Problematizing cultural difference: YouTube narratives about COVID-19 by South Korean and American Vloggers

ABSTRACT

In this study we compare 8 American and 8 South Korean vlogs by YouTubers documenting their experiences of COVID in 2020. We propose a nuanced approach to cross-cultural difference that rejects binaries and involves the study of how communicative resources are used by participants in specific events. We detail similarities and differences between the vloggers, from trends in the distribution of topics to use of visual and linguistic resources. Drawing on a narratives as practices approach and multimodal discourse analysis, we demonstrate that differences in the data are related to various scales that vloggers navigate in their experiences with COVID.

본 연구는 2020년에 코로나에 감염된 미국인 유튜버 8명과 한국인 유튜버 8명의 브이로그를 분석, 비교한다. 한국과 미국 유튜버들의 브이로그에서 다뤄지는 주제, 시각적 구성, 언어적 요소 등의 공통점과 차이점을 면밀히 살펴봄으로써, 보다 맥락에 민감하게 문화적 차이를 접근하는 방식을 제시하고자 한다. 내러티브를 실천으로서 바라보는 관점에서 브이로그를 접근하여 한국과 미국 유튜버들의 브이로그는 그들이 마주하는 서로 다른 사회문화적 현실의 여러 단면을 반영한다는 것을 확인하였다.

KEYWORDS:

• Storytelling
• narrative
• COVID-19
• intercultural communication
• social media
• vlogs

Introduction

The COVID-19 Pandemic that took over the world in 2020 and is still raging in many countries has been a watershed moment for the entire globe. Besides killing millions of people, the pandemic has unsettled social relations by creating profound social divisions and exasperating inequality (Bringel & Pleyers, 2022). It has altered every aspect of social life by confining people to their homes and isolating them from social interaction, while at the same time elevating the virtual to the main form of communication, and has brought entire countries to a halt. In that sense, COVID has been described as a truly global phenomenon (Aarts et al., 2021). Yet, the very fact that the pandemic has reached every corner of the world has also brought to light huge differences in government policies and perceptions by people in different countries. Not only have countries chosen entirely divergent ways of dealing with the pandemic, but their citizens have also shown remarkably different levels of acceptance of those measures and dissimilar behaviours vis-a-vis the new social conditions in which they find themselves.

A great deal has been written both at an academic level and in the popular press on these differences and, as we will discuss in our paper, much of the literature on this topic resorts to so called ‘cultural differences’ that divide countries into wide and unspecified categories such as collectivist or individualistic, or high and low context, authoritarian or liberal. Yet, little has been written on how such differences are manifested in concrete domains of communication and how they can be explained. In this study, we analyse vlogs filmed by eight American and eight South Korean YouTubers to document their experience as COVID patients, to identify specific patterns of similarity and difference between the vloggers. Unlike other studies of so-called ‘cultural differences,’ we consider not only the content of the stories told, but more in general the use of linguistic and multimodal resources by the vloggers and the specific ways in which they appropriate the affordances offered by YouTube. Thus, we propose a rich qualitative analysis of our data. In doing so, we aim to add a more nuanced understanding of the practices and material conditions that may be at the base of the difference in how the coronavirus was experienced by specific groups who live in these two national contexts. We will argue that what is generally described as ‘cultural difference’ is the result of the confluence of a variety of factors at different scales that include, amongst others, the national scale of policies by government, the local management of such policies, the concrete conditions in which the pandemic was experienced, the communicative practices shared by vloggers in those different contexts, and their particular ways of relating to their audiences. It is our contention that qualitative, context sensitive analyses can lead to a much greater and deeper understanding of the roots of ‘difference.’

The initial idea for this comparison came from a study of Korean vloggers’ narratives about COVID conducted by one of the authors, which promoted the question of whether American vloggers would offer different or similar accounts. We were also interested in this comparison given that, as mentioned above, the US and Korea are often associated with binary cultural frameworks of individualism vs. collectivism. In conducting a close analysis of the multimodal narratives told by these vloggers about their illness, we demonstrate that even in this relatively small set of data, there exist nuanced patterns of similarity and difference that require consideration of more factors than ‘national culture.’

Our paper is organised as follows: we first describe the general conditions of the response to the pandemic in the US and Korea. In sections 2 and 3, we introduce some theoretical and methodological background to our work. Specifically, we discuss our positioning vis-à-vis research and theorizations in intercultural and cross-cultural communication and our conception of storytelling as a semiotic and communicative practice, and we review studies of storytelling in the frame of the COVID-19 pandemic. In the following sections, we present our methodologies and data analysis. Section 6 is devoted to discussion and conclusions. Below we describe responses to the pandemic in Korea and the US.

The coronavirus pandemic in Korea and the United States

This study focuses on examining experiences of COVID documented during the first year of the pandemic. We chose to limit our dataset to vlogs that were filmed between March 2020 and January 2021 as the discourse surrounding the pandemic has changed dramatically since vaccines were made widely available in early 2021. In this section, we provide a short overview of what the pandemic looked like in both countries during 2020, in terms of their daily case count and the policies they implemented to prevent the spread of the virus.

The US and Korea both drew much international attention for how they dealt with the start of the pandemic, albeit for very different reasons. As is well documented in media reports all over the world, the US fared shockingly poorly in the early phase of the pandemic. The first case of COVID in the US was discovered on January 20th, 2020; only three months later, the US surpassed Italy as the global leader of deaths due to COVID (CDC, 2022). The number of cases declined somewhat between April and June, but spiked up again in July and December, concluding the year with a record high of 251,190 cases on December 18th.¹ Korea’s response, on the other hand, was evaluated as exceptionally successful. After the first case was discovered on January 20th, Korea saw a rapid uptick of cases; in response, the government quickly launched into action by implementing an extensive testing and contact-tracing system, and within a month was able to gain some control over the pandemic. In late February, Korea was second only to China with 800+ daily cases, but the case count dropped to approximately 100 by mid-March, and to less than 50 by April. The number of cases fluctuated somewhat for the rest of the year, but the count remained quite low in comparison to other countries, with its peak around only 1,000 cases per day towards the end of 2020.

