The Twitter Conference as a New Medium of Scholarly Communication (and How to Host One)

ABSTRACT The COVID-19 pandemic prompted a shift to online academic meetings such as the webinar and virtual conference. We add to the conversation about how these modes of knowledge mobilization may be more inclusive, accessible, and environmentally friendly than in-person conferences through a discussion of the Twitter conference—during which participants produce threaded tweets of their research and engage in both real-time and asynchronous scholarly discussion. In this article, we discuss how to host a Twitter conference; we claim that Twitter conferences require different skills and have different strengths and weaknesses than virtual conferences or webinars; and we recommend that they should be a permanent addition to the roster of academic knowledge-mobilization events.

T he onset of the COVID-19 global pandemic left academics scrambling to find ways to make virtual meetings possible (MacDonald 2020). At a loss for a safe way forward, hundreds of academic conferences were canceled in the spring, summer, and fall of 2020 (Viglione 2020). Although some in-person conferences have recently been held, some argue that this disruption has offered an opportunity to reimagine academic conferences toward more inclusive, accessible, and environmentally friendly formats (Dua et al. 2021;Fortais 2020). This article adds to the conversation on the future of academic meetings by evaluating the strengths, opportunities, and risks of one way to meet virtually: Twitter conferences. We discuss our experiences of hosting Twitter conferences on behalf of the Toronto chapter of Women in International Security-Canada (WIIS-C TO) and provide some lessons and guidance. 1 We argue that the Twitter conference is more than a means of presenting conference papers online: it has potential as a new medium of scholarly communication that requires different skills and has different benefits than a webinar.

BARRIERS POSED BY IN-PERSON CONFERENCING AND THE SHIFT TO ONLINE
As conferences were canceled during the COVID-19 pandemic, scholars lamented the loss of the unique opportunities afforded by in-person conferences, such as serendipity, social interaction, networking, ease of discussion, and the possibility for hands-on activities or demonstrations (Dua et al. 2021;Smith et al. 2021). However, the shift to online meetings also highlighted the ways that in-person conferences are inaccessible. Henderson and Moreau (2020) provide an overview of the literature on the numerous barriers created by in-person conferences, particularly their exclusionary nature on the basis of disability (Hodge 2014), deafness (O'Brien 2018), race (Ahmed 2012), class (Stanley 1995), and gender (Eden 2016;Bos, Sweet-Cushman, and Schneider 2019;Henderson and Moreau 2020). Caregivers, particularly those of young children, may also be limited in their ability to travel; when they do manage to make it to a conference, they typically face limited or nonexistent childcare options, hampering their ability to take full advantage of the benefits of in-person conference participation (Bos, Sweet-Cushman, and Schneider 2019). Further, Henderson and Moreau (2020, 81) show that academics with caring responsibilities of any nature are less likely to attend academic conferences. Strikingly, these barriers have real impacts on participation in academic conferences, with a study in Israel concluding that men are three times more likely to participate in academic conferences than women (Eden 2016).
The pandemic shift to online conferences showed the transformative potential that virtual networks, communities, and conferences may have in terms of inclusivity and accessibility. Most obviously, graduate students or precarious faculty may be able to take advantage of more opportunities to share their work and network than may have been accessible to them previously, given the sometimes prohibitively expensive travel costs of an in-person meeting (Fullick 2016;Hong 2018;Malloy 2020). Given the lack of physical barriers (Doshi 2014) and the presence of accessibility options like closed captioning or text readers, online options for participation may greatly ameliorate issues of accessibility and inclusion (De Picker 2020).
Yet, at the same time that webinars and virtual conferences have alleviated some barriers, they have also created new problems. Virtual webinars allow participation from around the world, which is a benefit, but at larger conferences, it can be very hard to synchronize time zones so that no one is presenting in the middle of the night. Technical difficulties and spotty internet access have shown that virtual conferences, too, have accessibility issues. Finally, even though the ability to participate in webinars all over the world that would otherwise be inaccessible has greatly enriched scholarly conversations, the sheer number of available conferences and webinars has brought the term "Zoom fatigue" into the lexicon, indicative of the tiring effect that regular and lengthy video calls may have on participants (Bailenson 2021).
Although a Twitter conference cannot solve all these problems of online gatherings, it can address some. The platform can act as a community hub or network for scholars facing types of barriers other than accessibility, such as exclusion because of race or gender. Sullivan and Brennan (2020), for example, found that the use of hashtags such as #AcademicTwitter, #WomensReality, and #WomeninHigherEd created a space and community around the struggles of being a woman in academia. A recent network analysis of political scientists at PhD-granting institutions who use Twitter found that there was little evidence that gender influenced who was central to that network (Bisbee, Larson, and Munger 2020). In February 2016 Beaulieu and colleagues (2017) launched "Women Also Know Stuff," a crowd-sourced initiative to showcase women's expertise: it has since accumulated a Twitter following of almost 30,000. Transparent and accessible networking opportunities, like Twitter, may be a way to ameliorate the exclusive elements of in-person conferences.
More importantly, however, a Twitter conference presentation is fundamentally different from a webinar conference presentation and both presenters and organizers require different skills than they do for webinars. Twitter has become important for knowledge mobilization (Klar et al. 2020;Scoble 2014), especially because "altmetrics," or alternative measures of impact, now consider social media presence (Burns, Blumenthal, and Sitter 2018). The field of political science seems to value traditional modes of scholarship to a greater degree than "new media," yet Esarey and Wood (2018) found that, under some circumstances, tools like social media are highly valued if "they meet the needs of political scientists." These community-building and knowledge-mobilization functions come together in the use of Twitter as a backchannel during in-person conferences. Synchronous use of Twitter during conferences has increased in recent years (Kimmins and Veletsianos 2016), and these Twitter-enabled backchannels have enhanced participant experiences in conferences (Priem 2011;Ross et al. 2014): they facilitate knowledge dissemination by micro-connecting diverse audiences and creating space for collaboration and the co-construction of knowledge (Honeycutt and Herring 2009;McNely 2009;Reinhardt et al. 2009;Ross et al. 2014). Conversely, one recent study found that Twitter use decreased during an online conference: participants used fewer activist hashtags and had fewer online social and interpersonal interactions than when they used Twitter during an in-person conference (Beste et al. 2022).
In addition, the inherently concise nature of Twitter can lead to the production of misinformation, which then is at risk of being widely disseminated given the networked nature of the platform. Despite this, Bombaci and coauthors (2015) found in their analysis of Twitter use during real-time conferences that Twitter backchannel conversations and retweets usually accurately captured the information presented.

