The value(s) of social media rituals: a cross-cultural analysis of New Year’s resolutions

ABSTRACT

New Year’s resolutions are acts of valuation where people express ideas about what is important and worthwhile in life. Although resolutions have a long history, the twenty-first century has transformed the practice into a social media ritual with greater visibility, interactivity, and reach. Using this unique event to explore the globalization of values, we analyze tweets about New Year’s resolutions in English, German, Italian, Japanese, and Korean. Combining network analysis (n = 160,592) and content analysis (n = 2000), we compare discursive topics, modes of ritual participation, and the values expressed in resolutions. Our findings indicate both that the ritual crosses cultures and that there are language-specific dynamics that do not map neatly onto established divisions between ‘Eastern’ and ‘Western’ value orientations. Instead, we identify three underlying tensions organizing the articulation of values: self-acceptance vs. self-improvement, public vs. private, and conformity vs. oppositionality. We discuss these in relation to an overarching tension between local contexts and global platform cultures. Finally, we explore the study’s broader implications for understanding the interaction between values, norms, and global communicative practices.

KEYWORDS:

• Digital culture
• globalization
• norms
• resolutions
• social media ritual
• values

New Year’s resolutions are not new. The ritual declaration of commitments crosses centuries, calendars, cultures, and modes of communication. Yet the twenty-first century has significantly changed the enactment of the practice, transforming it into a ‘social media ritual’ (Burgess et al., 2018) where people around the world publicly express what they care about in an interactive and highly-visible environment. Building on notions of social constructionism, which stress the role of communication in the articulation of shared understandings of the world (Berger & Luckmann, 1966), we contend that social media rituals not only express but actively participate in shaping what people consider important. Thus, New Year’s resolutions offer a unique site to explore the construction of values on social media.

Given our concern with communication, we adopt a pragmatist approach (Heinich, 2020; Hutter & Stark, 2015) that moves away from conceptualizing values as abstract mental representations (i.e., beliefs that someone has) towards concrete instances of valuation (i.e., practices that assign or establish worth). Research in this latter tradition has examined valuation through measurement, attachment, and judgment, although rarely in the context of social media. Conversely, cross-cultural investigations of values on social media tend to be unconcerned with valuation and instead operationalize values as predictive variables for other communicative practices (e.g., Park et al., 2014). Drawing on surveys of national values, such studies explore differences between so-called ‘Eastern’ and ‘Western’ countries (Hofstede, 2003). Yet the scale and global structure of social media platforms challenge the premise of discrete national cultures, necessitating examination of the complex relationships between global platforms and local practices (see Miller et al., 2016; Zidani, 2020).

Our study compares tweets about New Year’s resolutions in five languages: English, German, Japanese, Italian, and Korean. This design allows us to investigate global and local dimensions of this ritual by combining the most popular language on Twitter (English) with four languages tied to specific geographic regions that, according to existing research, represent different value systems. We begin by reviewing the literature on cultural values and social media rituals, highlighting the potential of New Year’s resolutions for a bottom-up comparative analysis of valuation. Next, we outline our method of data collection and analysis, which includes a combination of semantic network analysis and content analysis. Across languages, we compare conversational topics, ways of participating in the ritual, and the values expressed in resolutions. Through an integrative analysis of the results, we identify three tensions underlying the construction of values: self-acceptance vs. self-improvement, public vs. private, and conformity vs. oppositionality. We also posit an overarching tension between national and platform-related values to help explain the discrepancies between the cross-cultural differences we identify and the conventional division between Eastern and Western value orientations. Finally, we discuss the study’s broader implications for understanding the intersection of values, norms, and global communicative practices.

Literature review

The communication of cultural values on social media

The relationship between communication and culture is reciprocal, meaning that the ‘culture in which individuals are socialized influences the way they communicate, and the way that individuals communicate can change the culture they share over time’ (Gudykunst, 1997, p. 327). However, given the inherent complexity of such processes, research tends to focus on either how the culture of socialization shapes communication or how communication practices shape culture. To facilitate investigation, a significant body of research operationalizes culture through the concept of values, generally understood as core notions about the desirable shared by a group of people. Comparative analyses of cultural values are primarily concerned with the culture-to-communication route of influence where values predict communicative behaviors. Many of these studies focus on national groups, drawing upon large-scale survey research to establish the basic values (Schwartz, 2012) or cultural dimensions (Hofstede, 2003; Inglehart, 1990) associated with different countries.

A large portion of such cross-cultural analyses stress the opposition of Eastern and Western countries, typically characterized by the distinction between collectivistic (or other-directed) and individualistic (self-directed) value orientations (Hofstede, 2003). Differences in national culture are often established via Hofstede’s cultural dimensions, although other distinctions such as high-context vs. low-context communication styles (Acar & Deguchi, 2013) and tightness vs. looseness of culture (Ackland et al., 2019) also appear. This work offers evidence that national cultures shape how people use and perceive different aspects of social media such as emoticons (Park et al., 2014), self-presentation (Yoo et al., 2012), and selfies (Ma et al., 2017; Souza et al., 2015). Although cultural dimensions have some predictive power, there is no overarching account of how national cultures influence communicative practices on social media.

This lack of a clear story may be tied to theoretical issues with Hofstede’s model of cultural differences and its applicability to social media practices. As some studies readily acknowledge, Hofstede’s framework has been widely criticized for its essentialist notion of culture (Macfadyen, 2011) and the ethnocentrism of its analytic categories (Fougère & Moulettes, 2007). The context of social media further complicates the applicability of the national culture model (Shifman, 2016), as social media platforms facilitate the creation of geographically dispersed cultures of socialization, meaning that people may be exposed to values associated with different parts of the world. As such, social media provides us with the opportunity to examine whether the global conditions of communication contribute to the shaping of new, cross-national patterns of valuation.

