by ordinary people, but also and especially by politicians who seem to have found in Twitter a very effective means to communicate their thoughts and political ideas, not only personally or directly but also through false accounts or bots used to push their political agenda. One in particular is former president of the United States of America Donald Trump, who has declared in an interview that in truth it is thanks to Twitter that he managed to become president (Reference DemataDemata, 2018). Twitter was also one of the platforms that led to the decline of the former president, who was eventually banned from it after using it as a means to direct and encourage an assault on Capitol Hill, the metonymic but also actual site of the US government, in January 2022. From the beginning of his mandate Trump repeatedly used this platform to make official communications, while also using it to attack and denigrate the LGBTQ+ community. One of the most famous examples can be traced back to July 2017, when Trump announced a major change in the regulations for military service via a tweet. The tweet was posted in July 2017, and the actual law was reinforced in August. This law banned all trans people from being in the military. Before this, after he was elected in November 2016 and began his mandate as 45th president of the United States in January 2017, he had already attacked the trans community many times: in February 2016, by acting upon the law that allowed trans students to use toilets in public school, choosing them according to their gender identity; in July 2016, when the Department of Justice filed a legal brief on behalf of the United States in the US Court of Appeals, arguing that the 1964 Civil Rights Act does not prohibit discrimination based on sexual orientation or gender identity, implying that the administration should not prohibit this discrimination either; and lastly, in December 2016 when the staff at the Centers for Disease Control and Prevention was instructed not to use words such as “transgender,” “fetus” or “science-based” in any official documents. The growing disdain for transgender people that rippled across the USA and was mirrored by all of these actions led to a final move against the trans community culminating in October 2018, when the Department of Health and Human Services announced that they wanted to revise Title IX of the Federal Civil Rights law to elaborate and establish a legal definition of sex and gender identity according to which gender was a biological and immutable condition. A number of protests were raised across the country to fight this and say loud and clear that transgender people would not be erased, and a gathering organized by GLAADFootnote ⁴ was held in Washington Square Park, NY, on the same day this was announced, followed by a rally in front of the White House on the next days (10/22/2018). At the same time as Twitter was being used by the President of the United States to persecute a minority, this platform became the place for a counterattack, and the hashtag #wontbeerased became the symbol of this protest. I became interested in the way this counterattack was put forward on Twitter and decided to analyze the linguistic and discursive practices that were being used on the social media to pursue this protest. The following sections describe the data considered for this case study and the methodology used to analyze it, and discuss the analysis.

3.3 Data and Methodology

In order to carry out this analysis, the #wontbeerased corpus was created.Footnote ⁵ Tweets were scraped using the Twitter API Academic Research Access tool, which allows researchers to download public tweets for free.Footnote ⁶ The corpus was collected on a time span that stretches between October 2018 (the first tweet included the date October 21, when the protest started) and September 2021. The search terms used to download the data are based on the hashtag in all its forms, including #wontbeerased, #WONTBEERASED and #WontBeErased. The search was limited to the English language and includes the original posts, replies and retweets. Although replies and retweets are generally left out when building a corpus of tweets, as they can be seen as adding noise to the data, in this case study I am interested in observing the extent to which discourses around this hashtag are spread; therefore replicates of a Tweet or replies to it contribute to understanding how the hashtag spreads and how it is employed by Twitter users. Whether the use of the hashtag is relevant to the activist actions or is used in a reply to troll activists, this is relevant to the analysis of its use and to draw conclusions on the linguistic and discursive style of digital activists. The result of this search generated a corpus of 870,116 words that equals to 40,331 different texts. Given that the corpus is made on one text type and built around a single search string, it can be defined as a specialized corpus (Reference Koester, O’Keeffe and McCarthyKoester 2022); while not extremely large, it is still representative (Reference Gillings, Mautner and BakerGillings et al., 2023: 8) of the type of language that I am interested in investigating here (on adequate corpus sizes, see Reference Egbert, Biber and GrayEgbert et al., 2022). Reference Koester, O’Keeffe and McCarthyKoester (2022) specifies that while specialized corpora are not adequate for certain types of analyses, among which are lexis and phraseology, they offer a distinct advantage to delve into contextual information and are useful for exploring in more detail the link between “the corpus and the contexts in which the texts in the corpus were produced” (p. 49).