THE TWITTER CONFERENCE AS A MEDIUM FOR SCHOLARLY COMMUNICATION
We have now organized and moderated four Twitter workshops on behalf of the Toronto chapter of Women in International Security-Canada (WIIS-C TO) and have also advised on two other Twitter conferences in the field of international relations in Canada.
The Twitter conference borrows some of the ideas from the realtime backchannel and makes them the center of scholarly communication. As we have implemented it, a Twitter conference or workshop is a series of synchronous presentations, delivered by tweet, from a presenter's own Twitter account. Use of the same hashtag makes it possible to follow all the presentations. Another account-in our case, the @WIISToronto account-acts as moderator, providing introductory and logistical information about the conference, introducing and thanking presenters and discussants, opening up Q&A sessions, and closing with concluding remarks. The moderators also ensure that the hashtag is always in use, and they retweet with the hashtag when appropriate. They also can intervene in the case of off-topic or inappropriate commentary, although because our model has presenters tweeting from their own accounts, common Twitter strategies such as "block early, block Most obviously, graduate students or precarious faculty may be able to take advantage of more opportunities to share their work and network than may have been accessible to them previously, given the sometimes prohibitively expensive travel costs of an in-person meeting.
often" or "don't feed the trolls" are likely best implemented by the presenter (Center for Countering Digital Hate 2019).
Each presenter is given a 15-minute time slot and asked to prepare a presentation of about 6-12 tweets. During this timeslot, each participant tweets the presentation at the rate of approximately one tweet per minute 2 and ties them together by tweeting them as a thread. The moderator provides clear cues to begin the presentation ("Please go ahead, @handle") and asks each presenter to clearly mark the end of the presentation-with an /end marker, a 12/12 number, or another obvious signal-so that the moderator knows when to move on to the next presentation. After each presentation or each panel, there is a 15-minute discussion slot, during which other Twitter users can direct questions to the presenter(s) using their @handle and the conference hashtag. At the end of the morning and afternoon panels, we allow for 30 minutes of questions and answers.
Several features make the Twitter conference different from a webinar. The first and most obvious is that there is no face-to-face interaction between participants: it is entirely text-based. For some, this may be a preferred mode of communication, but we also know that women and people of color are disproportionately likely to be the targets of violence and toxicity on Twitter (Amnesty International 2018). A shift to digital spaces may only serve to re-create or perhaps even amplify existing inequalities (Burns, Blumenthal, and Sitter 2018).
Second, synchronous and asynchronous participation is seamless. In a webinar, audience members may be able to watch the recording later, but they will not be able to ask questions of the presenter and interact with other audience members. Because a Twitter conference relies on a hashtag, the audience and presenters can interact even after the conference has formally concluded. Although our workshops have always been presented live, it would also be possible for presenters in an inconvenient time zone to preschedule their tweets and then participate in the conversation several hours (or days!) later. This means that the Twitter conference can preserve the immediacy of a real-time interaction-and we would argue that most interactive Twitter conversations still do happen in a constrained period of time, before they are lost to the timeline -while still making it possible for the interaction to continue after the presentation has ended.
Third, the Twitter conference can have a wide impact for a very low cost, even if the individual Twitter users have comparatively small follower counts. A Twitter conference reaches the presenters' mesolevel follower-followee networks, followers of the conference hashtag, and macrolevel hashtag networks beyond the conference hashtag that link presenters to others in their research areas; for example, #cdnpoli, #natsec, #FemSecSt (Burns and Moe 2014). 3 Nonparticipants can easily jump in to comment on a presentation, and if the correct hashtag has not been used, a good moderator can retweet these questions or comments into the hashtag using the reply or quote tweet function. Importantly, presenters can immediately clarify any misinterpretation of their work, mitigating the risk that important details are lost amidst a series of 280-character tweets.
One potential challenge for a Twitter conference is the ephemeral nature of Twitter. As time passes, it can become more difficult to preserve the conference tweets. Some Twitter users use thirdparty tools to delete their tweets after a certain period of time, and others may choose to protect their accounts so that nonfollowers cannot access old tweets. 4 This ephemerality is similar to most conference presentations that are not reproducible. The Twitter platform makes the possibility of creating a permanent record tantalizingly appealing but ultimately imperfect. The use of a unique hashtag will help conference followers find all the presentations and discussions (we have used the same one for all the conferences in our series), but Twitter's search interface can still be confusing because it is a platform designed to be followed in real time. It is also possible to use a scraping tool to find historical tweets associated with a particular hashtag, but even though this creates an archive in the literal sense, it does not create a conference record that can be easily accessed without a certain amount of technical knowledge. Our solution to this problem is to use a thirdparty tool such as Thread Unroller 5 to capture the presentations so that they can be easily viewed (and archived as PDFs for posting elsewhere, as appropriate). However, this does not capture the conversation effectively.
During the second and third iterations of our Twitter workshop, we asked presenters to report to us the number of engagements (defined by Twitter as "clicks anywhere on the Tweet, including retweets, replies, follows, likes, links, cards, hashtags, embedded media, username, profile photo, or Tweet expansion") 6 on the first tweet of their presentation thread: there were an average of 95 engagements with these tweets, much larger than the audience one would get at a traditional conference. 7 The number of "impressions" (views of the tweet) was very much higher, averaging close to two thousand for the first tweet of each presentation thread. Clearly, there is real engagement here.
The Twitter conference can preserve the immediacy of a real-time interaction-and we would argue that most interactive Twitter conversations still do happen in a constrained period of time, before they are lost to the timeline-while still making it possible for the interaction to continue after the presentation has ended.
A good Twitter presentation is unintelligible as another mode of communication, such as a webinar, a blog post, or a scholarly paper.
The most striking thing we noticed after hosting several Twitter conferences was that participants naturally adapted their presentation style to the medium. As McCulloch (2019) has noted, the generations that have grown up with the internet understand that each form of social media has a particular lexicon and a particular mode of communication. A good Twitter presentation is unintelligible as another mode of communication, such as a webinar, blog post, or scholarly paper. It takes advantage of everything that is unique to social media and to Twitter and uses that to make something that communicates scholarship in a unique way. Hashtags, as discussed earlier, bring relevant networks into the conversation but can also be used emotively to add layers of communicative nuance (e.g., using #facepalm or #nothanks to express embarrassment or dislike). Memes use viral, intertextual combinations of images and text to connect the presentation topic to existing pop culture references that can clarify meaning. Similarly, GIFs convey emotional and cognitive reactions (Wagener 2020). For instance, Dr. Stéphanie Martel, a participant in our 2020 workshop, used Marvel movie references in every tweet in her presentation about the rules-based international order (RBIO) in Asia. The Marvel GIFs and references illustrated the RBIO in a way that clarified complicated theoretical concepts. By referring to memes that are already shared by the audience, scholarly material can be communicated more effectively (Wagener 2020, 8).
Twitter conferences are a new phenomenon, but our experience has been rewarding. They present an opportunity to overcome inperson conferences' accessibility barriers and offer a new outlet to showcase and disseminate scholarly work. We hope others can learn from our experience and that this medium will have a place in scholarly communication even as in-person conferences and workshops resume.