The complexity of digitally mediated life is evident in some studies that apply Hofstede’s model to social media and fail to find or find only partial corroboration for it, including the discovery that both presumably ‘collectivistic’ Japanese and ‘individualistic’ American students regularly post self-related and self-promotional tweets (Acar & Deguchi, 2013), as well as the lack of a meaningful relationship between the individualism score of a nation and the prevalence of selfies (Souza et al., 2015). These are not the only instances where cultural dimensions fail to predict communicative differences, but they highlight the importance of the specific norms of social media platforms and the complexity of communicative practices. In other words, cultural values associated with social media platforms such as individualism and authenticity (Senft & Baym, 2015) seem to intersect with the values of national cultures, affecting how people communicate online.

The interplay between globalizing platform-related values and local contexts has been recently invoked by Sangeet Kumar (2021) who connects three dominant modes of social media communication – the selfie, live-streaming vlogs, and influencer culture – to the core values of individualism, self-disclosure, and entrepreneurship that originate in Euro-American ideas of modernity. These values are closely aligned with the capitalist logics governing social media platforms, promoting notions of autonomous individuals acting freely to fulfill their goals. The propagation of these genres inevitably contributes to the globalization of the values associated with them. As Kumar (2021) explains, local modes of social media expression are ‘in dialectical interplay with and layered over the common genre of looking into a camera and broadcasting one’s life,’ (p. 161). Other accounts of the interplay between global genres and local contexts have been offered by Nissenbaum and Shifman (2018), who show how standardized meme templates allow for subversive expressions of emotions across different national contexts, and Zidani (2020), who shows how Arab youth remix popular culture to craft political messages and create community.

Following this emergent trend of treating social media genres as sites of globalization, we call for inductive approaches to the cross-cultural study of values on social media to complement existing top-down approaches. Thus, while this study builds on observations derived from prior work to select languages and evaluate our results, the main categories we use to analyze values have been developed inductively with sensitivity to the genre of communication.

Ritualized valuation through New Year’s resolutions

An interest in the social construction of values leads us to treat values as dynamic features enacted through communication rather than as things that are static and known in advance. This shift aligns with the movement from values to valuation in the sociological literature. Associated with the pragmatist tradition, the growing interest in valuation revives John Dewey’s insight that value is ‘a quality that has to be performed’ (Hutter & Stark, 2015, p. 2). Accordingly, studies of valuation tend to be qualitative and inductive, starting from specific evaluative situations (Lamont, 2012).

While evaluative situations are endemic to almost any human communication, rituals provide a compelling but under-analyzed site of valuation. From celebrations to commemorations, rituals express cultural values, helping to establish who people are and how they should relate to one another. Much academic interest in rituals can be traced to Durkheim’s foundational book The Elementary Forms of Religious Life, which analyzes how the enactment of rituals constitutes moral communities (Bellah, 2005). Where Durkheim was primarily concerned with religious rituals, Goffman (1982) extended the subject into more secular and mundane domains, focusing on the patterned ways people establish social worth in everyday life.

Rituals have conventionally been understood as place-bound, yet communicative technologies change the scale at which rituals can happen. Building on Dayan and Katz’s (1992) conceptualization of media events, Burgess et al. (2018) define social media rituals as ‘platform-specific social media conventions and patterned responses to acute events’ in which personal audience practices construct public discourse (p. 230). While social media rituals involve claims of solidarity and expressions of the good, the openness of social media platforms means that the purpose of the ritual and its governing norms are subject to negotiation (Highfield, 2016).

Among social media rituals, New Year’s resolutions are a good case study because they make valuation practices explicit. Resolutions emerge out of a rare period of reflection when people think about what worked well in the past year and what should be changed in the future. Values are formed and communicated in these instances of individual and group reflection. New Year’s resolutions are thus a perfect laboratory to explore the formation and dissemination of values.

Although academic research on New Year’s resolutions is scarce and typically focused on the US (e.g., Bernstein, 1977; Zeligs, 1964), it has identified enduringly popular topics such as smoking cessation (Norcross et al., 1989) and weight loss (Marlatt & Kaplan, 1972), both of which emphasize self-improvement. Despite the narrow geographic focus of academic research, the practice of making New Year’s resolutions is widespread. News outlets have documented the popularity of this ritual in each of the countries included in our study: Germany (e.g., Meldung, 2019), Japan (e.g., Nishioka, 2021), Korea (e.g., Kim, 2009), and Italy (e.g., Fraschini, 2020). Such widespread attention is perhaps unsurprising given that ‘New Year’s Day is the most celebrated holiday in the world, and the only truly global one,’ (George, 2020). The significant transnationality of New Year’s resolutions makes the ritual ideal for comparative research, with the expected commonalities of the event promising to place cultural differences in high relief.

In what follows, we approach the practice of sharing New Year’s resolutions on Twitter as a ritual of valuation through three questions. First, what do people talk about when they discuss New Year’s resolutions on social media? With this question, we wish to compare the discourse surrounding the ritual and identify common and culturally distinct topics. Second, how do people participate in the ritual? In particular, we wish to understand whether this ritual includes practices besides sharing serious resolutions such as conversations with others or critical reflections about the ritual itself. Finally, which values are constructed through this ritual? Here, we wish to understand the extent to which users across the globe converge on what they portray as important in life. An integrative analysis of these domains should expand our understanding of the relationship between social media rituals, values, and globalization.

Method

Data collection and sample

To comparatively investigate the practice of sharing New Year’s resolutions on Twitter, we employed a theoretically-driven sampling strategy designed to surface national cultural differences. We collected tweets in five languages: English, German, Italian, Japanese, and Korean. The multi-lingual dataset brings together the most popular language on Twitter (English), which is used globally, with four languages tied to specific geographic regions. Languages serve as important proxies for culture online (Matassi & Boczkowski, 2021; Rogers, 2013), especially when the language and region are strongly linked. Additionally, according to surveys of national values (Hofstede, 2003), these languages represent divergent cultures: Germany and Italy are associated with Western, individualistic value orientations, and Japan and Korea are associated with Eastern, collectivistic value orientations. Finally, to reduce the influence of other variables, each of the four languages are associated with countries that have democratic governments, high rates of Internet penetration, and relatively open Internet policies.