As for the theoretical framework to which I refer to carry out this analysis, I believe the work can be positioned within the broader field of Social Media Critical Discourse Studies (Reference Kopytowska and KopytowskaKhoshraviNik, 2017). Methodologically, I add a Corpus-Assisted Discourse Studies (CADS) perspective (Reference Gillings, Mautner and BakerGillings et al., 2023; Reference Partington, Duguid and TaylorPartington et al., 2013). In this sense, the tools of corpus linguistics were used here to simplify the processing of the data, while the analysis carried out can essentially be considered qualitative. Much could be said about the combination of these two approaches and the benefit that this synergy brings. I believe that this discussion is outside the scope of this Element, therefore the reader will find here a brief introduction to the two approaches. Social Media Critical Discourse Studies is a problem-oriented approach that addresses the complexities of digital discursive productions; in other words, it considers language in the way it is produced within digital technologies (Reference Kopytowska and KopytowskaKhoshraviNik, 2017; Reference KhoshraviNik and EspositoKhoshraviNik and Esposito, 2018) and how these new spaces of participatory communication influence the production and circulation of given discourses, ideas and values. The peculiarity of Social Media Critical Discourse Studies (SM-CDS) lies in the triangulation of the analyses of the processes of production, consumption and distribution of the communicative act within the cybersphere without leaving behind the impact of this communicative paradigm. Social Media Critical Discourse Studies considers power relations and how these are negotiated through dominant discourses and mainstream ideologies within social media platforms, mapping the different discourses and bearing in mind that they are shaped by technology (Reference KhosraviNikKhoshraviNik, 2022). Against this backdrop, the analysis in this section takes a “bottom-up” approach, which aims at identifying “recurring linguistic and rhetorical patterns in the texts included in the data set and aims to generate hypotheses a fortiori” (Reference RiboniRiboni, 2020: 58). In terms of the CADS approach, I align in this work with the definition provided by Reference Gillings, Mautner and BakerGillings et al. (2023) that defines CADS “as the examination of textual data, applying a corpus linguistics methodology, to explore the two-way relationship between discourse and society” (p. 5). In other words, I use this approach to automatically detect regularities (and irregularities) in the way in which words come together in tweets and try to explain how their relationship is situated within the sociocultural–political context – in the broader context of the web, to describe the extent to which discourses and representations are reproduced and shape mainstream conversations around the trans community (Reference Gillings, Mautner and BakerGillings et al., 2023; Reference BakerBaker, 2023). In particular, I make use of the frequency list tool, the concordance tool and the collocation tool (see, among others, Reference Brookes, McEnery, Adolphs and KnightBrookes and McEnery, 2020). More specifically, I started by creating a frequency list of the tokens included in my corpus for the purpose of numerically ordering them; this was useful to quickly identify the most frequent words in the corpus. Generally speaking, at the top of the list, it is common to find function words, but in many cases, it is possible to identify content words that have a predominant role in the corpus due exactly to their very high frequency, which puts them at the top of the frequency list. (More on the use of the frequency tool for discourse analysis can be read in Reference BakerBaker, 2023, among others.) At the same time, much can be learnt about the data by noting the words that are not very frequent or are altogether absent from one’s database (Reference Baron, Rayson and ArcherBaron et al., 2009). This can be seen as an initiatory exploration of the dataset that can lead to more extensive and detailed analyses that may involve statistical calculations, for example, the dispersion within the corpus of a given item. A frequency list may also highlight words that one could study more carefully using other tools of corpus linguistics. Concordances and collocation, differently from frequencies, take the analysis further by positioning a specific word within its context of use. A concordance analysis offers an in-depth overview of a given word within the corpus, providing a list of all the times it is found in the corpus, namely, its occurrences within its context of use; this word normally comes to be known as the node word, and in a concordance line it is presented at the center of the sentence with a few words occurring to the left and right. Concordance lines can be sorted in a variety of ways. This type of analysis is useful to highlight repeated language patterns that occur in relation to a given word or to understand the context of that word in more detail (Reference HunstonHunston, 2002). In particular, I used this tool to zoom in on specific elements of the corpus, for example to explore in more detail the verbal patterns reported in Section 3.7. Lastly, collocation analysis involves the statistical calculation of the association between lexical items. In other words, a collocation analysis generates a list of the most significant words which appear next to or nearby a given node word and can be defined as “the phenomenon surrounding the fact that certain words are more likely to occur in combination with other words in certain contexts” (Reference Baker, Hardie and McEneryBaker et al., 2006:36). A variety of measures can be used to calculate collocations, and their choice is based on the intent of the researcher (Reference BrezinaBrezina, 2018). Collocation analysis is useful to identify the discursive patterns around the use of a given word or phrase. In contrast to concordances, which do not involve any statistical measurement and focus on the direct context of occurrence, collocations are a way to have a more general overview of broader semantic patterns associated to the word or phrase (Reference Biber and ReppenBiber and Reppen, 2015). For the case study, I carried out a collocation analysis of the hashtag #wontbeerased.

To investigate the data, I use two different softwares: Wmatrix (Reference RaysonRayson, 2009), for an initial exploration of the main semantic domains found in the data, and CQPweb (Reference HardieHardie, 2012), for a more detailed analysis of the language used. When data is uploaded to Wmatrix it is automatically tagged by CLAWS (Reference Garside, Smith, Garside, Leech and McEneryGarside and Smith, 1997), a grammatical tagger for part-of-speech created at Lancaster University (each word is provided with information about its grammatical function) and semantically tagged by USAS (Reference Wilson, Rayson, Souter and AtwellWilson and Rayson, 1993), an English semantic tagger also created at Lancaster University. For this case study, I focused on the discourse fields generated by the second type of tagging. Simply put, each word is assigned a tag that indicates a discourse field that fits the meaning of that term, namely, “time” or “emotions.” Using log-likelihood as a standard value,Footnote ⁷ a frequency list of the tags is created and compared against the frequency list of the tags of a reference corpus; in this case, the reference corpus used is BNCwritten sampler (Reference BurnardBurnard, 1999), already available on Wmatrix, and this comparison generates a final list of the key semantic domains in my corpus.

As for CQPweb, the corpus was tokenized and tagged for part-of-speech using a Spacy-trained model and pipeline for English.Footnote ⁸ This software was used to carry out the different analyses mentioned earlier.

3.4 Hashtagging and Direct Tags

The initial exploration of the data through the use of the semantic analysis tool Wmatrix revealed a very specific pattern that involves the use of hashtags. Hashtagging, that is, the practice of using the symbol # in combination with words, phrases, or entire clauses as a support slogan or symbol of the issue described in the text of the tweet, or embedded within the tweet itself, is not a novel procedure; on the contrary, it is quite popular on this platform and in social media communication practices more generally. Hashtags have been defined as enabling users to embed metadata in their social media posts (Reference ZappavignaZappavigna 2018), and it has been recognized that they have gone beyond being simply useful ways to make talks searchable to devices through which people create networks, or systems of online ambient affiliation (Reference ZappavignaZappavigna 2012). However, the mode in which hashtags are used in this corpus takes their function a step forward. In fact, I argue that this can be considered as one of the main features of, and peculiar to the practices that revolve around LGBTQ+ digital activism. Differently from what we are usually used to seeing, the tweets that use the hashtag #wontbeerased have a tendency of including within the text not only this one hashtag but a high number of other hashtags as well. This is suggested both by the semantic analysis done with Wmatrix and by the collocation analysis of the term #wontbererased done via CQPweb, where all the most significant collocates of this node word were other hashtags.