The difference between the measures that were taken in the US and in Korea in response to the pandemic was as stark as the difference in their case counts. The contact tracing that Korea implemented was an effective system but also an invasive one, involving the review and public release of personal information of diagnosed patients. While many lauded the success of Korea’s contact-tracing system, several news articles (e.g. Thompson, 2020) also made note of the risk to privacy it could pose due to the amount of information released. In contrast, in the US, even less invasive measures such as masking and temperature checks were protested against by some as an imposition on freedom (Aratani, 2020). Due to the opposition against those measures, many US states chose not to implement state-wide masking or contact-tracing policies.

Importantly, the framing of freedom in these different responses to COVID was a theme picked up by many discussions comparing the policies of different countries. Korea was counted among the countries that were willing to sacrifice individual freedoms for societal security, while the US was taken as representative of a society in which people were fighting for their freedom in response to government-imposed restrictions. Many reporting on this issue described it as a distinction between the ‘East’ and the ‘Western world,’ or between ‘collectivist’ and ‘individualistic’ countries (e.g. Lewis, 2020; Thompson, 2020). However, this framing of cultural differences is quite simplistic. As we will discuss below, what is described simply as ‘cultural differences’ between the two countries’ responses to the pandemic is the product of a variety of divergences in social practices and experiences across multiple scales.

Review of relevant literature

Intercultural difference

In this section, we review literature on interculturality and present our own view of how cultural difference should be understood. Investigations that focus on comparisons between people’s perceptions and reactions to social phenomena in different countries are often placed in the category of cross-cultural or intercultural studies (Hall, 1973), depending on whether they deal with people in separate national contexts or in situations of interaction. And indeed, in a recent article, Kulich et al. (2021) articulate the need for scholars to re-examine intercultural relations in light of the COVID pandemic. To strengthen this point, they quote from a publication of the Canadian Psychological Association, stating that ‘the pandemic, along with the measures taken to combat it, is shaped in important ways by culture … . policymakers, healthcare workers, and the public at large should keep in mind that the pandemic experience may be very different for different people’ (Ryder et al., 2020).

Similarly, the few existing works in the social sciences that compare aspects of the pandemic in different countries immediately invoke cultural factors to explain the way discourses about COVID are organised and communicated. Very few socio or applied linguistic studies on this topic are available, but those we have found resort to wide ranging binary oppositions that have traditionally been used in cross-cultural and intercultural studies. For example, in an investigation of Facebook narratives by Filipino and American COVID survivors, Bautista (2021) resorts to contrastive rhetoric (Kaplan, 1966), arguing that narratives told by Filipinos adhere to a ‘high-context’ cultural style of writing which, in her words, is more detailed in that writers feel the responsibility of getting themselves understood. On the other hand, Americans write in a ‘low-context’ style, which gives the burden of understanding to the reader. High versus low-context is just one of the many binaries used in literature that deals with cross – or intercultural communication. In research on neologisms about the pandemic in Russia and the UK, Ponton and Davletshina (2022) resuscitate Hofstede’s (1980) distinction between individualistic and collectivist countries in explaining how populations in those nations conceptualised the pandemic through new words.

Research conducted in areas such as crisis management or health studies employ the same kinds of approaches. For example, Bajaj et al. (2021) who analysed leaders’ communications with publics about COVID in different countries, mention oppositions between high – or low – context cultures, face-saving or direct communication cultures, and self-independent or interdependent ones. Gelfand et al. (2021) suggest a relationship between the health outcomes of the pandemic and cultural ‘tightness-looseness,’ which refers to the willingness by citizens to follow social rules and ability of their governments to impose them. Similarly, Hahn and Bhaduri’s (2021) study on mask behaviour proposes a relationship between wearing masks and belonging to individualistic or collectivistic cultures. Although these investigations do not come from socio – or applied linguistics, they show the influence and perdurance of theories that reduce cultures to general norms and values and that identify cultural groups with national communities.

These overarching and binary conceptions of culture are still underlying inter and cross-cultural research even though critical approaches to intercultural studies – which have become more and more mainstream in socio and applied linguistics – caution researchers about the dangers of reducing cultures to nationality and of reproducing widespread stereotypes about national character. Dervin and Risager (2014) reject ‘discourses of the inseparability of language and culture, not least when understood in nationalistic and territorial terms,’ (p. 5) advocating instead for a nuanced and contextualised understanding of cultural identities. In a recent paper on this topic, Zhu et al. (2022) discuss the dangers of reductive views of culture in which gross generalisations about national groups are made. They argue that such generalisations are often based on stereotypes and presuppositions and prevent researchers from understanding the complexities of cultures. Worse, these ideas often directly lead to the marginalisation of minority groups by strengthening polarising oppositions between ‘us’ and ‘them’ and reinforcing ethnocentric divides (also see Jones, 2013 on this point).

Similar views have been expressed by non-essentialist, postmodernist thinkers such as Holliday (2022), who cautions against ‘grand narratives’’ of nation and civilization that have been put at the service of colonialist views of the world and have been used to divide the West from the rest of the world. In a more radical development, some scholars such as Baker (2022) have proposed to do without the term ‘intercultural’ which still evokes clearly separate and bounded groups and substitute it with the term ‘transcultural’ communication to focus on multicultural contexts across multiple spatial-temporal scales. All these conceptualizations point to a problematization of culture as a unified system of beliefs, ideologies, and practices.

Much of the critical perspectives reviewed above focus on multilingual contexts or at least on contexts where language is a central element for discussing diversity. Our research does not consider language per se, as we look at semiotic communication including discourse but also multimodal resources. We also do not deal with interactions between US and Korean vloggers, but rather study their vlogs separately. Our objective is to understand whether people belonging to specific social groups who live in different national contexts present their experiences with COVID in different ways, even though they all participate in the wide transnational domain of the internet and experience the same kind of health emergency. We ask whether there are intragroup commonalities in such discursive and semiotic presentations and in that case how they can be explained. Nonetheless, we take inspiration from critical perspectives on interculturality in embracing a non-reductionistic and complexity driven view of differences.