Since English is not tied to a single nation, we examined the national context of the tweets in the dataset. Looking at the profiles of a random sample of 150 tweets, we found 104 accounts that included city location information. Of these profiles, the majority were located in the U.S. (59%) or the U.K. (17%), with the remaining accounts (24%) located in 17 different countries including Australia, India, and Finland. We also examined the location of the tweets generated in German since this language is prevalent in several neighboring countries. An analysis of the 2024 profiles from the German dataset with geolocation information revealed that 83.2% percent of the accounts listed locations within Germany and 10.2% were from Austria or Switzerland. These analyses help contextualize our claims about the relationship between languages and national contexts on Twitter.

Despite our efforts to select languages associated with countries with comparable government structures and contrasting cultural orientations, it is important to note that Twitter users have distinct demographic qualities that may vary between countries (e.g., Barberá & Rivero, 2015). Twitter market research demonstrates the variable popularity of the study’s languages, ranging from 47.05 million English-speaking users to 2.35 million Italian-speaking users (Kemp, 2020). While there is no comprehensive comparative data on user demographics, we find consistent indicators that Twitter users are younger than the general population in the U.S. (Wojcik & Hughes, 2019), Germany (ARD/ZDF, 2021), Italy (Vaccari et al., 2013), Korea (Roh, 2020), and Japan (Ministry of Internal Affairs, 2020). There is also evidence that Twitter users tend to be more educated and politically left-leaning in the U.S. (Wojcik & Hughes, 2019) and Italy, at least among users that talk about politics (Vaccari et al., 2013). Although we cannot assume that the people who share New Year’s resolutions are statistically representative of the broader population in each country, we are interested in the culture of the platform and the factors that distinguish Twitter users make them well-suited for cross-cultural investigation. This approach has been adopted in prior research on Twitter, including cross-cultural comparisons of network engagement patterns (Garcia-Gavilanes et al., 2013) and emoticon use (Park et al., 2014).

A multilingual team consisting of native speakers of the study’s languages worked together throughout the project, employing a constant-comparative approach. We collected data on Twitter from 25 December 2019–3 January 2020 searching for the word ‘resolution’ in five languages using the software Mozdeh. We then selected tweets that contained a combination of the words ‘2020’ or ‘year’ in each language. The final dataset (N = 160,592) included 136,165 English tweets, 4768 German tweets, 7984 Italian tweets, 2606 Japanese tweets, and 9069 Korean tweets.

Analysis

Semantic network analysis

We employed semantic network analysis to compare the topics of tweets about New Year’s resolutions. To construct the networks, we generated a list of word frequencies for each language. After screening out common stop words, we individually reviewed each of the most frequent words, combining direct synonyms and alternative forms. This iterative process resulted in lists of the top 50 keywords in each language. Next, we identified the frequency of keyword co-occurrences using the R package Quantenda and entered the resulting matrix into Visone (https://visone.info/) to generate the maps. We improved the legibility of the networks by reducing the number of edges while maintaining the network structure following established procedures (Segev, 2020) and identified topical clusters using the Louvain modularity algorithm.

Content analysis

1. Modes of Participation. To investigate the second research question about modes of participation, we performed content analysis on a random sample of tweets (n = 1000, 200 per language). Coding was based on four categories detected in a preliminary investigation of the corpora: (a) resolution, (b) conversation, (c) metacommentary, and (d) other. ‘Resolution’ tweets contained at least one personal goal such as my New Year’s resolution is to read more books. ‘Conversation’ tweets featured interpersonal conversations about resolutions such as I hope that you are able to keep your resolution to read more books. ‘Metacommentary’ tweets featured public commentary on the general practice of New Year’s resolutions including humorous or sarcastic responses, such as my New Year’s resolution is to eat more books. All remaining tweets, including commercial promotions and generic holiday wishes, fell into the ‘other’ category. To assess reliability after a training period and pilot testing, five coders analyzed 50 tweets from the English dataset, resulting in an 0.77 score (Krippendorff’s alpha). We supplemented the coding with a close reading of the tweets.

2. Value Construction. To address our third research question on the construction of values, we created another subset of tweets (n = 1000, 200 per language). Tweets were randomly sampled and screened to include only sincere resolutions given the difficulties in conducting a systematic content analysis of humorous content. Although most survey research approaches values as principles, or the abstract criteria by which worth is established, people rarely invoke such principles explicitly in natural communication – a situation even more pronounced in short-form communication like tweets. Accordingly, we adopted a more tangible understanding of values drawing on Nathalie Heinich’s (2020) notion of values-as-goods: those (tangible or abstract) objects that have consistently been deemed worthy of appreciation such as art or friendship. We collaboratively developed the codebook of values based on prior schemas (Baden & Springer, 2017; Boltanski & Thévenot, 2006) and a preliminary investigation of the tweets. Our value codes capture what specific resolutions frame as important, or worthwhile, for the upcoming year. For example, a resolution about reaching 50,000 subscribers on YouTube frames recognition as important. Table 1 includes definitions of the codes and examples from the dataset, translated to English when relevant and lightly modified to protect users’ privacy (Markham, 2012). After a training period and pilot tests, five coders analyzed a sample of 50 English tweets, resulting in an intercoder reliability score of 0.80 (Krippendorff’s alpha). We proceeded to code the resolutions and conducted a cross-tabulation analysis with a posthoc analysis of a chi-square test to compare values across the five languages. Again, we supplemented the quantitative analysis with a close reading of the tweets.

   Table 1. Value coding scheme with definitions and examples.