Here I discuss the analysis done with Wmatrix. Once the data was tagged as per the methodology explained in the previous section, a list of domains was returned. As an entry point into my data, I focused on the most significant item. At the top of the list of the most frequent domains in my data is one labeled “unmatched” (Z99 in the labels used by Wmatrix). This group contains all the words that the tagger did not recognize. Interestingly, all the occurrences included in this domain are hashtags and @ signs, used on Twitter to tag a specific person or account within the post. In other corpora the terms included in this domain might be irrelevant, but in the case of a corpus collected from Twitter, these become key elements.

Table 1 shows the twenty most frequent elements that were found in this domain. As mentioned earlier, these results are also confirmed by the collocation analysis of the hashtag. In that case, collocations were calculated using Mutual Information as a statistical measure.Footnote ⁹ Scrolling through the top 100 collocates of this term, only nine of these are not hashtags or @ signs.

An example of how these hashtags are used can be seen in the following tweetsFootnote ¹⁰ (Examples 1 and 2) retrieved from the #wontbeerased corpus.

As these examples show, the hashtags make up for a great part of the tweet, and it is very common to see more than one or two hashtags used within the same tweet.

The centrality, both in terms of number of occurrences and statistical significance,Footnote ¹¹ of this specific feature of the corpus highlights two main aspects of the way in which activists employ language on Twitter in the case of the #wontbeerased corpus, the first one being by means of the use of hashtags, taking Reference ZappavignaZappavigna’s (2018) definition a step further. The combination of numerous different hashtags and their use as the primary text of the tweets is what is being done in the #wontbeerased corpus. This also allows users to launch slogans that become empowering phrases; see, for instance, Examples 11, 14, 17 and 20 in Table 1. Because Twitter only allows 280 characters per tweet (that is, since 2017; previous to this date, the limit was 140), the use of hashtags to support the statement being made becomes fundamental. It appears, in digital activism, that the hashtag acquires a semiotic function similar to that of an image, that is to say that meaning-making is embedded within the hashtag itself; one hashtag can carry a much wider meaning than the words that make up that hashtag would if it were standing alone. This feature of the hashtag frees the user from the need to write more and explain themselves, because that word or phrase preceded by the symbol # already says everything, already brings the meaning with it. The hashtag begins to work in the same way as an image, as, for example, an emoji or meme would, suggesting that online communication is becoming more and more summarized and iconographic. Along the same lines, the use of direct tagging (through the @ sign)Footnote ¹² becomes extremely important as it enables a sort of interaction between the politician or whomever the conversation is directed at – the main actor, and the member of the public who is using the social media to make a claim. It forces the person who is pursued by the tag to take a stance, in one way or another, that is, choosing to ignore a tag is a stance in itself. In this case, for example, as Table 1 shows, the main interlocutor is precisely Donald Trump (Table 1, Example 6). The tweet reported here (Example 3) is an illustration of this trend. Since this data was collected, Donald Trump’s account has been shut down in January 2022 as a consequence of the Capitol Hill events (Reference GounariGounari, 2022) and later reinstated in X in November 2022.

The practice of direct tagging, while recognized as one of the potential new features of online communication, acquires a very specific role when it comes to LGBTQ+ digital activism, as it enables the activists to take a stance and almost force their interlocutors to do the same, or at least to be publicly involved in some way. This aspect of the corpus will be discussed in more detail in Section 3.6.

All in all, from the point of view of the structure of these texts used by trans and trans-ally activists, we can conclude that one short tweet not only facilitates the users in expressing a much more complex idea that 280 characters would normally allow, but also permits an interactive communication with specific users. Both these features would not be allowed by what we can consider as traditional activism, namely, marches or sit-ins, for example.

As Reference Jackson, Bailey and WellesJackson et al. (2020:33) posit:

The exploration via Wmatrix also returned other domains that were relevant in the corpus, like “people” and “pronouns.” However, these outcomes overlap with those returned by the collocation and concordance analysis done using CQPweb; for this reason I decided to leave out the discussion related to the semantic analysis and discuss in more detail the investigation aided by the use of CQPweb.