Indeed, in our perspective it would be easy to attribute intragroup commonalities in discourse and semiotic practices to some kind of pre-supposed cultural homogeneity, as implied in much of the literature on COVID and inter or cross-culturality. However, we argue that understanding what can be perceived as ‘cultural difference’ in a non-essentialist way implies taking into consideration the converging influence of a variety of contexts at different scales. Such scales include, the local scale of the communicative practices, also involving the use of different semiotic resources that are common among vloggers in Korea and in the US, and the way vloggers relate with their audiences, the scale of the public health policies put forth by the national or local governments and the widely shared expectations about appropriate behaviour in the face of this particular health emergency, the scale of the concrete conditions in which they experience the pandemic, the physical environments in which they are isolated and their access to medical facilities.

In that sense, culture is always a mix of materialities and expectations, and it is always relative to specific groups even within the borders of a nation and cannot be equated with systems of beliefs, ideologies, and practices that homogeneously characterise all the citizens of a nation. Rather, the commonalities that define specific groups’ semiotic practices stem from the interaction of the group members with these different communicative and material contexts. Thus, instead of presupposing cultural differences, we investigate them from the bottom up and we also do not imply that the differences that we find between vloggers in the US and in Korea can be generalised to all health narratives by Korean and American citizens.

Narratives and storytelling

A thorough overview of linguistic research on illness narratives is beyond the scope of this paper, but as Jones (2015) has noted, discourse analytic research on illness narratives provides important insight into the ways in which speakers’ accounts of illnesses are socially occasioned and embedded. Jaworska (2018), for instance, examines online narratives about postnatal depression to demonstrate how storytelling allows women to appropriate and challenge hegemonic discourses about motherhood. Jones (2021) analyses Boris Johnson and Donald Trumps’ narratives of COVID-19 discussing how they draw on stances indexically associated with masculinity in their narratives to present themselves as strong leaders and shift attention away from the failures of their governments. Our study similarly explores how the online narratives of COVID patients shape and are shaped by the social contexts in which they are told.

Research on COVID discourse has thus far examined diverse contexts of the pandemic, ranging from the discursive production of ‘pandemic culture’ in paediatric care consultations (Ekberg et al., 2021), the use of COVID-related metaphors in Arab cartoons (Abdel-Raheem, 2021), to the expression of stances towards pandemic related chronotopes (see De Fina & Perrino, 2022). Studies such as these have vividly illustrated how various domains of discourse shape and reflect perspectives of the pandemic; however, they have largely focused thus far on political or media discourse (e.g. Billig, 2021; Sambaraju, 2022), perhaps due to the politicisation of the pandemic in many countries around the world. Likewise, many of the studies examining narratives about COVID have focused on storytelling in political discourse (e.g. Jones, 2021; Vásquez, 2021) or on exceptional experiences with the virus, such as ‘long COVID’ (e.g. Miyake & Martin, 2021). Relatively little attention has yet been paid to the ways in which COVID patients talk about their everyday experiences, which is what we turn to in this study.

Narratives in general are seen as important loci for the construction of identities (De Fina, 2015) and as tools that people use to make sense of experiences (Ochs & Capps, 2001). Contrary to biographical approaches that see narratives as more or less mediated expressions of the identities of individuals, interactionally oriented studies of narratives (see De Fina & Georgakopoulou, 2011) focus on their social functions and on the ways in which stories are embedded into concrete contexts of communication and therefore are the result of interactional processes involving tellers, audiences and a variety of contexts and affordances. In that way, storytelling is seen as a form of communicative and semiotic practice (De Fina, 2021), rather than as the expression of individual identity processes.

Taking this approach of analysing narratives as practices (see below for further elaboration), we look at narratives told by vloggers as involving curation by the teller who constructs them with their audiences in mind and also in constant interaction with them. Vlogs are not only multimodal in the sense that tellers use a variety of semiotic resources that include oral and written language, visual elements and sound to construct them, but also in the sense that tellers take advantage of the affordances offered by YouTube and follow the practices that are usual in the performance of this particular genre of narratives. While producing them, however, they also continuously alter and innovate so that those practices are in constant flux. Following this view, we take into consideration the content of the stories (what vloggers talk about), the how they tell them (resources and affordances), and the ways in which they relate to their audiences. By paying attention to ‘meta-talk’, that is talk about the activity of vlogging itself, we tap into how vloggers conceptualise what they are doing.

Data & methods

The data used in this study comprises 16 YouTube vlogs, all of which document vloggers’ experiences being diagnosed with and recovering from COVID from March 2020 to January 2021. As mentioned previously, we chose to collect vlogs from this time frame because experiences of the virus changed dramatically in early 2021 with the development of vaccines. Eight of the vlogs that we collected were posted by American vloggers, and eight by Korean vloggers. The eight vlogs that make up each group can be further divided in half based on the time frame in which they were uploaded to YouTube. Four vlogs in each group were filmed and posted during the early stages of the pandemic (March – July 2020 in Korea, May – August 2020 in the US) and the other four were from the later spike in both countries between November 2020 – January 2021. The time frames from which the vlogs were collected were based on the official COVID case count for both countries (i.e. based on when each country experienced a spike), then adjusted for how many vlogs were available for those time frames.

The vlogs we examine are a relatively small number; this is because very few vlogs on COVID were uploaded by Korean vloggers early in the pandemic, likely due to the severe stigma associated with being infected with the virus during that time. To address this issue to the extent that we could, we began our data collection by identifying as many vlogs as was possible for this subset, which ended up being four. The vlogs for the rest of the subsets were collected to match the first subset as closely as possible. All of the vloggers included in the dataset are in their 20s to 30s, and do not have any professional background in medicine. We also tried to match the number of male and female vloggers for each group, but it proved difficult to find male vloggers for the American subset. Another characteristic that was difficult to match across the dataset was the number of subscribers that the vloggers had; while most of the Korean vloggers had relatively few subscribers and few videos on their YouTube channels, the American vloggers seemed to have built up a substantial following over the long term.

Table 1 shows a summary of the relevant information for the vlogs used in this study. Based on ethical considerations of online data use, all of the vloggers have been assigned pseudonyms, and screenshots of the vlogs presented in the analysis have been edited to conceal the vloggers’ faces. Our use of this data was approved by Georgetown University’s Institutional Review Board (STUDY00004420).