CONFLICTS OF INTEREST
The authors declare that there are no ethical issues or conflicts of interest in this research. ▪ N O T E S 1. The workshops were not large; we hosted approximately five to eight presenters and discussants in each one. However, as we discuss later, they had substantial impact in the relevant Twitter communities, including inspiring others to host their own Twitter workshops.
2. This is opposite to the Twitter default user interface, which asks you to type in all your tweets beforehand and then release them at once. We have also learned from experience that it is essential for the moderator and the presenters to draft all their tweets ahead of time, so that they can simply copy and paste them into the Twitter app at the appropriate time. As moderators, we generated a master list of all the tweets from the moderator's account, including promotional tweets to go out for the call for papers and to advertise the workshop, introductory tweets, presenter bios, transitional tweets, and thank yous. Examples of these tweets can be seen by searching the hashtag #WIISTOTC on Twitter. 3. These are the user-developed, commonly used hashtags for identifying content related to Canadian politics, national security, and feminist security studies. 4. See Twitter, "About Public and Protected Tweets," for details on how protected tweets work (https://help.twitter.com/en/safety-and-security/public-and-protectedtweets).
7. Note that this does not necessarily mean that 95 different people engaged with the tweet. Although many tweets can only be done once per user (retweet, follow, like), others can represent more sustained engagement (following the user), and some can be completed multiple times per user (clicking the username or expanding the tweet).