3.5 Multimodal Approaches to Digital Activism

Using CQPweb, I continued my analysis by creating a frequency list sorted by parts of speech, this choice was dictated by the high presence of hashtags and other elements which prevented the realization of a classic frequency list. By classic frequency list I mean one compiled on the basis of all the tokens that comprise the corpus; in this corpus many words were joined together in the hashtags, therefore I believe the frequency list would not be as representative of all the words used in the data. The list presented some expected results, the most frequent part of speech was that of nouns, but some less expected ones as well. For instance, the second most frequent type of words consisted of proper nouns. These were then followed by verbs, punctuation, adverbs, pronouns, determiners, adjectives and so on. Given the nature of computer-mediated communication, which resembles speech more than writing (Reference YatesYates, 1996; Reference Knight, Baker and McEneryKnight, 2015), it seemed particularly interesting to explore in more detail the frequency list of the Punctuation category, which is not usually considered a peculiarity of oral production. Approximately 11 percent of the corpus is constituted by punctuation. The tagger employed to tag the corpus before uploading it to CQPweb includes in the category of punctuation not only full stops, commas and exclamation marks, that is, traditional punctuation, but also emojis, a graphical device, or “a graphicon” (Reference Herring and DainasHerring and Dainas, 2017), used to convey meaning on contemporary social media platforms. The high frequency of these symbols found in the data suggests that this can be identified as one of the features of the language used for LGBTQ+ digital activism in my corpus. The role played by these nonlinguistic features in digitally mediated communication has recently attracted extensive interest among academics (Reference Logi and ZappavignaLogi and Zappavigna, 2021). Reference Logi and ZappavignaLogi and Zappavigna (2021: 3222–3223), suggesting that emojis are widely used to express emotions, convey stances and negotiate interpersonal alignments. In fact, they draw from Reference Herring and DainasHerring and Dainas (2017), who suggest that emojis fulfill similar roles as emoticons, graphical devices formed from ASCII characters, which have been in use since the late 1970s. They (a) express emotions, (b) function as nonverbal signals, (c) function as tone management or indication of illocutionary force and (d) replace punctuation or structural markers (Reference Herring and DainasHerring and Dainas, 2017: 2185). It is important to point out some of the limitations of the use of emojis. One above all is related to the fact that different emojis are interpreted with different meanings by different people and that different viewing platforms render emojis in a variety of ways (Reference Herring and DainasHerring and Dainas, 2017). These small images have become so popular that by 2016 almost 50 percent of messages used in technology-mediated communication included one (Reference Logi and ZappavignaLogi and Zappavigna, 2021). We can now agree that they are evolving into a language of their own; Reference Herring and DainasHerring and Dainas (2017, p. 2185) go as far as to suggest that they can be considered “a complete grammatical system.”

Against this backdrop, interestingly enough, the most popular and straightforward emojis, those that serve the function of “tone modification” following Reference Herring and DainasHerring and Dainas’s (2017) categorization, do not occur frequently in the corpus and show a very low and not varied presence, the most common ones being the 😄 (smile), the 😍 (heart eyes), the 😘 (kissing heart) and the 🥰 (smiling face with hearts). These types of emojis are “nonverbal, paraverbal, or paralinguistic cue[s] as to how the text should be interpreted” (p. 2188), or more simply put, a way of expressing an emotion, such as the face emojis or the hearts. Beyond these, I found several other symbols that display a notably higher number of occurrences. These are listed in order of frequency in Table 2.

This feature of the data suggests a very strong iconographic element embedded within the tweet. In addition to the hashtags, whose function, I suggested, resembles that of a symbol in the way they carry meaning, here the emojis are the actual symbols being used. These seem to align in part with the function described by Reference DanesiDanesi (2016) of being adjunctive emojis; that is, they make meaning alongside language (among the ones in Table 2 we group in this category the following: 🖋, 🌈, 🏳, ✊, 🗳), but also with the other function described by this author of being substitutive emojis, in this case they replace the language altogether (this category includes among the emojis described the following ones: 📨, 📬, 📝). More specifically, the tweets that used the Fountain Pen, the Incoming Envelope, the Memo and the Open Mailbox emojis are employed in particular types of tweets that invite the reader to sign a petition to be delivered to local senators, or remind people to vote, therefore, to actively engage in more practical forms of protests pointed directly at those politicians who play a role in the marginalization of the trans community or who could help end these types of abuses. The function of these emojis can recall what Reference Herring and DainasHerring and Dainas (2017: 2189) define as “action,” “used to portray a (typically) physical action.” The rainbow, fist and white flag emojis I believe can be better described not according to their pragmatic function, as earlier, but through a social semiotic approach which takes into account the interaction between the ideational (constructing the experience of the world through language), interpersonal (enacting relationships through language) and textual (organization of discourse into a coherent and cohesive text) functions of language, following the Systemic Functional Linguistics’ tradition (Halliday, 1978). In this sense, I follow Zappavigna and Logi’s (2022: 233) taxonomy and suggest that the rainbow, fist and white flag emojis realize the discursive semantic feature of involvement. That is to say that these emojis are used to express solidarity “by employing semiotic resources to encode meaning that can only be interpreted by members of a particular community” and their allies. The rainbow recalling the LGBTQ+ community, the white flag to symbolize the end of persecution and discrimination against the trans community and the fist to remind everyone to keep fighting and resisting. In a way, each of these features recalls one of the aspects of digital activism highlighted in previous sections, in which scholars suggest that digital activism is useful in keeping a connection open between online and offline mobilization practices. In this sense, the use of these emojis responds to the most basic of the speech functions, the illocutionary, the locutionary and the perlocutionary effect (Reference AustinAustin, 1962) that the tweet aims at having; the emojis produce literal meaning, they have a social function, and they encourage the interlocutor to take action.

Moreover, as noted in the literature review (Reference Esposito and SinatoraEsposito and Sinatora, 2021), the multimodal component is deeply prominent in the tweets beyond the use of emoji. While conducting a more detailed analysis of the elements identified in the Punctuation category frequency list, through a concordance analysis that allowed the researchers to access links to each tweet, it was remarkably frequently observed that the tweets included either a photo or some kind of visual aid, a post or a video, as Examples 1, 2 and 3 in Section 3.4 also show. While this observation is purely speculative, the way the data was collected and stored does not allow one to test this hypothesis thoroughly; in line with the existent literature, it could be said that multimodality plays a unique and important role in online mobilization and in LGBTQ+ digital activism in particular.