Table 1. Summary of information on vlogs.

	Upload date	Video length	Gender	# of subscribers^2
KR early
Min	Mar 20, 2020	3m 3s	M	10.1K
Dahee	Apr 27, 2020	22m 32s	F	1.04K
Joonwoo	May 2, 2020	15m 32s	M	2.05K
Yoona	Jul 3, 2020	9m 54s	F	Unknown
KR late
Chan	Dec 15, 2020	6m	M	697
Sooji	Dec 18, 2020	3m 25s	F	182
Rina	Dec 20, 2020	10m 1s	F	303
Kim	Dec 24, 2020	25m	F	12.2K
US early
Ava	Jul 3, 2020	18m 4s	F	9.4K
Emily	Jul 5, 2020	21m 14s	F	95.4K
Julie	Jul 31, 2020	10m 31s	F	211K
Mark	Aug 2, 2020	8m 57s	M	1M 650K
US late
Nicole	Nov 21, 2020	11m 6s	F	209K
Vanessa	Dec 11, 2020	19m 29s	F	104K
Sarah	Dec 13, 2020	12m 22s	F	268K
Maria	Jan 11, 2021	20m 29s	F	326K

We use a combination of quantitative and qualitative methodology to analyse this data. On the quantitative side, we use coding of different elements through MAXQDA. In particular, we coded the number of times certain categories representing particular narrative units appeared in the data, the length of speech devoted to certain topics, the presence of specific linguistic elements. The quantitative analysis was used to see whether there were patterns of any kind between and within the two groups. We used qualitative analysis in order to see how each vlogger organised their narrative and how different communicative modes were employed by them in individual narratives.

For the MAXQDA analysis, we coded the transcripts for the topics they touched on (e.g. symptoms, food, etc.) and the time that each vlogger spent talking about those topics, the emotions expressed by vloggers (e.g. gratitude, concern, etc.), and relevant linguistic devices (e.g. modals, impersonal reference, etc.). Finally, we analysed vloggers’ metacommentary about their vlogs, i.e. their talk about how they had organised the vlog, and the explanatory comments they directed to their followers (see Appendix for a complete list of codings). We focused not only on the content of the narratives but also on the way semiotic resources were used by vloggers and on the context of their strategies of production of the vlogs and relationships with audiences. Both authors coded a small sample of the data independently of each other to check intercoder reliability, which was 95 percent. The rest of the data were divided between the two authors to code, and then swapped around for the other to review. The coding of any segments that prompted disagreement was decided through discussion, and the entire set of data was reviewed whenever there were any new codes that emerged.

We then conducted qualitative analyses of the data focusing on overall organisation and focus of the vlogs, conditions in which the vloggers were filming and the more detailed ways in which they used semiotic resources such as camera perspective or captions. Our qualitative analysis is based on a view of narratives as practices (De Fina, 2021), as previously discussed, combined with insights from multimodal perspectives on discourse and communication. The narrative as practices orientation, is a particular form of discourse analysis that combines an ethnographic perspective with an analysis of discourse at different levels: from the generic organisation of narratives, to their content, to the function of linguistic resources within particular narrative structures, to the analysis of interactional patterns and of how these are understood by participants. For this reason, the level of metacommentary is important in this study, because it allows us to get to participants’ views. Indeed, narrative practices are seen as involving narrators and audiences as both actual interlocutors (in face-to-face storytelling) and imagined ones (in virtual storytelling).

A narrative as practices orientation is sensitive to the use of particular narrative resources in different contexts, and, for this reason, we combine it with multimodal approaches to the study of communication. Multimodal analysis is a necessary tool for the study of vlogs, as the latter combine different modes and media. Indeed, as noted by Frobenius (2014), vloggers use ‘multimodal elements that are regularly part of spoken interaction, such as gaze shifts, shifts in posture, shifts in facial expression, shifts in voice quality and pitch and also pointing’ (p. 59). But they also use other modes such as written captions and camera movement, their own and other objects’ position in space to convey point of view and to establish relationships with their audience. For example, in her analysis of Instagram images, Zappavigna (2016) shows how image producers can create different interpersonal relationships with their viewers by representing themselves and the objects they interacted through a variety of camera angles and through manipulating their own position in the photographs. The use of these resources reflects the different objectives that they pursue through their narratives. In our case, the use of multimodal resources was important to characterise the focus of attention of the vlogger and the ways they conceived of their relationship with audiences. For example captions were particularly important for Korean vloggers but not for their American counterpart, while ways in which camera movement incorporated objects in the environment also revealed different communicative objectives for vloggers with regards to their audiences and foci of attention. Thus, as proposed by Jewitt, Bezemer, & O'Halloran (2016), we analysed modes separately at first and then tried to combine the analysis to understand how they contributed to meaning.

Analysis

Our analysis outlines the differences and similarities between the 16 vlogs. We begin with an overview of the overall distribution of topics discussed in the vlogs, then move on to look into vloggers’ use of semiotic resources such as video and captions, and linguistic devices such as modals. We conclude by examining vloggers’ meta-comments about their vlogs, which concern the relationships they construct with their audiences and the actions filmed within their vlogs.

Distribution of topics

The illness narratives told by American (hereafter US) and Korean (hereafter KR) vloggers touch on similar topics overall, but devote different degrees of attention to each of those topics. A summary of the distribution of topics in the data sample is given in Table 2. The table does not cover all the topics discussed in the dataset, but just the three most common topics: symptoms, food and descriptions of space/medical equipment. In the following subsections, we analyse the distribution of each of these topics in more detail, in order of their frequency in the vlogs.

Table 2. Overall distribution of topics.