   Value	Definition	Example
   Money	The valuation of a luxury lifestyle, or the ability to have and spend a lot of money.	New Year’s Resolution: making a lot of money for real and be successful.
   Culture	The valuation of making, consuming, and experiencing culture (e.g., books, sports, television, or video games).	It’s a day late but happy new year! This year’s resolution is to just draw …
   Self	The valuation of the self, typically in the form of caring for, loving, or expressing oneself.	This year I just believe in myself: That I am lovable, funny, eloquent and also valuable in my melancholy phases. #selflove
   Pleasure	The valuation of enjoyment or sensory pleasure.	My 2020 resolution is to chill as much as possible.
   Transcendence	The valuation of spiritual practices or religious organizations.	My New Year's Resolution is to keep praying and wait for good things to come.
   Relationships	The valuation of close personal relationships, typically in the form of initiating or strengthening connections.	Resolution for 2020: be more present for the people I care about by taking time away from work.
   Recognition	The valuation of fame, attention, or other forms of recognition.	Alright here is my new year resolution/goal for 2020! Hopefully getting 50k subscribers on YouTube.
   Collective Well-being	The valuation of shared or collective benefit, typically in the context of social or political causes.	My resolution for the year 2020: I want to go to Tohoku. Our debut year was the year of the earthquake so I want to contribute to those areas too.
   Effort	The valuation of focus, determination, and effort in pursuit of goals.	Happy new year 2020. My resolution: stay as motivated as possible.
   Body	The valuation of health and/or appearance.	This year’s resolution for the first time after living 20 odd years is to ‘become thin.’

Results

The universe of topics in resolution discourse

What do people on Twitter talk about when they mention New Year’s resolutions? Comparing the 50 most frequent keywords in each language (see appendix), four terms appear across all five languages. The core discourse frames the New Year as an opportunity for new beginnings (‘start’), a declaration of effort (‘work’), and a focus on well-being (‘health’ and ‘happy’). The thirteen keywords shared among four of the five languages add greater detail to the picture. New Year’s resolutions are organized around the individual (‘myself’) and focus on personal cultivation (‘book,’ ‘learn,’ ‘continue,’ and ‘read’), financial decisions (‘money,’ ‘buy,’ and ‘spend’), relationships (‘love’ and ‘friend’), the body (‘eat’ and ‘smoking’), and – reflecting a more recent development – social media use (‘Twitter’). The proportion of unique terms varies across the corpus, from 30% in English to 52% in Japanese. Some unique terms have clear place-based associations (e.g., K-pop supergroup BTS in Korean or praying at shrines in Japanese) while others are less specific (e.g., environmental politics in German or being positive in English).

Despite commonalities in vocabulary, the semantic network analysis (see Figure 1) allows us to identify unique structural features of resolution discourse in the five languages. The Japanese network is highly centralized, with all resolution topics organized around the central node of ‘myself.’ This is striking because Japanese is known for pronoun drops, where the focus is on context rather than the self (Kashima & Kashima, 1998). Yet alongside the explicit focus on the self, the strong connection of ‘myself’ with words such as ‘full effort’ and ‘work hard’ suggests that many resolutions are specifically about self-improvement, which may be motivated by social expectations. The English network is organized around two distinct nodes: ‘start,’ which is associated with activities (e.g., ‘exercise,’ ‘books’), and ‘myself,’ which is associated with states of being (e.g., ‘happy,’ ‘positive’). The Korean and German networks exhibit more decentralized structures. The Korean discourse is organized around ‘hard work,’ ‘study,’ ‘exercise,’ and ‘money,’ keywords that emphasize the exertion of effort and financial considerations. The German discourse is organized around more open-ended terms such as ‘learn,’ ‘realize/achieve,’ and ‘health.’ Finally, the Italian network is the most distributed, indicating less hierarchy among goals that range from losing weight to meeting new friends.

Figure 1. Semantic networks of keywords.

Note: Semantic networks of the 50 most frequent words used in 2020 New Year’s resolution tweets in five languages. The size of each node is proportional to its betweenness centrality (a measure of the extent to which it is central to paths in the overall network). The width of a tie between two words is proportional to the number of tweets containing both.

Modes of participating in the social media ritual

While semantic networks illustrate prominent topics of conversation, they do not reveal how people talk about these topics. To better answer our second question, we conducted a content analysis of different modes of participation. In all languages, sincerely sharing resolutions constituted the largest proportion of tweets. The emphasis on sincere resolutions is a distinguishing feature of the Japanese dataset at 63%. Metacommentary tweets were most frequent in the Italian dataset, comprising 30% of the sample. Beyond the differences in distribution, we noticed a distinction between serious and sarcastic versions of metacommentary. In the Korean, Japanese, and German datasets, almost all of the metacommentary was serious in tone, featuring reflections on the challenges of keeping resolutions and links to related news articles or social media content from other platforms. In contrast, we found frequent humorous or sarcastic responses in the English and Italian datasets, including complaints about new people crowding the gym or jokes about starting a pizza-only diet. The conversation category primarily involved people asking and responding to questions about other people’s resolutions. It was significantly more prominent in the German dataset, an unexpected result since prior research has found that people from so-called collectivistic countries are more likely to mention other Twitter users in posts than people from individualistic countries (Garcia-Gavilanes et al., 2013). Interestingly, conversations in the Japanese context were largely addressed to celebrities including fictional anime characters, most of whom are popular in Japanese subcultures. (Figure 2).

Figure 2. Modes of participation.

Note: Modes of participation in the 2020 New Year’s resolution social media ritual based on a sample of 200 tweets per language (n = 1000). The crosstabulation analysis shows that there is a significant difference between modes across five languages (χ²[12] = 73.23, p < .001). The results of the post hoc analysis are as follows: for resolution, Japanese is a major contributor to the chi-square value (standard residuals [SR] = 5.71, p < .001). For meta-commentary, Italian is a major contributor (SR = 4.20, p < .001) and the German score is a negative predictor (SR = −3.65, p < .001). Lastly, for conversation, German is a positive predictor to the chi-square value (SR = 3.96, p < .001) and Japanese is a negative contributor (SR= −1.71, p < .001).