3.6 People

In this section I look in more detail at another of the categories emerging from the frequency list sorted by parts of speech, that of Proper Nouns. This category, which I also found unexpected, includes names of people, cities, or public places. I decided to explore this category in more detail because I believe it can shed additional light on the social actors involved in my corpus and the people/ places that are included directly in the discourses surrounding #wontbeerased. In fact, the terms included in the frequency list for this category overlap in part with the discussion presented in Section 3.4, many of the proper names are preceded by @ signs. With this in mind, I sorted the frequency list to include only those terms that are premodified by the symbol @. This search returned a list that I have sorted to include only tags referring to people. Thanks to this type of selection, I was able to detect those accounts of very specific profiles among the social actors that are involved within the data of this corpus and who can be related to the cause put forward by the #wontbeerased slogan. Differently from the procedure followed in Section 3.4 where names and identities of the users were made anonymous, here, given the status of public figures that the people involved in this section of the analysis have, I did not remove personal information. Public figures choose, as part of their occupation, to make their ideas and thoughts publicly displayed; for this reason, I do not anticipate ethical conflict. Table 3 shows a list of the twenty most frequent subjects tagged in the tweets included in my corpus.

The items presented in Table 3, excluding number 5, correspond to the Twitter handle of a specific individual.Footnote ¹³ In the initial list that was produced by searching names preceded by the @ sign, some were related to accounts of newspapers. I decided to remove those, but on the contrary I kept the tag related to the Human Rights Campaign (@HRC, item 5 in Table 3). The Human Rights Campaign is one of the largest and most influential advocacy organizations for the LGBTQ+ community in the USA,Footnote ¹⁴ and while this handle does not belong to a single individual, unlike the others included in this table, I still view this account as presenting a reality that is very different from that of a newspaper and representative of one of the voices contributing to the narrative being built around the hashtag #wontbeerased. As a matter of fact, this account massively used the hashtag during and after the actual protest. This list, which only includes the top twenty most frequent accounts tagged in the tweets, is far from being the complete one, which is much longer, including over 7,400 items. It is interesting to note how, except for item number 5, already discussed, and item number 10 that belongs to author and public figure Anthony Oliveira,Footnote ¹⁵ all the other Twitter handles identified here belong to senators of the United States and to Donald Trump, who at the time of data collection was the president. This practice suggests that the tweets included in the corpus and used by trans and trans-ally activists are far from being impersonal or generic but, on the contrary, the individuals who produce these tweets make an effort to create very personal content and above all aim at bringing the discussion back to the people, at engaging directly with those who can really make change possible, who have the power to reverse the series of injustices and discrimination that the trans community must endure.

When looking at the data in more detail, I found a language pattern that was repeated very often: “I just delivered this letter from.” The concordances of this sentence revealed that all the tweets containing it were posted by the same account. Figure 2 displays an example of this type of tweet.

Figure 2 Open Letters account

These tweets were all posted by one specific account, the Open Letters account. This is a bot that automatically publishes copies of letters delivered to elected officials in the United States. It is connected to a free service called ResistbotFootnote ¹⁶ that allows a user to contact an official in a few small steps through the use of text messages and a chatbot. There are 4,330 tweets posted by this account in my corpus, which makes up almost 14 percent of all tweets (this equals to a similar percentage of words, 14.9%). While this is obviously not an actual person posting on Twitter, it serves the purpose of disseminating these actions, making them visible and possibly encouraging as many people as possible to follow this lead. In addition, while the posts are not created by a person directly, the bots are made by people with the intention of disseminating a specific topic widely; this could be considered in itself a form of digital activism. This practice puts these politicians in a situation in which they must respond in one way or another and position themselves publicly with regard to these issues, and when a politician is being called upon directly and publicly, even not taking a stance is actually a way of taking one.

Continuing the analysis by sorting the data to exclude all tweets posted by this account that not only follow the same structure as the example provided here but also are not representative of natural language production, I found three different discursive patterns employed by the users of the hashtag #wontbeerased. First, a group of tweets I label as celebratory tweets, used to proclaim some kind of victory or achievement (as displayed in Example 4).

Second, we have reporting tweets (see Example 5); these tweets are used to condemn discrimination committed against trans individuals or the community at large. As with the previous type of tweets, it is common to find an attachment to the tweet that allows the user to find more information via a newspaper article or a video.

These types of tweets, by embedding the link to a video, become in a way multimodal; even if the nonverbal content is not directly visible, it is attached to the tweet and is an integral part of the meaning-making of the text. In this sense, these types of tweets recall the linguistic practice described in Section 2.2.1 quoting Reference Chałupnik and BrookesChałupnik and Brookes (2022), who talk about multimodal communicative acts, that is, directives and declarations that serve the purpose of both providing information on the protests while also mobilizing protestors. In addition to these two categories of tweets, I retrieved another type of tweets in the data. While celebratory and reporting tweets are used by a variety of different accounts, this last type of tweets (see Example 6) are used mostly by trans people to claim their rights as human beings and use the space to fight their marginalization. These tweets written by trans people who use this social media serve to bring visibility to their identity and their community as whole. I define these as agency tweets; because they permit people to position themselves, they give them a space of existence, allowing trans people to actively change the way in which they are portrayed and perceived.

Agency tweets also have the function of reminding Twitter users that it is actual human beings that are at stake when laws are suddenly changed or representations are used to undermine a person’s survival. These tweets specify that while this issue is unequivocally concerned with laws and politics, it should, above all, be discussed in terms of what or whom it truly affects, that is, people’s lives that are being questioned.