	US vlogs			KR vlogs
	Avg	Min	Max	Avg	Min	Max
Symptoms	18.5% (n = 8)	8.03% (1m 27s)	31.29% (2m 48s)	9.57% (n = 8)	1.67% (10s)	19.21% (2m 59s)
Food	12.71% (n = 8)	1.02% (13s)	43.47% (8m 27s)	10.34% (n = 6)	0.59% (8s)	17.97% (1m 28s)
Space and medical equipment	2.85% (n = 4)	0.24% (3s)	9.13% (1m 39s)	14.29% (n = 6)	3.11% (29s)	32.38% (3m 14s)

Talk about symptoms

Given that our data consists of illness narratives, the most frequently featured topic was, predictably, vloggers’ talk about the COVID symptoms they experienced. Table 3 shows a summary of the average, maximum, and minimum amounts of talk about symptoms in each group of vlogs. We also divided the vloggers’ talk about symptoms into two sub-categories, based on whether the talk occurred before or after the vlogger received their COVID-19 diagnosis. We introduced these sub-categories because vloggers’ descriptions of their symptoms at these two points in time seemed to serve different purposes; vloggers’ talk about symptoms before their diagnosis usually provided the rationale for why they thought to self-isolate or get tested for the virus, while their talk about symptoms after their diagnosis seemed to provide documentation for the trajectory of their illness.

Table 3. Comparison of talk about symptoms.

		US vlogs		KR vlogs
		Early	Late	Early	Late
Before diagnosis	Avg	16.26% (n = 4)	2.86% (n = 4)	1.94% (n = 3)	5.49% (n = 4)
	Min	6.18% (1m 7s)	0.23% (15s)	0.97% (9s)	0.67% (4s)
	Max	29.80% (2m 40s)	4.72% (48s)	4.44% (1m)	16.67% (1m)
After diagnosis	Avg	2.13% (n = 4)	15.74% (n = 4)	8.06% (n = 4)	3.64% (n = 4)
	Min	0.48% (3s)	9.67% (1m 53s)	1.52% (9s)	0.56% (2s)
	Max	4.71% (1m)	25.61% (3m 10s)	18.24% (2m 50s)	8.13% (2m 2s)

Overall, US vlogs featured much more talk about COVID symptoms than KR vlogs. Both US and KR vlogs also showed differences between the early and late groups, in terms of the average amounts of time spent discussing symptoms before and after the diagnosis. For the KR vloggers, the early group had more talk about symptoms after the diagnosis than the late group, as most of the early vloggers had been diagnosed after flying to Korea and were asymptomatic at the point of diagnosis. The reverse was true for US vloggers, with the late group talking more about their symptoms after their diagnosis, which seems to reflect the fact that testing became quicker and more available later in the pandemic in the US.

Talk about food

Food was another topic that was frequently featured, as vloggers often documented the food they ate as an important part of how they took care of themselves during their illness. Table 4 shows the amount of time each group spent talking about food.

Table 4. Talk about food.

		US vlogs		KR vlogs
		Early	Late	Early	Late
Talk about food	Avg	3.48% (n = 4)	22.94% (n = 4)	6.36% (n = 2)	11.83% (n = 4)
	Min	1.02% (13s)	10.81% (1m 12s)	0.59% (8s)	6.11% (22s)
	Max	4.28% (23s)	43.47% (8m 27s)	12.12% (1m 12s)	17.97% (1m 28s)

US vloggers overall spent more time talking about food, although there were similarities between the groups in that the late groups for both KR and US vloggers spent more time talking about food than the early groups.

One qualitative difference between talk about food in the KR and US vlogs was that the KR vloggers tended to talk about food as a part of the medical treatment that they were given at the treatment centre. In most of the segments of talk on food in KR vlogs, the vloggers described receiving food as part of their daily treatment routines, as shown below:

(KR) Sooji:

8 시에 아침식사 하라는 방송을 해주셨어요,

At 8am they gave the announcement for breakfast,

샐러드와 빵, 사과, 차, 간식, 과자

Salad, bread, apple, tea, and snacks

이렇게 준비해 주셨습니다.

Are what they prepared for us.

On the other hand, none of the US vloggers were isolated in any sort of hospital or official facility, and so had to prepare their food themselves. Five out of the eight US vloggers talked about the food that they ordered through delivery apps, walking their audience through their thought process behind purchasing particular items:

(US) Vanessa:

Got black bean quinoa chilli?

It's a vegan, already made soup,

I wanted to get things that were already made,

and that would be really easy to reheat,

There were also some differences between the ways in which vloggers filmed the food they bought or ate. Half of the vlogs included video of food. As shown in Table 5, videos of food made up a slightly higher percentage of KR vlogs than US vlogs.

Table 5. Video of food.

		US vlogs	KR vlogs
Video of food	Avg	6.60% (n = 4)	7.25% (n = 4)
	Min	2.03% (22s)	0.67% (10s)
	Max	33.88% (6m 36s)	35.02% (3m 28s)

A more significant difference is the ways in which KR and US vloggers filmed their food, an example of which is shown in Figure 1. The image on the right shows a screenshot of a KR vlogger’s video footage of the food she received while in treatment at a hospital, while the image on the left shows a US vlogger showing off her purchases from a grocery delivery service.

Figure 1. How food is presented in a KR vlog (left) and in a US vlog (right).

The screenshot of the US vlog centres the vlogger even in their footage of food, whereas the screenshot of the KR vlog shows just the food without the vlogger. To draw on Zappavigna’s (2016) discussion of point of view in social media images, the video footage used in US vlogs present more visual cues such as vloggers’ faces or parts of their bodies that directly inscribe the subjectivity of the vlogger. The video footage in the KR vlogs, in comparison, include fewer such cues and tend to instead draw on captions (as in the above image) to signal the vloggers’ subjectivity. In their video footage, as in their talk about food, the US vloggers again seem more focused on documenting the vlogger’s personal experience of illness, whereas the KR vloggers seem to construct a ‘fly on the wall’ perspective of a patient in a treatment centre.

Talk about space and medical equipment

Another frequent topic featured in the videos was the vloggers’ descriptions of the spaces in which they stayed – their homes for some, the hospital or other treatment centres for others – and the medical equipment they used, such as oximeters and thermometers. On this topic, the difference between the KR and US vloggers was even more stark, as can be seen in Table 6.

Table 6. Talk about space and medical equipment.