Constructing values through New Year’s resolutions

In this section, we paint a broad picture of the values associated with sharing New Year’s resolutions on Twitter as outlined in Table 2. Overall, the primary values associated with New Year’s resolutions are effort (21.3%), understood as the focus and determination involved in pursuing goals; the self (20.9%), understood as caring for, loving, or expressing oneself; and culture (18.9%), understood as making or consuming cultural goods. The prominence of these values aligns with many of the keywords identified in the semantic network analysis, sometimes overlapping directly (‘effort,’ ‘self’), and in other cases, constituting a set of nouns (e.g., ‘books,’ ‘games’) and verbs (e.g., ‘draw,’ ‘write’) that specify what different values (e.g., ‘culture’) entail. Although slightly less prominent, resolution tweets also establish the importance of the body (13.9%) and relationships (10.5%). Given that most of the existing New Year’s resolutions research focuses on goals like smoking cessation and weight loss (Marlatt & Kaplan, 1972; Norcross et al., 1989), it is somewhat surprising that the body is not more prominent. Additionally, the body represents a particularly ambiguous object of value, encompassing concerns from mental health to physical appearance which align with a broad range of underlying value principles. Finally, and far less frequent, is the value of collective well-being (4.4%), which was significantly boosted by its prominence within the German corpus (see below).

Table 2. A cross-tabulation of values across five languages.

Values	German	Italian	Japanese	Korean	English	Total	Visualization
Effort	15.0 (−2.10)	14.4 (−2.80)**	25.5 (1.43)	34.7 (5.11)***	16.5 (−1.71)	21.3
Self	21.4 (0.17)	35.6 (6.03)***	17.0 (−1.33)	12.0 (−3.40)***	15.0 (−2.11)	20.9
Culture	12.8 (−2.13)	14.9 (−1.69)	19.4 (0.20)	22.1 (1.29)	25.1 (2.34)	18.9
Body	13.7 (−.09)	10.9 (−1.49)	10.9 (−1.25)	18.2 (2.02)	15.7 (0.81)	13.9
Relationships	6.8 (−1.72)	14.7 (2.41)	13.0 (1.21)	6.8 (−1.98)	10.1 (−0.18)	10.5
Collective Well-being	17.5 (9.54)***	0.6 (−3.41)***	1.2 (−2.39)	0.0 (−3.69)***	6.0 (1.23)	4.4
Pleasure	3.8 (0.98)	4.0 (1.39)	2.8 (0.05)	0.6 (−2.24)	2.6 (−0.15)	2.8
Money	0.4 (−2.08)	1.1 (−1.71)	4.0 (1.37)	2.6 (−0.04)	5.2 (2.63)	2.6
Recognition	3.8 (1.90)	1.1 (−1.19)	2.4 (0.40)	1.3 (−0.94)	2.2 (0.21)	2.1
Transcendence	0.9 (.58)	0.0 (−1.41)	0.4 (−.34)	0.3 (−0.57)	1.5 (2.01)	0.6
Other	3.8 (1.79)	2.6 (.57)	3.2 (1.18)	1.3 (−1.01)	0.0 (−2.39)	2.1

Note. χ² (40) = 306.76, p < .001. *** = p < .001, ** = p < .05. The left number in each cell is the percentage of resolutions adhering to the value from the total number of coded resolutions in a sample of 200 tweets per language (n = 1,000). Numbers written in parentheses are standardized residual scores based on the chi-square posthoc analysis. In the chart (right side), positive predictors of the chi-square value are expressed as blue colors, whereas red colors signify negative predictors of the chi-square value.

The least prominent values were pleasure (2.8%), money (2.6%), recognition (2.1%), and transcendence (0.6%). The functional absence of spiritual or religious resolutions invoking the value of transcendence is striking given our framing of New Year’s resolutions as a ritual. While early research on rituals presumed a connection to sacred or transcendent aspects of culture (Bellah, 2005), our results situate this ritual squarely within the realm of the secular and mundane (Goffman, 1982). The lack of importance designated to recognition, money, and pleasure is also interesting given the high-visibility of flaunting wealth and status on social media (Marwick, 2015). Despite these precedents, Twitter users themselves do not publicly claim to care about these values – at least in the context of personal goals.

Next, we compared the distribution of values between the five languages using a chi-square analysis, revealing statistically significant differences (χ² [40] = 306.76, p < .001), see Table 2. We then conducted a posthoc analysis of the chi-square test which showed that collective well-being in German is a significant positive contributor to the overall chi-square value – basically, a concern with collective well-being (typically through environmentalism) is significantly more prominent in German than the other languages. We also found that effort in Korean, valued in nearly 35% of resolutions, is a significant positive predictor of difference. Finally, the centrality of the self, invoked in over 35% of Italian resolutions, is a major contributor to the overall chi-square value. We found no significant differences in the patterns of valuation among the English or Japanese resolutions. In what follows, we discuss the implications of these findings in relation to existing literature on cross-cultural differences in value orientations.

Discussion

Whether sharing a goal, supporting a friend, or mocking other people’s behavior, Twitter provides a platform for a transnational social media ritual. We found strong commonalities in the enactment of the ritual and its associated values despite variations in the demographic profiles of Twitter users and our selection of languages designed to surface cross-cultural differences. The discourse surrounding New Year’s resolutions focuses on declarations about work and effort, alongside a focus on health and happiness in each language. We also found that sharing sincere resolutions is the most common way to participate while sarcastic responses to resolutions are largely limited to English and Italian tweets. Finally, we discovered that resolutions across languages consistently construct the importance of effort, the self, and the consumption and production of culture. Yet there are some statistically significant differences between languages: an emphasis on effort in Korean, the self in Italian, and collective well-being in the context of environmentalism in German.