3.7 Hate versus Love

The literature in the field of online/computer-mediated communication has often proven that there is a tendency for users, because of the internet’s new ways of communicating, to easily resort to hate speech, using unique linguistic techniques. Reference KhoshraviNik and EspositoKhoshraviNik and Esposito (2018), in particular, highlight three specific techniques that online haters employ, and these are anonymity, seen as the ability of the web to hide one’s identity; physical separation, which comes as a consequence of anonymity and is intended as the practice of distancing oneself from one’s online identity, resulting in lack of face-to-face interaction and acknowledgement of people’s humanity; and lastly the practice of deindividuation, which consists in relying on the group, or being part of a specific ensemble by reducing self-awareness and self-visibility. The data discussed so far, on the way in which digital activists use language, suggest that activists’ linguistic and discursive choices differ largely from these practices. On the contrary, these people try not to hide themselves, but make their bodies visible by posting selfies, for example, and making their stances loud and clear. A study on the perception of selfies in posts conducted by Reference Katz and CrockerKatz and Crocker (2015) suggested that adding a selfie to a post tends to elicit comments and encourages conversations between users who are, in turn, also stimulated to post a photo. In this sense, the practice can trigger a kind of chain reaction toward online mobilization.

While the academic literature sadly overflows with studies that tackle the way in which language is used to create hate speech online (see, among others, Reference Balirano and HughesBalirano and Hughes, 2020; Reference BrindleBrindle, 2016; Reference RasuloRasulo, 2023), the last portion of analysis that I wish to discuss here goes in a different direction and discusses a way to resist hate rather than contribute to it. The analysis relies on another of the part-of-speech categories derived from the frequency list presented earlier and reinforces the idea that digital activists, to put it in a more poetic way, spread love rather than hate. The category under scrutiny here is that of verbs. The frequency breakdown function on CQPweb allowed me to produce a list of all the verbs used in the corpus in order of frequency. It is very common when looking at a frequency list to find in the first positions of the list a number of auxiliary verbs, as it was in this case where we found verbs such as “to be,” “to have,” and “to do” in a variety of forms. These were excluded for the purpose of this analysis. We are, thus, left with a list of main verbs, those whose meaning is related to actions, events and states. The top thirty most frequent are listed in Table 4.

In this list of verbs, we find a variety of different verbal processes. In order to classify these verbs, I refer to SFL’s (see Section 3.5) taxonomy (Halliday, 1978). I then look at the context of use to determine the function that they carry out and the outcome of the choices made by the activists in using them. Halliday (1978) identifies six process types: material, mental, relational, verbal, behavioral and existential. Many of the items in Table 4 are material verbal processes, namely, verbs that express the process of doing something, of a physical action like stand, fight/fighting and support, join or protect. In particular these verbs recall a semantic domain that seems to be closer to violence than love, that of war language; it recalls many slogans popular in the United States related to, for example, joining the army, or remembering the veterans and soldiers who are fighting to protect the country, a type of “warspeak,” as defined by Reference MyersMyers (2019), that permeates the United States. However, if we look in more detail at the way these verbs are used in the tweets, we find that these are actually used to encourage people not to surrender to injustice, to tell trans people that they are not alone in facing these discriminations. This discourse is supported by phrases such as “fighting for rights,” “resisting erasure,” “deserving to be alive and to exist,” and “attack our basic human rights,” which despite recalling battlefield language – namely, fighting, resisting – present a very positive discourse. An illustration of this is displayed in Example 7.

1. 10) WE stand WITH the trans* community because y’all MEANS y’all! #nhpolitics #wontbeerased

2. 11) Black trans folks matter. Intersex folks matter. Disabled trans folks matter, indigenous trans folks matter, latinx trans folks matter, immigrant trans folks matter – Every single one of us matters. We will stand up against this discrimination! #TAGUrself #wontbeerased

3. 12) I cannot stand by as trans people are under attack. This is not about identity but about humanity. The fight for trans lives is a fight for recognition of all people’s full humanity. Stop targeting trans people @realDonaldTrump #WontBeErased #TDOR18 #TransDayofRemembrance

4. 13) To our transgender & non-binary youth, staff, friends, neighbors, & family members: we see you, we love you, and we stand by you. #wontbeerased

All these examples show how Twitter users are employing this hashtag to write messages of hope (recalling what Reference ClarkClark, 2016 suggests, as explained in Section 2.2.1) and support to the trans community. This same pattern repeats with the verb to fight; examples include “We fight for our trans brothers and siblings,” “fight attempts by those trying to erase the #LGBTQ Community,” or “I’m here for you and I will fight this b.s. with you.” In these examples we observe how the individual poster becomes an agentive actor, that is to say, they are always involved in the action described by the verb.

The other set of verbs that are strongly present in the data analyzed are existential verbal processes, which define the existence of something or someone. In particular, the list shows the frequent use of the verb “to exist”; these verbs are mainly included in tweets that acknowledge trans people’s existence in opposition to the attempt at erasing them, with posts by both trans individuals and trans allies.

3.8 Concluding Remarks

In this case study on the linguistic and discursive choices of Twitter users employing the hashtag #wontbeerased, I tried to observe the data to draw some conclusions with regard to the strategies put forward by LGBTQ+ digital activists. Some of these strategies overlap with those that had already been pointed out by other scholars, for example the use of visual and multimodal elements within the tweets, or the need to establish a connection between the online and the offline mobilization. Some other strategies seem to be peculiar to my set of data, although I wonder if more overlaps between strategies would be found if there were more studies with a linguistic perspective to compare with.