		US vlogs		KR vlogs
		Early	Late	Early	Late
Talk about space and medical equipment	Avg	3.71% (n = 3)	0.24% (n = 1, 3s)	5.87% (n = 3)	22.74% (n = 3)
	Min	0.74% (4s)		3.11% (29s)	6.07% (1m 31s)
	Max	9.13% (1m 39s)		10.28% (2m 19s)	32.38% (3m 14s)

Overall, there was much more discussion of medical equipment and space among the KR vloggers; there were only four US vloggers who mentioned the topic at all, compared to six KR vloggers. The KR and US vloggers also diverged in whether talk about space and equipment was more frequent in the early or late groups. Among US vloggers, the topic was mentioned more frequently in the early vlogs, likely because new equipment such as oximeters were introduced to the public early in the pandemic. The opposite was true for KR vloggers, who had more coverage of space and equipment in the late vlogs. Three out of the four vloggers in the late KR group had just recently been moved into a treatment centre at the time of filming, and were introducing their new surroundings in their videos.

The amounts of video footage of space and medical equipment mirrored the amount of talk. As shown in Table 7, there was much more footage in the KR vlogs; only one US vlog included relevant footage, which featured the vlogger showing her new oximeter to her audience.

Table 7. Video of space and medical equipment.

		US vlogs	KR vlogs
Video of equipment	Avg	8.49% (n = 1, 1m 32s)	15.38% (n = 4)
	Min		3.11% (29s)
	Max		24.39% (50s)

The video footage of space and equipment also showed differences similar to that found in video of food. Figure 2 presents two screenshots, one of a KR vlog’s footage of medical equipment and the other from a US vlog. KR vlogs appeared to focus on showing the equipment and space, whereas the only US vlog that included video of equipment showed the vlogger lifting up her oximeter near her face as she explained to her audience what it was for.

Figure 2. How equipment is presented in a KR vlog (left) vs. in a US vlog (right).

Even when KR vloggers appeared on screen alongside medical equipment, the vlogger did not engage directly with the audience to speak to them about the equipment. Rather, they appeared on screen only to demonstrate how the equipment was used, as shown in Figure 3.

Figure 3. Use of medical equipment in KR vlogs.

In terms of the model proposed by Zappavigna (2016) for analysing images, these depictions, like the ones we discussed in the case of the food presented, establish different interpersonal relationships of the vloggers with their audience. Indeed, by looking directly into the camera and by showing themselves next to the equipment, US vloggers seem to invite greater involvement by the viewers, while Korean vloggers seem to establish a more ‘objective’ relation with their viewers as they do not engage in eye contact when showing themselves and or the objects surrounding them. Rather, they seem to adopt the viewpoint of the audience observing them.

Linguistic details

Next, we discuss the linguistic resources used by vloggers to discuss these topics in their narratives, focusing specifically on their use of modals. We coded the data for other linguistic patterns as well, such as passive constructions or impersonal pronouns; however, we only include modals in our analysis because they appeared the most frequently and showed meaningful patterns of similarity and difference.

Modals

We examined the data for two groups of modals: ‘can’ and ‘must.’ Each of these groups also included negative constructions of those modals, as well as words or phrases that have a similar meaning such as ‘be able to’ for ‘can’ and ‘should’ for ‘must.’ Table 8 summarises the frequency of both types of modals in US and KR vlogs.

Table 8. Use of modals.

	US vlogs (n = 8)				KR vlogs (n = 8)
	Total	Min	Max	Median	Total	Min	Max	Median
Can/ cannot/be able to	72	3	18	8.5	87	1	29	8.5
Must/ should	34	1	8	4.5	28	0	14	2.5

Both types of modals were used similarly in terms of their overall frequency. However, there were some differences that are not visible in the above table, such as that there were more instances of ‘can’ used when talking about symptoms in the US vlogs (39 times) than in KR vlogs (26 times). Of the symptoms, the most frequently described were the vloggers’ temporary loss of smell & taste (US: 15 times, KR: 15 times), with the vloggers often talking about testing out their experience of specific kinds of tastes or food, as in the following example:

(US) Vanessa:

I can like taste the consistency of the peanut butter?

but I can't taste it.

Let me see the blueberries … .

I can taste the tartness,

That this particular symptom was mentioned so many times in both US and KR vloggers’ narratives indicate how ‘tellable’ a part of their experience of COVID it was.

Other instances of ‘can’ and ‘cannot’ that were not related to symptoms in the US and KR vlogs were fairly similar in terms of the kinds of actions they described (e.g. can’t open window, can’t meet people), but they focused on different factors constraining the narrators’ agency. In the KR vlogs, agency was often limited by official rules, whereas US vloggers were constrained by their lack of energy or exercising self-control:

(US) Nicole:

I need to open my window

<I can’t open my window.>

(KR) Dahee:

창문을 열 수 없기 때문에,

Because you can't open the windows,

환기가 안되는 그런 공간에 있습니다.

I'm in a space that can't be ventilated.

Uses of ‘must,’ on the other hand, did not show significant quantitative or qualitative differences between the two groups of vlogs.

Meta-comments about the vlog

We now turn to examine the ways in which the vloggers explicitly commented on various aspects of their vlogs. We first discuss how the vloggers’ constructed their relationship with their audience, for instance by answering their questions or giving advice. We also examine the vloggers’ comments on their actions presented in the vlog.

Relationship with the audience

As previously mentioned, one key difference between the vloggers included in the KR and US datasets was that the US vloggers on average had a much higher number of subscribers than the KR vloggers; KR vloggers had about 3800 subscribers on average, while US vloggers had an average subscriber count of 359,100. All of the US vloggers were already active as vloggers before documenting their experience with COVID, while COVID vlogs were the first videos uploaded on five of the eight KR vloggers’ channels.

Because this was the case, the ways in which the two groups of vloggers oriented to their audience were quite different. The vloggers’ meta-comments in which they addressed their audience directly (e.g. to comment on their recent lack of activity on YouTube or to thank their followers for watching), were quite similar in terms of their content across US and KR vlogs. Such meta-comments, however, were more frequently present in US vlogs than KR ones; for instance, three US vloggers commented on their lack of activity on YouTube prior to uploading their COVID vlog, as in the excerpt below:

(US) Emily:

It's been so long since I’ve vlogged

or talked to any of you guys.