While our analysis shows that New Year’s resolutions are surprisingly consistent across languages, the social media ritual is far from being monolithic as it encompasses competing practices of valuation. Synthesizing our findings, we identified three fundamental tensions underpinning the ritualistic expression of values: self-acceptance vs. self-improvement, public vs. private, and conformity vs. oppositionality. Interestingly, these tensions do not neatly correspond with previous findings on cross-cultural differences in value orientations, an observation we explain by way of a final overarching tension between national contexts and platform cultures.

Self-acceptance vs. self-improvement

The idea of change is at the heart of New Year’s resolutions. As the saying goes, ‘new year, new you.’ While prior research on resolutions has depicted change almost exclusively in terms of self-improvement (e.g., losing weight or quitting smoking), our data reveals an alternate understanding of change organized around the notion of self-acceptance. The orientations largely align with the two most prominent values in the corpus: effort and the self. Resolutions which frame effort as important, from getting organized to earning better grades in school, tend to promote the idea of self-improvement. On the other hand, resolutions which frame the self as important, from loving yourself unconditionally to ignoring what other people say, tend to promote self-acceptance.

Self-acceptance resolutions resist the pressure to change in order to meet normative standards and, in so doing, help establish new evaluative principles. This pushback against the imperative to improve oneself demonstrates how communication on social media not only reflects pre-existing values but also takes part in the dynamic construction of ideas about what is important and worthwhile. Thus, our data captures what can be seen as an ongoing framing battle not just about resolutions but about the meaning of the good life in the broader sense. As detailed below, this type of resistance to dominant paradigms of success or beauty resonates with other movements on social media and is part of a cultural trend on these platforms.

Public vs. private

Sharing visions of the good life on social media is inevitably shaped by communicative norms. Following Heinich’s distinction between public and private values, the second tension we identified distinguishes between values that are socially acceptable to talk about versus those that are not. As Heinich (2020) explains, ‘not all values stand on the same level regarding their ability to be justified. The so-called ‘moral values’ allow public expressions better than the so-called “interests,”’ (p. 60). In other words, people make evaluations according to a range of principles but are likely to directly invoke only those values which are widely endorsed in public discourse.

Although previous research suggests the distinction between public and private values is contextual and related to national repertoires of valuation (Lamont & Thévenot, 2000), we found surprising consistencies across languages. The most popular values (effort, the self, culture, the body, relationships) and least popular values (pleasure, money, recognition, transcendence) invoked in resolutions suggest a possible division between public and private values. Further research is needed to determine if the rejection of explicitly materialistic values like money and recognition holds in other aspects of Twitter’s public sphere or on other social media platforms. It is also worth examining whether such values manifest in more implicit modes of communication.

Conformity vs. oppositionality

The third tension in the expression of values reflects broad ways of participating in the social media ritual, distinguishing between conformist and oppositional stances. This tension relates to the construction of values through both the substance of media rituals and the modes of expression underpinning them. In three of the languages (German, Korean, and Japanese), people consistently shared, discussed, and commented upon resolutions in a serious and straightforward matter. Conversely, in English and Italian, numerous responses espoused a non-conformist stance to the ritual. This oppositional stance was evident in humorous and sarcastic metacommentary that challenged the importance or possibility of self-improvement. For example, among the many resolutions in English about the importance of reading, one user quipped about turning on television subtitles. Similarly, some Italian tweets subverted the general optimism of New Year’s resolutions by sarcastically proposing a much darker goal: to die. Where sincere resolutions call for self-improvement or self-acceptance, much of the metacommentary calls such values into question in favor of pleasure or personal autonomy through a non-conformist stance towards the ritual itself. Defiant Twitter users shared more in common with each other regardless of language than with people sincerely sharing and discussing resolutions.

Between national contexts and platform cultures

Alongside the aforementioned similarities, we found cross-cultural differences in the enactment of the ritual. At face value, some of these results do not align with the conventional distinction between the value orientations of Eastern and Western countries (Hofstede, 2003). First, the values of Japanese and Korean resolutions do not neatly align with each other, nor do the German and Italian ones. Since Twitter users around the globe use English, it does not allow the same kind of geographic comparison; still, the absence of any qualities distinguishing English from the other languages is striking.

Second, an evaluation of the statistically significant differences in value expression yields some complex observations regarding the notions of collectivist self-transcendence and individualist self-enhancement. The most explicit focus on the self is associated with a Western country (Italy) as might be expected, yet it lacked proportionate attention in the German and English resolutions. Even more curious is the strong focus on collective well-being in German resolutions, seemingly in defiance of an individualist orientation. Environmentalism, by definition, involves caring about something broader than oneself. However, since support for the German environmentalist movement tends to come from the young and educated (Wallis & Loy, 2021), posting environmentalist resolutions may represent a form of personal branding and self-enhancement. Nor is the phenomenon reducible to the context of Twitter resolutions since pro-environment discourse in Germany is prominent across social media platforms (Allgaier, 2020). A similar twist may be at work in the Korean emphasis on effort. At face value, the prominence of effort could be interpreted as evidence of self-transcendence, given the cultural construction of hard work, nationalism, and Confucianism during the economic boom in the second half of the twentieth century (Kim & Park, 2003). Yet a look at the resolutions themselves (and the semantic networks) shows instead that effort strongly relates to the fulfillment of self-enhancing goals such as achieving professional success.

While a full explanation of these findings is beyond the scope of this study, an initial reading may tie them to differential levels of socio-economic anxiety. Financial instability in the 1990s and 2000s has had major repercussions in South Korea (Kim & Park, 2006) and Italy (Bull, 2018), while Germany navigated this period relatively untouched (Bulmer, 2014). Building on Inglehart’s (1990) theory that economic development drives a shift towards outward-looking value priorities such as environmentalism, we cautiously propose that worsening socio-economic conditions may drive a shift towards inward-looking value priorities. Problematizing the teleology of progress embedded in Inglehart’s theory, we can speculate that relative socio-economic stability allows German-speakers to focus their resolutions on long-term societal goals while increasing pressure to function in a ‘flexible low wage economy’ (Kim & Park, 2006, p. 455) motivates Korean-speakers to focus on individual achievement. Similarly, the enduring effects of the Eurozone crisis (Bull, 2018) may be a factor behind a disproportional concern for self-care and the widespread use of dark humor in Italian tweets.