In particular, I explore four different patterns retrieved in the data. First is the practice of hashtagging, by which LGBTQ+ digital activists create meaning using hashtags. I argue that this modality of using many hashtags contemporarily can be seen as an evolution of the way in which these symbols were traditionally thought of and could be seen as a peculiarity of digital activism on social media. The second is the strong presence of multimodal elements. By multimodal elements I do not refer only to images or photos, as in the case study by Reference Esposito and SinatoraEsposito and Sinatora (2021) for example, but rather to a consistent and powerful use of emojis, which serves the purpose of reinforcing the point being made while encouraging other people to take action in the fight against the marginalization of trans individuals. This agentive push that I observe in the use of multimodal elements is reinforced by the practice of direct tagging, thus of giving people a central role in this mobilization. This observation stems from an analysis of the use of the @ sign; this also highlighted the strong presence of political personalities, both taking stances in support of the trans community but also being actively involved in the tweets. Lastly, I looked at verbal processes. Here, we see that LGBTQ+ digital activists employ a type of discourse that opposes online hate through the use of supportive and encouraging messages, making use of particular verbal processes. Although these discursive patterns might hint at a sort of “battlefield” language, the in-depth analysis reveals that the verbs are used in a positive way and the language is directed toward the production of a beneficial meaning.

The analysis of this case study demonstrates the importance and the value of LGBTQ+ digital activism in the fight for social change. First and foremost, it provides a space in which activists can feel safe to express their thoughts; secondly, it allows them to reach a much greater public that crosses national and international borders, echoing worldwide and receiving support. The #wontbeerased hashtag was launched in 2018 to counterattack one of the most violent aggressions perpetuated against the trans community at an official level, by the governing body; LGBTQ+ digital activists were able to have their voices heard, and they continue to do so through this hashtag even years later.

This case study obviously has some limitations, mainly in the fact that it only takes into account a limited number of tweets produced during a limited amount of time. Future research could enrich this case study by adding more data or looking at how other hashtags were used for similar purposes. The changes that were made to the platform now called X and formerly known as Twitter might limit this type of investigation; nonetheless, looking at different social media platforms would offer an important contribution to this area of inquiry.

4.1 Concluding Discussion

The rise of the Web 2.0 and of a new communicative paradigm (Reference KhoshraviNik, Flowerdew and RichardsonKhoshraviNik 2017) that fostered the rise of new linguistic strategies able to thrive in a digital environment has pushed scholars interested in the analysis of language use to try and capture the nuances of this novel communicative style. Most importantly, it has encouraged them to observe the way in which it impacts our society, which is increasingly mediated by digital technology to the point that attempts are being made to create a parallel digital universe of existence; see, for example, the ultimate work in terms of innovation technology, the metaverse,Footnote ¹⁷ a virtual reality social platform launched in 2022 by Mark Zuckerberg’s enterprise. While just a few years ago this would have been considered science fiction, the reality of our society is that life is increasingly being lived digitally. This is even more so after a global pandemic spread worldwide, forcing people to remain in their houses to stay alive. While individuals used to passively receive digital innovation, a number of scholars have now challenged this assumption and view people as having agency, as being active participants in the creation and development of digital environments (Reference AndroutsopoulosAndroutsopoulos, 2014; Reference Szabla and BlommaertSzabla and Blommaert, 2020; Reference Tagg, Seargeant and BrownTagg et al., 2017). The COVID-19 pandemic also helped to speed up this process. The active role of individuals is strictly connected to the growing wave of digital activism that permeates media discourses, from #BlackLivesMatter (Reference NarteyNartey, 2022b) to #MeToo (Bouvier, 2020; Reference Mendes, Ringrose and KellerMendes et al., 2018). New digital technologies have, in fact, helped make marginalized communities’ protests and social movements visible and create networks beyond national borders and on a global scale. As Reference NarteyNartey (2022a:2) maintains, “digital activism and the digital environment serves as a way to center the voice and agency of minority groups, including positive self-representation and their solidarity formation for group empowerment. Social media can be used to advance the goals of repressed groups in order to instigate progressive social change.” In line with this, I argue that LGBTQ+ and feminist digital activism can easily fall under the category of emancipatory discourses (Reference NarteyNartey, 2022a), seen as the way in which marginalized and discriminated groups have historically resisted and challenged the status quo (Reference Lazar, Ehrlich, Meyerhoff and HolmesLazar, 2014).

This Element sets out to open up the conversation about LGBTQ+ and feminist digital activism within the field of linguistics while trying to identify some preliminary, generalized, linguistic and discursive patterns employed to carry out mobilizations online. The discussion presented in Section 2 is meant to outline a sort of state of the art in the field, and it undoubtedly reveals a gap within the field of linguistics; most of the literature reviewed discusses the work of sociology, communication and gender studies scholars. While I value this scholarship and actively make use of the findings proposed by these colleagues, I wonder if an analysis using the tools of applied linguistics and discourse analysis could reveal even further valuable elements about the way LGBTQ+ and feminist digital activism is done, perhaps in relation to the content of the messages that these activists put forward rather than the modalities used. As a point in fact, as mentioned in Section 3.8, while some of the patterns retrieved in the #wontebeerased case study align and overlap with those discussed in the literature review, others are peculiar to this set of data. It is against this backdrop that I wonder whether offering a linguistic perspective in the analysis of digital activism movements might highlight more common patterns in the way that LGBTQ+ and feminist digital activism is done.

Contemporarily, in the attempt at delineating a possible set of patterns that could be associated in a more generalized way with the actions and words of LGBTQ+ and feminist digital activists, multimodality is one that definitely seems to be recurrent. Here, by multimodal I also mean multisemiotic; in fact, it is not only images that are often found along the sides of hashtags or other forms of digital communication for activist purposes, but also elements such as emojis or more elaborate graphicons (Reference Herring and DainasHerring and Daines, 2017) such as memes, alongside the use of posters and even complex products like videos or performances. Close at hand with multimodality, the use of hashtags and hashtagging also takes an unquestionable central role, to the point that digital activism comes to be known as hashtag activism as well. Hashtags not only aid in the creation and circulation of the message but also enable people to connect in the narration of shared lived experiences, emotions and actions. This last point has been identified as one of the benefits of digital activism, the possibility of identifying in other people’s narratives those who have lived a similar experience and are most likely to understand the struggle. At the same time, shared narratives prove to a greater public that discriminatory and marginalizing behaviors are not isolated incidents but are, unfortunately, somewhat of a daily reality for certain groups in our society. One last element which can be counted within the great innovation of digital activism, and of digital environments in general, is that of opening up new spaces for marginalized communities to express themselves, but also a space in which it is possible to create a connection between the online and the offline experience, enabling new ways of doing activism as well as communication online.