In addition, three other US vloggers mentioned in their vlogs that they had previously alerted their audiences on Instagram or other platforms that they had tested positive for COVID and were not able to upload vlogs as usual. In contrast, only one of the KR vloggers made reference to their video uploading schedule; the vlogger happened to have the most subscribers in the KR group, which suggests that this pattern is linked to how established the vloggers’ YouTube channels are.

The US vloggers also included much more advice to their audience in their vlogs. Seven US vloggers included advice, compared to four KR vloggers. In their advice, there were also differences in the content they covered. The KR vloggers only gave advice either on how to avoid getting infected or on what to bring to the health facilities, while US vloggers also included advice on medicine and on mental health of self and others, as in the example below:

(US) Sarah:

It's really really really lonely.

so I just want to say,

if you know someone who currently has COVID,

send them a text,

give them a call,

A similar pattern of difference was found in the vlogs’ use of questions by the audience. Overall, not many vloggers incorporated questions asked by the audience into their vlogs; only three US and two KR vlogs did so. There were thus no significant quantitative differences between the two groups, but the kinds of questions that the vloggers responded to were quite different. The questions that KR vloggers answered had to do with COVID symptoms, detection and with the facilities in which patients are interned:

(KR) Dahee:

정말 무증상이면 아무 증상이 없나요?

If you're asymptomatic do you really not have any symptoms?

어떤:

:—어떤 건가요, 라고 말씀을 해주시는데,

what:

:–what's it like, is what you asked,

In the case of the US vlogs, however, questions were much more varied, ranging from symptoms to how the vloggers were able to feed themselves, what kinds of medicine they took, or how they organised their time, as shown in the example below:

(US) Sarah:

<<Reading from phone>How did you spend

your time in quarantine,

and how did you keep a positive mentality.>

One last type of meta-comment about the relationship between the vlogger and their audience is the ‘epistemic disclaimer,’ which involves clarifications on the part of the vlogger on their epistemic status, i.e. on their knowledge about medical issues. In our data, these disclaimers took the form of declarations about the fact that the vloggers are not doctors and do not possess specialised knowledge. There was a clear difference between the US and KR vlogs in the use of such disclaimers; they were present in five out of eight US vlogs, but absent in all the KR vlogs. All instances of such disclaimers were very similar to each other, as shown in the following:

(US) Sarah:

I'm not a doctor, obviously.

I don't have all the answers,

I don't know.. every term, and whatnot.

This is what personally happened to me,

US vloggers seem to be worried to be the target of criticism or even legal action if they give dangerous or unfounded advice to others. This was evidently not a preoccupation shared by Koreans, at least not among the vloggers examined in this study.

Comments on actions in the vlog

Vloggers’ comments about their actions in the vlog were perhaps the most common type of meta-comment, with six US and seven KR vloggers using some form of such comments. These comments generally referred to background information on what was happening in the immediate context in which the vlogger was talking, orientations on time and day, explanations about what the vlogger was doing, etc. However, the type of comments and their range were quite different, with US vloggers covering a much wider spread of topics.

The US vloggers commented on topics such as their feelings, antecedents to an action, actions happening (e.g. driving to testing facilities), references to days of isolation, food and medicine ordered, technical aspects of vlogging, the environment (such as noise), their looks, or other people in the vlog:

(US) Vanessa:

I realised, after I got this call,

that I don't have any groceries.

<<Vlogger shows groceries delivered in bags at her front door>

So I – I actually did my first grocery delivery,

I've never gotten my groceries delivered,>

So I'm gonna show you what I ordered.

Korean vloggers’ comments covered a more reduced number of topics, such as technical aspects of the vlog, reasons for vlogging, and details about being hospitalised, as shown in the example below. Only one case each occurred of the vlogger commenting on their feelings or looks:

(KR) Joonwoo:

오늘은 입원한지 약 3주차 … 가량 된 날입니다.

and today marks around 3 weeks … at the hospital.

Discussion

We now move on to discuss some of the scales that seem to have shaped these patterns of similarities and differences between the US and KR vloggers’ narratives. The scales we discuss are the public health policies put forth by the national and local governments, the concrete conditions in which the vloggers were isolated, local conventions of vlogging, and the online communities the vloggers have created on their individual YouTube channels. These scales are not an exhaustive list of the relevant factors shaping the vloggers’ narratives, but rather a summary of some of the main factors that are often lumped together and dismissed as ‘cultural difference’ in discussions of diverse narratives and experiences of the pandemic.

Of these various factors, public health policies on both the national and local level, especially on the accessibility of COVID testing and of medical care, seem to have played an especially salient role in shaping the vloggers’ narratives. For instance, because KR patients were given professional care in hospitals and treatment centres, there was much more talk about space and medical equipment in the KR vlogs, and more use of the modal ‘can/cannot’ to talk about government policies or hospital regulations. Due to the slow and inaccessible testing in the US early in the pandemic, the early US vlogs included more talk about the vloggers’ symptoms before they received their diagnosis compared to other groups of vlogs. As testing later became more accessible, however, the distribution of topics in the US vlogs shifted to reflect that reality. The policies in these two countries had a clear impact on the ways in which the vloggers were able to respond to their diagnoses, and so seem to have dictated much about how their experience of COVID unfolded overall.

Related is the scale of the material environments from which the vloggers were filming, in the sense that the vloggers’ environments were shaped by how their governments decided to treat COVID patients, and how those policies were carried out on a local level. US patients isolated at home, while KR patients were put in hospitals if they had severe symptoms and in treatment centres if not. However, further variation also existed depending on whether the US vloggers were based in suburban or urban areas, or if the KR vloggers were in hospitals or treatment centres. If a KR vlogger was assigned to a treatment centre temporarily converted from a hotel or apartment, as was the case for most of the vloggers in the late KR group, they tended to devote more of the video to displaying the space itself. The vloggers’ meta-comments on their actions were also shaped by their physical context, in that the US vloggers were able to roam around a much wider range of spaces, such as to go to testing sites, and so commented more on their whereabouts, whereas KR vloggers were bound to their government-assigned rooms. The many points of similarity and difference in the vloggers’ narratives shaped by their physical environments speak to the extent to which it impacted their experiences of COVID.

