This section explains the criteria used for creating the corpus and selecting secondary literature, as well as the methodology employed to analyze both.
To assemble the corpus, the first step was collecting as many samples as possible using trending hashtags such as #antivaxx, #vaccineswork, #measles and further similar ones; which resulted in over 200 samples total. Then this selection was filtered again based on the hypotheses (excluding H5), which led to the final corpus, which became the base for the case studies cited in this work (to be found in the appendix).
In order to conduct a thorough qualitative analysis, the quality of the samples far outweighed the quantity, which in this case was disregarded given that the aim was not to gather quantitative data. On the other hand, this also presents a limitation: due to the research being qualitative and no quantitative not enough samples were used to be able to extrapolate the research to a bigger scale conclusion which applies to all discourse on both platforms by both main opposites of the spectrum involved. When discussing results in the summary, the conclusions are meant for what applies to these specific samples. The main aim in doing it this way is to be able to analyze the phenomenon of humor and to understand the users’ different strategies, as well as being able to establish a more detailed comparison of both platforms by looking at concrete examples. Due to the length of this work, it focuses on text-based materials only, which represents another limitation, since most mixed media like memes, reaction gifs and other imagery, which are core sources of online humor, were purposely not included. Further specifics about the corpus are that there is the same quantity of samples for both platforms, a mix of posts from anonymous as well as identifiable users, as well as content posted by both, anti-vaxxers and anti-anti-vaxxers. Despite the specific amount of reblogs per post not having been incorporated as a criterium to select the samples, the final corpus deemed heterogeneous in this aspect, too.
The samples were grouped in different, sometimes overlapping, categories (stated in chapter 4), and analyzed based on the core arguments of three main theories of the field of Pragmatics, namely Speech Act Theory, Grice’s maxims of conversation, and humor theory, which established a frame for analyzing the users’ linguistic strategies. For the first one, the focus was mainly in classifying speech acts present in the samples according to Searle’s taxonomy; regarding the maxims, the focus was on occasions of flouting and extracting implicatures; lastly, in the field of humor special attention was paid to semantic script-switch triggers and cues. Due to space restrictions, not every detail from every sample will be analyzed. How all three fields make up for the constituents of the linguistic strategies being researched is further explained in chapter 3.
Last but not least it is relevant to mention that one more criterium was employed for literature research, that being that the sources quoted – apart from seminal works from the field of Pragmatics – are from latest 1990 but preferably not older than ten years old, given that both platforms being researched are relatively new and increasingly dynamic in nature. In the next section, all important existent publications regarding Tumblr, Twitter, and the topic of anti-vaccination which inspired this work are briefly listed.
This section provides with an overview of the relevant existing literature regarding the topic of antivaccination, as well as existing works based on Twitter and Tumblr samples. Then, the crucial secondary sources of literature which were taken into consideration for this work shortly be mentioned. The fundamental texts of linguistic theory will be outlined in chapter 3.
As far as I could research, there seems not to be any existing studies regarding linguistic aspects of the discourse around antivaccination, which served as motivation to explore this particular area, apart from those already previously stated. There are a few studies centered on Tumblr and Twitter, but most of them do not address the linguistic features of the platforms’ discourse. There are plenty of different approaches to the platforms, but the following are considered to be core readings which led to this work’s conception.
Mccracken (2017) and Kanai (2017) wrote about youth subculture, almost exclusively regarding the ones in Tumblr. As to why the platform is particularly appealing to some specific subcultures, Renninger’s (2014) work on what he called counterpublic discourse proved particularly useful. Cho (2017) proposed the notion of default publicness, another concept which contributes the questions behind Renninger’s motivation, too. Morimoto (2018), as well as Click and Scott (2018), paid particular attention to the concept of fandom, a wide-spread online phenomenon which is strongly present in but not exclusive to Tumblr; nevertheless, it will not be addressed in this paper. Vásquez and Creel (2017) researched what makes a post “relatable” and how creative processes take place in Tumblr’s text posts based on the idea of double voicing. Cho, Maccracken, and Stein are considered prominent scholars in the field and they are working on A Tumblr Book which was announced will cover a broad and miscellaneous scope of essential Tumblr notions, which shows how this platform has slowly but steadily been receiving more attention from a linguistic point of view.
Gillen and Merchant (2013) wrote about communicative practices in Tumblr from a linguistic perspective. Argüelles Álvarez and Muñoz Muñoz (2012) provided with a comparison between tweets in Spanish and in English, extracting similarities regarding the users’ use of language in both languages. Wilkström wrote two papers on hashtags (2014) and animated reported speech (2014) on Twitter. The former enriches Daer’s (2014) research on metacommunicative functions of hashtags. Lastly, Reyes, Rosso, and Veale’s (2013) research regarding irony in Twitter provides a contrasting and enriching view on irony in Twitter, which features different strategies from the Tumblr ones researched by Vásquez and Creel.
The conclusion which can be drawn from the multifaceted studies –which have surfaced mainly over the last ten years, – is that these platforms provide interesting and rich material to a new type of computer-mediated communication; which, despite being in its infancy stages, has great potential for being studied linguistically, but most importantly, within the field of Pragmatics, too. Further reasons as to why these platforms were chosen and the details of their characteristics will be elucidated next.