Non-binary people continue to face structural invisibility, and research has only recently begun to examine how their experiences circulate on social media. Centred on visual culture, Instagram's affordances bring into view this community whose visibility has long been limited by institutional and social gatekeeping.This paper analyses how non-binary influencers represent their identities intersubjectively and multimodally on Instagram using an adaptation of Du Bois's (2007) Stance Triangle. The resulting Visual Stance Triangle (VST) operationalises evaluation, positioning and alignment across image-caption units, drawing on visual social semiotics (Kress & van Leeuwen, 2021) and visual intersubjectivity (Zhao & Zappavigna, 2018).Results identify six recurring thematic domains of non-binary visibility: POLITICAL, AESTHETIC, ORDINARY, VULNERABILITY, HOSTILITY and INTERSECTIONAL. Across these domains, visibility varies in explicitness, stance orientation and audience relation. Political posts construct collective visibility and address potential allies; vulnerable posts reframe personal disclosure as shared queer recognition; aesthetic and ordinary posts keep non-binary identity present without making it the central stance object. Overall, the findings show that non-binary visibility on Instagram is both plural and ambivalent: it can challenge binary and transnormative expectations by presenting different ways of being and appearing non-binary, while remaining exposed to misrecognition and hostile interpretations.
This article examines how social virtual reality (social VR) mediates discourse by reconfiguring access to the person, place, and time coordinates on which deixis depends. The primary framework is mediated discourse analysis, with deictic-field theory, frame and footing analysis, indexicality, multimodal interaction, and affordance research used as supporting concepts. The data comprise 48 headset-based sessions from VRChat and Horizon/Workrooms (approximately 10.9 h) and eight Zepeto post-thread units used as an avatar-centred social virtual-world boundary case (approximately 1.2 h of associated source media; approximately 86,700 words and captions overall), supplemented by researcher fieldnotes on platform operation. Analysis of demonstratives, pronouns, address forms, controller gestures, portals, mute events, and explicit ‘in real life’ (IRL) marking shows how participants establish, contest, and repair an avatar-centred interactional ‘here’. The virtual deictic field names this wider, mediated organisation of reference and participation. Building on blended-origo accounts of co-available physical and virtual coordinates, it brings referential access, identity ratification, participation rights, and repair into one analytic model. Observable orientation in talk and action is treated as interactional evidence, not as a direct measure of psychological presence. The findings show how platform resources support referential coordination, the assignment of participation roles, and identity ratification, shaping how discourse is produced and interpreted.
Danmu is a time-anchored commenting system where comments are anonymous and displayed on screen as videos unfold. Unlike conventional comment sections, danmu organises participation through playback-time alignment, scrolling visibility, and weak addressivity, making it a distinct online communicative context where interaction takes a highly fluid form. Drawing on the framework of ambient affiliation, this study traces the dynamic enactment, clustering, and negotiation of bonds over time in a sample of 9555 danmu comments collected from a food vlog on Bilibili, with the aim of exploring the interactional formation of communities in danmu. In the case analysed, six recurrent bonds are identified across three targets—i.e., vlogger(s), food, and vlog characters. As the vlog unfolds, these bonds cluster as transient peaks or sustained low-intensity currents. When plurality arises, bonds are negotiated either by rejecting alternatives and sharpening oppositional stances, or by sustaining multiple orientations without polarisation. The findings are interpreted in the context of liquefied sociality with respect to a bond-centred mechanism of community building shaped by danmu's particular affordances: fragmented and condensed evaluations become recognisable as bonds and then cluster and undergo negotiation in the comment flow.
This article focuses on curated lists of newsbites (hyperlinked headlines accompanied by ledes or images) sent in emails by the official account of The White House during the first administration of U.S. President Donald Trump (2017–2020). Building on previous research on how evaluative language constructs power and organizes sociality in mediated texts, I use the Appraisal framework to identify how news events are interpreted from both partisan and non-partisan media newsbites. Identifying patterns in comparisons highlighted by newsletter newsbites, I show that intensifying and specifying lexis in headlines and ledes bolsters positive, non-normative imaginings of the president or administration's activities. Findings suggest that newsbites' graduated evaluations increasingly constructed partisan notions of ‘real news’ from 2017 to 2020 by indexing chronotopes (timespaces) of an idealized American present that frame events like tax cuts, impeachments, and a global pandemic in upscaled evaluations similar to Trump's own linguistic style. By arguing that such graduated meanings emphasize temporal shifts that divide society similar to other populist and authoritarian discursive strategies, this study contributes to scholarship on how newsbites in particular can facilitate the strategic deployment of chronotopic references by those in power to deepen political polarization.
This article examines asynchronous mukbang (eating shows) as sites of video-mediated interaction that constitute a popular part of digital culinary practices. By analyzing the fine-grained sequencing of embodied eating and talk, our approach considering discourse in its mediated and para-interactional qualities shows how the mukbang performers choreograph a continuous flow of talk-like activity despite the bodily constraints of eating and despite the absence of an actual co-present interlocutor, thereby constructing the felt experience of shared mealtime para-interaction. Drawing on qualitative sequential analysis of selected examples drawn from a corpus of asynchronous mukbang videos posted on YouTube, we analyze speech, food, and eating-related action entanglements that are performed for the mediated entertainment of others. Our results show that mouth actions and in particular chewing serve important multifunctional purposes, as they are used for/as 1) a central performative element and turn emulation, 2) embodied and mediated tasting, and 3) the creation of suspense and structuring of narrative. Eating and tasting sequences as performed jointly with speech in mukbang are transformed into publicly accountable and sensorially accentuated performances that are integral to the format of eating shows as digitally mediated interaction. In addition, the article introduces a novel multimodal transcription system that is tailored to the transcription of speech and food actions in mediated discourses.