Another relevant local scale is that of the conventional communicative practices among US and KR vloggers. As discussed in the analysis, the two groups of vloggers seemed to have different conventions on how to film their food and medical equipment, and how much to center themselves in such video footage. Another difference visible in the screenshots presented in the analysis is the heavy use of captions in KR vlogs, sometimes to the point of vloggers choosing to replace speech altogether with captions in narrating their experiences. The US vlogs in our dataset, in contrast, only rarely used captions, usually to label the date at which the video was filmed. This particular practice is likely a carry-over from Korean entertainment programs, which tend to rely heavily on captions as a means to build a clear narrative arc (Lee, 2020). Yet another aspect of our data that may have been impacted by the vloggers’ conventional practices is the lack of male vloggers in the US dataset. The relative rarity of male US vloggers may have to do with common perceptions of which subjects are considered appropriate for female and male vloggers based in the US and KR. These patterns in the data highlight that the vloggers’ narratives of their experiences of COVID need to be understood in terms of not only how their experiences unfolded, but also how they choose to narrate those experiences in different ways.

Of course, the communicative practices of the vloggers must also be understood in context of the scale of the communities that the vloggers have built around their individual YouTube channels. The vloggers we examined varied in terms of how much they spoke directly to their audience, or what kinds of topics they mentioned in their advice to their audiences. US vloggers generally seemed to touch on more personal topics, such as mental health struggles, and more often addressed their audiences directly; however, KR vloggers who had a more substantial following on their channels also showed a relative tendency to act in a similar manner to US vloggers. This pattern suggests that the vloggers who had established a personal connection to their community of viewers were constructing their narratives differently from the vloggers who did not. To take this idea further, the specific kinds of identities the vloggers have established, such as ‘lifestyle vlogger’ or ‘travel vlogger,’ may also have shaped their narratives in subtle ways, although a detailed discussion of those distinctions is beyond the scope of our analysis.

Our discussion has provided a brief overview of some of the scales that may have impacted the ways in which the US and KR vloggers’ narratives were constructed; while limited in scope, it nonetheless illustrates how a complex web of material and communicative conditions come together to shape the vloggers’ narratives of their experiences of COVID. While the vlogs are too few to make consistent comparisons of two national cultures, the data is sufficient to demonstrate how various scales come together to form nuanced patterns of difference between the two groups of US and KR vloggers, as well as within each group, for instance between the ‘early’ and ‘late’ groups of vloggers. As previously mentioned, it would be easy to talk about the narratives of these vloggers in terms of broad cultural orientations such as ‘individualism vs. collectivism’; some of the points of difference that we have identified, such as the relative focus on personal decisions and emotions in US vlogs and the ample coverage of community rules in KR vlogs, certainly lend well to such distinctions. However, to talk about the data in those terms would obscure the many factors that play a role in shaping experiences and perceptions of COVID-19, such as local policies on the pandemic, patients’ material environments, and narrators’ multimodal storytelling practices. By tracing the impact of these various scales in relation to the patterns of similarity and difference identified in the data, our study highlights the necessity of conducting context-sensitive, detailed analyses in examining cultural difference.

As previously discussed, it is our general view that communicative practices are inserted within and therefore are shaped by and shape social contexts at different scales that are characterised by specific material conditions, social relations, power dynamics and ideologies. In the case of the data we have analysed, narrative vlogs were shaped and at the same time helped shape the ways in which the COVID-19 pandemic was experienced in different countries, shared ideologies about the origins of illness, its effects and the modalities for dealing with it, the social position of vloggers in relation to other actors. At a more local level, these communicative practices are also shaped by and shape the conventions of vlogging and current ways of curating and presenting the self within those practices. In our view, a theory of discourse needs to show how local meaning making relates to all these contexts at different scales and ultimately to what extent it represents a break from or a solidification of expected patterns.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Jungyoon Koh

Jungyoon Koh is a PhD candidate in Linguistics at Georgetown University. Her research interests include identity construction in social media discourse and human-computer interaction.

Anna De Fina

Anna De Fina is Professor of Italian Language and Linguistics in the Italian Department and affiliated faculty with the Linguistics Department at Georgetown University. Her interests and publications focus on discourse and narrative, identity, chronotopes, migration and super diversity. She has published extensively on these topics including many articles in internationally recognized journals and 12 volumes between authored and edited books. She is co-editor of the book series Encounters for Multilingual Matters and Discourse, Narrative and Interaction for Routledge. Her latest publications are the Cambridge Handbook of Discourse Studies, coedited with Alexandra Georgakopoulou (2020, Cambridge University Press) and Exploring (Im)mobilities: Language Practices, Discourses and Imaginaries, co-edited with Gerardo Mazzaferro (2021, Multilingual Matters).

Notes

1 All case counts are from the COVID-19 database provided by JHU CSSE.

2 Subscriber count accessed on May 11, 2022.

Appendix

Complete coding system for MAXQDA analysis.

• COVID knowledge

  • Medical professional

  • Vlogger’s knowledge

    • ▪ Vlogger’s lack of knowledge

  • Others’ experience

    • ▪ Hearsay

    • ▪ Others’ lack of knowledge

• Emotions

  • Relief/ Happiness

  • Frustration

  • Scared/Concerned

  • Gratitude

• Path to diagnosis

  • Receiving results

  • Exposure to/of other people

  • Guesses on path of infection

  • Activity prior to test

  • Precautions

  • Contact-tracing

  • Cause for test

• Symptoms

  • Before diagnosis

  • After diagnosis

• Treatment

  • Interactions with other people

    • ▪ Video of other people

  • Official policies

    • ▪ Established practices

    • ▪ Official rules

  • Food

    • ▪ Video of food

  • Medical equipment

    • ▪ Video of equipment

  • Other kind of equipment or space

  • Medicine

    • ▪ Side effects of medicine

• Meta-information

  • Meta-comments about the audience

  • Epistemic disclaimer

  • Questions from the audience or others

  • Meta-comment on actions in the vlog

  • Why vlog?

  • Structure of video

• Closing

  • Advice to audience

• Modals

  • Would

  • Want

    • ▪ Wish/hope

  • Can/ cannot/be able to

    • ▪ Could-hypothesis

  • Have to/ must/should

• Generic reference

• Passive