While the research carried out for this Element allowed me to better grasp the state of digital activism both in general and with a focus on gender identity and sexuality, it still raised a number of questions that have been partially introduced throughout the previous section and by some of the scholars whose work was discussed here. These questions emerge from the reflection on the downsides of digital activism. One of these is related to the diverse accessibility of people to digital technology, which is a twofold issue. I argue that this point has two sides to it because diverse accessibility (see Section 2.1.2) can be caused either by lack of economic resources on the part of the user or the country in general, or by lack of freedom of speech. How can research in digital activism contribute to ending these types of limitations? Identifying the gaps is one step. Countries in which research in digital activism is not documented are probably those in which activism is more difficult to carry out due to the first or the second, or both, of the reasons just listed. What can be done to support activists who face these types of limitations? Perhaps highlighting countries in which difficulties emerge from lack of economic resources can initiate a chain of support and economic aid toward those parts of the world that have little access to technology or digitalization. The discourse changes drastically for those countries in which perhaps access to technology is overdeveloped but there is strict censorship; what can be done to ensure the security of digital activists? As discussed in Section 2.2.2, moving away from predetermined Western epistemologies of doing activism might be one solution, as Reference PhillipsPhillips (2014, Reference Phillips2020) shows through his work. But there are, unfortunately, countries in which censorship and persecution of minorities are much stricter than the reality of Singapore. For this reason and many others, these questions remain unanswered in this context.

Another issue that is most likely to remain unanswered has to do with the concept of marketization of LGBTQ+ and feminist protests. Some scholars (see, among others, Reference MikulakMikulak, 2019; Reference RiceRice, 2023) warn the public of the possibility that marketization and monetization of these protests becomes so important that it could surpass the scope of the protest to the point that the very aim of the protest is forgotten. Can this be seen the other way around? Perhaps it is time that the linguistic community suggests some strategies that can be used to exploit the marketization of LGBTQ+ and feminist digital activism to its own benefit. Changing this perspective with effective linguistic strategies could represent a turning point for LGBTQ+ and feminist digital activism, especially in those countries or cultures in which these discourses still struggle to get through.

These are just two of the downsides of digital activism that have already been brought to the fore and that I try once again to highlight. Linguistic scholarship in this area could potentially play a role in addressing them. In fact, as Reference NarteyNartey (2022a:5) posits, it is our job as applied linguists and (critical) discourse analysts to “expose and resist the inequalities and injustices in society, [while adopting] an activist-scholar posture in order to push for positive social change. By so doing, our research will not only be done on social groups, but more importantly, for and with them.”

4.2 On Moving the Field Forward

This Element aims at being a starting point for scholars in applied linguistics to explore LGBTQ+ and feminist digital activism further. However, the first step in this process lies in actually recognizing that a lot of the work that linguists have already been doing is focusing on LGBTQ+ and feminist digital activism, but simply not being labeled using this definition. It is time to categorize studies in LGBTQ+ and feminist digital activism as such. The first sections of this Element clearly show the abundance of the literature that looks at the use of hashtags or, more generally, language employed for the purpose of fighting discrimination. At the moment, studies that fall under this description can be found grouped by methodological approach, rather than scrutinized by platform. Grouping them under the label of LGBTQ+ and feminist digital activism could be useful not only to retrace common patterns and identify more clearly the gaps that need further investigation, but also to weave a more intricate web with other disciplines rather than simply taking from them.

Additionally, like in many other circumstances, the study of English language content prevails over other languages. While this is related to the fact that data in English is more abundant and more easily retrievable, the field could really benefit from perspectives into other idioms and cultures. This is especially so if we think of how much could be gained from this type of investigation. On the one hand, this could lead to answers to some of the questions related to downsides of digital activism that were discussed earlier; on the other, these perspectives could open up new modalities of doing digital activisms that the Western world has not yet conceptualized.

As the case study presented in Section 3 and the review of the existing scholarship in Section 2 have shown, multimodal and multisemiotic elements play a seminal role in digital activism. For this reason, another way of enhancing the field would be to focus more extensively on these elements, to explore the role of this novel way of using language and the reason why it is so prominent in online mobilization. Is it simply mirroring the way society is changing toward a more fluid way of communicating or does it contribute to digital activism in a specific and necessary way?

There are many directions in which this field can grow and expand. Here I discuss just a small illustration of them. The hope is for this Element to be the first step toward a more in-depth and extensive exploration of LGBTQ+ and feminist digital activism.

In a time (November 2022) in which Twitter, one of the most commonly used platforms for digital activism, is undergoing a radical adjustment due to its unexpected change in ownership (Reference Clayton and HoskinsClayton and Hoskins, 2022), and the rules that regulate the use of free – or otherwise known as hate (Reference Zottola, Balirano and HughesZottola 2020b) – speech are being completely overturned, studies on the language used in LGBTQ+ and feminist digital activist are gaining a renewed importance. The linguistic patterns and discursive choices that will be used for digital activism will play, unmistakably, a crucial role in distinguishing between discrimination and the fight against it.