Alongside these differences, the similarities between the languages in the enactment of the social media ritual suggest the existence of another cultural force acting beyond familiar and well-studied national contexts: the culture of the platform. While the values expressed on Twitter are as diverse as the people using it, the platform itself is not value-neutral. Following Kumar’s (2021) aforementioned analysis, as well as Burgess and Baym’s (2020) pioneering biography of Twitter, we may cautiously speak of self-expression and, relatedly, the expression of unique or subversive voices as core values of this platform.

Against this backdrop, the evaluative pushback of ‘love yourself’ resolutions can be read as complementing other prevalent social media movements that aim to resist dominant social imperatives. Such resolutions echo, for instance, body positivity activists who challenge hegemonic beauty standards (e.g., Retallack et al., 2016). In a similar vein, one can easily associate the snarky metacommentary about the ritual with the irreverent culture of humor, parody, and satire evident in many reports about (English-speaking) Twitter (e.g., Highfield, 2016; Holton & Lewis, 2011). Interestingly, we only found this trend in the Italian and English datasets. While a full explanation of the absence of sarcasm in German, Japanese, and Korean resolutions would require further investigation, this finding may be related to the intersection of the particularities of the social media ritual and the national contexts, as prior research has found confrontational political humor in Germany (Bogerts & Fielitz, 2019) and Korea (Park, 2013). Another possible explanation might relate to cultural differences in navigating conflict given the well-documented avoidance of direct expressions of conflict in Japan (Abe, 1995). Overall, it seems that the enactment of social media rituals is influenced by multiple and overlapping ideas of culture, including local (or, in this case, national) contexts, global platform contexts, and changing socioeconomic conditions.

Conclusion

When we began collecting data during the final days of 2019, COVID-19 was just a string of characters for most people on this planet. Twitter users were playfully moaning about crowded gyms not knowing that within two months, open gyms could not be taken for granted. Yet this collective sharing of dreams and frustrations provided an opportunity to examine the complex intertwining of global and local dimensions of social media rituals, shedding light on how such rituals express and shape who we are and what we care about.

An analysis of tweets about New Year’s resolutions shared in English, German, Italian, Japanese, and Korean revealed significant commonalities in what people talked about, how people participated in the ritual, and the kinds of things people said were important in life, along with some distinctive cultural characteristics. We identified three fundamental tensions organizing the expression of values in the ritual: self-acceptance vs. self-improvement, public vs. private, and conformity vs. oppositionality. While these dimensions helped surface some differences between the five languages, the differences did not map neatly onto established theories of a cultural divide between East and West. We proposed a final overarching tension between national contexts and platforms cultures to explain the unexpected results.

By way of conclusion, we wish to discuss the contributions of our findings to the study of values on social media. First, the context of social media requires a reevaluation of the distinction between norms and values in the literature. Our study highlights the dialectical relationship between values and communicative norms on social media, where each shapes the conditions of possibility for the other. Although values and norms have always been entwined, the dynamics of this relationship are intensified on social media where every message is potentially public. The explicit articulation of some values such as effort or the self seem to have broad transnational legitimacy, while others such as money and recognition fall outside the norms of expression associated with this ritual. This perception of communicative norms as key to the construction of values leads to our second observation: that the distinction between collectivist and individualist value orientations is problematized when exploring actual communication on social media. This is readily exemplified through the environmentalist resolutions of German-speaking Twitter users: while the content of such resolutions is other-directed, their expressive context may lead to their interpretation as self-directed. A similar tension appears in the prevalent expression of sarcastic metacommentary in English and Italian: while the content of these tweets is subversive, their collective expression may be seen as conformist since it follows the platform’s norms. In this sense, these modes of communication resemble the monumental scene from Monty Python’s Life of Brian when the crowd echoes enthusiastically, in one voice, the ‘prophet’s’ words: ‘Yes! We’re all individuals! Yes! We ARE all different!’ (and a single man replies: ‘I’m not’).

Our study also offers three methodological contributions. First, the modes of participation we identified can be tested in the context of other social media rituals. Similarly, our typology of values can be adapted for other investigations of social media content, helping to systematize the analysis of valuation practices. Finally, our study demonstrates the utility of inductive cross-cultural comparisons for understanding the construction of values in digital spheres. The public expression of values is shaped by communicative norms, yet norms are context-specific: what is suppressed in one context might be an acceptable form of expression in another. Therefore, cross-cultural comparison of social media content offers a promising path to uncover the norms governing the construction of values.

While the scope of this project is much broader than any research on New Year’s resolutions conducted to date, it has some limitations. First, five languages fall far short of representing the cultural complexities of globalization. Further research should examine other regions and dynamics, such as the division between the Global North and South. In addition, our analysis is focused on Twitter, a primarily textual platform that does not account for the multimodal construction of values on social media. The analysis of images may shed further light on the implicit expressions of values where, for example, on a platform such as Instagram people might refrain from explicitly discussing beauty even as they promote the value through the photos they share. Finally, in examining a transnational social media ritual, we chose to focus on an arena with a strong potential for conveying cultural similarities since the practice of making New Year’s resolutions is intrinsically connected to the idea of a fresh start. While such a process of comparing apples to apples was necessary for this exploratory foray, in future research we aim to also investigate expressive rituals or genres associated with specific geographic areas (such as ‘Karen’ videos in the United States or mukbang in Korea).

Notwithstanding these limitations, our study provides compelling evidence about prominent modes of value articulation on Twitter, revealing both widespread practices of public valuation and more localized ones. In particular, the strong focus on the self, especially the trajectory of self-acceptance, has been identified as a transnational component that resonates with what we can provisionally define as a global platform culture. At the same time, our findings suggest that some of the existing ideas about the culture of the platform, including the idea of Twitter as a site for sarcasm and confrontation, may only apply to some language communities or ritual types. We hope that future studies will build on the empirical observations and analytical tools offered in this project to deepen our understanding of the globalizing digital public sphere.

Acknowledgement

We are grateful to the two anonymous reviewers, Paul Frosh, Avishai Green, and CJ Reynolds for their helpful comments on this paper. We would also like to thank the members of the DigitalValues advisory board Tartelton Gillespie, Yuko Kawai, Eun-Ju Lee, Carsten Reinemann, and Roberta Sassatelli for their insight and support. Finally, we would like to thank Anna Balestieri, Katharin Denisova, Adam Ota, Tim Park, and Rebecca Zola for their assistance in coding the data.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme [grant agreement No 819004].

Notes on contributors

Blake Hallinan

Blake Hallinan is a postdoctoral fellow in the Department of Communication and Journalism at the Hebrew University of Jerusalem researching how technologies such as ratings, reviews, engagement metrics, and recommendation systems shape cultural notions of value and worth [Email: [email protected], Twitter ID: @blakeplease].

Bumsoo Kim

Bumsoo Kim is a member of the digital values team, and his research also includes incivility and communication network heterogeneity. He is an assistant professor at Joongbu University in South Korea [Email: [email protected], Twitter ID: @bkim29].

Saki Mizoroki

Saki Mizoroki is a doctoral student at the University of Tokyo and a Visiting Research Fellow at the Hebrew University of Jerusalem, Israel. Her research focuses on feminist media studies, drawing on her experience as a veteran journalist [Email: [email protected], Twitter ID: @SMizoroki].

Rebecca Scharlach

Rebecca Scharlach is a PhD Candidate in the Department of Communication and Journalism, the Hebrew University of Jerusalem. Her research focuses on critical platform studies [Email: [email protected], Twitter ID: @rscarlets].

Tommaso Trillò

Tommaso Trillò is a postdoctoral researcher at the Department of Communication and Journalism of the Hebrew University of Jerusalem. Tommaso holds a PhD in Political Sciences from the University of Łodz, Poland, where he worked as part of the Marie Skłodowska-Curie ITN GRACE [Email: [email protected], Twitter ID: @TommyTrillo].

Mike Thelwall

Mike Thelwall is Professor of Data Science and leader of the Statistical Cybermetrics Research Group at the University of Wolverhampton, UK. His current research includes identifying and analysing social web phenomena using quantitative-lead research methods. Mike has developed the Mozdeh suite of free tools for gathering and analysing social web data [Email: [email protected], Twitter ID:@MikeThelwall].

Elad Segev

Elad Segev is an Associate Professor in International Communication and the Head the Department of Communication at Tel Aviv University. He studies the relationship between information and power from cross-national perspectives, employing data mining and network analysis techniques [Email: [email protected]].

Limor Shifman

Limor Shifman is a Professor at the Department of Communication and Journalism, the Hebrew University of Jerusalem. She is currently the PI of the ERC-funded project DigitalValues: The Construction of Values in Digital Spheres [Email: [email protected], Twitter ID: @limor_shifman].

Appendix Shared

and unique keywords found in New Year’s resolution tweets

Table

Shared keywords	English	Italian	German	Korean	Japanese
Happy	14,210 (10.4%)	177 (2.2%)	65 (1.4%)	141 (1.6%)	56 (2.1%)
Start	8,627 (6.3%)	243 (3.0%)	89 (1.9%)	36 (0.4%)	209 (8.0%)
Health	3,703 (2.7%)	77 (1.0%)	108 (2.3%)	155 (1.7%)	115 (4.4%)
Work/Job	5,242 (3.8%)	426 (5.3%)	79 (1.7%)	342 (3.8%)	133 (5.1%)
Myself	4,461 (3.3%)	142 (1.8%)	45 (0.9%)	–	347 (13.3%)
Friend	3,061 (2.2%)	211 (2.7%)	95 (2.0%)	–	27 (1.0%)
Book	2,197 (1.6%)	109 (1.4%)	28 (0.6%)	–	46 (1.8%)
Love	4,339 (3.2%)	150 (1.9%)	112 (2.4%)	69 (0.8%)	–
Buy	1,381 (1.0%)	43 (0.5%)	47 (1.0%)	43 (0.5%)	–
Twitter	3,962 (2.9%)	173 (2.2%)	201 (4.2%)	241 (2.7%)	–
Continue	1,364 (1.0%)	107 (1.3%)	–	102 (1.1%)	155 (5.9%)
Eat	3,020 (2.2%)	108 (1.4%)	–	92 (1.0%)	59 (2.3%)
Money	1,744 (1.3%)	67 (0.8%)	30 (0.6%)	133 (1.5%)	–
Read	3,025 (2.2%)	207 (2.6%)	80 (1.7%)	92 (1.0%)	–
Smoking	788 (0.6%)	43 (0.5%)	35 (0.7%)	31 (0.3%)	–
Learn	3,377 (2.5%)	–	63 (1.3%)	229 (2.5%)	72 (2.8%)
Spend	1,397 (1.0%)	55 (0.7%)	–	22 (0.2%)	78 (3.0%)
Examples of unique resolutions in each language	God, Mental, Positive, Single, Fitness	Balance Sheets, Graduation, Anxiety, TV, Instagram	Nazis, Politics, Climate Protection, Right-wing party, Environment	BTS, Certificate, Lottery, Physical Energy, Chinese	Feel, Dream, Shrine, Collaboratively, Oshi,

Note. Percentages were calculated based on the total number of tweets by language. This table includes keywords shared in four or more languages - the full list of keywords is presented in the appendix.